Developing novel antibacterials targeting the glyoxalase system using kinetic target guided synthesis Carrie-Anne Molyneaux Thesis submitted for the degree of Doctor of Philosophy Supervised by Dr Jess Healy Department of Pharmaceutical and Biological Chemistry School of Pharmacy University College London 20171 I, Carrie-Anne Molyneaux, confirm that the work presented in this thesis is my own. Where information has been derived from other sources, I can confirm that this has been indicated in the thesis. 2 Dedicated with love to my parents and grandparents. 3 Acknowledgements I would like to thank my supervisor, Dr Jess Healy, for this amazing opportunity to work on such an interesting project and for all her help and encouragement. To Dr Geoff Wells, for your help with docking studies and general lab questions. To Prof Frank Kozielski and Dr Sandeep Talapatra, for always being willing to answer my biology questions and giving me help when I needed it. To Dr Abil Aliev for help with the 19F-NMR. To all the members of G25: Sophie, Niall, Mohammed, Nikolaus, Shaista and Po-Chang. I’ve really enjoyed going through this journey with you all. The conversations, support and laughs helped every day. To the newer members of the Healy group, Ewelina, Jingyuan and Weixian, it was a pleasure to get to know you all and I wish you well in your studies. To Tammy, Milan, my extended family and friends for all their encouragement and support. To my parents and grandparents, to whom this thesis is dedicated. Thank-you for teaching me to work hard, to keep persevering, believing in me and supporting me to work towards my dreams. To Mark, for your encouragement and support every single day, I couldn’t have done this without you. Thank-you! 4 Abstract Antibiotic multidrug resistant infections are increasingly challenging to treat. To avoid entering a post antibiotic era novel antibiotics acting via new or underexploited mechanisms are required. In the human body bacteria are exposed to numerous stresses, e.g. methylglyoxal, and detoxification is essential for bacterial survival and infection progression. Glutathione and the glyoxalase (Glx) system are essential to this detoxification process (Fig. A). Fig. A. Schematic of the glyoxalase pathway. We hypothesise that inhibitors of bacterial GlxI and/or GlxII will disrupt this protective pathway leading to bactericidal activity, and may act alone or synergistically with existing antibiotics and the host immune system. To identify potent and selective chemical tools to investigate our hypothesis kinetic target guided synthesis (kTGS), where the enzyme generates its own inhibitors by acting as a catalyst to ‘click’ fragments together within the binding site (Fig. B), was employed in 2 ways; 1) a biased and 2) an unbiased approach. The chemistry used was a Huisgen [3+2] cycloaddition between azides and alkynes to form triazoles and hit compounds were identified by LC-MS/MS. Fig. B. Schematic of kTGS. In the biased approach an azido derivative of glutathione (GAz) was designed and synthesised. kTGS with GlxI and GlxII using 96 alkyne fragments to probe the hydrophobic pocket resulted in hit rates up to 28%. The unbiased approach used azide and alkyne fragment libraries to identify non-peptidic inhibitors and to investigate the potential of kTGS as a tool for fragment based drug discovery. Using 2 azide and 2 alkyne fragments and the corresponding fragment library resulted in hit rates up to 11%. Identified peptidic and non-peptidic hits were synthesised and biophysical evaluation demonstrated binding to and inhibition of the glyoxalase enzymes, thereby demonstrating the potential of kTGS for lead discovery and the opportunity to develop these hits into novel antibacterial agents. 5 Impact statement The research carried out in this thesis has the potential to have an impact in a number of ways: towards antibacterial drug research; in the further development of kinetic target guided synthesis as a tool for chemical biology and drug discovery research; towards publications; towards funding opportunities for post-doctoral and PhD students to carry on this research; the potential to become a spin-out company based on this technology. Novel antibacterials, which act via new or underexploited mechanisms of action, are essential if we are to avoid entering a ‘post-antibiotic’ era. The glyoxalase system was identified as a target of interest due to its role in detoxification to allow the bacteria to survive and the infection to progress. Compounds targeting the glyoxalase pathway could be used in combination with current antibacterials or as a single therapy. Our research has shown that the absence of the glyoxalase genes resulted in the resensitisation to existing antibiotics and other toxins and thus the glyoxalase system could be a valid target. We have identified a starting point to design chemical tools which could disrupt the glyoxalase pathway through similar binding interactions as the essential co-factor, glutathione. Kinetic target guided synthesis is an exciting technique that allows the rapid exploration of a binding site and the generation of hit compounds through in situ linking due to the biological targets ability to act as a catalyst to link building blocks together that are a best fit within the binding site. In our first biased kTGS approach we designed and synthesised an azido derivative of the natural substrate glutathione to carry out a biased kTGS with Glyoxalase I and II enzymes to identify novel inhibitors. Our second approach, an unbiased kTGS, has the potential to have a huge impact on drug discovery. The unbiased kTGS aligns with fragment based drug discovery where the in situ fragment linking could not only identify small novel inhibitors but also have a significant impact on the time from fragment to lead identification, allowing drugs to enter the market earlier than current research projects allow. The unbiased kTGS would have a second impact in terms of identifying compounds against biological targets where no chemical tools are known, this would enable drug discovery projects to be explored on new or unknown biological targets. The hit compounds identified in my research using both biased and unbiased kTGS approaches have been synthesised and evaluated against glyoxalase I and II, resulting in a number of compounds being selected for development, using traditional medicinal 6 chemistry techniques, from lead compounds into novel antibacterial agents. The results of this research should lead to at least 3 publications and will have an impact on funding applications for post-doctoral and additional PhD’s to continue this research. The methodologies and standard operating protocols optimised for chemical synthesis, analysis and biological assays during my PhD is having an impact of other members of the Healy group to carry out their research. 7 Contents Acknowledgements ....................................................................................................... 3 Abstract ........................................................................................................................ 4 Impact statement .......................................................................................................... 5 Contents ....................................................................................................................... 7 List of Tables .............................................................................................................. 11 List of Figures ............................................................................................................. 15 List of Schemes .......................................................................................................... 23 List of Abbreviations ................................................................................................... 24 Chapter 1: Introduction ............................................................................................... 28 1.1 Background ....................................................................................................... 28 1.1.1 A brief introduction and history of antibacterial agents ................................ 28 1.2 Multidrug resistance .......................................................................................... 33 1.3 Methylglyoxal and the Glyoxalase System ........................................................ 35 1.3.1 Glyoxalase I ................................................................................................ 41 1.3.2 Glyoxalase II ............................................................................................... 44 1.3.3 Glutathione ................................................................................................. 46 1.4 Target guided synthesis .................................................................................... 49 1.4.1 Dynamic combinatorial chemistry ................................................................ 50 1.4.2 Kinetic target guided synthesis ................................................................... 51 1.5 Fragment based drug discovery ........................................................................ 53 1.6 Hypothesis and Aims ......................................................................................... 55 Chapter 2: Exploration of the glyoxalase pathway and inhibitors for antibacterial drug discovery .................................................................................................................... 57 2.1 Introduction ....................................................................................................... 57 2.1.1 The glyoxalase system ............................................................................... 57 2.1.2 Known glyoxalase inhibitors ........................................................................ 58 2.1.3 Synthesis of glutathione S-conjugates and derivatives ................................ 59 2.2 Protein expression and purification optimisation ................................................ 65 2.2.1 Glyoxalase II: Optimisation of growth, expression and purification .............. 65 2.2.2 Glyoxalase I: Optimisation of growth, expression and purification ............... 67 2.3 Preliminary SAR study ...................................................................................... 72 2.3.1 Assay optimisation: Glyoxalase II ................................................................ 72 2.3.2 Assay optimisation: Glyoxalase I ................................................................. 74 2.3.4 Evaluation of glutathione analogues using DSF .......................................... 76 8 2.4 Evaluation of antimicrobial activity of S-bromobenzyl glutathione esters ........... 82 2.4.1 Design of esters .......................................................................................... 83 2.4.2 Minimum inhibitory concentration determination and antibacterial and resistance modulator activity studies.................................................................... 83 2.5 Antibiotic susceptibilty studies with strains lacking GlxI or GlxII ......................... 86 2.6 Summary and conclusion .................................................................................. 90 Chapter 3: Biased kinetic target guided synthesis ....................................................... 92 3.1 Towards the development of inhibitors of GlxI/II using kTGS ............................. 92 3.1.1 Design of the biased azido peptide scaffold for kTGS ................................. 93 3.1.2 Synthesis of azido peptide scaffold γ-glutamyl-L-azidoalaninylglycine (GAz) ............................................................................................................................ 95 3.1.3 Biophysical evaluation of γ-glutamyl-L-azidoalaninylglycine (GAz) .............. 96 3.2 Towards the development of inhibitors of GlxII using biased kTGS ................... 99 3.2.1 Pilot experiment to establish biased kTGS conditions ............................... 100 3.2.2 Biased kTGS with GlxII ............................................................................. 103 3.2.3 Synthesis of alkyne building blocks and triazoles ...................................... 108 3.2.4 Biophysical evaluation of hits .................................................................... 112 3.3 Towards the development of inhibitors of GlxI using biased kTGS .................. 130 3.3.1 Biased kTGS with GlxI .............................................................................. 130 3.3.2 Synthesis of alkyne building blocks and triazoles ...................................... 135 3.3.3 Biophysical evaluation of the GlxI hits ....................................................... 137 3.4 Competition kTGS ........................................................................................... 149 3.5 Copper addition to the kTGS assay ................................................................. 154 3.6 Summary and future work suggestions ............................................................ 158 Chapter 4: Unbiased kinetic target guided synthesis ................................................. 163 4.1 Unbiased approach ......................................................................................... 163 4.1.1 Fragment based drug discovery ................................................................ 164 4.1.2 Fragment library design ............................................................................ 164 4.2 Towards the development unbiased kTGS methodology ................................. 167 4.2.1 Stage 1: Preliminary fragment screening .................................................. 167 4.3 Stage 2: Unbiased kTGS assay ....................................................................... 170 4.4 Stage 3: Synthesis and biophysical evaluation of the unbiased kTGS hits ...... 172 4.4.1 Synthesis of the unbiased triazoles ........................................................... 172 4.4.2 Biophysical evaluation of the unbiased triazoles ....................................... 176 4.4.3 Docking studies ........................................................................................ 186 4.4.4 Summary .................................................................................................. 189 9 4.5 Fragment screening for future unbiased kTGS ................................................ 190 4.5.1 DSF screening with GlxII........................................................................... 190 4.5.2 DSF screening with GlxI............................................................................ 192 4.5.3 Summary .................................................................................................. 196 4.5.4 STD NMR ................................................................................................. 197 4.5.5 Further investigation into the DSF conditions with Alkyne library 1 and GlxII .......................................................................................................................... 197 4.6 Conclusions and future work ........................................................................... 200 Chapter 5: Fluorine NMR assay development ........................................................... 203 5.1 19F-NMR for ligand binding analysis ................................................................ 203 5.1.1 Introduction to NMR for ligand binding ...................................................... 203 5.1.2 Glutathione S-transferase as a model system ........................................... 205 5.1.3 Protein expression and purification optimisation GST ............................... 206 5.1.4 Ligand design and synthesis ..................................................................... 207 5.1.5 Preliminary binding experiments ............................................................... 208 5.1.5.1 DSF screening results ........................................................................ 208 5.1.5.2 Preliminary 19F NMR experiments ...................................................... 210 5.1.6 Quantitative Determinations ...................................................................... 212 5.1.7 Discussion ................................................................................................ 214 5.2 Application of the 19F NMR to the kTGS hits .................................................... 215 Chapter 6: Conclusions............................................................................................. 220 Chapter 7: Experimental ........................................................................................... 224 7.1 Biological materials and methods .................................................................... 224 7.1.1 Chemicals and Buffers .............................................................................. 224 7.1.2 Growth media ........................................................................................... 224 7.1.3 Plasmids ................................................................................................... 225 7.1.4 Strains ...................................................................................................... 226 7.1.5 Primers ..................................................................................................... 226 7.2 Plasmid techniques ......................................................................................... 227 7.2.1 Transformation of competent cells ............................................................ 227 7.2.2 Mini plasmid purification ............................................................................ 227 7.2.3 Midi plasmid purification ............................................................................ 228 7.2.4 Agarose gel .............................................................................................. 229 7.2.5 Glycerol stocks ......................................................................................... 229 7.2.6 Long-term storage..................................................................................... 229 7.2.7 Short-term storage .................................................................................... 229 10 7.3 Protein Purification Techniques ....................................................................... 229 7.3.1 Growth and over expression of proteins .................................................... 229 7.3.2 Protein Purification .................................................................................... 230 7.3.2.1 Purification Buffers ............................................................................. 230 7.3.2.2 Protein Purification ............................................................................. 235 7.3.2.3 Analytical Protein Techniques ............................................................ 237 7.3.2.3.1 SDS-PAGE Gel ............................................................................... 237 7.3.2.4 Spectroscopic Techniques ................................................................. 237 7.3.2.4.1 UV/Visible Absorbance Spectroscopy ............................................. 237 7.3.2.4.2 Bradford Assay to calculate protein concentration ........................... 238 7.3.2.4.3 MALDI-TOF ..................................................................................... 238 7.3.2.4.4 Trypsin Digest ................................................................................. 239 7.4 Biophysical and biochemical assays ............................................................... 239 7.4.1 Differential scanning fluorimetry (Thermal shift) ........................................ 239 7.4.2 Enzyme kinetics ........................................................................................ 240 7.4.3 Kinetic target guided synthesis (kTGS) ..................................................... 242 7.4.4 19F-NMR binding assay ............................................................................. 243 7.4.5 STD-NMR ................................................................................................. 244 7.4.6 Disk diffusion assay .................................................................................. 244 7.4.7 Molecular modelling .................................................................................. 244 7.5 General Chemistry Experimental ..................................................................... 245 7.6 Experimental data of synthesised compounds ................................................. 248 7.6.1 Synthesis of Glutathione S-conjugates...................................................... 248 7.6.2 Synthesis of Glutathione S-conjugate esters ............................................. 259 7.6.3 Synthesis of γ-Glutamyl-L-azidoalaninylglycine (GAz) .............................. 264 7.6.4 Synthesis of Alkynes identified in the biased kTGS ................................... 270 7.6.5 Synthesis of triazoles hits from the biased kTGS assay ............................ 276 7.6.6 Synthesis of Alkynes identified in the unbiased kTGS ............................... 303 7.6.7 Synthesis of Azides identified in the unbiased kTGS................................. 304 7.6.8 Synthesis of triazoles identified in the unbiased kTGS .............................. 306 References ............................................................................................................... 312 Appendix .................................................................................................................. 319 11 List of Tables Chapter 2. Table 1. Yield of optimised conditions for GlxII purification. Table 2. Optimisation of expression conditions for the purification of GlxI. Table 3: Summary of purification optimisation and yield for GlxI. Table 4. Thermal shift with respect to GlxI and GlxII with S-blocked analogue (3) (ΔTm, °C). Table 5. Thermal shift with respect to GlxI and GlxII with glutathione S-conjugates (ΔTm, °C). Table 6. Thermal shift with respect to GlxI and GlxII with GSH fragments (ΔTm, °C). Table 7. Thermal shift with respect to GlxI and GlxII with ester prodrugs (ΔTm, °C). Table 8. Synthesised ester analogues Table 9. Minimum inhibitory concentrations of various antimicrobial agents, in the presence (+) and absence (-) of 3, for E. coli and Proteus sp. Table 10. Minimum inhibitory concentrations of various antimicrobial agents, in the presence (+) and absence (-) of 29, for E. coli and Proteus sp. Table 11. Minimum inhibitory concentrations of various antimicrobial agents, in the presence (+) and absence (-) of 30, for E. coli and Proteus sp. Table 12. Antibiotic and toxin concentrations for use in disk diffusion susceptibility studies. Chapter 3. Table 13. Thermal melt temperature (Tm, °C) and thermal shift (ΔTm) in the presence of GAz and GSH relative to that of GlxI alone. Table 14. Thermal melt temperature (Tm, °C) and thermal shift in the presence of GAz and GSH relative to that of GlxII alone. Table 15. Thermal melt temperature (Tm, °C) and thermal shift in the presence of GAz and GSH (from ampule) relative to that of GlxII alone. Table 16. Structure and identifying codes of alkyne fragments in library 1 as positioned in the 96-well plates. Table 17. Structure and identifying codes of alkyne fragments in library 2 as positioned in the 96-well plates. Table 18. Summary of alkyne synthesis conditions. SM = starting material. Table 19. Summary of synthesis of biased 1,4-triazole hits for GlxII. Table 20. Melt temperature and thermal shift of GlxII in the presence of the biased triazoles. Table 21: Summary of % inhibition data for biased triazoles with GlxII measure at 412 nm. Table 22. Specific activity and % Inhibition for GAz (32) with GlxII. Table 23. Specific activity (U/mg) and % inhibition of GlxII with biased 1,4-triazoles. Table 24. Overall summary of the % triazole formation, DSF, % inhibition at 412 nm and from specific activity at 240nm. 12 Table 25. Summary of alkyne synthesis conditions. Table 26. Summary of the 1,4-triazole synthesis for the GlxI hits. Table 27. Summary of melt temperature and thermal shift of GlxI in the presence of the biased triazoles. Table 28. Specific activity data for GlxI and % inhibition by a range of GAz concentrations. Table 29. Summary of specific activity and % inhibition of GlxI. Table 30. Summary of % triazole formation in kTGS, DSF, specific activity and % inhibition of GlxI. Table 31. Groups of alkynes for competition kTGS with GAz in GlxI or GlxII. Chapter 4. Table 32. Summary of physicochemical properties of azide and alkyne fragment libraries generated by KNIME. Table 33. Structure and identifying codes of azide fragments in library 1 as positioned in the 96-well plates. Table 34. Structure and identifying codes of azide fragments in library 2 as positioned in the 96-well plates. Table 35. Thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from Alkyne library 1. Table 36. Thermal melt temperatures (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from azide library 1. Table 37. Selected fragments for the unbiased kTGS due to the combination of ΔTm and aqueous solubility. Table 38. Unbiased ttriazole synthesis conditions. Table 39. Thermal shift (ΔTm, °C) binding results for synthesised unbiased hits to GlxII. Table 40. Thermal shift (ΔTm, °C) binding results for synthesised triazoles and corresponding alkyne and azide fragments to GlxII. Table 41: % Inhibition of GlxII by the unbiased triazoles at 10 and 100 μM, measured at 412 nm. Table 42: % Inhibition of GlxII at 10 and 100 μM, measured at 412 nm after preincubation. Table 43: Activity in U/mg and % Inhibition GlxII at 10 and 100 μM. Unbiased triazoles measure at 240nm. Table 44: Summary of % triazole formation during kTGS, ΔTm, % inhibition and specific activity data for unbiased triazoles. Table 45: Thermal shift data of fragments from alkyne library 2 with a positive ΔTm in the presence of GlxII. Table 46: Thermal shift data of azide library 2 in the presence of GlxII. Table 47: Thermal melt temperatures (Tm, °C) with respect to GlxII of selected fragments. 13 Table 48: Positive ΔTm data of fragments from alkyne library 1 in the presence of GlxI. Table 49: Positive thermal shift data of fragments from alkyne library 2 against GlxI. Table 50: Positive thermal shift data of fragments from azide library 1 against GlxI. Table 51: Positive thermal shift data of fragments from azide library 2 against GlxI. Table 52: Thermal shift (ΔTm, °C) in the presence of the selected fragments relative to that of GlxI alone. Table 53. Thermal shift (ΔTm, °C) in the presence of fragments from alkyne library 1 at 1mM and 5 mM with and without 0.01 % TWEEN20 with respect to GlxII alone. Chapter 5. Table 54. Thermal shift in the presence of fluorine containing ligands, 3 and GSH relative to GST alone (ΔTm, °C). Table 55. Average mole fraction bound % for S-(2-trifluoromethybenzyl)-glutathione (7) Table 56. Overall summary of data for F-analogues including: DSF, % inhibition at 412 nm and from specific activity at 240nm. Chapter 7. Table 57. Plasmid details. Table 58. Bacterial strain details. Table 59. Primer details. Table 60. Summary of expression conditions. Appendix. Table 61. Structures of Antibiotics discussed in Chapter 1. Table 62. DNA sequencing results for pTrccgloBH6. Table 63. DNA sequencing results for pET3a-gloA. Table 64. Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Alkyne library 1. Table 65: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Azide library 1. Table 66: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Alkyne library 2. Table 67: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Azide library 2. Table 68: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Alkyne library 1. Table 69: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Alkyne library 2. 14 Table 70: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Azide library 1. Table 71: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Azide library 2. Table 72. DNA sequence of pGEX. 15 List of Figures Chapter 1. Fig. 1. Schematic of Gram positive and Gram negative bacteria cells. Fig. 2. The five major clinically validated antibacterial targets/pathways. Fig. 3. Schematic of sugar units and cross-linking peptide chains forming the backbone to a bacterial cell wall. Fig. 4. Cartoon summary of resistance mechanisms in bacteria. Fig. 5. Schematic of selected antibiotic drugs from discovery to development of resistance. Fig. 6. Schematic of a glycation reaction forming AGE’s; CMdG, DEdG, MG-H1, CEL and MOLD. Fig. 7. Schematic of the glyoxalase system and its relationship with the efflux of potassium ions (Kef) in the bacterial cell. Fig. 8. Pymol representation of the X-ray crystallographic structure of the active Ni2+-bound form of the E. coli GlxI (PDB:1F9Z). Fig. 9. Pymol representation of the metal binding site of the active Ni2+-bound form of the E. coli GlxI (PDB:1F9Z). A: N-terminal, B: C-terminal. The water molecules sit above the nickel ions. Fig. 10. Structure of S-(N-hydroxy-N-p-iodophenylcarbamoyl)glutathione. Fig. 11. Pymol representation of the inhibitor S-(N-hydroxy-N-p-iodophenylcarbamoyl)glutathione bound to human GlxI (PDB: 1QIN) showing the “Y” conformation in metal binding site and the two oxygen’s on the substrate replacing the two water molecules in the active metal binding site. Fig. 12. Pymol representation of the X-ray crystallographic structure of GlxII in S. typhimurium (PDB:2QED). Fig. 13. Pymol representation of the metal binding site for GlxII in S. typhimurium (PDB:2QED), where the water molecules sit above the iron ions. Fig. 14. Pymol representation of a X-ray crystal structure of (A) bound GSH in human GlxII (PDB:1QH5)1 and conserved sites with predicted binding in (B) S. typhimurium (PDB:2QED). Fig. 15. Glutathione (GSH) biosynthesis. Fig. 16. GSH dependant enzymatic reduction of hydrogen peroxide, R = H. Fig. 17. Structures of glutathione disulphide (GSSG) and ethylendiaminetetraacetic acid (EDTA). Fig. 18. Representation of dynamic combinatorial chemistry (DCC). Fig. 19. Kinetic target guided synthesis (kTGS) representation utilising a click reaction. Fig. 20. First FDA approved FBDD drug Vemurafenib/Zelboraf. Chapter 2. Fig. 21. S-blocked glutathione conjugates 16 Fig. 22. Schematic of the glutathione analogues, where R1 are the S-blocked compounds, R2 the ester prodrugs and R3 the N-blocked compounds. Fig. 23. Structure of GSH and the fragments ɣ-Glu-Cys and Cys-Gly. Fig. 24. Example of ester analogues Fig. 25. SDS-PAGE Gel of GlxII Fig. 26. SDS-PAGE gel of GlxI after column Fig. 27. SDS-PAGE gel of GlxI after buffer exchange. Fig. 28. SDS-PAGE Gel of purified GlxI Fig. 29. Protein concentration and well volume optimisation. DSF curve and results for GlxII. Fig. 30. Protein and dye concentration optimisation. DSF curve and results for GlxII. Fig. 31. Results for buffer optimisation against GlxII. Fig. 32. Results for buffer optimisation for GlxI. Fig. 33. Protein concentration and well volume optimisation. DSF curve and results for GlxI. Fig. 34. Protein and dye concentration optimisation. DSF curve and results for GlxI. Fig. 35. Bar graphs of thermal shift with respect to A) GlxI and B) GlxII (ΔTm, °C) for S-(4-bromobenzyl)-glutathione. Fig. 36. Bar graphs of thermal shift with respect to A) GlxI and B) GlxII (ΔTm, °C) for glutathione S-conjugates. Fig. 37. Bar graphs of thermal shift with respect to A) GlxI and B) GlxII (ΔTm, °C) for glutathione fragments. Fig. 38. Bar graphs of thermal shift with respect to A) GlxI and B) GlxII (ΔTm, °C) for ester prodrugs. Fig. 39. Transport rates across the human cell membrane. Fig. 40. Example of ester analogues. Fig. 41. Difference in zones of inhibition for antibiotics between the parent strain (BW25113) and the gloB knock out strain (JW1643). Fig. 42. Difference in zones of inhibition for antibiotics between the parent strain (BW25113) and the gloB knock out strain (JW0202). Fig. 43. Difference in zones of inhibition for toxins between the parent strain (BW25113) and A) gloA knock out strain (JW1643) and B) gloB knock out strain (JW0202). Chapter 3. Fig. 44. Proposed screening cascade for biased kTGS. GAz is shown in the predicted binding position in GlxI (PDB: 19FZ). Fig. 45. A. Human GlxI from (PDB:1QIN) with bounds S-(N-hydroxy-N-p-iodophenylcarbamoyl) glutathione (1) and B. Human GlxII (PDB: 1QH5) with bound S-(N-hydroxy-N-bromophenylcarbamoyl) glutathione (31).1 Visualised with Chimera. Fig. 46. Structure of glutathione (GSH) and γ-glutamyl-L-azido-alaninylglycine (GAz). 17 Fig. 47. A. Predicted binding of GAz (32) in the G-site of GlxI (PDB: 1F9Z). Green lines indicated H-bond between GAz and the residues. B. Predicted binding of GAz (32) in the G-site of GlxII (PDB: 2QED). Visualised in Chimera. Fig. 48. DSF to evaluate binding of GAz to GlxI. A. Melt curve and B. Bar graphs of thermal shift with respect to GlxI (ΔTm, °C). Fig. 49. DSF to evaluate binding of GAz to GlxII. A. Melt curve and B. Bar graphs of thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII alone. Fig. 50. DSF to evaluate binding of GAz and stocks of GSH to GlxII. A. Melt curve and B. Bar graphs of thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII. Fig. 51. Representative data for ‘hit’ triazole for AlkF3. SIM scan on the left, ESI scan on the right. Fig. 52. ‘Hit’ compounds identified in LC-MS/MS spectra from pilot biased kTGS experiment with GlxII, Gaz (32) and alkyne library 1. Fig. 53. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 1 in MOPS buffer. Fig. 54. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 1 in PBS buffer. Fig. 55. Compounds identified in biased kTGS with GlxII, GAz (32) and Alkyne library 1 in PBS. Fig. 56. LC-MS/MS trace overlays for the synthesis, protein, BSA and Buffer of A. AlkF3 as a 1,4-triazole and B. AlkC2 as a 1,5-triazole. Fig. 57. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 2 in PBS buffer. Fig. 58. Compounds identified in biased kTGS with GlxII, GAz (32) and Alkyne library 2 in PBS buffer. Fig. 59. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 2 in MOPS buffer. Fig. 60. Compounds identified in biased kTGS with GlxII, GAz (32) and Alkyne library 2 in MOPS buffer. Fig. 61. Structures of alkynes purchased from chemical suppliers. Fig. 62. Bar graphs of thermal shift (ΔTm, °C) in the presence of the ligand relative to that of GlxII alone. Fig. 63. Overlay of human GlxII (1QH5 in blue) and S. enterica (2QED in grey). Fig. 64. (A) human GlxII (1QH5 in blue) and S. enterica (2QED in grey) overlay and the (B) active site of 2QED. Fig. 65. Vina docking scores for 1,4 and 1,5 triazoles in Glyoxalase II (2QED). Fig. 66. Chimera representation of 56 predicted docking in GlxII (2QED). Fig. 67. Chimera representation of 55 (A) and 66 (B) predicted docking in GlxII (2QED). Fig. 68. Chimera representation of 54 (A) and 65 (B) predicted docking in GlxII (2QED). 18 Fig. 69. Schematic of the glyoxalase pathway under analysis. Fig. 70. OD at 412 nm vs time for: A) Variation of GlxII concentration (0 – 20 µM). B) GlxII concentration at 0.2 and 2 µM. C) Variation of SLG concentration (0.1 – 0.9 mM). D) negative controls. Fig. 71. Structure of ligands evaluated by enzyme kinetics. Fig. 72. Graph of OD vs time for A) 3 and B) GAz (32) over 0 – 2000 µM. Fig. 73. Normalised slope of OD vs time curve used to calculate % inhibition of 3 and 32 over 0 - 2000 µM. Fig. 74. A) Normalised slope calculated from OD (412 nm) vs time graph and B) % inhibition at 10 and 100 µM for 3 and GAz (32). Fig. 75. % Inhibition of relative GlxII activity for biased synthesised triazoles. Fig. 76. % Inhibition of GlxII with GAz (32) at 0 – 1000 µM measured at 240 nm. Fig. 77. % Inhibition bar graphs of GlxII due to 1,4-triazoles at 10 and 100 µM measured at 240 nm. Fig. 78. Hit compounds for identified for further development targeting GlxII. Fig. 79. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz (32) and alkyne library 1 in HEPES buffer. Fig. 80. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 1 in HEPES buffer. Fig. 81. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz (32) and alkyne library 1 in PBS buffer. Fig. 82. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 1 in PBS buffer. Fig. 83. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz and alkyne library 2 in PBS buffer. Fig. 84. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 2 in PBS. Fig. 85. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz (32) and alkyne library 2 in HEPES buffer. Fig. 86. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 2 in HEPES. Fig. 87. Summary of alkyne from chemical suppliers. Fig. 88. Bar graphs of thermal shift (ΔTm, °C) in the presence of the ligand with respect to GlxI alone. Fig. 89. Schematic of the glyoxalase pathway under analysis. Fig. 90. % Inhibition of GlxI by GAz (32) over 0 – 1000 µM. Fig. 91. % Inhibition bar graphs of triazoles at 10 and 100 µM with GlxI. Fig. 92. Hit compounds for further development targeting GlxI. Fig. 93. Overlay of GlxI from E. coli 1F9Z (purple) and from human 1FR0 (grey). 19 Fig. 94. (A) E. coli GlxI 1F9Z (purple) and human GlxI 1FR0 (grey) overlaid with a bound benzyl-glutathione conjugate (79) (cyan blue) and (B) the active site of E. coli 1F9Z. (C) H-bonds in glutathione binding site of E. coli 1F9Z. (D) Bound benzyl-glutathione ligand (79) in E. coli 1F9Z. Visualised in Chimera. Fig. 95. Vina docking scores for 1,4 and 1,5 triazoles in GlxI (PDB: 1F9Z). Fig. 96. Predicted docking of 74 in GlxI (19FZ). Visualised in Chimera. Fig. 97. Predicted docking of A) 71 and B) 77 in GlxI (19FZ). Visualised in Chimera. Fig. 98. Bar graphs of % triazole formation in the competition assay group GlxI-Alk1. Fig. 99. Bar graphs of % triazole formation in the competition assay group GlxI-Alk2. Fig. 100. Bar graphs of % triazole formation in the competition assay group GlxII-Alk1-1. Fig. 101. Bar graphs of % triazole formation in the competition assay group GlxII-Alk1-2. Fig. 102. Bar graphs of % triazole formation in the competition assay of group GlxII-Alk2. Fig. 103. Bar graphs of % triazole formation between AlkE3 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 104. Bar graphs of % 1,4-triazole formation between AlkA7-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 105. Bar graphs of % 1,5-triazole formation between AlkA7-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 106. Hit compounds for further development targeting GlxII identified in the biased kTGS. Fig. 107. Hit compounds for further development targeting GlxI identified in the biased kTGS. Fig. 108. Proposed screening cascade for future biased kTGS hit identification experiments. Chapter 4. Fig. 109. Proposed screening cascade for unbiased kTGS. Fig. 110. Bar graphs of thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from Alkyne library 1. Fig. 111. Thermal melt temperatures (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from azide library 1. Fig. 112. LC-MS/MS trace overlays for the synthesis, protein, BSA and Buffer of AzG2 and AlkD1 in the SIM and MS scan. Fig. 113. ‘Hit’ compounds identified in the unbiased kTGS. Fig. 114. Bar graphs of % triazole formation for the four unbiased kTGS pilot experiments with GlxII. Fig. 115. Structure of 83 and 90 with the interactions under examination numbered. Fig. 116. HMBC of 83. Fig. 117. HMBC of 90. Fig. 118. Thermal shift (ΔTm, °C) for unbiased synthesised triazoles and their corresponding fragments with respect to GlxII. 20 Fig. 119. A) OD (412 nm) vs time for unbiased synthesised triazoles. B) normalised slope and C) % Inhibition of relative GlxII activity for unbiased synthesised triazoles at 10 and 100 µM. Fig. 120. A) Normalised curve and B) % Inhibition of relative GlxII activity for unbiased synthesised triazoles. Fig. 121. Bar graphs showing % inhibition of the unbiased triazoles against specific GlxII activity. Fig. 122. 89 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. Fig. 123. 84 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. Fig. 124. 83 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. Fig. 125. 89 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. Fig. 126. 85 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. Fig. 127. 86 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. Fig. 128. Predicted Vina docking scores (kcal/mol) for the unbiased 1,4-triazoles. Fig. 129. Bar graphs of thermal shift (ΔTm, °C) for fragments from alkyne library 2 with greater that 1 °C ΔTm in the presence of GlxII. Fig. 130. Bar graphs of thermal shift (ΔTm, °C) of fragments from azide library 2 in the presence of GlxII with a ΔTm greater that 1 °C. Fig. 131. Bar graphs of thermal shift (ΔTm, °C) of fragments from alkyne library 1 with a greater that 1 °C ΔTm against GlxI. Fig. 132. Bar graphs of thermal shift (ΔTm, °C) for selected fragments with ΔTm > 1 °C from alkyne library 2 in the presence of GlxI. Fig. 133. Bar graphs of thermal shift (ΔTm, °C) for fragments from azide library 1 with ΔTm > 1 °C in the presence of GlxI. Fig. 134. Bar graphs of thermal shift (ΔTm, °C) for fragments from azide library 2 with ΔTm > 1 °C in the presence of GlxI. Fig. 135. Bar graphs of thermal shift (ΔTm, °C) in the presence of fragments (5 mM) from alkyne library 1 with ΔTm >1 °C relative to that of GlxII alone. Fig. 136. Bar graphs of thermal shift (ΔTm, °C) in the presence of fragments (1 mM + 0.01% TWEEN20) from alkyne library 1 with ΔTm >1 °C relative to that of GlxII alone. Fig. 137. Bar graphs of thermal melt temperature (Tm, °C) of GlxII with and without 0.01% TWEEEN20. Chapter 5. 21 Fig. 138. Schematic demonstrating a non-bound and bound F-analogue resulting in peak broadening, loss of signal or peak shift. Fig. 139. Pymol representation of GST from Schistosoma Japonicum complexed with glutathione (PDB: 1U87) bound in the G-site. Fig. 140. SDS-PAGE Gel showing fractions containing glutathione S-transferase after purification and buffer exchange. Fig. 141. Example of fluorine analogues (4-10, 17). Fig. 142. Bar graphs of thermal shift (ΔTm, °C) in the presence of fluorine containing ligands, 3 and GSH relative to GST alone. Fig. 143. 19F NMR for fluorine ligands (4, 5, 6, 7, 8, 9) showing (A) F-ligand, protein and competitive inhibitor (3) (1:1:10) (B) F-ligand and protein (1:1) and (C) the F-ligand in buffer. Fig. 144. 19F NMR for fluorine ligands (10, 17) showing (A) F-ligand, protein and competitive inhibitor (3) (1:1:10) (B) F-ligand and protein (1:1) and (C) the F-ligand in buffer. Fig. 145. 19F NMR for the titration spectra for S-(2-trifluoromethybenzyl)-glutathione (7) Fig. 146. [L]/[P] vs % bound, where [P] is 100 μM. Kd 2.2 μM for 2-CF3. Fig. 147. 19F NMR for 60 showing (A) the di-F-ligand in buffer (100 μM) (B) di-F-ligand, GlxII and competitive inhibitor (100 μM: 14.5 μM: 1000 μM) (C) di-F-ligand and GlxII (10 μM: 14.5 μM). Fig. 148. 19F NMR for 72 showing (A) the F-ligand in buffer (50 μM) (B) F-ligand, GlxI and competitive inhibitor (50 μM: 25 μM: 500 μM) (C) F-ligand and GlxI (50 μM: 25 μM). Fig. 149. 19F NMR for 84 showing (A) F-ligand and GlxI (50 μM: 10 μM) (B) F-ligand, GlxI and competitive inhibitor (50 μM: 10 μM: 500 μM) (C) the F-ligand in buffer (50 μM). Fig. 150. 19F NMR for 85 showing (A) F-ligand and GlxI (100 μM: 10 μM) (B) F-ligand, GlxI and competitive inhibitor (100 μM: 10 μM: 1000 μM) (C) the F-ligand in buffer (100 μM). Chapter 6. Fig. 151. Structure and data summary of hit compounds selected for further development targeting GlxII. Fig. 152. Structure and data of hit compounds selected for further development targeting GlxI. Fig. 153. Unbiased hit compounds prioritised for further development targeting GlxII. Appendix. Fig. 154. Bar graphs of % triazole formation between AlkF3 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 155. Bar graphs of % triazole formation between AlkC1 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 156. Bar graphs of % triazole formation between AlkB4 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. 22 Fig. 157. Bar graphs of % triazole formation between Alk5 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 158. Bar graphs of % triazole formation between AlkD2-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 159. Bar graphs of % triazole formation between AlkC1-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 160. Bar graphs of % triazole formation between AlkB1-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 161. Bar graphs of % 1,4-triazole formation between AlkA5-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 162. Bar graphs of % 1,5-triazole formation between AlkA5-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Fig. 163. Bar graphs of thermal melt temperature (Tm, °C), thermal shift (ΔTm) with respect to GlxII with Alkyne library 1 at 1 mM. Fig. 164. Bar graphs of thermal melt temperature (Tm, °C) and thermal shift (ΔTm) with respect to GlxII with Alkyne library 1 at 5 mM. Fig. 165. Bar graphs of thermal melt temperature (Tm, °C) and thermal shift (ΔTm) with respect to GlxII with Alkyne library 1 at 1 mM + Tween20. Fig. 166. Bar graphs of thermal shift (ΔTm, °C) with respect to GlxII with Azide library 1. Fig. 167. Bar graphs of thermal shift (ΔTm, °C) of Alkyne library 2 in the presence of GlxII. Fig. 168. Bar graphs of thermal shift (ΔTm, °C) of Azide library 2 in the presence of GlxII. Fig. 169. Bar graphs of thermal shift (ΔTm, °C) of Alkyne library 1 in the presence of GlxI. Fig. 170. Bar graphs of thermal shift (ΔTm, °C) of Alkyne library 2 in the presence of GlxI. Fig. 171. Bar graphs of thermal shift (ΔTm, °C) of Azide library 1 in the presence of GlxI. Fig. 172. Bar graphs of thermal shift (ΔTm, °C) of Azide library 2 in the presence of GlxI. 23 List of Schemes Chapter 2. Scheme 1. Synthesis of S-(4-bromobenzyl)-glutathione. Scheme 2. Synthesis of fluorinated ligands. Scheme 3. Synthesis of γ-glutamyl-4-fluoro-L-phenylalaninylglycine (4FPhe) ligand. Scheme 4. Synthesis of ɣ-Glu-Cys. Scheme 5. Synthesis of Cys-Gly. Scheme 6. Synthesis of diesters analogues. Chapter 3. Scheme 7. Synthesis of the azido transfer reagent. Scheme 8. Synthesis of γ-glutamyl-L-azidoalaninylglycine (32, GAz). Scheme 9. Synthesis of the biased 1,4-triazole hits for GlxII. Scheme 10. Formation of GSH monitored by the reaction with Elman’s reagent. Scheme 11. Synthesis of the biased 1,4-triazole hits identified with GlxI. Chapter 4. Scheme 12. Synthesis of azide fragments. Scheme 13. Synthesis of alkyne fragments. 24 List of Abbreviations A absorbance Å Angstrom ACN/MeCN Acetonitrile Amp Ampicillin AGE advanced glycation end-product Arg Arginine Asn Asparagine br s broad singlet (spectral) BSA bovine serum albumin c concentration CDCl3 deuterated chloroform CEdG N2-(1-carboxyethyl)-deoxyguanosine CEL N-ε-(carboxyethyl)lysine ChlorAmp Chloramphenicol CMdG N2-carboxymethyl-deoxyguanosine CTD C-terminal domain Cys Cysteine d doublet (spectral) D aspartic acid Da Dalton DCC Dynamic combinatorial chemistry dd doublet of doublets (spectral) ddd doublet of doublet of doublet (spectral) dec. decompose DHAP dihydroxyacetone phosphate DIPEA diisopropylethylamine – Hunig’s base DMAP 4-(dimethylamino)pyridine DMF N, N-dimethylformamide DMSO dimethylsulfoxide DMSO-d6 deuterated dimethylsulfoxide D2O deuterated water DSF differential scanning fluorimetry dTMP deoxythymidylate kinase DTNB (5,5'-dithio-bis-[2-nitrobenzoic acid]), Ellman’s reagent d-water deionised water E. coli Escherichia coli EDTA ethylenediaminetetraacetic acid 25 Et ethyl FBDD Fragment based drug discovery FDA Food and drug administration Fmoc Fluorenylmethyloxycarbonyl g grams G-site Glutathione binding site GAz γ-glutamyl-L-azido-alaninylglycine G3P glyceraldehyde 3-phosphate GCS ϒ-glutamylcysteine synthetase Glu glutamic acid Gly glycine Glx Glyoxalase GlxI/GloA Glyoxalase I GlxII/GloB Glyoxalase II GSH reduced L-glutathione: ϒ-glutamylcysteinylglycine GSSG glutathione disulfide GST glutathione-S-transferase GS-X glutathione S-conjugates H+ proton HAGH hydroxyacylglutathione HBTU O-Benzotrizole-N,N,N’,N’-tetramethyl-uronium-hexafluorophosphate HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid His Histidine HPLC High performance liquid chromatography H-site Hydrophobic binding site HTA Hemithiolacetal HTS High throughput screening IPTG Isopropyl β-ᴅ-1-thiogalactopyranoside kDa kilo daltons Kd Dissociation constant kTGS kinetic target guided synthesis L litre l pathlength LB Lysogeny broth/ Luria broth/Lennox broth/ Luria-Bertani medium LC-MS liquid chromatography-mass spectrometry Lys lysine ITPG Isopropyl β-D-1-thiogalactopyranoside m multiplet (spectral) or medium (IR) 26 M molar MDR Multi drug resistance Me methyl MG-H1 glycation adduct hydroimidazolone mg milligrams MG methylglyoxal MHz megahertz MIC Minimum inhibitory concentration MOLD methylglyoxal–lysinedimers μL microliter μmol micromoles μM micromolar min minutes mL millilitre mM millimolar mmol millimoles MOPS 3-(N-morpholino)propanesulfonic acid mp melting point MS mass spectrometry m/z mass to charge ratio nm nanometre or wavelength NMR nuclear magnetic resonance OD Optical density PAGE polyacrylamide gel electrolphoresis PBS phosphate buffer solution PDB protein database pH potential of hydrogen to specify acidity or basicity of aqueous solution PMSF phenylmethylsulfonyl fluoride ppm parts per million q quartet (spectral) qn quintet (spectral) Rf retention factor ROS reactive oxygen species RT room temperature s singlet (spectral) or strong (IR) S. typhimurium Salmonella typhimurium S. enterica Salmonella enterica SAR Structure activity relationships 27 SIM single ion monitoring SLG S-lactoyl glutathione sp septet (spectral) STD saturation transfer difference (NMR) sx sextet (spectral) TB Terrific broth TBTA Tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine tBu tertiary butyl TGS target guided synthesis THF tetrahydrofuran, tetrahydrofolate THP tetrahydropyrimidine TFA Trifluoroacetic acid TLC thin layer chromatography Tm melt temperature ΔTm thermal shift TMS trimethylsilane TRIS tris(hydroxymethyl)aminomethane V volume VOC vicinal oxygen chelate w weak (IR) WHO World Health Organisation 28 Chapter 1: Introduction 1.1 Background The treatment of multidrug resistant infections in the clinic is becoming increasingly challenging. Antibiotic drug resistance threatens both effective treatment of infectious diseases and the ability to manage complications from chemotherapy and surgeries. Antibiotic drug resistance can be described as the resistance of bacteria to an antibiotic drug that was once effective against that organism. According to the World Health Organisation (WHO) the number of new cases of antibiotic resistance reported increases every year. In 2017 the WHO published a list of the top 12 antibiotic resistant priority pathogens and the antibiotics they are resistant to.2 The top 3 critical pathogens: Acinetobacter baumannii, Pseudomonas aeruginosa and Enterobacteriaceae, are carbapenem resistant, which together with cephalosporin’s are the best antibiotics for the treatment of multidrug resistant bacteria. The 6 high priority pathogens are: Enterococcus faecium which is resistant to vancomycin, Staphylococcus aureus which is methicillin resistant, Helicobacter pylori is clarithromycin resistant, Campylobacter spp. and Salmonellae are fluoroquinolone resistant and Neisseria gonorrhoeae which is resistant to both cephalosporin and fluoroquinolone. There are also 3 pathogens in the medium priority category: Streptococcus pneumoniae which is penicillin resistant, Haemophilus influenza which is ampicillin resistant and Shigella spp. which is resistant to fluoroquinolone. Resistance threatens medical standards and if not addressed will put patients’ lives at risk and exert a greater toll on medical and economic resources. Currently, in 2017, it is estimated that at least 700 000 people die every year of antimicrobial drug resistance. During a review on antimicrobial resistance3,4 it has been projected that by 2050 10 million people could die every year and cost the global economy 100 trillion US dollars. The development of resistance to antibiotics is a natural process, however the increasing challenge is the rate at which bacteria are developing resistance compared to the rate at which new drugs are being discovered.5 Novel antibacterials acting via new or under exploited mechanisms are, therefore, essential if we are to avoid returning to a pre-antibiotic era. 1.1.1 A brief introduction and history of antibacterial agents An antibacterial compound, or antibiotic, is defined as a compound that acts against bacterial life and inhibits the growth and reproduction of bacteria. 29 Bacteria are classified into two distinct classes, Gram-positive and Gram-negative. The class to which a bacterium belongs can be easily identified using the Gram visible staining method as described by Christian Gram in 1884,6 where Gram-negative bacterial cells are easily decolourised to red but Gram-positive bacterial cells resist the decolourisation and remain purple due to differences in cell wall structure. A simplified schematic is shown in Fig. 1 to highlight the extra thick peptidoglycan layer found in Gram-positive bacteria and the additional periplasmic space and outer membrane found in Gram-negative bacteria. Gram-positive bacteria have an outer membrane containing teichoic acids, which consist of glycerol and ribitol phosphates linked via phosphodiester bonds that form a rigid cell wall that attracts magnesium and sodium ions and an extra thick peptidoglycan layer, where they can retain the violet dye. Gram-negative bacteria have an extra outer membrane consisting of phospholipids, lipoproteins and proteins, this additional barrier presents a challenge in transporting antibacterials and dyes from the outer membrane across the peptidoglycan layer to the inner membrane and across the cell wall. Fig. 1. Schematic of Gram positive and Gram negative bacteria cells. The development of synthetic antibiotics and the discovery of natural products with antibacterial properties began in Germany in the 1880’s when Paul Ehrlich observed that different dyes could bind to bacterial, animal or human cells, and after screening hundreds of dyes against different organisms, in 1910 he discovered a dye that was selectively bound to bacteria resulting in bacterial death, this compound was called salvarsan, or arsphenamine, and exists as a mixture of a trimer and pentamer.7,8 The structures of all the antibiotics mentioned in this section are detailed in Table 61 of the Appendix. During this time, Ehrlich developed the ‘Principle of Chemotherapy’ which is defined as a chemical that can interfere directly with the growth and reproduction of microorganisms at concentrations that could be tolerated by the host and became known as the magic bullet. In 1935 prontosil was discovered, it was a red dye effective against streptococcal infections, and was later discovered to be a precursor for the sulphonamide 30 class of antibacterials as well as the first drug that could be used for bacterial infections in the bloodstream.9 Penicillins were the first natural antibiotics discovered (Alexander Fleming in 1928) and are produced by Penicillium fungi. Penicillins, containing a β-lactam core, are toxic chemicals that are produced as a defence mechanism by the fungus to kill surrounding bacteria ensuring survival of the fungus as the bacteria competes for the same nutrients. However, it wasn’t until Howard Florey and Ernst Chain found a way to isolate penicillin in the 1940’s that it became the antibiotic of choice soon replacing both salvarsan and prontosil.9 Soon after, in 1944, streptomycin was isolated from soil organisms, which they named Streptomyces griseus,10 and was used to treat Gram-negative bacteria and tubercle bacillus and became the first compound in the aminoglycoside class of antibacterials.9 The period between the 1940’s and 1960’s is often referred to as the ‘golden age’ of antibiotic discovery. During this time, several new classes of antibiotics were discovered: peptides (e.g. chloramphenicol) and cyclic peptides (e.g. cycloserine), tetracyclines (e.g. chlortetracycline), macrolides (e.g. erythromycin), glycopeptides (e.g. vancomycin), polymyxins (e.g. colistin) and additional antibiotics which belonged to existing classes, e.g. cephalosporin C to the β-lactam class, were discovered. In 1962 the first quinolone antibacterial was discovered called nalidixic acid, which was replaced by a second-generation compound in 1987 called ciprofloxacin.11 This golden-age of antibiotic discovery not only identified novel classes but also led to semisynthetic antibacterials being discovered through medicinal chemistry approaches. In the 2000’s, new classes of antibacterials were discovered, including Oxazolidinone (e.g. linezoid12), ketolide (e.g. telithromycin, which is a semi-synthetic derivative of erythromycin), glycycline (e.g. tigecycline13), lipoglycopeptide14 (e.g. telavancin, which is a semi-synthetic derivative of vancomycin) and fidaxomicin, which has a macrocyclic core.11 In 2015 a new antibiotic called teixobactin was identified from an uncultured soil bacteria screen and found to inhibit the cell wall synthesis.15 Teixobactin is the first in a new class of lipid II binding antibiotics and has been found to have a low susceptibility to resistance, however, clinical trials are still awaited. Antibiotics are grouped either by class, Gram classification, bactericidal versus bacteriostatic or by their mechanism of action. There are five main mechanisms of action that have been clinically validated11 (Fig. 2), which are summarised below: 31 Fig. 2. The five major clinically validated antibacterial targets/pathways. 1. Inhibition of cell wall synthesis: Vancomycin-type glycopeptides and antibiotics containing the β-lactam ring such as penicillin and cephalosporin’s inhibit aspects of cell wall biosynthesis. This results in weakening of the cell wall allowing an influx of water and ultimately leading to cell lysis. Bacterial cell walls are made up of two types of sugar units, N-acetylmuramic acid (NAM) and N-acetylglucosamine (NAG), in a parallel series to form a backbone, peptide units are cross-linked between NAM sugar units by a tetrapeptide chain (Fig. 3), which occurs through the displacement of a glycine with a D-alanine. Penicillin’s and cephalosporin’s inhibit this final peptide cross-linking chain ultimately resulting in water entering the cell leading to the cell swelling and therefore bursting. Fig. 3. Schematic of sugar units and cross-linking peptide chains forming the backbone to a bacterial cell wall. Penicillin’s are active against Gram-positive bacteria as the outer membrane around the Gram-negative bacteria cells make it difficult for penicillins to interact with the cell wall. Certain β-lactam antibiotics including ampicillin, amoxicillin, carbenicillin and cephalosporin’s are broad-spectrum antibiotics with activity against Gram-positive bacteria and Gram-negative bacteria that do not produce β-lactamases. 2. Inhibition of protein synthesis: aminoglycosides, tetracyclines and chloramphenicol disrupt protein synthesis by inhibiting bacterial ribosomes and therefore protein biosynthesis. Aminoglycosides, e.g. streptomycin, bind to a ribosomal subunit (30S) and 32 inhibit the growth of the protein chain. Tetracyclines, e.g. chlortetracycline, are broad-spectrum antibiotics and bind to the same 30S ribosomal subunit and prevent binding of amino-acyl-tRNA. Chloramphenicol binds to the 50S ribososomal subunit and inhibits the movement of ribosomes along mRNA and extension of the peptide chain. 3. Inhibition of DNA or RNA synthesis: Quinolone’s such as nalidixic acid and fluoroquinolones inhibit nucleic acid function, preventing both cell division and the synthesis of essential enzymes, and inhibit DNA gyrase, a bacterial enzyme involved in quaternary folding in double stranded DNA preventing DNA transcription and replication. Quinolones are primarily active against Gram-negative bacteria. Enoxacin and ciprofloxacin are second-generation quinolone compounds with broad-spectrum activity. Aminoacridine, i.e. proflavine, intercalates between DNA base pairs thereby inhibiting DNA transcription. 4. Inhibition of folate synthesis: Sulphonamides, e.g. prontosil, and antifolate drugs, e.g. trimethoprim, are usually used in combination to block two sequential steps in folate biosynthesis which prevent cell growth and division. More specifically, sulphonamides prevent the biosynthesis of folic acid by inhibiting the enzyme required for linking the component parts of folic acid. The antifolate drugs in turn inhibit the biosynthesis of tetrahydrofolate (THF), which prevents the availability of the one carbon unit required for the deoxythymidylate kinase (dTMP) building block in DNA synthesis. This does not lead to bacterial cell death but slows cell growth and division thus giving the body’s own defence mechanism time to kill the bacteria, this class of drug is known as bacteriostatic. 5. Membrane disruption: Polymyxins, i.e. colistin, and tyrothricins interact with the plasma membrane, disrupting membrane integrity and leading to uncontrolled transport of ions which is fatal for the cell. An alternative mode of action has been investigated which was unrelated to the target whereby a central mechanism generates reactive oxygen species (ROS) leading to bacterial DNA damage and cell death. The molecular details of how cell death is achieved is not yet fully understood. In a study by Kohanski et al.16 they demonstrated ciprofloxacin, chloramphenicol, ceftazidime and piperacillin increased the levels of O2- in Gram negative and Gram positive strains of bacteria that were sensitive to these antibiotics resulting in oxidative stress. These results identified a connection between antibiotic susceptibility and ROS.17 Kanamycin, ampicillin and norfloxacin were shown not to increase ROS to lethal levels.18 Imlay et al.19 and Collins et al.20 have both 33 suggested that combination treatments could improve the efficacy of antibiotics and the susceptibility of bacteria to oxidation by compromising the detoxification pathways within the bacteria therefore increasing the ROS levels. 1.2 Multidrug resistance Multidrug resistance (MDR) can be described as the resistance of a bacterial pathogen to multiple antibiotic drugs, with different mechanisms of action, that were once effective against that organism. A bacterial pathogen is classified as being MDR when more than one antibiotic is no longer effective against that organism. Bacterial pathogens rapidly evolve resistance to the antibacterial agents designed to kill them. They do this via three major mechanisms: changing the permeability of cell membranes; mutations either to the biological target or to the antibiotic; up regulation of target expression.9,11 The bacterial cell can alter membrane permeability in a number of ways, for example, by increasing the expression of efflux pumps therefore preventing the drug from entering the cell. Alternatively changes to the porin structure, by modifying the transmembrane passages, can decrease the ability of antibacterial agents to enter the cell. In addition to changes in permeability the bacterium might counteract the effect of a drug by generating an excess of the target, acquiring a mutation in the target that prevents drug binding or by producing enzymes to destroy the drug. The classic example of the latter is the production of β-lactamases to open the β-lactam ring of penicillin’s and cephalosporin’s thereby degrading the antibiotic and neutralising its bactericidal activity. Another mechanism of resistance is via plasmid transfer, where information is transferred through mutations utilising plasmids encoded with the relevant resistance mechanism information. Transferring this plasmid DNA to other cells allows recipient cells to acquire resistance to the drug. A cartoon summary of resistance mechanisms encountered in bacteria is shown in Fig. 4. 34 Fig. 4. Cartoon summary of resistance mechanisms in bacteria.21 Resistance to an antibiotic drug is inevitable, irrespective of mechanism of action, as can be seen in the schematic of antibiotic drug discovery to development of resistance (Fig. 5). Wright et al.22,23,24 have published numerous papers on the antibiotic resistome, where they report antibiotic production and resistance evolve alongside each other for survival. The resistome consists of genes involved in resistance to pathogens found in the clinic and the environment. These genes have been grouped into proto-resistance genes, which have the potential to become resistant genes through mutations and increased expression levels, and silent genes, which share homology with resistance genes but do not contribute to phenotypic resistance as they are currently unexpressed. The conversion of the proto or silent genes to become functional can occur through a horizontal gene transfer, similar to plasmid transfer, where functional genes are required from other bacteria. This research highlights the ability of bacteria to develop resistance to antibiotic drugs, whether they are derived from natural or synthetic sources. As a result, there is a necessity to develop new antibacterials with novel mechanisms of action or antibacterials that can be administered in combination with current drugs to overcome resistance through a synergistic effect with the host system. 35 Fig. 5. Schematic of selected antibiotic drugs from discovery to development of resistance. In addition to resistance mechanisms developed by the bacterium there are also socio-economic factors which have contributed to the rise of resistance, these include the wide use of antibacterials in human medicine, veterinary medicine and animal feeding to increase animal weight, the ease of obtaining antibiotics without a prescription and the use of antibiotics when not required, e.g. to treat the common cold.2 1.3 Methylglyoxal and the Glyoxalase System The glyoxalase system is a detoxification pathway that enzymatically converts reactive α,β-dicarbonyl compounds, e.g. methylglyoxal (MG), and α-oxo-aldehydes to ᴅ-hydroxyacids utilising intracellular thiols, such as glutathione, to protect the cell from modification of proteins, lipids and nucleic acids. In the absence of detoxification, the accumulation of such modifications can ultimately lead to disease and cell death. MG and the glyoxalase system have been most extensively studied in mammalian systems. In humans MG, a reactive glycating agent, is formed as a by-product during metabolic degradation of the triosephosphate glyceraldehyde 3-phosphate (G3P) and dihydroxyacetone phosphate (DHAP) by methylglyoxal synthases and is a precursor of advanced glycation end-products (AGE’s). In E. coli, MG synthase enzymatically produces MG from DHAP.25,26 MG is produced as a by-product of glycolysis, when methylglyoxal synthase reacts with sugars due to the accumulation of phosphorylated glycolytic intermediates. The activity of methylglyoxal synthase is dependent on sugar keyDiscovery/useResistanceSulfonamidesGolden Age of Antibiotic drug discovery1930 1940 1950 1960 1970 1980 1990 2000 2010 2020PenicillinStreptomycinChloramphenicolTetracyclineCephalosporinsErythromycinVancomycinMethicillinAmpicillinNalidixic acidFluoroquinolonesLinezolidDaptomyconRetapamulinFidaxomicinTeixobactin36 metabolism and, therefore, an imbalance within the cell and subsequent degradation of DHAP, triosephosphates and G3P.27 Only 0.1-0.4% of this glucosetriose flux forms MG but without detoxification MG would accumulate to toxic levels within the cell and inhibit cell growth.28-29 AGE’s have been found on lipids, proteins and DNA consequently resulting in degradation and inactivation in proteins, DNA damage through strand breaks, mutagenesis and cytotoxicity, all of which ultimately lead to harmful effects on aging and diseases. The schematic in Fig. 6, shows the formation of MG and the resulting AGE’s. MG makes irreversible modifications on mammalian DNA by reacting with deoxyguanosine and forming stable products, CMdG (N2-carboxymethyl-deoxyguanosine) and CEdG (N2-(1-carboxyethyl)-deoxyguanosine), causing genotoxic effects leading to diabetes, renal failure and increased risk of cancer (Fig. 6).30 Glycation, also known as the Maillard reaction, occurs in all living systems where reactive aldehydes make modifications to amino acids in proteins. An example of a common, albeit minor, glycation reaction is the N-terminal of glucose reacting with a lysine side chain forming fructosamine residues and N-fructosyl-lysine derivative, known as the Amadori product resulting in the formation of CEL [N-ε-(carboxyethyl)lysine] and MOLDs (methylglyoxal–lysinedimers). N-fructosyl-lysine derivatives can undergo additional irreversible reactions to form stable AGE’s and have been linked to protein dysfunction and impairment (Fig. 6). MG can also modify the guanidine side chain of arginine residues forming the major glycation adduct hydroimidazolone residues (MG-H1) as well as argpyrimidine and tetrahydropyrimidine (THP).30, 31 The deglycation of N-fructosyl-lysine residues are catalysed by fructosamine 3-phosphokinase thus repairing the damage, but there is no known mechanism for de-glycation of the MG-H1. Hydroimidazolone formation is reversible and the formation is observed during increased GlxI expression. A reason that MG glycation is so damaging to the cell is that modifications are directed at the arginine residues which are located at functional sites of proteins and MG-H1 is one of the main modifications to affect the health/disease of the proteome.30, 31 The relationship between the highly reactive dicarbonyl metabolite, MG, and the formation of AGE’s with proteins, nucleotides and basic phospholipids during glycation reactions in vivo and ultimately cell diseases indicates that detoxification is essential to 37 the survival of a bacterial cell and therefore the glyoxalase system is a key area of research.32 Fig. 6. Schematic of a glycation reaction forming AGE’s: CMdG, DEdG, MG-H1, CEL and MOLD. The thiol-dependent glyoxalase system is an efficient enzymatic detoxification system which plays an important role in most cells by neutralising the reactive electrophile MG, other α-oxoaldehydes metabolites and glyoxal-derived AGE’s thereby protecting the 38 biological system in humans from damage or disease and in bacteria by allowing cell survival. MG is detoxified by a catalytic conversion to ᴅ-lactate via the intermediate S-ᴅ-lactoylglutathione. Detoxification is accomplished by two thiol-dependant enzymes as shown in Fig. 7; glyoxalase I (GlxI, S-ᴅ-lactoylglutathione methylglyoxal lyase) and glyoxalase II (GlxII, S-2-hydroxyacylglutathione hydrolase) as well as a catalytic amount of reduced glutathione (GSH). The first step of the conversion is the non-enzymatic reversible formation of hemithioacetal from MG and GSH. GlxI irreversibly catalyses the isomerisation of hemithioacetal to S-ᴅ-lactoylglutathione (SLG) via an enediolate intermediate. GlxII catalyses the hydrolysis of the intermediate SLG to ᴅ-lactate and regenerates GSH.33,34,35 The glyoxalase system is found in the cytosol of all mammalian cells, bacterial organisms, e.g. E. coli and Salmonella typhimurium, parasitic organisms, i.e. Plasmodium falciparum and yeast, e.g. S. cerevisiae. In Gram-negative bacteria, e.g. E. coli, the efflux of potassium ions is linked to SLG as part of an emergency system termed Kef. Therefore, one of the methods of transport of potassium ions in E. coli can be inhibited by glutathione and activated by SLG, these potassium efflux systems are called KefB and KefC (Fig. 7). To balance the efflux of potassium ions, hydrogen ions enter the cell, this decreases the pH inside the cell and protonates nucleophilic groups on DNA and proteins thus preventing reactions with electrophilic toxins such as MG. Conjugation between an electrophilic species and GSH results in the efflux channel opening as the GS-conjugate is transported out of the cell, thereby allowing cell protection and survival. Electrophiles, such as MG, are activators of the KefB efflux system and provide the cell protection from MG poisoning and therefore allow cell survival. By identifying compounds that can inhibit the glyoxalase system, the potassium efflux system is likely to be affected.27, 36-37 39 Fig. 7. Schematic of the glyoxalase system and its relationship with the efflux of potassium ions (Kef) in the bacterial cell. GlxI are key enzymes for antiglycation as they are directly related to the metabolism of MG and consequently, suppression of glycation reactions. Therefore inhibition of the glyoxalase system leads to accumulation of MG and induces apoptosis as there is increased cytotoxicity due to glycation which results in antitumor or antimicrobial activity.38 Gene knock out studies in E. coli by Ozyamak et al.39 have shown increased sensitivity to MG when lacking GlxI and that GlxI is not induced when the bacterial cells are exposed to sub lethal doses of MG (< 0.1 mM), however it is induced as a protective system when the MG tolerance does not have a major physiological role or when MG concentration is > 0.4 mM. The regulation of GlxI has been correlated with MDR in some tumour cells which is suggestive of MG being involved in cell signalling. MG has been implicated in the antitumor effect of cisplatin by activation of protein kinase C δ (PKCδ) thus affecting cell signalling in response to DNA damage and ultimately resulting in apoptosis. The upregulation of glycolysis in tumour cells causes an increased glucose consumption resulting in increased GlxI expression and activity to prevent MG toxicity and promote growth.30 The overexpression of GlxI has been correlated with multidrug resistance in cancer chemotherapy due to increased metabolism of electrophiles (including drugs), proving that although numerous inhibitors have been developed many antitumor drugs recognize MG as being cytotoxic, however, there are currently no therapeutic agents to target GlxI-mediated multidrug resistance.40,29,41 Metformin, a type-2 diabetes, obesity and polycystic ovarian syndrome drug has shown potential to be an ovarian cancer treatment through the repression of GlxI expression and as GlxI is known 40 to be overexpressed in ovarian cancer cells suggests a link between overexpression of GlxI and resistance to other treatments such as cisplatin. These results indicate that metformin could be used to resensitise ovarian cancer cells to other chemotherapy treatments if used in combination.42 The mechanism of multidrug resistance is currently unknown but the overexpression of GlxI is possibly connected to the increase in MG formation. Inhibition of GlxII leads to decreased concentrations of GSH being regenerated resulting in reduced GSH adducts being formed during detoxification which directly impacts on the accumulation of SLG and therefore increases the concentration of MG. Inhibition of GlxII has been observed to not affect the rate of detoxification but it does inhibit the activation of KefB thereby inhibiting the efflux of potassium ions.26 Gene knock out studies in E. coli by Ozyamak et al.27 have shown a reduced ability to detoxify MG when lacking GlxII but did not increase the sensitivity of the cell to MG, even above concentrations of 0.4 mM of MG, due to the increased SLG pools, as a product of GlxI and HTA, being detoxified through Kef activity. Knocking out both the GlxII and Kef genes did however increase the sensitivity of the bacterial cell to MG. GlxII is implicated in cancer MDR through the upregulation of GlxII by transcription factors p63 and p73, binding to hydroxyacylglutathione (HAGH) resulting in overexpression of GlxII in the cytosol but not in mitochondria. Increased MG concentrations lead to a depletion of GSH and an accumulation of SLG during the first stage of the glyoxalase cycle. Therefore, overexpression of GlxII results in the replenishing of GSH concentration from SLG and an overall decrease of MG within the cell.43 There have been observations of higher GlxII expression for invasive tumour stages without an increase in GlxI expression.30 In summary, the glyoxalase system is an important enzymatic detoxification process in both the host mammalian cells and bacterial cells to prevent covalent modifications caused by the dicarbonyl metabolite MG on DNA and proteins. Inhibition of the glyoxalase system leads to accumulation of MG leading to cellular damage and cell death. The knock out studies carried out by Ozyamak et al.27,39 are suggestive of an overexpression of GlxI / II under conditions of stress in the E. coli as was implicated in cancer MDR, therefore a link between MDR in cancer chemotherapy and GlxI / GlxII overexpression is suggestive of a link between MDR and microbial infections.43 The mechanism of cell death as a result of MG accumulation is presently unknown. One possibility in humans is that the MG modifies mitochondrial proteins in humans including 41 the permeability transition pore which stimulates the release of cytochrome c, a water soluble heme protein found on the inner membrane of mitochondria essential for electron transport, causing mitochondrial swelling and lysis. Ferguson et al.44 have shown that high concentrations of MG resulted in destruction and degradation of E. coli DNA. However in bacterial cells, reaction of MG with cysteine residues may be enough to inhibit cell growth through protein and DNA synthesis modifications and subsequent cell impairment.26 1.3.1 Glyoxalase I As previously described, GlxI catalyses an irreversible reaction converting MG to HTA and then SLG. GlxI has been characterised in a variety of organisms including mammals, plants, yeasts, bacteria and protozoan parasites.30 The GlxI enzyme isolated from Pseudomonas putida45 and human is a metalloenzyme which exhibits a zinc (Zn2+) dependence although the GlxI enzyme isolated from Gram-negative species E. coli,9 Yersinia pestis and Neisseria meningitides46-47 as well as the GlxI enzyme isolated from Gram-positive species Clostridium acetobutylicum, Trypanosome protozoa and Leishmania major48 require Ni2+ and Co2+ ions for activation. GlxI can therefore be divided into two distinct classes; Zn2+-activated and non-Zn2+-activated.49 Honek et al.49 solved the X-ray crystal structure of GlxI from E. coli and found it to be homodimeric, consisting of two subunits each with an active site in which each metal centre is important for catalysis. Located between the interfaces of the two subunits is the glutathione binding site. GlxI is a member of the vicinal oxygen chelate (VOC) superfamily of proteins where the topological fold consists of two βαβββ units.50 Each monomer consists of two domains, the N-terminal domain which has a β1α1β2β3β4 arrangement from residues 3-60 and the C-terminal domain which has a β5α2β6β7β8 arrangement from residues 72-126, with these domains connected by an extended loop formed by residues 61-71. The four β-sheets within each subunit curve to form a catalytic pocket at the dimeric interface (Fig. 8). The site around the metal ion is formed by residues from both the N-terminal domain (Leu41, Phe43, Val54) and C-terminal domain (Ala76, Val103, Ile110, Phe112, Ile124). At the base of the catalytic pocket are four residues, consisting of His5, Glu56, His74 and Glu122, and two water molecules which form the metal binding site in an octahedral geometry (Fig. 9).51 42 Fig. 8. Pymol representation of the X-ray crystallographic structure of the active Ni2+-bound form of the E. coli GlxI (PDB:1F9Z). A B Fig. 9. Pymol representation of the metal binding site of the active Ni2+-bound form of the E. coli GlxI (PDB:1F9Z). A) N-terminal, B) C-terminal. The water molecules sit above the nickel ions. The overall fold of the structure between bacterial and mammalian is conserved.52,49,53 Literature has reported that there is a 52 - 57% identity match between the mammalian and bacterial glyoxalase I and 70% match between amino acid sequences.54,55 However, the major differences between the human and E. coli GlxI are the activating metal and number of amino acids, where the human GlxI is zinc activated and consists of 184 amino acids, whereas the E. coli is nickel activated and consists of 135 amino acids. 29 of the additional amino acids in the human GlxI are at the N-terminal forming an arm which wraps around the subunit. In addition, there are further loops between the two βαβββ units which are lacking the bacterial homolog. The active site in E. coli GlxI is larger due to an additional hydrophobic pocket consisting of 15 amino acids as an α helix and loop. Studies have shown similar structural conformations with conserved sites around the C terminalN terminalNi2+Ni2+Val103Ile110Phe112Ile124Ala76Val54Phe43Leu41Ni2+Glu56His5N terminal C terminalNi2+Glu122His7443 metal ion. As there are no published ligands bound into E. coli GlxI in the PDB, the inhibitor S-(N-hydroxy-N-p-iodophenylcarbamoyl)glutathione) (1, Fig. 10), which was developed by Honek et al.56 was studied in human with the assumption that due to the conservation between bacterial and mammalian, this ligand could be anticipated to bind in a similar fashion in E. coli. S-(N-hydroxy-N-p-iodophenylcarbamoyl)glutathione) uses a hydroxamate linkage to stabilise the analogue against hydrolysis by Glyoxalase II (GlxII), this improved stability enhanced the duration of action of the inhibitor. The X-ray crystal structure of the human Zn2+-activated GlxI (PDB: 1QIN)57 carried out by Cameron et al.53 indicated bound S-(N-Hydroxy-N-p-iodophenylcarbamoyl)glutathione) (1) was chelating the zinc ion and the hydroxy interacting with Glu172 through a hydrogen bond as shown in Fig. 11. The two hydroxyl functionalities in the inhibitor have the potential to replace the two water molecules in the metal binding site thereby conserving the octahedral coordination around the metal centre through hydrogen bonds. As in many GSH binding proteins, the peptide adopts a “Y” type conformation with the carboxylate group of the ϒ-glutamate in the tripeptide backbone interacting through hydrogen bonds with Arg122. The tripeptide backbone interacts with the phenyl rings of Phe162. The benzyl group of the inhibitor lies in the hydrophobic pocket at the back of the active site but doesn’t fill the pocket, indicating further ligand development opportunities. Fig. 10. Structure of S-(N-hydroxy-N-p-iodophenylcarbamoyl)glutathione (1). Fig. 11. Pymol representation of the inhibitor S-(N-hydroxy-N-p-iodophenylcarbamoyl)glutathione bound to human GlxI (PDB: 1QIN) showing the “Y” conformation in metal binding site and the two oxygen’s on the substrate replacing the two water molecules in the active metal binding site.57 Zn2+Glu172Phe162Arg12244 1.3.2 Glyoxalase II As described previously, Glyoxalase II (GlxII) catalyses the hydrolysis of the thio-ester SLG to regenerate GSH and form 2-hydroxycarboxylic acid or D-lactate. GlxII has a high specificity for the glutathione moiety but can also react with other thioesters.30 It exists in both the cytosol and mitochondria of animals and in the cytosol in bacteria. Characterisation of GlxII has been carried out in humans, plants, yeast (S. cerevisiae), bacteria (E. coli and Salmonella typhimurium) and the protozoan parasite P. falciparum.30, 48 GlxII is a metalloenzyme that has a binuclear active site for each monomer and can be activated by Fe2+, Fe3+, Zn2+ and Mn2+ ions. The different metal ions are important for catalysis and ligand binding, they don’t have a significant effect on the enzymatic activity but play a role in the cellular mechanisms in controlling the enzymatic activity.58 GlxII is a member of the metallo-β-lactamase superfamily of proteins, which have a αβ/βα fold and a conserved metal binding site. The N-terminal domain is formed by a conserved metallo-β-lactamase four-layered β-sheet sandwich, which consists of two β-sheets and is lined with α-helices; the C-terminal domain is formed by α-helices and is essential for the substrate binding. There are two metal binding sites located in the N-terminal domain on the edge of the β-sheet sandwich.48 The X-ray crystal structure of GlxII from S. typhimurium (PDB:2QED), as represented in Fig. 12, shows the β-sheets and α-helices surrounding the two metal binding sites containing Fe atoms. One of the Fe atoms interacts with His53, His55 and His110 and a water molecule whilst the other Fe2+ ions co-ordinates to Asp57, His58, His165 and a water molecule; these two metal atoms are bridged by another water molecule and Asp127, forming an octahedral coordination shell around the metal centre as with GlxI, as represented in Fig. 13. 45 Fig. 12. Pymol representation of the X-ray crystallographic structure of GlxII in S. typhimurium (PDB:2QED).59 Fig. 13. Pymol representation of the metal binding site for GlxII in S. typhimurium (PDB:2QED), where the water molecules sit above the iron ions. Due to the lack of bacterial crystal structures in the PDB the key residues within the binding site can be identified by superimposing the structures of human GlxII with bound GSH (PDB:1QH5) and the apo structure of GlxII from S. typhimurium (PDB:2QED) (Fig. 14), which revealed a number of conserved residues within the active site. The key residues in the human were: Lys252, Arg249, Lys143, Tyr145 and Tyr175. These residues corresponded to the following residues in S. typhimurium: Lys248, Arg245, Arg136, Phe138 and Tyr167. The carboxylate of the glycine moiety of GSH interacts with FeAsp127His110His55His53His58Asp57His165FeFe46 the Arg245 and Lys248 in the C-terminal; the amide nitrogens of the glycine and cysteine amino acids and the thiol from GSH interact with the hydroxyl group of Tyr167. The carbonyl oxygen of the central cysteine in GSH interacts through hydrogen bonds with Arg136, Lys143 is the equivalent conserved basic residue in human GlxII. The carbonyl oxygen in the ϒ-glutamate moiety of GSH interacts with Tyr145 in the human GlxII and the equivalent residue Phe138 in S. typhimurium GlxII. It has been proposed that the substrate recognition occurs through the GSH tripeptide backbone and interaction of the thioester group with the metal ions.49 Apart from an interaction between the ϒ-glutamate and Lys143, ϒ-glutamate in GSH appears to be unnecessary for binding.1 As E. coli GlxII has a 78% sequence identity with GlxII in S. typhimurium these results give an indication of how GSH will bind in E. coli GlxII.59 A B Fig. 14. Pymol representation of a X-ray crystal structure of (A) bound GSH in human GlxII (PDB:1QH5)1 and conserved sites with predicted binding in (B) S. typhimurium (PDB:2QED). 1.3.3 Glutathione Glutathione (ϒ-L-glutamyl-L-cysteinylglycine or GSH) is a low molecular weight tripeptide and the most abundant thiol found in mammalian and E. coli cells, with intracellular ZnLys252Arg249Lys143Tyr145Tyr175GSHFeLys248Arg136Arg245Phe138Tyr16747 concentrations of up to 10 mM in the cytosol and mitochondria in eukaryote cells and the cytoplasm of Gram-negative bacteria.60,61 Of the Gram-positive bacteria, only streptococcus and enterococcus produce GSH. Alternative low molecular weight thiols have been identified in Gram-positive bacteria which include mycothiol, bacillithiol and ergothioneine.62-63 GSH is an essential antioxidant in mammals, bacteria, plants and fungi and also has roles in the glyoxalase, glutathione S-transferase and Kef detoxification pathways, to neutralise reactive species and electrophiles such as MG as well as in maintaining intracellular potassium levels.60 GSH is synthesised in vivo by two cytosolic enzymes in two ATP-dependent reactions (Fig. 15).62 The first reaction is rate limiting and involves the cytosolic enzyme, ϒ-glutamylcysteine synthetase (GCS), which catalyses the formation of the amide bond between the ϒ-carboxyl group of the glutamate (Glu) residue and the amino group of the cysteine (Cys) residue. The second enzyme, GSH synthetase (GS), subsequently catalyses the formation of an amide bond between glycine (Gly) and the cysteine carboxyl group of the ϒ-glutamylcysteine dipeptide resulting in the formation of GSH. In both steps activation of the carboxyl groups is carried out by ATP so as to allow the amino functional group to carry out a nucleophilic attack to form the amide bond. Fig. 15. Glutathione (GSH) biosynthesis. An alternative pathway, which is found in a number of Gram-positive bacteria, involves a glutathione fusion protein (GshF) which performs both steps in the GSH biosynthesis. GshF consists of an N-terminal ϒ-glutamylcysteine ligase (GshA) and a C-terminal ATP dependent domain which form the amide bonds to synthesis the glutathione tripeptide.64 GSH can be degraded extracellularly by ϒ-glutamyl transpeptidase (ϒ-GT), which catalyses the hydrolysis and transpeptidation of the ϒ-glutamyl residue. The products from these reactions are substrates of ϒ-glutamyl cyclotransferase and are subsequently converted to 5-oxoproline and the amino acids, cysteine and glycine. 5-Oxoproline is then catalytically converted to L-glutamate by 5-oxoprolinase in an ATP-dependent process thus completing the cycle to resupply amino acid precursors for subsequent GSH biosynthesis. In bacteria, the synthesis of GSH is limited by availability of L-cysteine.65 This degradation reaction is also a mechanism of cell protection against a 48 wide range of electrophiles whereby during a catalytic reaction with glutathione-S-transferase (GST), GSH carries out a nucleophilic attack to produce GSH S-conjugates. These S-conjugates are then metabolised by ϒ-GT to the amino acids, Glu, Cys and Gly. The key functions of GSH are in detoxification and metabolism where it can act both enzymatically and non-enzymatically. The free sulfhydryl group of the cysteine is nucleophilic in nature and due to the electrophilic nature of compounds such as reactive oxidative species thiol-S-alkylation chemistry, reduction and conjugations can occur to protect the cell from free radicals and reactive species that may be formed during metabolism or encountered in the environment. An example of how GSH protects the cell from reactive species involves GSH acting as an electron donor in the enzymatic reduction of peroxidases, i.e. hydrogen peroxide (H2O2), to produce glutathione disulfide (GSSG), which is subsequently reduced back to GSH by NADPH-dependent glutathione reductase (GR) (Fig. 16).63 In E. coli 99.5% of GSH is found in its reduced form and 0.17 – 0.33% as GSSG.65 GSH has also been demonstrated to neutralise reactive nitrosative stress through conjugation of nitric acid, which has been found to initiate and inhibit apoptosis. GSH regulates the susceptibility of cells to nitric oxide by forming hydroxylamine and GSSG and is therefore sensitive to nitric oxide as the GSH levels are compromised by oxidative stress. GSH can also convert formaldehyde to S-formylglutathione and acts as a cofactor in the glyoxalase system. Fig. 16. GSH dependant enzymatic reduction of hydrogen peroxide, R = H. GSH also plays a vital role in metal transport, storage and metabolism and can conjugate metals through production of complexes via non-enzymatic reactions. The thiol in GSH is a source of Cys for metal binding and acts as a co-factor in redox reactions that involve metals. The Cys thiol has a high affinity for mercury, silver, cadmium, lead, zinc and copper with which it can form stable complexes. A function of GSSG is the regulation of Zn2+ content in the cells and zinc metabolism. GSSG is a versatile ligand for metals due to its two amine groups and four carboxylate groups providing a slightly larger coordination site than ethylendiaminetetraacetic acid (EDTA), which is a hexadentate ligand that chelates metal ions (Fig. 17). The amine and carboxylate of the Glu moiety is the primary metal binding position whilst the Gly residue’s role is to stabilise the complex. The Zn2+-GSSG system is relatively stable when formed in vivo.66 It has been 49 observed that bacterial cells exposed to toxic metals have decreased level of thiols due to the GSSG complexing to the metals, implying the importance of GSH towards bacterial cell survival.65 Fig. 17. Structures of glutathione disulphide (GSSG) and ethylendiaminetetraacetic acid (EDTA). During infection, the cell experience oxidative stress resulting in protein S-gluthionylation, where the Cys thiol forms mixed disulphides with protein thiols thus protecting the Cys residues against irreversibly oxidation to sulfonic acids. Protein S-gluthionylation therefore plays a major role in regulation of oxidation and nitrosative stresses, and stores GSH to prevent GSSG distribution under oxidative stress.67 Another essential role of GSH in E. coli is in the regulation of potassium ion transport by means of the gate closing of ion channels via the Kef system. GS-X conjugates formed between GSH and electrophilic species activate KefB and KefC resulting in rapid efflux of potassium ions from the bacterial cell and concomitant influx of H+ ions. The resulting modulation of the intracellular pH protects the cells from the toxic effects of electrophiles.65 In summary, the purpose of GSH is to protect the cell through nucleophilic reactions, conjugations and metal complexing resulting in detoxification and metabolism of reactive species to allow cell survival. GSH is an essential part of the glyoxalase system as the co-factor to commence the first enzymatic step with GlxI. Bacterial cells that lack GSH have been observed to be more sensitive to toxic species, such as MG, and electrophiles implying the importance of GSH to the bacterial cell for protection and survival as well as the importance of inhibition of GSH pathways in relation to multidrug resistance.65 1.4 Target guided synthesis It has been reported68 that approximately 99% of compounds from a combinatorial library are inactive in biological screens meaning that more efficient methods for the discovery of lead compounds and drugs are required to reduce the number of compounds being screened and the time taken for screening and analysis. An alternative combinatorial 50 approach is target guided synthesis (TGS), which was first conceptualised 30 years ago by Rideout et al.69 working on the in situ self-assembly of cytotoxic hydrazine for tumour selectivity. TGS employs the biological target in the self-assembly of the desired ligands from the pool of fragments within the protein binding site. This process combines synthesis and screening into a one procedure. These methods remove the need to carry out high throughput screening of large libraries of compounds, as the biological target generates its own inhibitors from the choice of building blocks provided resulting in only hit compounds being generated, and therefore fewer compounds requiring synthesis. The reactions that have been utilised in a self-assembly screen to form C-N bonds are: Huisgen [3+2] cycloaddition between azides and alkynes to form triazoles;70 reductive aminations between amines and aldehydes; amide bond formations between carboxylic acids and amines; alkylation’s between amines and alkyl bromides. In order to form C-S bonds, S-alkylation reactions between thiol compounds and alkyl halides, nucleophilic substitutions using thiol compounds to open epoxide rings and between thiol containing compound and vinyl containing compounds to form thio-ethers are most common.71 TGS can be divided into two main types: Dynamic combinatorial chemistry72,73 (DCC) and kinetic target guided synthesis68 (kTGS), which employ reversible and irreversible reactions respectively. TGS has been used to explore the binding sites of a number of proteins in therapeutic areas such as CNS, cancer, infectious and metabolic diseases and is becoming increasingly appealing for ligand identification.74 1.4.1 Dynamic combinatorial chemistry Dynamic combinatorial chemistry (DCC) is a thermodynamic or equilibrium approach whereby the reactive fragments are reversibly combined with the favourable pairs binding most tightly to the enzyme thereby skewing the equilibrium towards those products.72,73 The results of DCC are amplified towards compounds with the highest affinity for the enzyme. DCC is a longer process than kTGS; the first step involves screening a library of building blocks with the desired protein to find the fragments that bind to the enzyme; the second step involves the selected building blocks to be linked together using conventional combinatorial chemistry and reversible reactions to carry out the formation of disulphides, hemithiolacetals, hydrazine’s and imines; the third step involves screening these compounds using conventional assays to confirm the hit compounds (Fig. 18). 51 Fig. 18. Representation of dynamic combinatorial chemistry (DCC). 1.4.2 Kinetic target guided synthesis The non-equilibrium or the kinetic target guided synthesis approach (kTGS) ligates the building blocks irreversibly, forming a ligand with additive bioactivity, and selects one or more products over others due the brief acceleration of reaction rate by the target. The kTGS methodology uses the biological target to carry out the synthesis by combining the fragments that best fit the binding pockets, implying that the biological target ‘catalyses’ the reaction to identify hit compounds The function of the biological target and the compounds produced are independent of each other as the formation of a product is dependent on the fragments being held in position and being in close proximity to the binding site of the protein to undergo the reaction, one example of an irreversible reaction is the Huisgen [3+2] cycloaddition to form triazoles (Fig. 19).75-76 Fig. 19. Kinetic target guided synthesis (kTGS) representation utilising a click reaction. The Huisgen [3+2] cycloaddition reaction between azides and alkynes to form triazoles was pioneered by Sharpless and colleagues70 to generate acetylcholine esterase inhibitors and is an example of a click chemistry reaction.70 It was found to be an advantageous reaction to use for target guided synthesis and has become the gold standard method in kTGS known as in situ click chemistry. Other reactions that have been explored using kTGS include:77-78 the formation of C-S bonds through S-alkylation reactions between thiols and alkyl halides, nucleophilic substitutions using thiols to open epoxide rings and between thiols and vinyl containing compounds to form thio-ethers; 52 the formation of C-N bonds through reductive aminations, between amines and aldehydes, and amide bond formations between carboxylic acids and amines, as well as alkylations between amines and halides. Jaegle et al.79 have recently reported the use of amidation reactions. Kolb, Sharpless and Finn80,81 have also explored other click reactions, such as Diels-Alder, ring openings, oxime ether formations and oxidation reactions. The reasons that the Huisgen [3+2] cycloaddition reaction has become the standard synthesis for kTGS is that the reactive fragments involved in the synthesis are unlikely to react with other species including the biological target, the reaction rate is slow at room temperature allowing the enzyme to kinetically control the reaction and the reaction can be carried out in physiological conditions. These are also requirements for other reactions utilised in kTGS, including the ability to synthesis the ligands using standard synthetic chemistry. The azide group is a unique example of a reactive 1,3-dipole compound that is stable upon storage and used in a variety of synthetic transformations. The alkynes are generally small, non-polar fragments and once they have been activated, either by a metal catalyst or the enzyme, to a strained electron deficient species the reaction with the azide occurs with fast kinetics to form the triazole products. It was assumed that false positive hits (formed by the combination of the reactive fragments without the enzyme) would be minimised due to the positive hits (formed by the simultaneous binding of two fragments in the presence of the enzyme) overcoming high kinetic barriers with the aid of the enzyme and binding tightly with the target. Triazole compounds are advantageous as the dipole moment is equivalent to that of an amide bond and can participate in hydrogen bonding with the receptor as well as π-stacking interactions with other amino acid residues.74 This process does not use reagents that have the potential to disrupt the target enzyme or protein, i.e. catalysts. The advantages of a reaction utilised in kTGS are: synthetic tractability; the ability of reactions to be carried out in an aqueous environment at physiological pH and temperature; the stability of reactive fragments and products; minimal side product formation from reactive building blocks; the ease of monitoring reactions and obtaining useful and meaningful data; the possibility to accelerate the drug discovery process by removing the step-wise nature and by combining synthesis and screening into one step. A diverse set of fragments can be tested allowing unexpected products to be identified as knowledge of the receptor is not always required in order to carry out these in situ reactions.75, 82 Subsequent profiling on target generated compounds to determine affinity 53 and inhibitory data, as well as pharmacokinetic properties can be carried out on a smaller number of synthesized compounds, which improves the efficiency of the drug discovery process. The disadvantages of kTGS include challenges identifying small, on occasion trace, quantities of compounds formed in situ using LCMS and the resynthesis of hit compounds for confirmation using biophysical assays. The 1,3-dipolar cycloaddition between azides and alkynes to form triazoles requires harsh thermal conditions if no catalyst (copper or ruthenium) is present leading to potentially long incubation times between a few hours, to days or weeks at 37 °C in order to observe product formation and identify ‘hit’ compounds.71 Another disadvantage is the fragment choice is limited to the commercial availability or synthetic tractability of the fragments.70 To overcome challenges with low binding affinity of ligated compounds, a synthetic handle can be used so as to promote the reaction and allow exploration of the binding landscape and protein conformations. The aim of kTGS is to get information on positive and negative hit compounds with the target rather than to isolate the compounds. kTGS has been used to expand the ‘affinity range’ of known ligands so that these ‘hit’ compounds could give an understanding of the chemical space in the binding site leading to a smaller set of compounds that need to be synthesised to be profiled further in order to develop a lead compound. kTGS has been previously used to explore the binding landscape and generate inhibitors of acetylcholine esterases,68 carbonic anhydrases,83 HIV proteases,84 chitinases,85 a mycobacterial transcriptional regulator86 and an insulin degrading enzyme.87 kTGS provides the opportunity to probe the biological target with a diverse set of fragments with a range of functional groups without the need to design and synthesis numerous chemical analogues. 1.5 Fragment based drug discovery Fragment based drug discovery (FBDD) is an approach to drug discovery that identifies small fragments that have a weak binding affinity to a desired biological target, these fragments are then linked and optimized to lead compounds with high affinity and selectivity. Traditional drug discovery programs usually identify hit compounds through library screening, such as high throughput screening (HTS), and develop the hit into a lead through the synthesis of many analogues over months to years. FBDD began from publications by Jencks88 in 1981, where he described the theoretical basis of additivity 54 through change in Gibbs free energy of a bound compound (A-B) being equal to the combination of change in Gibbs free energy of A, B and a connection energy, and Ariëns et al.89 in 1982, where they described that active peptide drugs were ligands required for activity linked by pharmacodynamically inert spacers. In 1996 Shuker et al.90 published their research on the identification of ligands with high affinity by a method they called ‘SAR by NMR’, where they linked building blocks that had optimal binding for subsites within the protein. The building blocks, or fragments, were screened by NMR so as to identify fragments that bound in different pockets of the binding site. The fragments were then linked resulting in a high affinity compound. FBDD has since developed into an industry standard approach to drug discovery with many compounds entering clinical trials in the past few years due to the approach being a rapid and efficient method to identify hit compounds. A number of compounds are currently undergoing Phase 1, 2 and 3 evaluation and two compounds have FDA approval. The first FDA approved drug developed using FBDD being Vemurafenib/Zelboraf (2), Plexxikon / Roche’s cancer inhibitor targeting B-Raf (Fig. 20).91 Fig. 20. First FDA approved FBDD drug Vemurafenib/Zelboraf (2). The advantages of FBDD over HTS include; the smaller size of the building blocks (typically 100 – 300 Da) enables exploration of a greater proportion of chemical space with smaller libraries leading to higher fragment hit rates; fragments generally have lower affinity for the biological target but bind with higher ligand efficiency than hits from HTS campaigns; linking fragments88,89 results in compounds with higher affinity and binding efficiency compared to bigger HTS compounds with weaker binding efficiency; typically fewer compounds are synthesised compared to HTS. Another advantage is the availability of commercial fragment based libraries and solubility of the fragments. FBDD does have some disadvantages, due to the smaller size of fragments binding affinities are weaker as fewer interactions are made with the biological target. As a result fragment screening requires the use of sensitive biophysical methods and high ligand concentration to identify the required fragments for linking.92,93,94,95 55 As FBDD developed, cheminformatic properties were observed and the ‘Rule of three’ was introduced.96,97 The Rule of 3 for fragment-like compounds is similar to Lipinski et al.’s Rule of 5 for oral drug-like compounds,98 where the key properties were identified as being MW < 300, clogP ≤ 3, H-bond donors and H-bond acceptors ≤ 3, number of rotatable bonds ≤ 3, polar surface area = 60 Å.99 Once fragments have been identified, it is then necessary to design linkers to join two or more fragments or to grow and evolve the fragments into drug-like molecules.92,100-101 Fragment linking is therefore an important design element to take fragments to a lead compound that preserves the binding modes of the fragments to maintain the binding efficiency. Fragment linking and evolution through traditional medicinal chemistry approaches and computational design has however proved challenging therefore DCC has been used employed as a strategy to limit the disadvantages of traditional chemistry approaches.93,102 kTGS has also been used as a strategy for in situ covalent linking however the small number of chemical reactions that are suitable is a limiting factor.74, 102 1.6 Hypothesis and Aims The rapid evolution and spread of multidrug resistance and a dearth of new antibiotics leaves us in a precarious position. If urgent action is not taken to discover new antibiotics, adjuvants and resistance breakers we risk returning to a pre-antibiotic era in which modern medicine simply cannot function. What is needed to tackle this problem are new approaches, new biological targets and ultimately new drugs. Bacteria are unicellular organisms and so experience changes to their environment in a rapid and dramatic manner and rapidly detoxify stresses, such as reactive oxidative species and electrophiles, to ensure survival. This provides an opportunity to investigate whether inhibition of pathways associated with detoxification will provide a route to the discovery of novel antibiotics. It has long been suggested that the glyoxalase system could be a target for antimicrobial infections. There is evidence that inhibitors of this pathway have antimalarial, antifungal and antimicrobial activity but this hasn’t been capitalised on.103,104-105,106 As a result one of the main aims of this research was to demonstrate that the glyoxalase pathway plays a role in antibacterial drug resistance and is a viable target for antibacterial drug discovery. Glutathione and GSH dependant detoxification enzymes such as glyoxalase I & II (which detoxifies α-oxoaldehydes such as methylglyoxal) are essential to the detoxification 56 process. We hypothesise that inhibitors of glyoxalase I (GlxI) and/or glyoxalase II (GlxII) in bacterial pathogens will disrupt this protective pathway leading to bactericidal activity and may act synergistically with existing antibiotics and the host immune system. To investigate this hypothesis potent chemical probes were required. The initial small molecule probes were based on the glutathione tripeptide backbone to act as a handle in the binding site. Novel, potent and selective inhibitors of GlxI and GlxII were identified by kinetic target guided synthesis (kTGS), an exciting addition to the medicinal chemistry toolbox which involves the biological target acting as a catalyst to assemble ligands from a diverse set of fragments within the protein binding site thus allowing for rapid screening and optimisation. To provide a scaffold for kTGS optimisation using a diverse library of alkyne fragments an azido derivative of GSH, γ-glutamyl-L-azido-alaninylglycine (GAz), was designed and synthesised. Further optimisation was expected to lead to the identification of novel triazole containing mimetics based upon the glutathione conjugates that have a high affinity for the GlxI and GlxII enzymes involved in the detoxification pathway leading to further profiling into their anti-bacterial properties and function whether alone or in combination with current antibacterials. The methodology of kTGS was further developed with the aim of providing an alternative hit identification screen towards the advancement of fragment linking in fragment based drug discovery, to identify non-peptidic hits for future development into GlxI and GlxII inhibitors for antibacterial drug discovery. To support compound profiling and provide an assay which can qualitatively and quantitatively confirm the hit compounds, we have also designed a number of 19F containing GSH S-conjugates and are in the process of validating an assay system for evaluating novel inhibitors by 19F-NMR. The ultimate aim was to investigate the role of glutathione and GSH dependant detoxification enzymes in MDR and validate whether these enzymes can act as targets for antibacterial drug discovery. 57 Chapter 2: Exploration of the glyoxalase pathway and inhibitors for antibacterial drug discovery 2.1 Introduction 2.1.1 The glyoxalase system To explore if the glyoxalase system was a good target for antibacterial drug discovery, we began by comparing literature with the mammalian glyoxalase system, as this system has been more extensively studied as a drug target. As discussed in Section 1.3, there are reports that there is a 52 - 57% identity match between the mammalian and bacterial glyoxalase I and 70% match between amino acid sequences,54,55 however there are also key differences, for example an additional hydrophobic pocket in the active site in E. coli GlxI, in which to gain selectivity. In the active site of GlxI (human) the two Zn2+ ions are surrounded by amino acid residues (His127, Glu173, Gln34, Glu100) and water molecules in an octahedral metal coordination. In E. coli, the active site is surrounded by (His74, Glu122, His5, Glu56) and a water molecule in an inactive octahedral coordination around the Zn ion or an active trigonal bipyramidal coordination around the Ni2+ ion. The overall fold in GlxI is conserved. With regards to glyoxalase II, a 78% sequence identity between GlxII in S. typhimurium and E. coli, and a comparative study with mammalian X-ray crystal structures have given us an indication of how GSH will bind in E. coli GlxII.59 In the active site of GlxII (human) the Zn2+ ions are surrounded by amino acid residues (Tyr175, Tyr145, Lys143, Arg245, Lys252) whereas in GlxII (S. typhimurium) the amino acids surrounding the Fe2+ are as follows: Tyr167, Phe138, Arg136, Arg245, Lys248), however the binding mode of GSH is conserved as is the overall fold with Zn2+ metallohydrolase protein families.35 Given the sequence similarities, the mammalian studies were used as a starting point for research with the aim of obtaining selectivity for bacterial targets. To probe the SAR of bacterial glyoxalases known mammalian inhibitors were explored as a starting point. As discussed in Section 1.3. Vince,107-108 Thornalley,109 Creighton,110,111,112 Honek,113 D’Silva104,105 et al. have all shown that glutathione S-conjugates act as inhibitors of GlxI and II in mammalian and parasitic systems. With a view to explore if the activity of the glutathione S-conjugates observed against GlxI / II in humans and parasites is retained against E. coli, they were used as a starting point to ultimately lead to the discovery of novel antibacterials. To determine if the glyoxalase system was a valid target for antibacterial drug discovery, disk diffusion experiments using knock out genes from the 58 Keio collection were explored to determine how essential the glyoxalase system was as a natural defence mechanism and whether disruption of this pathway could lead to antibacterials acting via a new mechanism of action. In addition, treatment of glyoxalase inhibitors in antibiotic susceptibility studies were also carried out to determine if glyoxalase inhibitors could act synergistically with existing antibiotics and the host immune system to resensitise current antibiotics for combination therapies. 2.1.2 Known glyoxalase inhibitors To identify a starting point for antibacterial ligand design known glyoxalase inhibitors were explored. Previous studies carried out by Vince and Warn108 showed that S-blocked glutathione compounds were potent inhibitors of human GlxI, with potential for the treatment of tumours. S-(4-Bromobenzyl)-glutathione analogue (3, Fig. 21) showed a Ki of 9 µM and GSH a Ki of 3.1 mM. S-(N-hydroxy-N-4-bromophenyl carbamoyl)-glutathione (1, Fig. 21) showed a Ki of 1.2 µM and was described as a slow substrate for GlxII.112 Fig. 21. S-blocked glutathione conjugates Thornalley et al. 109,114 found that although these glutathione S-conjugates were potent human GlxI inhibitors (0.16 μM for S-(4-bromobenzyl)-glutathione) they were unable to cross the cell membrane. This problem was overcome using a pro-drug approach, in which the two carboxylate moieties were converted to esters. Once inside the cell esterase’s catalysed the hydrolysis of the di-ester pro-drugs to the active bis-acid compounds. These diester derivatives were also shown to have potent antitumor activity in vitro and in vivo due to the inhibition of the glyoxalase detoxification pathway resulting in the resensitisation of antitumor drugs and disabling MDR.43 In addition, Douglas et al.115 reported N-blocked glutathione analogues were weak inhibitors of GlxI due to the N-R group being anticipated to extend out of the binding site but could still have minor steric effects on the binding of the compounds. Creighton et al.116 suggested that the glutathione analogues could not only act as inhibitors of GlxI but also substrates for GlxII and therefore act as competitive inhibitors during the hydrolysis of SLG by GlxII. D’Silva et al.105 showed the S-blocked analogues inhibited protozoal glyoxalases and 59 demonstrated antiparasitic activity. It was again necessary to employ a prodrug strategy and it was found that the optimal ester carbon length was 4 – 6 whilst the cyclohexyl ester provided optimal antiparasitic activity with an ED50 of 0.21 µM. The N-blocked analogues were shown to be important for toxicity against amastigotes stage of the trypanosomes life cycle although were later dismissed as antiprotozoal inhibitors due to the free α-amino being required for binding.105 A schematic summarising the glutathione analogues is shown in Fig. 22. Fig. 22. Schematic of the glutathione analogues, where R1 are the S-blocked compounds, R2 the ester prodrugs and R3 the N-blocked compounds. Given the extensive literature and that S-(4-bromobenzyl)-glutathione has been employed as an inhibitor of GlxI and II and that prodrugs of this inhibitor have demonstrated both anticancer and antiparasitic activity, ligands were initially designed based on the glutathione S-conjugates (R1), in particular the known inhibitor S-(4-bromobenzyl)-glutathione, to determine whether binding to or inhibition of GlxI and/or GlxII would result in any antibacterial activity being observed. Analogues of S-(4-bromobenzyl)-glutathione would allow investigation into the SAR of the bacterial glyoxalases and the ester derivatives would provide tools to investigate the effect of glyoxalase inhibitors in bacterial survival and synergy with existing antibiotics. N-blocked analogues (R3) were not focused on at this time due to our aim to design glutathione ligands which could be developed to explore the hydrophobic pocket and into the metal active site of GlxI and GlxII and not extending out of the glutathione binding pocket. 2.1.3 Synthesis of glutathione S-conjugates and derivatives To explore the SAR around S-blocked glutathione S-conjugates (R1), where R2 and R3 were hydrogen, a number of compounds were synthesised. Synthesis of the S-conjugate, S-(4-bromobenzyl)-glutathione (3), was carried out from L-GSH reduced and 4-bromobenzyl bromide using standard alkylation conditions (Scheme 1).117 60 Scheme 1. Synthesis of S-(4-bromobenzyl)-glutathione. Reagents and conditions: (i) GSH (1 eq), 1M NaOH (aq), EtOH, 4-bromobenzyl bromide (1 eq), RT, 3 h, 82%. A set of fluorine containing glutathione S-conjugates, based on the inhibitor, S-(4-bromo-benzyl) glutathione (3), were designed and synthesised for the development of a 19F-NMR assay, described in Chapter 5. A fluorine atom was an essential element of design for the fluorine NMR experiments to observe the 19F peak in the NMR spectra. In order to explore size, potential interactions and SAR around the H-site a F atom and CF3 group were used to replace the H in all positions around the aromatic ring. The desired seven fluorinated ligands, in different positions to probe the environment (2-F, 3-F, 4-F, 2-CF3, 3-CF3, 4-CF3, 3,5-diF) (4 - 10), were obtained from L-GSH reduced and the appropriate fluorinated benzyl halide using standard alkylation conditions with yields ranging from 12 – 65% (Scheme 2).117 Scheme 2. Synthesis of fluorinated ligands. Reagents and Conditions: (i) 1M NaOH (aq), EtOH, 4.5 h, RT. 12 - 65% yield. In addition to the 7 S-alkylated analogues, an example in which the central cysteine moiety was replaced with a 4-fluoro-phenylalanine was designed and synthesised to investigate how essential the sulfur was for binding affinity. The synthesis of γ-glutamyl-4-fluoro-L-phenylalaninylglycine (17) was achieved in 4 steps from an amide coupling between N-fluorenylmethyloxycarbonyl-4-fluoro-L-phenylalanine (11) and glycine-tert-butyl ester hydrochloride (12), followed by Fmoc deprotection, and subsequent amide coupling with N-Boc-L-glutamic acid-α-tert butyl ester (15). Global deprotection furnished the desired probe γ-glutamyl-4-fluoro-L-phenylalaninylglycine (17) (Scheme 3). 61 Scheme 3. Synthesis of γ-glutamyl-4-fluoro-L-phenylalaninylglycine (4FPhe) ligand. Reagents and conditions: (i) HBTU, DIPEA, DMF, RT, 3.5 h. 84% (ii) 10% v/v piperidine/DMF, RT, 35 min. 95% (iii) HBTU, DIPEA, DMF, RT, 3.5 h. 60% (iv) Triethylsilane, 80% v/v TFA/DCM, RT, 4 h, 12%. To investigate the relative importance of the amino acids in the GSH backbone for binding and recognition within the GlxI and GlxII binding sites two fragments were also synthesised namely; ɣ-Glu-Cys (18) and Cys-Gly (19) (Fig. 23). Fig. 23. Structure of GSH and the fragments ɣ-Glu-Cys (18) and Cys-Gly (19). The synthesis of these compounds was carried out by André Filipe da Silva Campaniço118 an Erasmus student and resynthesised by Jingyuan Zhou a MSc student at UCL, School of Pharmacy. Synthesis of the ɣ-Glu-Cys fragment was achieved using standard solution phase peptide chemistry. Firstly, N-Fmoc-S-trityl-L-cysteine (20) was converted to the t-butyl ester under copper catalysed conditions119 to give (21). Fmoc deprotection under standard conditions afforded (22). Amide coupling with N-Boc-Glu-OtBu using HBTU gave the protected ɣ-Glu-Cys dipeptide (23). Global deprotection with 50% TFA/DCM in the presence of a scavenger furnished the ɣ-Glu-Cys fragment (18) in moderate yield (Scheme 4). 62 Scheme 4. Synthesis of ɣ-Glu-Cys. Reagents and conditions. (i) N,N’-dicyclohexylcarbodiimide (4.6 eq), t-butanol (6.3 eq), copper (I) chloride (0.05eq), RT, 3 days. (ii) N-Fmoc-S-trityl-L-cysteine (1 eq), RT, 3 h. (iii) 20% piperidine/DMF (v/v), RT, 2 h, quantitative. (iv) S-trityl-L-cysteine tert-butyl ester (1eq), N-Boc-Glu-OtBu (1.05 eq), HBTU (1.05 eq), DMF, DIPEA (3 eq), RT, 3.5 h, 32%. (v) 50% TFA/DCM (v/v), Triethylsilane (4 eq), RT, 9 h, 55%. Synthesis of the Cys-Gly peptide was achieved through an amide coupling between N-Fmoc-S-trityl-L-cysteine (20) and glycine tert-butyl ester using HBTU which gave (24). Fmoc deprotection of tert-Butyl-Fmoc-S-trityl-L-cysteinylglycinate (24) provided (25), which then underwent a further deprotection to furnish the Cys-Gly dipeptide (19) (Scheme 5). Scheme 5. Synthesis of Cys-Gly. Reagents and conditions. (i) N-Fmoc-S-trityl-L-cysteine (1.00 eq), glycine tert-butyl ester hydrochloride (1.05 eq), HBTU (1.05 eq), DMF, DIPEA (3 eq), RT, 3.5 h, 63%. (ii) 20% piperidine/DMF (v/v), RT, 2 h, 75%. (iii) 20% TFA/DCM (v/v), Triethylsilane (2 eq), RT, 2 h, 93%. To explore the effect of the ester prodrugs (R2), where R1 was S-(4-bromobenzyl) and R3 a hydrogen, a number of compounds were synthesised. Based on Thornalley et al.’s 63 observations as discussed in section 2.4, initial plans were to synthesis ethyl, n-propyl and cyclopentyl diesters analogues of the known inhibitor S-(4-bromobenzyl)-glutathione (3) and test the binding affinity, minimum inhibitory concentrations and synergistic effects in the presence of known antibacterials of the diester analogues compared to the parent bis-acid S-(4-bromobenzyl)-glutathione (Fig. 24).109, 120 Fig. 24. Example of ester analogues. The glutathione S-conjugate S-(4-bromobenzyl)-glutathione (3) was synthesised as previously described (Scheme 4).117 With this building block in hand, the synthesis of the ethyl, propyl and cyclopentyl di-esters were attempted using acid catalysed esterification with a catalytic amount of H2SO4 and the desired alcohol as a solvent (ethanol, propanol, cyclopentanol respectively).108 After refluxing for 16 hours the TLC showed 2 major products had been formed, further analysis by NMR and LCMS confirmed a mixture of mono and di-esters. The two products presented a significant challenge to separate by reverse phase chromatography and of particular note was the instability of the isolated di-ester which exhibited partial hydrolysis back to the mono-ester. Partial hydrolysis to the mono-ester was also observed during concentration in vacuo in a 40 °C water bath, during work up with sodium bicarbonate (saturated) and dichloromethane, during purification with Dowex ion exchange and finally during reverse phase C18 column chromatography. These instability issues hampered the isolation of a pure sample of the diester material. When the reaction was carried out in the presence of thionyl chloride at room temperature for 3 days the TLC showed 1 spot relating to the di-ester but due to the instability, outlined above, during work up or purification the di-ester hydrolysed in part affording a mixture of the mono- and di-esters. A literature search revealed the pKa strength of the glycine and glutamate carboxylic acids of the tripeptide were different, pKa of the glutamate α-COOH was 2.12 whilst the glycine ϒ-COOH was 3.73. This difference in pKa can be attributed the glutamate α-NH2, which as a result of an electron withdrawing effect stabilises the adjacent α-COOH thus 64 lowering the pKa. Amine protection was expected to reduce the acidity of the α-COOH and produce a higher yield of diester due to this loss in preference of esterification.105,121 With a new strategy in place synthesis of the diesters, started once again with the preparation of the intermediate S-(4-bromobenzyl)-glutathione (3). The free α-NH2 was protected with N-(9-fluorenylmethoxycarbonyloxy) succinimide furnishing (26) in excellent yield. The di-esters were synthesised by dissolving S-(4-bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione in ethanol, n-propanol or cyclopentanol respectively and dropwise addition of thionyl chloride. The final step to afford the desired di-esters was the deprotection of the Fmoc group, which was carried out in the presence of 10% v/v piperidine/DMF. The desired di-ethyl (29) and di-n-propyl (30) esters were produced in good yields with no stability issues observed during concentration in vacuo or during recrystallisation. The cyclopentyl ester however remained as a mixture of the mono and di-ester derivatives, which were inseparable by further column or ion-exchange chromatography. The optimisation and isolation of a pure di-cyclopentyl ester was, therefore, put on hold (Scheme 6). Scheme 6. Synthesis of diesters analogues. Reagents and conditions: (i) 4-Bromobenzyl bromide, 1 M NaOH (aq), EtOH, 4 h, RT, 82%. (ii) 10% Na2CO3 (aq), acetone, 0°C, 2.5 h - RT, 16 h, 90% yield. (iii) R-OH, thionyl chloride, RT, 23 h, R2 = Et = 27 70%, R2 = nPr = 28 83% (iv) 10% v/v piperidine/DMF, RT, 1.5 h, R2 = 29 = Et 86%, R2 = 30 = nPr 53%. 65 Intermediate (26) also allows a preliminary evaluation of the N-blocked analogues, where R1 is the S-(4-bromobenzyl), R2 a hydrogen and R3 the Fmoc group. 2.2 Protein expression and purification optimisation To allow evaluation of the compounds synthesised in Section 2.1 against the glyoxalases from E. coli it was necessary to express and purify both GlxI and GlxII. Optimisation of the conditions to obtain the proteins in sufficient yield and purity in our hands are described below. 2.2.1 Glyoxalase II: Optimisation of growth, expression and purification The expression strain BL21(DE3) was transformed with pTrcgloBH6. Transformed cells were grown (LB supplemented with 50 μg/L ampicillin) to an optical density (OD600nm) of 0.70 at 37 °C and induced with ITPG (0.5 mM) for 4 hours at 25 °C. Cell lysate was prepared by sonication and the addition of PMSF (1 mM). The protein was purified by Ni-NTA affinity chromatography (50 mM Tris pH 8.0, 300 mM NaCl, 10 mM imidazole, gradient elution 25 – 300 mM imidazole), followed by buffer exchange (10 mM MOPS, 0.2 M NaCl, pH 7.2) according to conditions by Suttisansanee.122 Under these conditions purified GlxII was obtained in yields of 7.06 - 9.14 mg/ pellet batch. Fractions were analysed with SDS-PAGE were GlxII was identified as a band migrating approximately the same distance as the 30 kDa marker (Carbonic Anhydrase) in the size standard (Fig. 25). Fig. 25. SDS-PAGE Gel of GlxII after column purification and buffer exchange. 5 optimisation runs were carried out for the purification protocol of GlxII and the observed results are described. Increasing the time and number of cycles of sonication from 40 sec on and 30 sec off, repeating for 5 cycles, to 60 sec on and 30 sec off, repeating for 10 cycles greatly improved the intensity of the protein band at ~30 kDa on the SDS-PAGE gel, indicating that the cell lysis proceeded more effectively under the optimised conditions. Increasing the column size of Ni-NTA agarose gel from 1 mL to 2 mL gave 66 improved separation of protein from the impurities. Incubating the lysed protein solution overnight with 0.5 mL of Ni-NTA agarose gel prior to purification under standard conditions caused degradation of the protein and decreased the purity. Slowing down the gradient by gradually increasing the concentration of imidazole over a longer time period gave improved separation and removed the need for a second purification before buffer exchange. Combining these observations gave an optimised protocol for the purification of GlxII on a larger scale as described above. Table 1 shows the optimised conditions and yield per pellet. Table 1. Yield of optimised conditions for GlxII purification. Sonication cycle Column size Ni-NTA Wash buffer 50 mM Tris pH 8, 300 mM NaCl Elution buffer 50 mM Tris pH 8, 300 mM NaCl 2nd column Yield after buffer exchange 1 40 sec on 30 sec off 5 times 1 mL 20 mL 50 mM Imidazole 5 mL x 300 mM Imidazole 1 mL Ni-NTA 10 mL 10 mM Imidazole 2 mL x 50, 150m 300 mM Imidazole 0.31 mg / 500 mL expressed pellet 2 60 sec on 30 sec off 10 times 1 mL 10 mL 10 mM Imidazole 2 mL x 50, 150, 300 mM Imidazole 1 mL Ni-NTA 10 mL 10 mM Imidazole 2 mL x 50, 150m 300 mM Imidazole 0.26 mg / 500 mL expressed pellet 3 60 sec on 30 sec off 10 times 2 mL 30 mL 10 mM Imidazole 4 mL x 50, 150, 300 mM Imidazole glutathione sepharose 4B column 1.08 mg / 167 mL expressed pellet 4 60 sec on 30 sec off 10 times Incubation for 16 h in 2 mL Ni-NTA at 4 °C 30 mL 10 mM Imidazole 4 mL x 50, 150, 300 mM Imidazole glutathione sepharose 4B column 1.88 mg / 167 mL expressed pellet 5 60 sec on 30 sec off 10 times 2 mL 30 mL 10 mM Imidazole 4 mL x 25, 50, 75, 100, 150, 300 mM Imidazole n/a 2.05 mg / 167 mL expressed pellet 6 60 sec on 30 sec off 10 times 2 mL 30 mL 10 mM Imidazole 4 mL x 25, 50, 75, 100, 150, 300 mM Imidazole n/a 7.02 mg / 1000 mL expressed pellet To obtain the DNA sequence, the cloning system DH5α was transformed with pTrcgloBH6. The DNA was purified by a Qiagen Midi prep and sequencing carried out using a pTrcReverse primer. The sequence reported was a 73% match with E. coli strain 493/89 probable hydroxyacylglutathione hydrolase (ECs0208) gene DNA sequence, complete sequence on BLAST. The full DNA sequence is shown in the Appendix, Table 62. A sample was submitted for MALDI-TOF analysis, mass expected was calculated 67 using UniProt and ExPASy123 to be 29257.22 Da. The observed mass was 29156.4 – 29551.3 Da with a centroid mass of 29439.7 Da. Machine error is 0.05%, 29141.8 – 29556.0 Da. Sequencing and MALDI-TOF analysis confirmed GlxII was obtained. 2.2.2 Glyoxalase I: Optimisation of growth, expression and purification The expression strain, BL21(DE3), was transformed with pET3a-gloA, a C-terminally His-tagged construct. Transformed cells were grown (LB and 50 μg/L ampicillin) to an optical density of 0.60 at 37 °C and induced with ITPG (1.0 mM) for 4 hours at 25 °C. Cell lysate was prepared by sonication and the protein purified using Ni-NTA affinity chromatography (50 mM Tris, pH 8, 0.5 M KCl and 20 mM imidazole, gradient elution 50 – 300 mM imidazole), followed by buffer exchange (50 mM Tris, pH 8 and 0.5 M KCl) according to conditions by Suttisansanee.124 Under these conditions a weak band was observed in the SDS PAGE (Fig. 26) in the appropriate MW range for GlxI (14 kDA), however it was impure. Attempts to isolate a pure sample of GlxI using a second Ni-NTA column were unsuccessful. Fig. 26. SDS-PAGE gel of GlxI after column 1. Where S = standard, FT = flow through, W = wash, 1 – 12 = elution fractions. Further attempts to optimise the conditions included: 1) addition of hen egg white lysozyme (1 mg/mL) to aid cell lysis 2) addition of EDTA (4 mM) to prevent aggregation and 3) Ni-NTA affinity chromatography by FPLC proved unsuccessful as shown in (Fig. 27), where a weak band was observed in the SDS PAFE in the appropriate MW range for GlxI (14 kDA) along with two other impurities at ~30 and 60 kDa. 68 Fig. 27. SDS-PAGE gel of GlxI. After buffer exchange. Where S = standard, 1 – 13 = fractions. The low intensity bands observed at the 14 kDa marker were thought to be GlxI protein, bands observed at the 28 kDa marker were hypothesised to be the dimer mass and bands observed at the 64 kDa marker impurities. In order to ascertain if the band ~30 kDa was the dimer or an impurity, the bands at ~15 kDa and ~30 kDa were cut out for trypsin digest analysis. However, trypsin digest showed no mass consistent with GlxI or that the 2 bands were related to each other. A sample was also submitted for MALDI-TOF analysis where the mass was expected to be 15056.9 Dalton (calculated using UniProt and ExPASy), but no mass relating to GlxI was found. In conclusion, all 3 attempts to purify GlxI were not successful. The MALDI-TOF and trypsin digest proved that no GlxI had been isolated. A literature review125,126 revealed that there are common contaminants observed when purifying histidine tagged proteins with immobilised metal ion affinity chromatography (IMAC) such as Ni-NTA. This is due to proteins produced by E. coli containing non-consecutive histidine residues on the surface of the structure which bind to the metal resin and are copurified. Contaminants are also dependent on the expression conditions, (temperature, cell density at induction, the media and oxygen) the strain and the plasmid. A number of contaminants have been identified over the years. Three such contaminants are SlyD, ArnA and GlmS. SlyD is a cytoplasmic protein which is required to induce lysis and migrates at the 27 kDa marker. Sly refers to sensitivity to lysis. GlmS migrates at the 67 kDa marker, it is a precursor for components of peptidoglycan and lipopolysaccharides and as such is an essential step in the biosynthesis of the bacterial cell wall. ArnA is an enzyme involved in lipid A phosphate modifications with aminoarabinose and migrates near the 64 kDa marker. These contaminants are typically eluted with 0 – 150 mM imidazole. Solutions to preventing contaminants are in optimisation of growth and expression conditions, size exclusion chromatography instead of IMAC and changing the E. coli strain.125,126 69 Given the problems outlined above a small screen of conditions, different temperatures, length of time for growth and expression as well as growth media were investigated to grow and express GlxI. The conditions and observations are described. The transformed BL21(DE3)/pgloA cells were grown in 100 mL LB or TB media in the presence of ampicillin (50 μg/L) at 37 °C and induced with IPTG (1.0 mM) as summarised in Table 2. Table 2. Optimisation of expression conditions for the purification of GlxI. Media Growth temp Growth time OD600nm Induction temp Induction time 1 LB 37 °C ~ 4 h 0.6-0.7 25 °C 20 h 2 TB 37 °C ~ 4 h 0.6-0.7 25 °C 4 h 3 LB 25 °C 18 h n/a 25 °C 4 h A mini Ni-spin column was attempted on a portion of all 3 of the batches but no discernible bands were observed by SDS-PAGE gel for the wash or elution fractions. The purification as described above was carried out on all 3 batches. SDS-PAGE results indicated a protein mass corresponding to ~30 kDa and a weak band at ~15 kDa, suggestive of the same contaminants mentioned previously and not the desired GlxI protein. Another method provided by Dr Stefanie Frank, University of Kent using BL21(DE3)pLysS as the competent cell in order to carry out the transformation and the addition of Ni(II)Cl2 in the growth and expression to aid the protein functionality was carried out. The expression system BL21(DE3)pLysS is a better competent cell for transformation with pET3a-gloA due to the presence of both the DE3 lysogen, to allow for low levels of expression of the T7 polymerase by DE3 upon IPTG induction, and also the plasmid pLysS, which produces T7 lysozyme to reduce the basal expression of recombinant genes through inhibition of T7 polymerase and allowing the expression of gloA but not the contaminants observed in earlier attempts. The use of BL21(DE3)pLysS also requires the addition of chloramphenicol in addition to the ampicillin antibiotic, as pLysS plasmid has a chloramphenicol resistant gene and pET3a-gloA has an ampicillin resistant gene. In studies carried out by Honek et al.127 it was found that the non-Zn2+ GlxI metalloenzyme class of Gram-negative bacterials are activated by Ni2+ or Co2+. Metal activation profiles showed that the enzyme GlxI was activated by Ni2+ to 100% and Co2+ to 40%. Further metal titration studies showed the enzyme had maximum activity with 1 mol of metal per mol of enzyme. Honek et al.127 also observed that the presence 70 of the His tag could cause protein aggregation because of oligomers occurring due to interactions with the Ni2+ ions in the growth media which could result in a lower yield, but the His tag had no effect on the activity of the enzyme. The expression strain BL21(DE3)pLysS was transformed with pET3a-gloA. The transformed cells were grown in LB, supplemented with ampicillin (50 μg/L), chloramphenicol (25 μg/L) and nickel chloride (1 mM) to optical density (OD600nm) of between 0.5 – 0.7 at 37 °C and induced with ITPG (0.4 mM). Cell lysate was prepared by sonication and the protein purified using Ni-NTA affinity chromatography (20 mM Tris pH 8.0, 500 mM NaCl, 10 mM imidazole, elution gradient 100 – 400 mM imidazole). Fractions identified by SDS-PAGE Gel, 14 kDa marker, and the Bradford assay were pooled and loaded onto a PD-10 column to undergo buffer exchange with 50 mM Tris pH 8.0 and 100 mM NaCl to afford purified glyoxalase I in a yield of 24.70 mg/ 250 mL pellet batch. Fig. 28 shows the purified GlxI SDS PAGE gel. Fig. 28. SDS-PAGE Gel of purified GlxI after column purification. To optimise the purification conditions described above took 3 attempts of adjusting the elution gradient. No gradient (elution with 400 mM imidazole) resulted in GlxI eluting in fraction 2 in a yield of 5.95 mg / 250 mL pellet. A slower gradient resulted in improved separation from impurities however GlxI eluted over 14 fractions resulting in a yield of 5.15 mg / 250 mL pellet. The third attempt included a faster gradient which resulted in much-improved yield (24.70 mg/ 250 mL pellet batch) and fewer fractions to undergo buffer exchange. Table 3 shows the summary of yield of GlxI obtained during optimisation. x 71 Table 3: Summary of purification optimisation and yield for GlxI. Sonication cycle Column size Ni-NTA Binding buffer Wash buffer 20 mM Tris, pH 8, 500 mM NaCl Elution buffer 20 mM Tris pH 8, 500 mM NaCl Yield after buffer exchange 1 60 sec on 30 sec off 10 times 1 mL 20 mM Tris pH 8, 500 mM NaCl, 10 mM Imidazole 25 mL 100 mM Imidazole 15 mL 400 mM Imidazole 5.95 mg/ 250 mL pellet batch 2 60 sec on 30 sec off 10 times 1 mL 20 mM Tris pH 8, 500 mM NaCl 20 mL 10 mM Imidazole 4 mL x 50, 100, 150, 200, 300 mM Imidazole 5.15 mg/ 250 mL pellet batch 3 60 sec on 30 sec off 10 times 1 mL 20 mM Tris pH 8, 500 mM NaCl 20 mL 10 mM Imidazole 4 mL x 50 mM Imidazole 2 mL x 100, 200, 300 mM Imidazole 4 mL x 400 mM Imidazole 24.70 mg/ 250 mL pellet batch To obtain the DNA sequence, the cloning strain of DH5α was used due to its properties of being highly transformable and having a low recombination rate which allows for the stable maintenance of plasmid DNA.128 The transformation was carried out with pET3a-gloA and purified using a Qiagen Midi prep to obtain the pure DNA. DNA sequencing was carried using T7 primer and the E. coli DNA sequence was found to be a 95% match with pET-His Expression vector (from E. coli) DNA sequence, a 95% match with bacterial expression vector pET-3a, complete sequence on BLAST and a verified match for the expected sequence provided. The DNA sequence is in the Appendix, Table 63. A sample was submitted for MALDI-TOF analysis, mass expected from the sequencing information provided by Dr Stefanie Frank, University of Kent was calculated using UniProt and ExPASy123 to be 15847.8 Da. The observed mass was 15780.9 – 15930.2 Da with a centroid mass of 15849.5 Da. Machine error is 0.05%, 15773.0 – 15938.1 Da. The amino acid sequence obtained from UniProt and ExPASy123 gave an expected mass of 15742.7 Da. Sequencing and MALDI-TOF analysis confirmed that GlxI was obtained. In summary, the transformation into BL21(DE3)pLysS and addition of nickel chloride to the growth media resulted in the growth, expression and purification of GlxI. Other factors which could also have affected the growth and expression of GlxI included: longer induction time at a lower temperature, 18 h at 18 °C compared to the initial attempts at 25 °C for 4 hours, to influence the protein solubility and improve the protein folding; lower 72 IPTG concentration, 0.4 mM compared to the initial 1 mM, to improve the protein folding.128 2.3 Preliminary SAR study With GlxII and GlxI in hand, in order to evaluate the binding of the glutathione S-conjugates, fragments and ester prodrugs to provide some insight into the SAR for the bacterial proteins a differential scanning fluorimetry (DSF) or thermal shift assay was proposed. DSF is a screening method used to determine the binding and stabilisation of ligands to proteins by monitoring the thermal unfolding of the protein in the presence of a fluorescent dye such as SYPRO orange. The fluorescence of the dye increases as the protein unfolds with increased temperature due to the affinity of the dye for hydrophobic portions of the protein which are exposed upon unfolding. The midpoint or 50% of the unfolded protein is the melt temperature (Tm) of the protein. In the presence of a ligand the melt temperature can shift depending on the binding affinity of the ligand. A positive Tm shift (ΔTm) indicates the ligand stabilises the protein indicating binding. A negative ΔTm shift indicates the ligand could be binding to the denatured state of the protein. This screening method allows ranking of ligands depending on the ΔTm.129 2.3.1 Assay optimisation: Glyoxalase II To obtain the optimum experimental conditions a number of preliminary assays were performed varying each of the components in the assay: protein concentration (2, 5, 10, 15, 20 and 50 μM), total volume (25 μL and 40 μL), SYPRO orange dye concentration (1x, 2x and 5x) and buffer conditions. The first experiments focused on varying the protein concentration (2 μM, 5 μM and 10 μM) and well volume (25 μL and 40 μL), SYPRO orange dye (1x) and buffer (10 mM MOPS, pH 7.2, 0.2 M NaCl) remained constant (Fig. 29). The fluorescence signal for 2 μM GlxII concentration was too weak at 25 μL and 40 μL. At 5 μM GlxII concentration and 40 μL total well volume gave measurable fluorescence and was comparable to 10 μM GlxII concentration and 25 μL total well volume but was 5-fold weaker than 10 μM GlxII concentration and 40 μL total volume per well. Overall, 10 μM GlxII concentration and 40 μL total volume per well gave the best sigmoidal curves with the highest fluorescence signal, resulting in a more accurate determination of Tm.129 Optimum well volume was found to be 40 μL as the observed melt curve was more consistent than with 25 μL. 73 Fig. 29. Protein concentration and well volume optimisation. DSF curve and results for GlxII. In a second experiment, the optimal SYPRO orange dye concentration (2x and 5x) was evaluated with GlxII concentration at 10 μM and 20 μM. Well volume (25 μL) and buffer (10 mM MOPS, pH 7.2, 0.2 M NaCl) remained constant. The results showed 5x SYPRO orange dye and 10 μM GlxII concentration gave comparable results to 2x SYPRO orange dye and 20 μM GlxII concentration with a slightly lower fluorescence readout than the 5x SYPRO orange dye and 20 μM GlxII concentration (Fig. 30). 5x dye was found to be preferred due to lower concentrations of dye occasionally giving an initial high fluorescence or inconsistent melt curves. The optimal protein concentration was confirmed as 10 μM. Fig. 30. Protein and dye concentration optimisation. DSF curve and results for GlxII. The final variable investigated was buffer composition as this can have a significant effect on protein stability and therefore the observed Tm. Three buffers were prepared; 10 mM MOPS, pH 7.2, 0.2 M NaCl; 10 mM PBS, pH 7.4, 0.15 M NaCl; 10 mM HEPES, pH 7.5, 0.15 M NaCl. Protein concentration (10 μM), dye (5x) and well volume (40 μL) remained constant. MOPS and HEPES buffers were observed to give similar Tm for GlxII with PBS giving a slightly lower melt temperature. As MOPS buffer is employed in the purification and storage of GlxII and it gave comparable results to HEPES buffer it was used in all subsequent DSF experiments (Fig. 31). 74 Fig. 31. Results for buffer optimisation against GlxII. Efforts to optimise the assay conditions (concentration, well volume, dye concentration, buffer, ligand concentration and detergents) led to increased confidence in the data produced. In summary, the optimal conditions in our hands were found to be: protein GlxII at 10 μM concentration, 5 x SYPRO orange dye, 40 μL total well volume and MOPS buffer. Lysozyme was used as a positive control in the assay to confirm that the run gave consistent readouts and to ensure quality of data. 2.3.2 Assay optimisation: Glyoxalase I In order to obtain the optimal conditions for GlxI a number of preliminary DSF assays were performed varying each of the components in the assay: buffer conditions (Tris, MOPS, PBS and HEPES), protein concentration (2, 5, 10, 15 and 20 µM) and SYPRO orange dye concentration (2x and 5x). Well volume was kept constant at 40 μL after previous analysis with GlxII resulted in a more accurate determination of Tm compared to 25 μL. Firstly, the investigation of the effect of different buffers on the observed fluorescent signal and stability of GlxI was carried out with protein concentration (15 μM), dye (5x) and well volume (40 μL) remaining constant. Four buffers were prepared; 50 mM Tris, pH 8.0, 0.1 M NaCl; 10 mM MOPS, pH 7.2, 0.2 M NaCl; 10 mM PBS, pH 7.4, 0.15 M NaCl; 10 mM HEPES, pH 7.5, 0.15 M NaCl. The observed difference in melt temperatures was 1.17 °C. As HEPES had the most accurate Tm readout it was used in the thermal shift assays (Fig. 32). 75 Fig. 32. Results for buffer optimisation for GlxI. Secondly, to determine the optimum protein concentration a range of concentrations of GlxI (2 μM, 5 μM, 10 μM, 15 μM and 20 μM) were evaluated, with dye (5x), well volume (40 μL) and buffer (HEPES) remaining constant (Fig. 33). The fluorescence signal with 2 μM GlxI was too weak, the observed intensity of the fluorescence signal at 5 μM was also low. 10 μM and 15 μM GlxI gave comparable results. 20 μM showing the highest fluorescence. An increase is GlxI concentration resulted in an increase of the intensity of the fluorescence in the sigmoidal curves however Tm decreased. The results showed that 10 μM GlxI concentration gave a robust fluorescence signal and sigmoidal curve for Tm measurements. Fig. 33. Protein concentration and well volume optimisation. DSF curve and results for GlxI. Lastly, to investigate the optimal dye concentration (2x and 5x) two protein concentrations were analysed for GlxI (10 μM and 15 μM) (Fig. 34). Buffer (HEPES and well volume (40 μL) remained constant. The results showed 5 x SYPRO orange dye gave a higher intensity fluorescence compared to 2x SYPRO orange dye and was selected for further experiments. 76 Fig. 34. Protein and dye concentration optimisation. DSF curve and results for GlxI. In summary, the optimal conditions in our hands were found to be as follows: 10 μM GlxI, 5 x SYPRO orange dye, 40 μL total well volume in 10 mM HEPES, pH 7.5, 0.15 M NaCl buffer. 2.3.4 Evaluation of glutathione analogues using DSF With DSF assay conditions established for GlxI and GlxII the glutathione S-conjugates and fragments could be evaluated for binding. The known inhibitor, S-(4-bromobenzyl)-glutathione (3) was evaluated using the optimised DSF conditions with GlxI and GlxII. S-(4-Bromobenzyl)-glutathione (3) was observed to increase the Tm of GlxI by 1.32 °C indicating binding and a similar stabilisation effect to GSH, where the ΔTm was 1.16 °C (Fig. 15. A, Table 4). S-(4-bromobenzyl)-glutathione (3) was observed to increase the Tm of GlxII by 2.03 °C indicating a stabilisation effect and therefore binding (Fig. 35. B, Table 4). A: GlxI B: GlxII Fig. 35. Bar graphs of thermal shift with respect to A) GlxI and B) GlxII (ΔTm, °C) for S-(4-bromobenzyl)-glutathione. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). 77 Table 4. Thermal shift with respect to GlxI and GlxII with S-blocked analogue (3) (ΔTm, °C). GlxI GlxII Compound Structure ΔTm (°C) ΔTm (°C) GSH 1.16 -0.83 3 1.32 2.03 The known S-blocked glutathione inhibitor, S-(4-bromobenzyl)-glutathione (3), was shown to have a stabilising effect on both GlxI and GlxII using the optimised thermal shift conditions and could therefore be used as a starting point to design inhibitors of the bacterial glyoxalase system. The small library of fluorinated GS-X analogues were evaluated with GlxI and GlxII using the optimised DSF conditions. In the presence of 1 mM of the fluorinated ligands and GlxI (10 µM) a ΔTm of between -0.27 to 2.76 °C was observed. The removal of the cysteine sulfur, 17, resulted in a ΔTm of -0.27 °C indicating the cysteine sulfur is important for binding. The 4-CF3 analogue (9) and 3,5-diF (10) were observed to have the biggest stabilizing effect on GlxI with a ΔTm of 2.76 °C and 2.59 °C respectively, indicating binding to GlxI, as well as being the only two analogues with a ΔTm greater than GSH (ΔTm 1.72 °C). The remaining 19F ligands resulted in a stabilization of melt temperature for GlxI with a flat SAR. In the presence of the fluorinated ligands (1 mM) and GlxII (10 µM) a ΔTm of between 1.79 to 3.73 °C was observed which was an indication of binding. The 4-CF3 analogue (9) and the 2-F analogue (4) were observed to have the biggest stabilizing effect on GlxII with a ΔTm of 3.73 °C and 3.13 °C, respectively. One possible reason for the relatively flat SAR is that the binding of these compounds could be primarily due to the peptide backbone and that the change in benzyl groups is not contributing significant differences in the H site (Fig. 36, Table 5). The competitive inhibitor (3) showed equivalent stabilization to the fluorine analogues. 78 A. GlxI B. GlxII Fig. 36. Bar graphs of thermal shift (ΔTm, °C) with respect to A) GlxI and B) GlxII for glutathione conjugates. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 5. Thermal shift with respect to GlxI and GlxII with glutathione S-conjugates (ΔTm, °C). GlxI GlxII ID Structure ΔTm (°C) ΔTm (°C) GSH 1.72 -0.83 3 1.33 2.03 4 0.69 3.13 5 0.73 2.16 6 1.17 2.58 7 1.30 2.26 79 GlxI GlxII ID Structure ΔTm (°C) ΔTm (°C) 8 1.43 1.79 9 2.76 3.73 10 2.59 2.00 17 -0.27 1.90 To observe what effects the peptide fragments had on binding and recognition in GlxI and GlxII, additional DSF experiments were carried out (Fig. 37, Table 6). In the presence of GlxI, GSH was observed to result in a ΔTm of 0.02 °C. The ɣ-Glu-Cys (18) peptidic fragment increased the observed Tm by 1.49 °C, whereas the Cys-Gly (19) peptidic fragment resulted in a negative ΔTm of -2.68 °C indicating binding to the denatured GlxI or no stabilisation of GlxI. These results indicate that the ɣ-glutamic acid residue is more important for binding and recognition in GlxI than the glycine residue. These results are a confirmation of the anticipated binding of the GSH tripeptide binding and recognition to GlxI from the structural studies carried out by Cameron et al.130 In the presence of GlxII, GSH resulted in a negative ΔTm of -1.07 °C, the ɣ-Glu-Cys peptidic fragment (18) resulted in a negative ΔTm of -4.07 °C and the Cys-Gly peptidic fragment (19) stabilised GlxII as observed by a ΔTm of 1.27 °C. These results indicate that the glycine residue is more important for binding and recognition in GlxII than the glutamic acid residue which is a confirmation of the anticipated binding of the GSH tripeptide binding and recognition to GlxII from the structural studies carried out by Cameron et al.1 80 A. GlxI B. GlxII Fig. 37. Bar graphs of thermal shift with respect to A) GlxI and B) GlxII (ΔTm, °C) for glutathione fragments. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 6. Thermal shift with respect to GlxI and GlxII with GSH fragments (ΔTm, °C). GlxI GlxII ID Structure ΔTm (°C) ΔTm (°C) GSH 0.02 -1.07 ɣ-Glu-Cys (18) 1.49 -4.07 Cys-Gly (19) -2.68 1.27 The binding data discussed above indicated that in GlxI that the glutamate-cysteine residues were important for binding and recognitions whilst in GlxII the cysteine-glycine residues were important for binding and recognition. These differences could present an opportunity to gain selectivity between GlxI and GlxII in the future, but in terms of designing ligands that could be used to probe both GlxI and GlxII it was decided to retain the full tripeptide backbone for this research to optimise methodologies and understand and directly compare different ligand changes with a single compound against both GlxI and GlxII. The evaluation of the ester prodrugs with DSF and GlxII (Fig. 38, Table 7), showed there was no significant difference in affinity between the acid and ester compounds. On incubation with 1 mM of the parent S-conjugate (3) with GlxII a ΔTm of 2.03 °C was observed, whilst unexpectedly the esters showed an equivalent level of stabilisation with a ΔTm of 2.10 °C and 2.79 °C observed for the di-ethyl ester (29) and di-n-propyl ester (30) respectively. This suggests that the carboxylate was not essential for binding within 81 the binding site due to glycine carbonyl interaction being conserved with either the Lys248 or Arg245 and the glutamate not being required for binding and recognition within the GlxII binding site. A. GlxI B. GlxII Fig. 38. Bar graphs of thermal shift with respect to A) GlxI and B) GlxII (ΔTm, °C) for ester prodrugs. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 7. Thermal shift with respect to GlxI and GlxII with ester prodrugs (ΔTm, °C). GlxI GlxII ID Structure ΔTm (°C) ΔTm (°C) GSH 1.16 -0.83 3 1.32 2.03 29 0.31 2.10 30 0.77 2.79 No clear structure activity relationship was observed with the peptide analogues against GlxII, suggesting that the ester groups did not have a detrimental effect on binding affinity. The ester analogues screened against GlxI showed the anticipated trend for differences in ΔTm between the bis-acid and ester analogues, suggesting that the carboxylates were important for binding and the ester derivatives are less tightly bound as correlated with an observed decrease in ΔTm. Unfortunately, due to solubility 82 problems it was not possible to evaluate the N-Fmoc intermediate (26). Therefore, the N-blocked analogous were not pursued at this time. In summary, important structural features for binding and recognition to GlxI include the sulfur cysteine, glutamic acid residue and carboxylate moieties. For GlxII the glycine residue is required for binding and recognition and a CF3 group can replace the Br in all cases. 2.4 Evaluation of antimicrobial activity of S-bromobenzyl glutathione esters GSH has a short half-life and due to the bis-carboxylic acid and the nature of the tripeptide it has difficulty permeating the cell membrane, which results in high doses being administered to obtain sufficient intracellular concentrations. This suggests that evaluation of the effect of glyoxalase inhibition in bacteria using the glutathione analogues, synthesised above, could be challenging. Both mono- and di-esters were employed by Thornalley et al.109 as a prodrug delivery system for GSH and GS-X into cells, where the ester is hydrolysed by non-specific esterases back to the parent bis-acid. This study by Thornalley et al.109 investigated various ester analogues of S-(4-bromobenzyl)-glutathione (3) and observed that whilst the dimethyl ester was toxic in mice the diethyl ester was not and rapidly converted to the mono-ethyl ester due to hydrolysis by esterases in the plasma. Human blood plasma does not have GS-X diester-α-esterase activity so the GS-X diester is therefore stable at the cellular level. The diethyl ester (29) was found to be an efficient delivery agent and diffused more rapidly into the cell compared with the mono ester analogue due to increased hydrophobicity of the di-ester groups compared to the mono-ester analogue (Fig. 39). In the same report, it was also found that the n-propyl diester (30) and cyclopentyl diester were also efficient prodrugs and were among the most potent analogues tested. D’Silva et al.105 also used this prodrug approach for the delivery of their antiprotozoal inhibitors where the n-propyl diester (30) had an IC50 of 1.9 µM. An important feature of the esters was to be resistant to cleavage by plasma or extracellular plasma membrane bound non-specific esterases and the susceptibility of the ester groups to hydrolysis by non-specific esterases in the cytosol of the target cells so sufficient intracellular concentrations could be achieved (Fig. 19). Given the precedence described above we proposed to use the synthesised diesters (29 and 30) and the parent S-4-bromobenzyl glutathione (3) analogue, to investigate the effect of treatment of a selection of bacterial pathogens in the presence and absence of 83 existing antibiotics to determine if glyoxalase inhibitors could act synergistically with existing antibiotics and the host immune system to resensitise current antibiotics for combination therapies. Fig. 39. Transport rates across the human cell membrane. 2.4.1 Design of esters To enable the investigation of bacterial cell permeability and determine if glyoxalase inhibitors exhibit a synergistic effect with existing antibiotics, it was first necessary to synthesise a number of ester prodrugs of S-(4-bromobenzyl)-glutathione (3). Based on Thornalley et al.’s observations as discussed in section 2.4, initial plans were to synthesis ethyl, n-propyl and cycopentyl diesters analogues of S-(4-bromobenzyl)-glutathione and test the binding affinity, minimum inhibitory concentrations and synergistic effects in the presence of known antibacterials of the diester analogues compared to the parent bis-acid S-(4-bromobenzyl)-glutathione (3, Fig. 40).109, 120 Synthesis and preliminary binding evaluations are described in Sections 2.1.3 and 2.3.3 respectively. Fig. 40. Example of ester analogues. 2.4.2 Minimum inhibitory concentration determination and antibacterial and resistance modulator activity studies With the parent glutathione S-conjugate (3) and two ester derivatives (diethyl, 29 and dipropyl, 30 (Table 8)) in hand, evaluation of their minimum inhibitory concentrations (MICs) against a variety of bacterial strains was evaluated. The MICs were also evaluated in the presence of existing antibiotics to determine if glyoxalase inhibitors could act synergistically with antibiotics and the host immune system. 84 Table 8. Synthesised ester analogues. ID Structure 3 29 30 Minimum inhibitory concentration (MIC), which is defined as the lowest concentration of the antimicrobial agent that inhibits visible growth of the microorganism over a given time period, were determined in 96-well microplates in Iso-Sensitest broth, using the following strains which are susceptible to most antimicrobial agents; E. coli NCTC 10418; Staphylococcus aureus, Citrobacter freundii, Enterococcus faecalis, Bacillus subtilus, E. coli G69, Proteus sp 10830P, Psuedomonas aeruginosa. E. coli G69 and Proteus sp are multi-drug resistant clinical isolates. The concentration strain range used was 1024 – 1 µg/mL. Compounds 3, 29 and 30 were evaluated over the concentration range of 256 to 0 g/mL and the MICs were found to be >256 g/mL for all organisms. To determine if glyoxalase inhibitors exhibit synergistic effect with existing antibiotics. The antibacterial and resistance modulatory activity of 3, 29 and 30 against a sensitive strain, E. coli NCTC 10418, and two resistant strains, E. coli G69 and Proteus sp. 10830P, was evaluated. The following antibiotics, with a range of mechanisms of action, were used: amoxicillin (cell wall synthesis inhibitor), chloramphenicol (protein synthesis inhibitor), kanamycin (a protein synthesis inhibitor), rifampicin (RNA synthesis inhibitor), nitrofurantoin (inhibitor of multiple targets), norfloxacin (DNA synthesis inhibitor), trimethoprim (folate synthesis inhibitor). The concentration range of the antibiotics was 1024 to 0 g/mL in the presence and absence of 3 (209 µM), 29 (187 µM) and 30 (178 µM), at a fixed concentration of 100 g/mL. In all cases the concentration ranges were two-fold serial dilutions and the quantity of DMSO did not exceed 1%. 85 3 was found to decrease the MIC value of kanamycin, a protein synthesis inhibitors, from >1024 to 512 g/mL in Proteus sp. 10830P, which is a MDR clinical isolate. No decrease in MIC was observed on treatment with 3 and all other antibiotics tested against the susceptible organism (E. coli NCTC 10418) or multi-drug resistant isolates (E. coli G69 or Proteus sp) (Table 9). Table 9. Minimum inhibitory concentrations of various antimicrobial agents, in the presence (+) and absence (-) of 3, for E. coli and Proteus sp. (* controls didn’t grow). -, Absence of 3; +, presence of 3 (100 g/mL) Organism Minimum inhibitory concentration (g/mL) Amoxicillin Chloramphenicol Kanamycin Rifampicin Nitrofurantoin Norfloxacin Trimethoprim - + - + - + - + - + - + - + E. coli NCTC 10418 4 4 1 4 8 8 4 2* 1 1 1 2* 1 1 E. coli G69 MDR clinical isolate 256 512 512 512 4 4 8 8 64 128 >1024 >1024 1 1 Proteus sp. 10830P MDR clinical isolate 512 1024 128 128 >1024 512 2 2 32 32 1 1 1024 1024 No decrease in MIC for all antibiotics tested was observed on treatment with the diethyl ester prodrug (29) in either a susceptible organism (E. coli NCTC 10418) or multi-drug resistant isolates (E. coli G69 or Proteus sp) (Table 10). Table 10. Minimum inhibitory concentrations of various antimicrobial agents, in the presence (+) and absence (-) of 29, for E. coli and Proteus sp. -, Absence of 29; +, presence of 29 (100 g/mL) Organism Minimum inhibitory concentration (g/mL) Amoxicillin Chloramphenicol Kanamycin Rifampicin Nitrofurantoin Norfloxacin Trimethoprim - + - + - + - + - + - + - + E. coli NCTC 10418 4 4 1 4 8 8 4 8 1 4 1 1 1 1 E. coli G69 MDR clinical isolate 256 512 512 512 4 4 8 8 64 64 >1024 >1024 1 1 Proteus sp. 10830P MDR clinical isolate 512 512 128 128 >1024 >1024 2 2 32 64 1 1 1024 1024 86 A decrease in the MIC value of kanamycin, from 8 to 4 g/mL, in the sensitive strain E. coli NCTC on treatment with the di-n-propyl ester prodrug (30) was observed. However, as before no decrease in MIC was observed with the remaining antibiotics tested against both a susceptible organism (E. coli NCTC 10418) or multi-drug resistant isolates (E. coli G69 or Proteus sp) (Table 11). Table 11. Minimum inhibitory concentrations of various antimicrobial agents, in the presence (+) and absence (-) of 30, for E. coli and Proteus sp. -, Absence of 30; +, presence of 30 (100 g/mL) Organism Minimum inhibitory concentration (g/mL) Amoxicillin Chloramphenicol Kanamycin Rifampicin Nitrofurantoin Norfloxacin Trimethoprim - + - + - + - + - + - + - + E. coli NCTC 10418 4 8 1 4 8 4 4 8 1 2 1 1 1 1 E. coli G69 MDR clinical isolate 256 512 512 512 4 4 8 8 64 128 >1024 >1024 1 1 Proteus sp. 10830P MDR clinical isolate 512 512 128 128 >1024 >1024 2 2 32 64 1 1 1024 1024 To summarise, the diethyl ester 29 did not convert antibiotic-susceptible isolates to sensitivity or modulate antibiotic resistance phenotypes. The bis-acid, (S-(4-bromobenzyl)-glutathione), 3 did however show it could decrease the MIC value of kanamycin, a protein synthesis inhibitor, from >1024 to 512 g/mL in Proteus sp. 10830P which is the MDR clinical isolate and the dipropyl ester 30 altered the MIC value of kanamycin from 8 to 4 g/mL in E. coli NCTC. These results indicate a GlxI or GlxII inhibitor could resensitise the protein synthesis inhibitor kanamycin if used as a combination therapy. However, due to the challenge of treatments with glutathione analogues not being cell membrane permeable, future work should include repeating 3, 29 and 30 at a higher concentration as well as evaluating the permeability of these compounds. 2.5 Antibiotic susceptibilty studies with strains lacking GlxI or GlxII To provide further confidence in the glyoxalase pathway as a novel target for antibacterial drug discovery, strains with the genes gloA and gloB which encode for GlxI and GlxII respectively, were obtained from the Keio collection, Keio University, Japan.131 The Keio collection is a set of E. coli K-12 organisms with a nonessential single gene deletion 87 allowing for the analysis of gene networks and testing of mutagenic effects to give insight into E. coli biology. Antimicrobial susceptibility studies using disk diffusion and a range of antibiotics that act via a range of mechanisms were used to interrogate the importance of GlxI and GlxII for survival on exposure to ‘stress’. In addition to antibiotics a number of toxins including ROS and MG were also investigated. These studies were anticipated to show that upon removal, or inhibition, of gloA or gloB would lead to bacterial cell death, thereby validating the glyoxalase system as a target for antibacterial discovery. These experiments were carried out by Mona Belaid, a UCL School of Pharmacy summer placement student. The following strains were used: Parent E. coli K12 strain BW25113, gloA knockout strain JW1643 and gloB knockout strain JW0202. The concentration of the antibiotics and toxins were determined using the MICs as a guide to determine the susceptibility of the parent and knockout strains to the antimicrobials (Table 12).132 Table 12. Antibiotic and toxin concentrations for use in disk diffusion susceptibility studies. Antibiotic Concentration (µg/mL) Toxin Concentration Amoxicillin 10 Plumbagin 0.5 m g/mL Gemifloxacin 5 Hydrogen peroxide 5 mM Imipenem 10 Methylglyoxal 5 mM Streptomycin 10 Cefotaxime 30 Meropenem 10 Chloramphenicol 30 Norfloxacin 10 To evaluate the effect each of the antibiotics and toxins had on the parent (BW25113) and knockout strains (gloA knockout strain JW1643 and gloB knockout strain JW0202) a disk diffusion assay was used to determine the zone of inhibition, in mm, around the disc containing the antibiotic or toxin.133-134 The observed zones of inhibition were used to determine which antibiotics and toxins the strains were sensitive to. The observed size of the zones ranged from 9.7 to 36.5 mm for the parent strain BW25113, 11.3 to 39.2 mm for gloA knockout strain JW1643 and 11 to 38.7 mm for gloB knockout strain JW0202. The parent strain, containing both gloA and gloB, was found to be most sensitive to antibiotics meropenem (36.5 mm), cefotaxime (35.3 mm) and imipenem (33 mm), all of which are cell wall synthesis inhibitors. 88 The difference between the zone of inhibition of the knock out strains relative to the parent strain gives an indication of the effect inhibition of GlxI or II would have and whether this would increase sensitivity to existing antibiotics if dosed in combination. In JW1643, the gloA knockout strain, apart from ampicillin and streptomycin, all antibiotics and toxins showed a bigger zone of inhibition compared to the parent strain indicating that inhibitors of GlxI could increase susceptibility of bacteria to these antibiotics and toxins to lead to antibacterial activity (Fig. 41). These changes were significant for meropenem, imipenem and norfloxacin indicating re-sensitisation of bacteria to inhibition of cell wall synthesis, DNA replication and MG build up leading to bacterial death upon removal of gloA. This ΔgloA, JW1643, still contains the gloB gene, however these these results indicated that the removal or inhibition of gloA re-sensitised the bacteria to the actions of the antibiotics. No increase in sensitivity was observed for ampicillin and streptomycin in the gloA knock out strain. Further work is required to gain insight into the rationale for the results observed and the increased sensitivity to a range of antibiotics in the ΔgloA strain. Fig. 41. Difference in zones of inhibition for antibiotics between the parent strain (BW25113) and the gloB knock out strain (JW1643). Significance of changes evaluated by a Student’s t-test (where ** p ≤ 0.01, * p ≤ 0.05). In the gloB knock out strain, JW0202, only gemifloxacin, chloramphenicol and norfloxacin showed bigger zones of inhibition compared to the parent strain indicating that inhibitors of GlxII could increase susceptibility of the bacteria to these antibiotics to lead to antibacterial activity (Fig. 42). These changes were significant for norfloxacin indicating resensitisation to DNA replication upon removal of gloB. No increase in sensitivity was observed to ampicillin, cefotaxime, meropenem, imipenem, and streptomycin in the gloB knock out strain. Further work is required to gain insight into the rationale for the results 89 observed and the increased sensitivity to norfloxacin in the ΔgloB strain. This ΔgloB, JW0202, still contained gloA which could imply gloA has a greater ability to catalyse the neutralisation of the toxin/antibiotic in the glyoxalase pathway, as discussed in section 1.3. Fig. 42. Difference in zones of inhibition for antibiotics between the parent strain (BW25113) and the gloB knock out strain (JW0202). Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001). Exposure to ROS and toxins in the macrophage results in oxidative stress, cell damage and ultimately cell death, therefore neutralisation though pathways such as the glyoxalase system is essential for the bacterial cells survival. Studies16,17,19,18,20 have identified a connection between antibiotics, such as chloramphenicol, and increased O2- levels in Gram-negative bacteria. Therefore, to evaluate if the removal of gloA or gloB resulted in increased zones of inhibition, indicating re-sensitisation of the bacteria to toxins, thereby validating the glyoxalase system as a target. To evaluate the effect of the toxins namely; hydrogen peroxide which produces a reactive oxygen species (ROS), methylglyoxal a known toxic electrophile and established substrate of the glyoxalase pathway and plumbagin, a toxic and mutagenetic compound originally isolated from the plant genus Plumbago, which can generate ROS. In the gloA knockout strain, JW1643, and gloB knocked out gene strain, JW0202, hydrogen peroxide and plumbagin similar increases in the zone of inhibition were observed. These results indicated that on removal of gloA and gloB the bacteria are susceptible to the actions of ROS. The toxin, MG, had a greater and significant zone of inhibition for ΔgloA, JW1643, of 9.2 mm compared to 1.0 mm for ΔgloB, JW0202 (Fig. 43). This result is unsurprising for two 90 reasons; 1) MG is a substrate for GlxI in the glyoxalase pathway, however the increase in zone of inhibition upon removal of gloB indicates that inhibition of either GlxI or II leads to bacteria being susceptible to the effects of MG. 2) Gene knock out studies carried out by Ozyamak et al. 39,135 showed increased sensitivity to MG when lacking gloA and a reduced ability to detoxify MG but did not increase the sensitivity to MG when lacking gloB. A. B. Fig. 43. Difference in zones of inhibition for toxins between the parent strain (BW25113) and A) gloA knock out strain (JW1643) and B) gloB knock out strain (JW0202). Significance of changes evaluated by a Student’s t-test (where ** p ≤ 0.01, * p ≤ 0.05). In summary, these gloA and gloB knock out studies have shown the glyoxalase pathway is important for neutralisation of drugs and toxins and the inhibition of GlxI results in increased sensitivity to antibacterials and toxins, including methylglyoxal. The inhibition of GlxII results in increased sensitivity to toxins but not all antibiotics. These observations suggest that GlxI and GlxII are promising targets for antibacterial drug discovery, either as a single therapy or combination with existing antibiotics. 2.6 Summary and conclusion In summary, conditions were optimised for the expression and purification of GlxI and GlxII. DSF conditions were optimised with both GlxI and GlxII to enable the measurement of binding affinities and allow the evaluation of known glutathione S-conjugates, fragments and ester prodrugs on the bacterial glyoxalases. The S-blocked compound, S-(4-bromobenzyl)-glutathione (3), was shown to have similar binding affinity for the GlxI and improved affinity for GlxII compared to GSH suggesting it can be used as a starting point to design inhibitors of bacterial glyoxalases. Further evaluation of the glutathione tripeptide fragments using DSF revealed that the ɣ-glutamate-cysteine peptide (18) was important for binding to and recognition of GlxI and the glycine-cysteine peptide (19) was important for binding to GlxII. However, to evaluate both GlxI and GlxII the tripeptide 91 backbone was maintained with a view that in the future the fragment approach could be used to gain selectivity between GlxI and II. Adopting the ester prodrug approach to improve cell permeability and increase the intracellular concentration of the peptide inhibitors with bacterial strains was not successful under the conditions tested, however S-(4-bromobenzyl)-glutathione (3) altered the MIC value of kanamycin in Proteus sp. 10830P, an MDR isolate, and S-(4-bromobenzyl) di-propyl ester glutathione (30) was found to alter the MIC value of kanamycin in E. coli NCTC suggesting that these compounds could resensitise kanamycin if used in a combination therapy. In addition, the decrease in binding stabilisation in the DSF assay of the esters for GlxI and the comparable effects in GlxII resulted in these compounds not being taken further in the project at this time. To provide further validation of the glyoxalase pathway as a target for antibacterial drug discovery susceptiblity studies using Keio strains, lacking gloA and gloB genes, were carried out. These experiments indicated the glyoxalase pathway is vital for the neutralisation of drugs and toxins and that GlxI and GlxII are valid targets for antibacterial drug discovery, either as a single therapy or combination to resensitise bacteria to existing antibiotics. In order to apply these observations with the kinetic target guided synthesis approach, the glutathione tripeptide backbone was used to design a ligand with a modification to the S-blocked moiety to test and optimise the methodology to identify inhibitors to explore the hydrophobic pocket and the metal active site within the GlxI and/or GlxII binding sites ultimately resulting in inhibition of the bacterial glyoxalase pathway. 92 Chapter 3: Biased kinetic target guided synthesis 3.1 Towards the development of inhibitors of GlxI/II using kTGS Previous work has demonstrated the utility of kTGS to generate potent inhibitors and rapidly explore the binding landscape of a wide range of biological targets including: acetylcholine esterases,68 carbonic anhydrases,83 HIV proteases,84 chitinases,85 a mycobacterial transcriptional regulator,86 an antithrombotic target blood coagulation factor Xa79 and an insulin degrading enzyme,87 to name a few examples. To rapidly identify ‘hit’ compounds that have a high affinity for the bacterial glyoxalase system, GlxI and GlxII, and to build on the data described in the previous chapter two approaches were envisioned using a biased and an unbiased kTGS approach: 1) The biased approach required an azido peptidic scaffold to act as a handle in the binding site and to promote the Huisgen [3+2] cycloaddition reaction and 2) the unbiased approach for in situ fragment linking to identify non-peptidic novel ‘hit’ compounds. The unbiased approach can be viewed as a fragment based drug discovery tool in comparison to the current state of the art in the TGS field which is focused on a biased methodology. This chapter will deal with the biased approach and how it was used to validate this methodology as a tool to identify ‘hit’ compounds for GlxI and GlxII and to further explore the SAR of our target proteins. In order to initiate the biased kTGS, an azido derivative (GAz) of the peptide substrate GSH was designed and synthesised to act as a handle in the glutathione site (G-site). To test this biased kTGS approach in our hands the following screening cascade (Fig. 44) was designed where GAz was co-incubated with a diverse set of alkynes to rapidly explore the binding landscape in the hydrophobic site (H-site) utilising the Huisgen [3+2] cycloaddition to form triazoles. Due to the biased nature of the experiment, in regards to the high alignment of the tripeptide backbone of GAz with the GSH co-substrate acting as a handle in the GlxI and GlxII binding sites, it was expected that a relatively high ‘hit’ rate would be observed and it was anticipated that several ‘hit’ compounds would be identified. The compounds identified using LC-MS/MS would then be individually synthesised and biophysical evaluation carried out with DSF, enzyme kinetics and NMR to validate this biased approach with GlxI and GlxII. 93 Fig. 44. Proposed screening cascade for biased kTGS. GAz is shown in the predicted binding position in GlxII (PDB: 2QED). 3.1.1 Design of the biased azido peptide scaffold for kTGS As discussed in section 1.3, during the course of neutralisation of reactive electrophiles, i.e. MG, by the glyoxalase system, GSH reversibly reacts with MG to form a hemithioacetal (HTA) from which GlxI catalyses the formation of SLG. GlxII subsequently catalyses the hydrolysis of SLG to regenerate GSH and form neutralised D-lactate. As the GSH tripeptide backbone is conserved throughout the glyoxalase cycle and an indication of binding had been observed in the DSF data in section 2.3.4, it was therefore identified as a starting point to design an azido derivative to ensure compounds would retain high affinity within the G-site of GlxI and GlxII, thus enabling exploration of the adjacent H-site using a library of alkyne fragments in the in situ kTGS experiments. An examination of the literature136 revealed known inhibitors for the human homologues of GlxI (PDB: 1QIN) and GlxII (PDB: 1QH5), S-(N-hydroxy-N-p-iodophenylcarbamoyl) glutathione (1) and S-(N-hydroxy-N-bromophenylcarbamoyl) glutathione (31) respectively. Both inhibitors, 1 and 31, retained the tripeptide backbone and St e i se T S St e Synthesis of hits St e 3 io hysic e tion 94 functionalised the Cys thiol, with the carbamate moiety interacting with the metal ions in the active site and the phenyl ring occupying the H-site. These, and other reported glutathione S-conjugates including the DSF data in section 2.3.4, indicated the cysteine residue was an appropriate site for the modification with an azide (Fig. 45). Subsequent reaction with the azide resulted in triazoles, whereby comparable interactions with the metal ions and H-site could be observed. 1,2,3-Triazoles can chelate metals through the nitrogen’s therefore allowing this moiety to interact with the metal ions in the active site and the functionalised groups to occupy the H-site.137 A. B. Fig. 45. A. Human GlxI from (PDB:1QIN) with bounds S-(N-hydroxy-N-p-iodophenylcarbamoyl) glutathione (1) and B. Human GlxII (PDB: 1QH5) with bound S-(N-hydroxy-N-bromophenylcarbamoyl) glutathione (31).136 Visualised with Chimera. Taking these considerations into account, i.e. maintaining the tripeptide backbone to act as a handle in the G binding site, the azido derivative was designed with an azide functional group in place of the thiol. The azide moiety is a small and non-disruptive modification and so GAz (32) was anticipated to have similar binding affinity to the parent compound GSH (Fig. 46). Fig. 46. Structure of glutathione (GSH) and γ-glutamyl-L-azido-alaninylglycine (GAz). To investigate the binding mode, docking of GAz (32) to GlxI (PDB:1F9Z in E. coli) and GlxII (PDB: 2QED in Salmonella thyphimurium) was performed using Vina based on the binding orientation of GSH. The predicted binding modes were visualised with Chimera (with Dr Geoff Wells, UCL School of Pharmacy). 95 GAz (32) was predicted to bind in the G-site of GlxI (PBD: 1F9Z) with the tripeptide backbone in a Y-conformation held in position by inter-molecular hydrogen bonds between the Gly carboxylic acid and a water residue (HOH 256.B), Glu-NH2 and asparagine (Asn 60.A), the Glu carboxylic acid and arginine (Arg 9.A) and a water residue (HOH 1296.A) (Fig. 47 A). GAz (32) is also predicted to bind in the G-site of GlxII (PDB: 2QED) with the tripeptide backbone adopting the Y-conformation (Fig. 47 B). Encouragingly, in both predictions, the azide moiety was observed pointing towards the hydrophobic binding site and the metal binding site. A B Fig. 47. A. Predicted binding of GAz (32) in the G-site of GlxI (PDB: 1F9Z). Green lines indicated H-bond between GAz and the residues. B. Predicted binding of GAz (32) in the G-site of GlxII (PDB: 2QED). Visualised in Chimera. 3.1.2 Synthesis of azido peptide scaffold γ-glutamyl-L-azidoalaninylglycine (GAz) With our target peptide scaffold for kTGS designed, the synthesis of γ-glutamyl-L-azidoalaninylglycine (GAz, 32) was initiated with the preparation of a diazo-transfer reagent imidazole-1-sulfonyl azide hydrochloride (33) by slowly adding sulfuryl chloride to sodium azide, followed by the portion wise addition of imidazole. Ethanolic hydrochloric acid was added to the crude intermediate to form the HCl salt of the desired imidazole-1-sulfonyl azide (33) in comparable yields to those reported (Scheme 7).138, 139 Scheme 7. Synthesis of the azido transfer reagent. Reagent and conditions: (i) NaN3 (1 eq), MeCN, 0 °C – RT, 19 h. (ii) Imidazole (2 eq), 0 °C – RT, 3 h, 70% yield. The key step in the synthesis of GAz (32) was the preparation of the unnatural amino acid Fmoc-L-azidoalanine (36). This was achieved using a Hoffman rearrangement (i) 96 from Fmoc-L-Asn (34), followed by a diazo transfer140 (ii) as per the route published by Spring et al.141 (Scheme 8). In step (iii) Fmoc-L-azidoalanine (36) underwent an amide coupling under standard conditions followed by Fmoc deprotection under basic conditions (iv) to afford intermediate L-azidoalaninylglycine tert-butyl ester (38). The final amide coupling, to afford the assembled peptide incorporating a γ-glutamyl linkage (39), was achieved under standard conditions as before. Global deprotection under acidic conditions with the addition of a scavenger furnished the desired product γ-glutamyl-L-azidoalaninylglycine (32, GAz) in 11% yield over 6 steps. Scheme 8. Synthesis of γ-glutamyl-L-azidoalaninylglycine (32, GAz). Reagent and conditions: (i) [bis(trifluoroacetoxy)iodo]benzene (1.5 eq), DMF, water, RT, 15 min, pyridine, RT, 21 h, 64% yield. (ii) water, methanol, DCM, CuSO4.5H2O (0.006 eq), 33 (3.1 eq), pH 9, RT, 23 h, 77% yield. (iii) glycine-tert-butyl ester (1.05 eq), HBTU (1.05 eq), DMF, DIPEA (3 eq), RT, 3.5 h, 55% yield. (iv) 10% v/v piperidine/DMF, RT, 30 min, 70% yield. (v) N-Boc-L-glutamic acid-tert butyl ester (1 eq), HBTU (1.05 eq), DMF, DIPEA (3 eq), RT, 3 h, 84% yield. (vi) triethylsilane (6 eq), 80% v/v TFA/DCM, RT, 4 h, 74% yield. 3. .3 io hysic e tion of γ-glutamyl-L-azidoalaninylglycine (GAz) Before employing GAz (32) as the biased ligand in our kTGS experiments we first sought to demonstrate that our azido peptide bound to both GlxI and GlxII. DSF was used to obtain some initial indication of the binding affinity of GAz (32) relative to GSH for both GlxI and GlxII. An increase in the Tm of GlxI was observed in the presence of both 1 mM of GAz (32) and GSH with a ΔTm of 1.59 °C and 0.92 °C respectively, therefore these 2 compounds were deemed to have a similar stabilising effect on GlxI and are therefore likely to have similar affinities. This was encouraging and provided some evidence that modification of the peptide to include the azide group in place of the cysteine thiol did not significantly disrupt peptide binding and that GAz was, therefore, a suitable scaffold for use in biased kTGS experiments (Fig 48, Table 13). 97 A B Fig. 48. DSF to evaluate binding of GAz to GlxI. A. Melt curve and B. Bar graphs of thermal shift with respect to GlxI (ΔTm, °C). Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, ** p ≤ 0.01). Table 13. Thermal melt temperature (Tm, °C) and thermal shift (ΔTm) in the presence of GAz and GSH relative to that of GlxI alone. Compound Tm (°C) ΔTm ± GlxI 60.80 0.00 0.05 GSH 61.72 0.92 0.07 GAz, 32 62.39 1.59 0.41 An increase in the Tm of GlxII was also observed in the presence of 1 mM of GAz (32) with a ΔTm of 0.85 °C. A decrease in the Tm of GlxII with wide variability was observed in the presence of 1 mM of GSH with a ΔTm of -1.33 °C (Fig. 49, Table 14), this variability and spread of data had also been observed in section 2.3.4 and could be due to a number of factors including the lower affinity of GSH for GlxII due to it being the product of the enzymatic reaction or the purity of the material due to GSH’s potential to oxidize. In an effort to reduce the variability, fresh stock solutions were prepared on the day of use and single use ampules of GSH were used. A B Fig. 49. DSF to evaluate binding of GAz to GlxII. A. Melt curve and B. Bar graphs of thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII alone. Significance of changes evaluated by a Student’s t-test. 98 Table 14. Thermal melt temperature (Tm, °C) and thermal shift in the presence of GAz and GSH relative to that of GlxII alone. Compound Tm °C ΔTm ± GlxII 65.23 0.00 0.93 GSH 63.89 -1.33 4.4 GAz, 32 66.08 0.85 0.7 Fresh ampule stocks of GSH did provide the expected stabilization of GlxII however wide variability was still observed in this DSF screen. An increase in the Tm of GlxII was observed in the presence of both 1 mM of GAz (32) and GSH, from a new ampule, with a ΔTm of 0.53 °C and 1.07 °C respectively, therefore these 2 compounds were deemed to have a stabilising effect on GlxII (Fig. 50, Table 15). Thus, similarly to GlxI, these results provided some evidence that modification of the peptide to include the azide group in place of the cysteine thiol did not significantly disrupt peptide binding and was therefore suitable for use in biased kTGS experiments. A B Fig. 50. DSF to evaluate binding of GAz and stocks of GSH to GlxII. A. Melt curve and B. Bar graphs of thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII. Significance of changes evaluated by a Student’s t-test (where * p ≤ 0.05). Table 15. Thermal melt temperature (Tm, °C) and thermal shift in the presence of GAz and GSH (from ampule) relative to that of GlxII alone. Compound Tm (°C) ΔTm ± GlxII 66.53 0.00 0.40 GSH 67.60 1.07 2.80 GAz, 32 67.06 0.53 0.43 99 3.2 Towards the development of inhibitors of GlxII using biased kTGS With purified GlxII and our peptide scaffold, GAz (32) in hand, we could begin to evaluate the biased kTGS approach for the identification of novel inhibitors. To allow optimisation of conditions, a pilot experiment with GlxII, GAz (32) and the 45 alkyne fragments (library 1) was performed. Several conditions were taken into account including: controls, synthetic standards, analysis conditions (including column length and run time) and sample preparation. To enable the kTGS experiments an alkyne fragment library was assembled from commercial sources and consisted of 97 alkynes which were split into 2 libraries consisting of 45 and 52 fragments. The fragments were renamed according to their position in the 96-well plate and stored as 100 mM stocks in DMSO at -20 °C. Tables 16 and 17 show the structures of all alkyne fragments in the library. Table 16. Structure and identifying codes of alkyne fragments in library 1 as positioned in the 96-well plates. AlkA1 AlkA2 AlkA3 AlkA4 AlkA5 Alk1 AlkB1 AlkB2 AlkB3 AlkB4 AlkB5 Alk2 AlkC1 AlkC2 AlkC3 AlkC4 AlkC5 Alk3 AlkD1 AlkD2 AlkD3 AlkD4 AlkD5 Alk4 AlkE1 BLANK AlkE3 AlkE4 AlkE5 Alk5 AlkF1 AlkF2 AlkF3 AlkF4 AlkF5 Alk6 AlkG1 AlkG2 AlkG3 AlkG4 BLANK Alk7 AlkH1 AlkH2 AlkH3 AlkH4 BLANK Alk8 100 Table 17. Structure and identifying codes of alkyne fragments in library 2 as positioned in the 96-well plates. AlkA1-2 AlkA2-2 AlkA3-2 AlkA4-2 AlkA5-2 AlkA6-2 AlkA7-2 AlkB1-2 AlkB2-2 AlkB3-2 AlkB4-2 AlkB5-2 AlkB6-2 AlkB7-2 AlkC1-2 AlkC2-2 AlkC3-2 AlkC4-2 AlkC5-2 AlkC6-2 AlkC7-2 AlkD1-2 AlkD2-2 AlkD3-2 AlkD4-2 AlkD5-2 AlkD6-2 AlkD7-2 AlkE1-2 AlkE2-2 AlkE3-2 AlkE4-2 AlkE5-2 AlkE6-2 AlkF1-2 AlkF2-2 AlkF3-2 AlkF4-2 AlkF5-2 AlkF6-2 AlkG1-2 AlkG2-2 AlkG3-2 AlkG4-2 AlkG5-2 AlkG6-2 AlkH1-2 AlkH2-2 AlkH3-2 AlkH4-2 AlkH5-2 AlkH6-2 3.2.1 Pilot experiment to establish biased kTGS conditions The pilot biased kTGS experiment was set up as follows: GlxII was incubated with the azido peptide, GAz (32), and groups of up to 5 alkynes for 3 days at 37 °C. Control wells contained BSA or buffer incubated with the azide, GAz (32), and alkynes.68 The buffer and BSA controls were used to determine that the hit compounds synthesised in the experiment were the result of a protein templated effect and not the intrinsic reactivity of the fragments or non-specific binding. The alkynes were pooled into groups based on the molecular weights of the triazole products to enable identification of hits using LC-MS/MS. The final concentrations in each well were as follows: GAz (32) 1 mM, alkynes 1 mM, GlxII 10 µM; the total volume was 100 µL with <5% DMSO. Analysis of the crude 101 mixtures was carried out using the Agilent TripleQuad LC-MS/MS with an Aquasil C18 (100 x 4.6 mm, 5 µm particle size) column. Expected triazole masses were considered a ‘hit’ compound if observed in the GlxII wells with an area under the curve of at least double compared to the control wells. The sensitivity of the mass spectrometer was markedly increased when the single-ion-monitoring (SIM) scanning method was applied. An example is shown in (Fig. 51) where an overlay of protein and control LC-MS traces identifies a hit triazole for AlkF3. 13 triazoles were identified out of a possible 45 (excluding regioisomers) giving a hit rate of 28.9% (Fig. 52). Due to the large and open nature of the hydrophobic binding site in GlxII and the biased nature of the experiment, it was not surprising that a number of compounds with a wide diversity and sizes of R groups were identified as hits in the LC/MS-MS. Fig. 51. Representative data for ‘hit’ triazole for AlkF3. SIM scan on the left, ESI scan on the right. 102 Fig. 52. ‘Hit’ compounds identified in LC-MS/MS spectra from pilot biased kTGS experiment with GlxII, Gaz (32) and alkyne library 1. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. A few observations were made during this pilot biased kTGS experiment. Firstly, the length of time to analyse each sample on the Aquasil C18 (100 x 4.6 mm, 5 µm particle size) was 14 min. To reduce the analysis time per sample to 7 min and to increase the number of samples analysed per session, a shorter column (50 mm) was subsequently used. Secondly, it was observed that the retention times of peaks associated with triazoles did not always overlap between protein and control samples nor between MS2 scan and SIM methods. This resulted in method optimisation of the analysis conditions to ensure that the retention time of GAz (32) as an internal standard within each sample was the same. The addition of 2 min to the end of each run to allow the column time to re-equilibrate back to starting solvent conditions reduced the variability in retention time. The third point to address was that as the samples were being analysed crude this allowed buffer salts to go through the column and onto the LC-MS/MS source, so in order to extend the column life and remove the salts from building up on the source, each sample was extracted with acetonitrile. This more labour intense process that required each sample to be incubated with acetonitrile, centrifuged and extracted twice to allow the aggregation of protein and the extraction of fragments and triazoles into the organic layer. Due to a concern that the peptidic triazoles were not soluble in organic conditions which could result in hit compounds being missed, ultimately it was decided to not proceed with the acetonitrile extraction and a method to backwash the column was used to extend the life of the column. The one addition that made a really significant impact on the ease of hit identification in the kTGS assay was the introduction of the synthesis 103 plate,87 where fragments were incubated under synthetic conditions (copper sulfate in water:butanol (2:1)) with shaking for 3 days at RT to provide synthetic standards of the 1,4-triazoles. These standards were analysed using the LC-MS/MS conditions outlined above. The retention times and MS of the peaks were used as a guide to identify and confirm hits observed in the kTGS as well as to provide some insight into whether the 1,4 or 1,5 regioisomer was synthesised in the in situ optimisation experiment. These optimisations were applied to future experiments and resulted in a reliable and reproducible LC-MS readout. 3.2.2 Biased kTGS with GlxII To investigate whether different buffer conditions influenced the results of the kTGS experiment, the pilot experiment described above was repeated using the optimised conditions. The selected buffers were phosphate buffer and MOPS, which was selected as the buffer in the DSF experiments for GlxII in section 2.3.1. The 45 alkyne fragments from library 1 were pooled into the same 9 mixes of 5 fragments per well and each alkyne mix was incubated with GAz (32) in the presence of GlxII, BSA or buffer as before. The two buffers used were: 50 mM PBS, pH 7.4, 0.15 M NaCl and 10 mM MOPS, pH 7.2, 0.2 M NaCl. In the pilot kTGS experiment in section 3.2.1, 13 hits, 28.9% hit rate, were identified (Fig. 52). The alkyne codes identified were as follows: AlkB3, AlkC2, AlkE5, AlkH4, AlkD3, Alk6, AlkE1, AlkB2, AlkB1, AlkH2, AlkF3, AlkA3, AlkE3. In the experiment performed in MOPS buffer 8 hit compounds were identified: AlkC2, AlkD3, AlkE5, AlkB2, AlkF3, AlkH2 and Alk8 which corresponds to a 17.78% hit rate (Fig. 53). It is worth noting that 7 of the hits identified were observed in the previous pilot kTGS experiment and Alk8 was an additional hit. Fig. 53. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 1 in MOPS buffer. 0.005.0010.0015.00AlkB3 AlkC2 AlkD3 AlkE5 Alk8 AlkB2 AlkF3 AlkH2Protein 0.05 0.84 0.07 0.65 10.24 0.10 0.06 0.40BSA 0.00 0.31 0.00 0.32 3.78 0.01 0.03 0.04Buffer 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00% Triazole formation104 8 hit compounds, AlkB3, AlkC2, AlkD3, AlkE5, Alk8, AlkB2, AlkF3 and AlkH2, were identified in the experiment performed in PBS buffer, which corresponds to a 17.78% hit rate. Although AlkH2 does have similar percentage triazole formation in the PBS buffer control and protein, biophysical analysis of the synthesised compound should be studied to determine whether it is a hit compound and whether a variation in buffer affects the identification of hits due to the lack of percentage triazole formation in MOPS buffer (Fig. 54). The alkyne fragment structures are shown in Fig. 55. Fig. 54. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 1 in PBS buffer. Fig. 55. Compounds identified in biased kTGS with GlxII, GAz (32) and Alkyne library 1 in PBS. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. From these experiments, it appears that the buffer composition does not significantly affect the hits identified, although an improved % triazole formation in the presence of 0.002.004.006.008.00AlkB3 AlkC2 AlkD3 AlkE5 Alk8 AlkB2 AlkF3 AlkH2Protein 1.21 4.05 0.08 1.07 7.64 0.11 0.03 0.50BSA 0.07 0.05 0.00 0.13 0.00 0.07 0.01 0.29Buffer 0.08 0.03 0.00 0.25 0.00 0.00 0.01 0.53% Triazole formation105 GlxII relative to the controls was observed in PBS buffer. Some signals in the MS, initially thought to correspond to triazole masses, were identified in all protein samples suggesting that they were masses associated with the storage buffer and protein and not the kTGS assay. For example, m/z 420 was a commonly identified mass which could correspond to the mass of triazoles formed from Alk1, Alk4 and Alk2. Due to this reason, these alkynes were removed from the hits identified. Another alkyne which raised some concern was Alk8 as a high % triazole formation was observed which was up to 15 times greater in GlxII than that observed for other triazoles. The triazole formation is templated by the protein and the environment around the binding site dictates whether the 1,4 or the 1,5-regioisomer is formed in the in situ experiment. To investigate this further and to gain some insight into whether the triazoles formed in situ were the 1,4 or 1,5 regioisomer the LC-MS/MS traces were overlaid with the synthesis plate traces, where a 1,4 triazole should have the same retention time in the synthesis plate, the protein and the control wells providing an indication of regioisomers. An example of a 1,4-triazole is AlkF3, where there is a clear overlay between the experiment wells and the synthetic standard (Fig. 56 A). With AlkC2, however, the synthetic standard does not overlay with the experiment wells, suggesting a 1,5-triazole (Fig. 56 B), although these regioisomer indications would need to be confirmed through synthesis and biophysical evaluation. A. B. Fig. 56. LC-MS/MS trace overlays for the synthesis, protein, BSA and Buffer of A. AlkF3 as a 1,4-triazole and B. AlkC2 as a 1,5-triazole. The overlays between GlxII, controls and synthetic standard suggested that the triazole hits derived from the following alkynes were 1,4-triazoles: Alk6, Alk8, AlkB1, AlkB2, AlkB3, AlkD3, AlkE3, AlkF3, AlkH2, and AlkH4. The trace observed with the triazole derived from AlkE5 indicated that a mixture of 1,4 and 1,5-triazoles may have formed in 106 the in situ experiment. The alkynes, AlkC2 and AlkE1, were suggested to form 1,5-triazoles from the trace overlays. With promising results obtained with fragments from alkyne library 1, the biased kTGS with GlxII was expanded and carried out with the 51 alkyne fragments from alkyne library 2 in both MOPS and PBS buffer. The fragments were pooled into 11 mixes, with up to 5 fragments per well, and each alkyne mix incubated with GAz (32) in the presence of GlxII, BSA, buffer or synthetic conditions as before. 6 triazoles were identified in PBS buffer which corresponded to an 11.8% hit rate: AlkG6-2, AlkH5-2, AlkH6-2, AlkC1-2, AlkC6-2 and AlkA4-2 (Fig. 57 and 58). Fig. 57. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 2 in PBS buffer. Fig. 58. Compounds identified in biased kTGS with GlxII, GAz (32) and Alkyne library 2 in PBS buffer. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. 0.002.004.006.008.00AlkG6-2 AlkH5-2 AlkH6-2 AlkC1-2 AlkC6-2 AlkA4-2Protein 1.65 1.54 2.88 1.88 4.20 6.45BSA 0.00 0.00 1.10 0.40 0.50 0.43Buffer 0.00 0.31 0.81 0.35 0.00 0.97% Triazole formation107 In the kTGS experiment in MOPS buffer 6 hit compounds were also identified: AlkC2-2, AlkC7-2, AlkH6-2, AlkF6-2, AlkD2-2, AlkC6-2 which corresponded to a 11.8% hit rate (Fig. 59 and 60). Fig. 59. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxII, GAz (32) and alkyne library 2 in MOPS buffer. Fig. 60. Compounds identified in biased kTGS with GlxII, GAz (32) and Alkyne library 2 in MOPS buffer. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. In this second set of kTGS experiments with alkyne library 2 in PBS and MOPS buffers, six triazoles were identified in each experiment although only two were crossover hits, namely the triazoles derived from alkynes AlkH6-2 and AlkC6-2. Although the triazoles formed from alkyne fragments AlkC2-2, AlkC7-2, AlkF6-2 were observed in both the MOPS and PBS experiments they had a very high % triazole formation of 22 – 40% in the presence of GlxII, but not the controls, in the PBS experiment suggesting an anomalous result and were therefore discounted as hits. However, to confirm whether these compounds should be discounted, the triazole derived from GAz (32) and AlkF6-0.002.004.00AlkC2-2 AlkC7-2 AlkH6-2 AlkF6-2 AlkD2-2 AlkC6-2Protein 2.87 0.47 0.02 2.29 0.65 0.12BSA 0.56 0.00 0.00 0.11 0.02 0.00Buffer 0.31 0.00 0.00 0.11 0.04 0.00% Triazole formation108 2 was included in the list of candidates for synthesis. AlkD2-2 was observed in both PBS and MOPS experiments but had approximately 1% triazole formation in both GlxII and controls in the PBS experiment. AlkG6-2, AlkH5-2, AlkC1-2 and AlkA4-2 were observed in PBS, however, were not classified as hits in the MOPS experiment due to the same % triazole formation being observed in both protein well and the buffer control. A summary of indicated regioisomers from the LC-MS/MS trace overlays between the synthesis, GlxII and control results were as follows. For the PBS experiment the triazoles formed from the following alkynes, AlkA4-2, AlkC1-2, AlkC6-2 and AlkH5-2 were 1,4-triazoles. The trace overlay for the triazole formed between GAz and AlkH6-2 suggested that a mixture of 1,4- and 1,5-triazole may have formed in the in situ experiment. For the triazole derived from AlkG6-2 the overlay indicated that the 1,5-regioisomer formed in the kTGS experiment. For the MOPS experiment all the triazoles were indicated to be 1,4-regioisomers. In order to evaluate whether the hit compounds identified in the kTGS experiments were inhibitors of GlxII it was necessary to synthesise the 1,4-triazoles from GAz (32) and the corresponding alkyne fragments. From alkyne library 1: AlkB3, AlkC2, AlkE5, AlkH4, AlkD3, Alk6, AlkE1, AlkB2, AlkB1, AlkH2, AlkF3, AlkA3, AlkE3 and from alkyne library 2: AlkG6-2, AlkC1-2, AlkH6-2, AlkF6-2, AlkD2-2. All triazoles identified in library 1 were synthesised, however, due to the cost or lack of commercial availability a large number of the alkynes identified in either the MOPS or PBS experiments for library 2 were not included in the synthesis list. Two alkynes from the MOPS results (AlkF6-2, AlkD2-2), two from the PBS results (AlkG6-2, AlkC1-2) and one from both assays (AlkH6-2) were included in the synthesis list. The majority of the alkynes were synthesised, several were purchased and the remaining alkynes were not included. 3.2.3 Synthesis of alkyne building blocks and triazoles Synthesis of the hit compounds identified in the in situ kTGS experiments was required to allow biophysical analysis and confirmation of hits. As the triazole synthesis in the in situ experiment is performed in the absence of a metal catalyst the regiochemistry of the triazole produced is dictated by the protein environment. Thus, it was proposed to synthesise both the 1,4 and 1,5-triazoles of all identified hits to confirm whether the regioisomers could be distinguished by overlaying the LC-MS/MS traces with the 1,4-synthetic standard as discussed above or if any lessons could be learned for future kTGS experiments. The required alkyne fragments were obtained in one-step using alkylation, 109 acylation and urea formation conditions respectively, as summarised in Table 18 or purchased from chemical suppliers, as summarised in Fig. 61. Table 18. Summary of alkyne synthesis conditions. SM = starting material. SM1 SM2 Conditions, % yield Alkyne fragment a, 42% AlkC2, 40 a, 59% AlkB2, 41 a, 53% AlkA3, 42 a, 85% AlkF3, 43 b, 33% AlkE5, 44 c, quantitative AlkE1, 45 d, 45% AlkD3, 46 d, 37% AlkB1, 47 e, 7% AlkH6-2, 48 f, 18% AlkD2-2, 49 Reagents and conditions: (a) K2CO3 (2 eq), DMF, RT, 16 h. (b) NaH (1.1 eq), DMF, 0 °C-RT, 16 h. (c) DCM, 0 °C-RT, 3.5 h. (d) DCM, 0 °C-RT, 6 h. (e) Et3N (1.2 eq), THF, 0 °C-RT, 4 h. (f) CHCl3:H2O (1:1), K2CO3 (3 eq), RT, 16 h. 110 AlkB3 AlkH2 Alk6 AlkC1-2 AlkE3 AlkH4 AlkG6-2 AlkF6-2 Fig. 61. Structures of alkynes purchased from chemical suppliers. With the alkyne and GAz (32) building blocks in hand, it was proposed to synthesise the 1,4 regioisomers of the triazole hits using established copper catalysis conditions142 (Scheme 9). However, poor yields were obtained and purifications were challenging. To try address these problems, the amount of copper (II) sulfate pentahydrate was increased from 2 mol% to 0.75 eq and the number of equivalents of (+)-sodium L-ascorbate was increased from 15 mol% to 1 eq. This change did improve yields however purification was still challenging with conditions (g) (Table 19). A literature search revealed that the addition of 2 mol% tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine (TBTA)143,144 which stabilises the copper I transition intermediate could lead to improved yields. Indeed, this was found to be the case in our hands with conditions (h), with improved yields and ease of purification (Table 19). Synthesis of the 1,4-triazole with Alk8 was attempted, however the TMS group was removed under the reaction conditions and the compound subsequently degraded and could not be isolated so was not taken further. All other hits shortlisted for synthesis in the biased kTGS experiments with GlxII were successfully synthesised as the 1,4-regioisomer with GAz (32). The summary of 1,4-triazoles synthesised is shown in Table 19. Scheme 9. Synthesis of the biased 1,4-triazole hits for GlxII. Reagents and conditions: (g) GAz (32) (1 eq), alkyne (1 eq), H2O:t-BuOH (2:1), sodium ascorbate (1 eq), Cu(II)SO4.5H2O (0.75 eq), RT, 16-20 h. (h) GAz (1 eq), alkyne (1 eq), H2O:t-BuOH (2:1), sodium ascorbate (15 mol%), Cu(II)SO4.5H2O (2 mol%), TBTA (2 mol%), RT, 16-20 h. 111 Table 19. Summary of synthesis of biased 1,4-triazole hits for GlxII. Alkyne fragment Structure Yield, ID Alkyne fragment Structure Yield, ID AlkC2 (40) g, 34%, h, 33% 50 AlkB1 (47) h, 29% 51 AlkB2 (41) g, 14% 52 AlkB3 h, 26% 53 AlkA3 (42) h, 23% 54 AlkE3 h, 65% 55 AlkF3 (43) g, 15% 56 AlkH2 h, 74% 57 AlkE5 (44) h, 50% 58 AlkH4 h, 85% 59 AlkE1 (45) g, 34% 60 Alk6 h, 33% 61 AlkD3 (46) g, 0%, h, 60% 62 AlkH6-2 (48) h, 25% 63 112 Alkyne fragment Structure Yield, ID Alkyne fragment Structure Yield, ID AlkG6-2 h, 12% 64 AlkF6-2 h, 66% 65 AlkC1-2 h, 60% 66 AlkD2-2 (49) h, 50% 67 To synthesise the 1,5-triazoles, so as to provide insight into whether the synthetic LC-MS/MS traces were an indication of the preferred regioisomer formed in the in situ experiment step, a feasibility study to identify suitable synthetic conditions was undertaken. A series of test experiments considered a number of variables including: alkynes (AlkD3, AlkB1, AlkB3, AlkH4, AlkF3), anhydrous vs. non-anhydrous reactions, solvent variations (dioxane, toluene, acetonitrile, DMF, toluene/water), controlling the order of addition, not exceeding 90 °C temperature to prevent the decomposition of the ruthenium catalyst, microwave vs. thermal conditions, varying the ruthenium catalyst (RuCp*Cl(PPh3)2145 and Ru(COD))146-147 and mol% thereof. However, no product formation was observed under any of these conditions tested. GAz contains a number of potential ruthenium catalyst co-ordination sites, such as the unprotected Glu-amine. To evaluate if this was the problem a number of test reactions were carried out using the fully protected GAz (32), however again no reaction was observed under any conditions attempted. Due to the challenges encountered with the synthesis of the 1,5-triazole regioisomers hits the decision was made to proceed with the evaluation of the 1,4 regioisomers. Future attempts should include bespoke 1,5-triazole synthesis from the beginning of the GAz (32) synthesis, work currently being carried out in the group is showing promise with a 1,5-triazole being incorporated at the dipeptide stage. 3.2.4 Biophysical evaluation of hits To evaluate the binding to and inhibition of GlxII by the triazole hits identified in the kTGS experiments, DSF and enzyme kinetics assays were employed. 113 Due to conservation of the tripeptide backbone the majority of the triazoles were expected to retain comparable binding affinities for GlxII compared to GAz (32). Therefore, the variation in binding affinities depended on the triazole substituent and their interactions with the hydrophobic pocket. Apart from the 56, 55, 64, 66 and 63 which had ΔTm’s > than 1 °C, the ΔTm’s observed did not provide an enhancement of binding affinities compared to GAz (ΔTm 0.46 °C). This has been observed in other kTGS studies148,102 and was hypothesised to be due to an open and flat binding site. Binding sites with more enclosed binding pockets could organise reactive fragments on the surface of the enzyme more successfully due to the close proximity of groups. The DSF data is shown in Fig. 62 and Table 20. In the presence of 1 mM of the 1,4-triazoles, GlxII was found to have ΔTm’s in the range of -5.39 to 2.28 °C. Five of the compounds tested 56 (2.28 °C), 55 (1.03 °C), 64 (1.55 °C), 66 (1.05 °C) and 63 (1.08 °C) showed a positive ΔTm providing evidence of significant stabilisation of GlxII and therefore binding. 57 (0.15 °C), 50 (-0.72 °C), 52 (-0.44 °C), 62 (-0.75 °C), 53 (-0.59 °C), 59 (-0.22 °C) and 61 (-0.03 °C) did not demonstrate much effect on the melt temperature of GlxII indicating weak or no binding. For ligands 58 (-1.58 °C), 60 (-3.61 °C), 51 (-1.08 °C), 65 (-5.39 °C), 67 (-3.65 °C) and 54 (-5.62 °C) a significant decrease in ΔTm was observed implying binding to the denatured GlxII.129 The triazoles predicted to be in the 1,5 regioisomer from the LCMS overlays during hit identification were AlkC2 (40), AlkE1 (45) and AlkG6-2. A negative ΔTm was observed in the presence of the 1,4-triazoles, 50 and 60, derived from AlkC2 (40) and AlkE1 (45), suggesting that the 1,5-triazoles might be the preferred regioisomer. However, with the 1,4-triazole, 64, derived from AlkG6-2 showed a higher Tm relative to the protein alone thereby indicating the 1,4-triazole was tolerated in this case. These observations could be due to the aromatic and rigid nature of 50 and 60 compared to the flexible sp2 nature of 64, this flexibility could possibly allow the tetrahydropyran or alcohol moiety to interact with residues in the open H-site. 114 Fig. 62. Bar graphs of thermal shift (ΔTm, °C) in the presence of the ligand relative to that of GlxII alone. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 20. Melt temperature and thermal shift of GlxII in the presence of the biased triazoles. Tm (°C) ΔTm ± GlxII2%DMSO 67.12 0.00 0.36 50 66.40 -0.72 0.11 51 66.04 -1.08 1.43 52 66.69 -0.44 0.08 53 66.54 -0.59 0.67 54 61.50 -5.62 1.12 55 68.15 1.03 0.45 56 69.40 2.28 0.66 57 67.27 0.15 0.24 58 65.54 -1.58 0.51 59 66.90 -0.22 0.28 60 63.51 -3.61 3.28 61 67.10 -0.03 0.20 62 66.37 -0.75 0.53 63 68.21 1.08 0.35 64 68.68 1.55 1.33 65 61.73 -5.39 0.92 66 68.18 1.05 0.75 67 63.48 -3.65 1.10 115 To provide further insight into the DSF results described above and to investigate this further, docking studies were employed to observe how these triazoles were predicted to bind with GlxII. Docking studies were carried out in collaboration with Dr Geoff Wells (UCL School of Pharmacy). AutoClickChem149 was used to generate an in silico library of all possible triazole products between GAz and alkyne library 1 and library 2. To date no crystal structure of GlxII from E. coli has been published, therefore the crystal structures of GlxII from human and Salmonella enterica were obtained from the Protein Data Bank (PDB). LeDock and Vina150 were used to dock the ligands with GlxII and Chimera was used to visualise that data. Prior to docking, the apo structure of GlxII from S. enterica (PDB: 2QED) was overlaid with the human crystal structure (PDB: 1QH5) to identify the important binding site residues, as discussed in greater detail in section 1.3.2. The overlay shown in Fig. 63 revealed the overall 3D fold is conserved and that a number of key residues with the active site are conserved. Fig. 63. Overlay of human GlxII (1QH5 in blue) and S. enterica (2QED in grey). The important residues around the 2 Fe metals in the active site were His111, His54, His56, His59, Asp58, Asn13, His166, Asp128 and 3 water molecules (Fig. 64). 116 A. B. Fig. 64. (A) human GlxII (1QH5 in blue) and S. enterica (2QED in grey) overlay and the (B) active site of 2QED. GSH was used as a model to evaluate to how the triazoles were bound in the binding site and it was observed that Vina gave a better prediction of docking for the triazoles orientation compared to LeDock. The Vina docking scores ranged from -7.4 to 4.8 kcal/mol, where the lower docking score indicated a better fit for the structure. Fig. 65 shows the predicted binding affinity for each triazole and those hits identified in the kTGS assay are highlighted in red. The hits are scattered throughout the docking score graph and so there does not appear to be a correlation between docking score and affinity. Fig. 65. Vina docking scores for 1,4 and 1,5 triazoles in Glyoxalase II (2QED). The compound with the highest ΔTm, 56 (ΔTm 2.28 °C) was predicted to bind in the glutathione binding site with a Vina docking score of -6.9 kcal/mol, forming a hydrogen bond interaction between the carbonyl oxygen and the amine of Gln11 as well as an intramolecular hydrogen bond. The bromophenyl group of the triazole is predicated to form a π stacking interaction with Tyr168 in the hydrophobic pocket (Fig. 66). -7.7-7.2-6.7-6.2-5.7Docking score (kcal/mol)Vina Score kTGS Hits117 Fig. 66. Chimera representation of 56 predicted docking in GlxII (2QED). Two other 1,4-triazoles, 55 (ΔTm 1.03 °C) and 66 (ΔTm 1.05 °C), that demonstrated measurable stabilisation of the GlxII melting temperature and good Vina docking scores of -7.0 and -7.2 kcal/mol, respectively, were both predicted to bind with the tripeptide backbone in the glutathione binding site as expected and the functionalised triazole filling part of the hydrophobic pocket (Fig. 67). A B Fig. 67. Chimera representation of 55 (A) and 66 (B) predicted docking in GlxII (2QED). Two examples of triazoles resulting in a negative ΔTm for GlxII were 54 (ΔTm -5.62 °C) and 65 (ΔTm -5.39 °C), however these compounds were observed to have low docking scores of -6.7 and -6.0 kcal/mol, respectively. Both compounds were predicted to bind with the tripeptide backbone twisted to partially occupy the glutathione and hydrophobic binding site and the functionalised triazole filling the remaining part of the hydrophobic pocket or coming out of the binding pocket (Fig. 68). G-site H-site H-site G-site G-site H-site 118 A B Fig. 68. Chimera representation of 54 (A) and 65 (B) predicted docking in GlxII (2QED). The docking studies gave an interesting insight as to how the hits identified in the kTGS experiments were predicated to bind in GlxII and how these compounds could be developed further if required. However, the wide spread of docking scores for the kTGS hits either suggests an improved docking model is required or that kTGS identifies hits that would otherwise be missed through traditional screening methods. To further explore the difference between the predicted binding affinities and quantitative biophysical data and to understand the effect the synthesised hit compounds were having on the enzymatic activity of GlxII, established enzyme kinetic assays were employed.151,152,153 Fig. 69. Schematic of the glyoxalase pathway under analysis. The first assay, a colorimetric assay, involved the use of 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB or Ellman’s reagent) to monitor the production of GSH (Fig. 69).151,152 Thiols react with Ellman’s reagent cleaving the disulfide bond to give 2-nitro-5-thiobenzoate, which under alkaline pH and aqueous conditions ionizes to the anion resulting in a yellow colour. This allows the concentration of thiols to be measured by monitoring absorbance at 412-420 nm over time (Scheme 10). Scheme 10. Formation of GSH monitored by the reaction with Elman’s reagent. G-site H-site H-site G-site 119 The second assay employed directly monitored the decrease in SLG concentration by a reduction in absorbance at 240 nm over time as it was hydrolysed to GSH by GlxII.153 Two methods were used to determine the effect of the triazoles on GlxII activity, firstly the initial rate analysis and secondly time to steady state. The initial rate analysis was used to determine the % inhibition and the change in absorbance over time was used to calculate the specific activity of GlxII with and without the incubated triazoles. To obtain the optimal experimental conditions using Ellman’s reagent to evaluate the compounds as inhibitors of GlxII, optimisation of assay conditions was required. The relative inhibition of GlxII was calculated from the slope of the absorbance (OD) at 412 nm vs time curve. A number of preliminary assays were performed varying the protein and SLG concentration in the assay. A range of concentrations of GlxII (0, 2, 5, 10 and 20 µM) were evaluated with 0.9 mM S-D-lactoyl glutathione, 0.2 mM DTNB, 100 μL total volume (buffer: 40 mM PBS, pH 7.5, 0.26 M NaCl). The results showed that 2 µM GlxII allowed the initial linear slope of the curve to be observed and a similar time to Vmax was detected compared to higher enzyme concentrations (Fig. 70. A). An additional experiment was carried out to observe if a lower concentration could be used, 0.2 and 2 µM GlxII were monitored and an improved observation of the initial slope could be measured with 0.2 µM (Fig. 70. B). A publication by Reiger152 suggested that 0.2 mM DTNB was optimal and this was employed in all experiments. A subsequent set of experiments evaluated the effect of SLG concentration (0.1, 0.3, 0.4, 0.6 and 0.9 mM) with 0.2 µM GlxII and 0.2 mM DTNB. The results showed that 0.4 mM SLG allowed the initial slope of the curve to be observed and a similar time to Vmax to calculate the effect of the ligands on relative GlxII activity (Fig. 70. C). 0.4 mM SLG was also found to be the optimal concentration reported by Reiger et al.152 To test that the production of GSH was due to hydrolysis of SLG by GlxII a number of negative controls were performed. These included: combinations of buffer, SLG and DNTB, which showed no GSH formation. GSH formation was only observed when both SLG and GlxII were present (Fig. 70. D). 120 A. B. C. D. Fig. 70. OD at 412 nm vs time for: A) Variation of GlxII concentration (0 – 20 µM). B) GlxII concentration at 0.2 and 2 µM. C) Variation of SLG concentration (0.1 – 0.9 mM). D) negative controls. The optimal conditions in our hands were found to be as follows: 0.2 μM Glyoxalase II protein, 0.4 mM S-D-lactoyl glutathione, 0.2 mM DTNB, 100 μL total volume in 40 mM PBS, pH 7.5, 0.26 M NaCl and 2% DMSO and these conditions were used in all subsequent experiments. To find the optimal ligand concentrations for the evaluation of the hit triazoles, 2 compounds were chosen, S-(4-bromobenzyl) glutathione (3) and GAz (32) and evaluated over a range of concentrations from 0 – 2000 μM. 3 was chosen as it was a known human glyoxalase inhibitor (as described in Chapter 2) and demonstrated binding to GlxII in the DSF studies (ΔTm 2.03 °C). Gaz (32) the biased peptide scaffold employed in the kTGS experiments (Fig. 71) also demonstrated binding to GlxII in the DSF studies (ΔTm 0.85 °C). Fig. 71. Structure of ligands evaluated by enzyme kinetics. The raw data obtained is summarised in Fig. 72, where the slope, which was calculated form the initial rate in the OD (at 412 nm) vs time curve, was normalised to GlxII + 2% 121 DMSO = 1, where 100% GlxII activity is assumed, and the slope from the various ligands concentrations were converted to a ratio of slope (ligand) / slope (GlxII + 2% DMSO). The % inhibition was calculated as [slope (GlxII+2% DMSO) – slope (ligand)]*100 as shown in Fig. 73. A. B. Fig. 72. Graph of OD vs time for A) 3 and B) GAz (32) over 0 – 2000 µM. Fig. 73. Normalised slope of OD vs time curve used to calculate % inhibition of 3 and GAz (32) over 0 - 2000 µM. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). To use this normalised data to evaluate 3 and GAz (32) 2 concentrations were chosen. From the OD vs time (Fig. 72) and normalised slope vs concentration graphs (Fig. 73) the 2 concentrations chosen for comparison for the remaining ligands was 10 and 100 µM (Fig. 74). A. B. Fig. 74. A) Normalised slope calculated from OD (412 nm) vs time graph and B) % inhibition at 10 and 100 µM for 3 and GAz (32). Significance of changes evaluated by a Student’s t-test (where * p ≤ 0.05). 122 With the assay conditions and ligand concentrations optimized, the assay was applied to evaluate the synthesized triazoles hits from the biased kTGS experiments. The calculated % inhibition of relative GlxII activity is summarized in Fig. 75 and Table 21. From this analysis, it was found that 83.3% of the triazoles from the biased kTGS experiments showed inhibition of relative GlxII activity at 10 µM and 88.9% at 100 µM. Six compounds demonstrated a concentration dependent increase in inhibition upon increase from 10 to 100 µM and seven showed no concentration dependence. 63 did not exhibit measurable inhibition of GlxII under the conditions tested. 53 demonstrated inhibition at 100 µM and 54 appeared to inhibit GlxII at 10 µM however due to the variability in the data it was not significant. This inhibition data, although exhibiting wide variability in the spread of data, was encouraging. Fig. 75. % Inhibition of relative GlxII activity for biased synthesised triazoles. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 21: Summary of % inhibition data for biased triazoles with GlxII measure at 412 nm. % Inhibition % Inhibition Structure Alkyne and Triazole ID 0μM 00μM Structure Alkyne and Triazole ID 0μM 00μM GAz 32 12.40 26.18 AlkH4 59 44.34 44.27 AlkC2 (40) 50 57.15 35.29 AlkE1 (45) 60 50.66 57.33 123 % Inhibition % Inhibition Structure Alkyne and Triazole ID 0μM 00μM Structure Alkyne and Triazole ID 0μM 00μM AlkB1 (47) 51 49.25 36.75 Alk6 61 52.53 62.22 AlkB2 (41) 52 46.23 50.09 AlkD3 (46) 62 50.79 53.79 AlkB3 53 -14.09 14.60 AlkH6-2 63 -29.08 -16.25 AlkA3 (42) 54 47.00 -43.41 AlkG6-2 64 31.28 30.07 AlkE3 55 21.62 40.98 AlkF6-2 65 76.98 74.85 AlkF3 (43) 56 52.31 55.04 AlkC1-2 66 20.52 41.43 AlkH2 57 11.50 43.30 AlkD2-2 (49) 67 73.99 57.57 AlkE5 (44) 58 40.78 68.76 In summary, seven compounds demonstrated similar % inhibition at both 10 and 100 µM; 56 (52 and 55%), 52 (46 and 50%), 60 (50 and 57%), 62 (50 and 53%), 59 (44%), 64 (31 and 30%) and 65 (76 and 74%). Six demonstrated a concentration dependent increase in % inhibition with increase ligand concentration (10 to 100 µM): 58 (40 to 68%), 53 (-14 to 14%), 55 (21 to 40%), 57 (11 to 43%), 61 (52 to 62%) and 66 (20 to 41%). Three compounds showed a decrease in % inhibition with increased ligand concentration: 50 (57 to 35%), 51 (49 to 36%) and 67 (73 to 57%). 124 The wide spread of data points observed during this DTNB assay was a source of concern and could be due to a number of factors including:154,155 interference by compounds or the biological sample which affect absorption at 412 nm; oxidation of GSH which effects the DTNB reaction resulting in measurement errors; the known instability of DTNB where maintaining a pH <8 is very important to limit degradation which leads to background absorption. In an effort to improve the reproducibility of the assay it was repeated with the addition of a 10 min preincubation step (ligand, GlxII and SLG) at room temperature prior to the addition of DTNB and OD measurement at 412 nm, which would allow for measurement of the endpoint of the reaction. To analyze the results OD readings at 5.6 min where normalized and used to calculate the % inhibition. However, the results were inconsistent and the reproducibility was not improved, therefore the data was not included. Given the challenges observed with the DTNB assay a second enzyme kinetics assay was employed in which the decrease in SLG concentration (as a result of hydrolysis to GSH) was monitored directly by a reduction in absorbance at 240nm over 20 min. The change in absorbance over time was used to calculate the specific activity of GlxII using Beers Law, where the change in molar absorption coefficient was Δε240nm = -3.10 mM-1.cm-1, the relative specific activity was subsequently used to calculate the % inhibition in the presence of the triazoles ligands. The specific activity calculated was in units/mg (U/mg). The protocol described by Thornalley et al.153 was used to measure the glyoxalase activities. Where SLG (3 mM) in buffer (100 mM Tris/HCl pH 7.4) and water at 37 °C was added to GlxII (0.2 µM) with or without a ligand, and the decrease in SLG concentration was monitored by absorbance at 240 nm over time. As before, a range of concentrations of GlxII were evaluated including: 0, 0.2, 0.4, 1, 2 and 5 µM. Where 0.2 µM gave a similar profile to 1, 2 and 5 µM, and so was used in all subsequent assays to minimise protein consumption. GAz (32) was once again used to evaluate appropriate ligand concentrations for use in the inhibition assays (Table 22, Fig. 76). As was observed with the previous spectroscopic studies, the spread of data was quite wide leading to a lack in significance. It also appeared that above 50 µM concentration, GAz (32) activated the GlxII enzyme as opposed to inhibition. However, to enable direct comparison of results between the two assays the compound concentrations used for analysis were 10 and 100 µM. 125 Table 22. Specific activity and % Inhibition for GAz (32) with GlxII. Specific Activity % Inhibition from activity GlxII_2%DMSO 1.39 0.00 1 µM 0.94 32.76 5 µM 1.47 -5.13 10 µM 1.23 12.06 50 µM 1.70 -22.05 100 µM 2.13 -52.49 200 µM 2.09 -50.20 500 µM 1.98 -42.27 1000 µM 2.11 -51.40 0 M1 M5 M10 M50 M100 M200 M500 M1000 M-1 0 0-5 005 0G A z% inhibition GlxII* Fig. 76. % Inhibition of GlxII with GAz (32) at 0 – 1000 µM measured at 240 nm. Significance of changes evaluated by a Student’s t-test (where * p ≤ 0.05). The 1,4-triazoles from the biased kTGS were evaluated at 10 and 100 µM using the optimised assay conditions as described above. Where 72% of the triazoles tested were found to inhibit GlxII activity at 10 µM and 83% at 100 µM (Fig. 77, Table 23). 126 Fig. 77. % Inhibition bar graphs of GlxII due to 1,4-triazoles at 10 and 100 µM measured at 240 nm. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 23. Specific activity (U/mg) and % inhibition of GlxII with biased 1,4-triazoles. ǂ Data not included. GlxII Specific Activity % Inhibition Compound ID 10 µM 100 µM 10 µM 100 µM GlxII_2%DMSO 1.16 0.99 0.00 0.00 50 0.48 0.61 58.81 38.10 51 1.04 0.52 10.39 47.33 52 8.33 0.34 ǂ 65.60 53 0.79 0.60 31.91 38.66 54 2.64 0.29 -126.75 70.22 55 0.83 0.68 29.15 30.86 56 0.77 0.67 33.73 32.45 57 1.04 1.30 10.76 -31.63 58 0.52 0.49 55.77 50.01 59 1.17 0.67 -0.39 32.39 60 1.10 0.78 5.48 21.30 61 0.89 0.59 23.19 40.19 62 0.85 0.22 27.03 77.79 63 1.56 1.01 -34.15 -2.72 64 0.98 0.80 15.87 19.45 65 0.94 0.61 19.00 37.81 66 2.36 0.92 -102.35 7.33 67 0.93 0.96 20.04 2.94 127 Overall, five of the triazoles tested demonstrated no change in % inhibition with increase of ligand concentration from 10 to 100 µM: 56 (33 and 32%), 58 (55 and 50%), 53 (31 and 38%), 55 (29 and 30%) and 64 (15 and 19%). Six compounds demonstrated a concentration dependent increase in % inhibition: 60 (5 to 21%), 62 (27 to 77%), 51 (10 to 47%), 59 (-0.3 to 32%), 61 (23 to 40%) and 65 (19 to 37%). 50 demonstrated a decrease in % inhibition when the concentration of the ligand was increased from 10 to 100 µM with 58 to 38%, as did 67 with 20 to 2%, which could be due to compound interference as observed with GAz above 50 µM. Three compounds demonstrated inhibition at one concentration only: 52 inhibited GlxII activity by 65% at 100 µM, 57 by 10% at 10 µM and 54 by 70% at 100 µM. No inhibition of GlxII activity was observed with 63 at either concentration tested, which is consistent with the data in the assay with DTNB. All the data from this section, including the % triazole formation, DSF binding data, % inhibition and specific activity of GlxII is summarised in Table 24. From this it is clear that the compounds with the best combination of binding affinity and inhibition of GlxII across both enzyme kinetics assays are 56, 55 and 64 (Fig. 78). These 3 compounds show potential to be further developed into inhibitors of the bacterial GlxII using traditional medicinal chemistry approaches. Fig. 78. Hit compounds for identified for further development targeting GlxII. 128 Table 24. Overall summary of the % triazole formation, DSF, % inhibition at 412 nm and from specific activity at 240nm. ǂ Data not included. % formation in kTGS DSF % Inhibition 412 nm Specific Activity U/mg % Inhibition 240 nm Compound MOPS PBS ΔTm 10µM 100µM 10µM 100µM 10µM 100µM GlxII_2%DMSO N/A N/A 0.00 0.00 0.00 1.16 0.99 0.00 0.00 GAz (32) N/A N/A 0.46 12.40 26.18 1.23 2.13 12.06 -52.49 50 0.84 4.05 -0.72 57.15 35.29 0.48 0.61 58.81 38.10 51 0.09 nd -1.08 49.25 36.75 1.04 0.52 10.39 47.33 52 0.10 0.11 -0.44 46.23 50.09 8.33 0.34 ǂ 65.60 53 0.05 1.21 -0.59 -14.09 14.6 0.79 0.60 31.91 38.66 54 nd nd -5.62 47.00 -43.41 2.64 0.29 -126.7 70.22 55 nd nd 1.03 21.62 40.98 0.83 0.68 29.15 30.86 56 0.06 0.03 2.28 52.31 55.04 0.77 0.67 33.73 32.45 57 0.40 0.49 0.15 11.50 43.30 1.04 1.30 10.76 -31.63 58 0.65 1.07 -1.58 40.78 68.76 0.52 0.49 55.77 50.01 59 0.25 0.21 -0.22 44.34 44.27 1.17 0.67 -0.39 32.39 60 3.41 1.54 -3.61 50.66 57.33 1.10 0.78 5.48 21.30 61 nd nd -0.03 52.53 62.22 0.89 0.59 23.19 40.19 62 0.07 0.08 -0.75 50.79 53.79 0.85 0.22 27.03 77.79 63 2.88 0.02 1.08 -29.08 -16.25 1.56 1.01 -34.15 -2.72 64 N/A 1.65 1.55 31.28 30.07 0.98 0.80 15.87 19.45 65 2.29 N/A -5.39 76.98 74.85 0.94 0.61 19 37.81 66 N/A 1.88 1.05 20.52 41.43 2.36 0.92 -102.3 7.33 67 0.65 N/A -3.65 73.99 57.57 0.93 0.96 20.04 2.94 The biophysical data of the synthesised triazoles (Table 24) was not only used to identify compounds to be developed further into novel GlxII antibiotics but also to determine which buffer conditions used in the kTGS experiments provided preferential hits for future application to kTGS experiments. Hit compounds only identified from the kTGS experiment in PBS were AlkG6-2/64 and AlkC1-2/66. These 2 triazoles demonstrated a stabilisation in the DSF screen by a shift in melt temperature of ΔTm 1.55 and 1.05 °C, respectively. 64 and 66 also demonstrated 129 inhibition at 10 and 100 µM. The hit compounds only identified from the MOPS kTGS were AlkF6-2/65 and AlkD2-2/67 which demonstrated a binding to the denatured GlxII by a decrease in melt temperatures in the DSF screen, ΔTm -5.39 and -3.65 °C, respectively. However, both compounds demonstrated inhibition at 10 and 100 µM. These results indicated that even though the hits identified in the MOPS screen inhibited GlxII, they did not provide stabilisation of GlxII in the DSF screen, these observations provided an indication that to identify compounds with a good balance of binding affinity and inhibition, the PBS buffer should be used in future experiments. Analysis of the results in Table 24 revealed that the % triazole formation did not provide a good indication of binding or inhibition. An example is AlkF6-2/65, which was observed to have a large % triazole formation in PBS of 2.29% and ΔTm -5.39 °C. High % triazole formation is therefore an indication of fragment reactivity and should not be used as a criteria for compound synthesis. During the pilot kTGS 6 hits (AlkH2, Alk6, AlkE1, AlkB1, AlkB1, AlkA3, AlkE3) were identified that were not identified during the repeat kTGS experiments with MOPS or PBS, these compounds were however shown to inhibit GlxII but only 2, AlkH2/57 and AlkE3/55, were shown to stabilise the GlxII melt temperature, by a ΔTm of 0.15 and 1.03 °C, to indicate binding. These results raised some questions such as: were these hits misidentified initially; was there a change in the assay which resulted in the hits being identified as false negatives during the repeat kTGS assays; as none of these compounds have been identified as hits to develop further is there a way to remove them from the synthesis list by an addition to the proposed screening cascade prior to synthesis. To address these questions, competition and copper addition assays were designed, which is further discussed in 3.4 and 3.5. Overall, from 97 alkyne fragments and the biased peptide scaffold GAz, hit rates of 14.4 to 19.5% (excluding regioisomers) were observed. Three compounds have been identified as having potential to be developed further, which gives a validated hit rate of 3.1% for the in situ screen. Considering it has been reported68 that 99% of a HTS combinatorial library will result in inactive compounds, these biased kTGS assays have shown an improvement without the need to synthesis all 97 compounds for screening. 130 3.3 Towards the development of inhibitors of GlxI using biased kTGS 3.3.1 Biased kTGS with GlxI As described in 3.2.2 for the biased kTGS with GlxII, the corresponding experiment was performed with GlxI in two buffers to explore if a change in buffer conditions affected the kTGS results. The selected buffers were HEPES, which was selected as the buffer in the DSF experiments for GlxI, and PBS, the preferred kTGS buffer to maintain consistency between kTGS experiments. The two buffers used were: 50 mM PBS, pH 7.4, 0.15 M NaCl and 10 mM HEPES, pH 7.5, 0.15 M NaCl. Hits were identified using the optimised LC-MS/MS analysis methods. Library 1, which contains 45 alkyne fragments, were pooled according to molecular weight of the triazole products into 9 mixes of 5 fragments per well. Each alkyne mix was incubated with GAz (32) in the presence of either GlxI, BSA, buffer or synthetic conditions. From the kTGS experiment performed in HEPES 6 hit compounds were identified: AlkD5, AlkE3, Alk4, AlkA3, AlkC1, AlkG4 (Fig. 80). This corresponds to a 13.33% hit rate (excluding regioisomers). The % triazole formation of the hits identified in the presence of GlxI and the controls is shown in Fig. 79. Fig. 79. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz (32) and alkyne library 1 in HEPES buffer. 0.000.501.00AlkD5 AlkE3 Alk4 AlkA3 AlkC1 AlkG4Protein 0.25 0.06 0.93 0.20 0.93 0.23BSA 0.08 0.01 0.31 0.05 0.12 0.04Buffer 0.06 0.01 0.34 0.05 0.08 0.06% Triazole formation131 Fig. 80. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 1 in HEPES buffer. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. For the corresponding experiment performed in PBS buffer 7 hits were identified: AlkA2, AlkE3, Alk4, AlkA3, AlkC1, AlkG4, AlkD4 (Fig. 82), which corresponds to a 15.56% hit rate (excluding regioisomers). The % triazole formation for each of these hits is summarised in Fig. 81. Fig. 81. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz (32) and alkyne library 1 in PBS buffer. 0.000.200.400.600.80AlkA2 AlkE3 Alk4 AlkA3 AlkC1 AlkG4 AlkD4Protein 0.42 0.06 0.55 0.58 0.74 0.13 0.26BSA 0.21 0.01 0.16 0.06 0.19 0.03 0.08Buffer 0.19 0.02 0.29 0.15 0.20 0.03 0.09% Triazole formation132 Fig. 82. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 1 in PBS buffer. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. Comparing the results described above demonstrated that there is significant overlap between the 5 hits identified in both buffers. In the HEPES buffer one unique hit was identified, AlkD5, and in the PBS buffer two additional hits were identified, AlkD4 and AlkA2. Overlaying the LC-MS/MS traces for the hits and the 1,4-triazole synthetic standard revealed the identified triazoles derived from the following alkynes were suggested to be 1,4-triazoles; Alk4, AlkA3, AlkE3, AlkG4 and AlkD4. The trace observed for the triazole derived from AlkA2 indicated that a mixture of 1,4 and 1,5 regioisomers may have formed in the in situ experiment. The triazoles formed from AlkC1 and AlkD5 were predicted to be 1,5-triazoles due to a slight difference in retention time of the peaks in the traces compared to the synthetic standards. With a total of 8 hits being identified with alkyne library 1, the experiments were repeated with alkyne library 2 which contained 51 alkyne fragments. The fragments were pooled into 11 groups according to the molecular weight of the triazole products, as before, of up to 5 fragments per well. Each alkyne mix was incubated with GAz (32) in the presence of either GlxI, BSA, buffer or synthetic conditions. As with alkyne library 1, two buffers were used and the results compared. From the kTGS experiment in PBS buffer 8 triazole hits were identified which corresponds to a 15.68% hit rate: AlkG4-2, AlkC2-2, AlkA6-2, AlkF5-2, AlkB3-2, AlkC1-133 2, AlkD4-2 and AlkC6-2. AlkF5-2 and AlkC1-2 (Fig. 84). The % triazole formation of the hits identified in the presence of GlxI and the controls is presented in Fig. 83. Fig. 83. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz and alkyne library 2 in PBS buffer. Fig. 84. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 2 in PBS. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. For the corresponding experiment in HEPES buffer 10 hit compounds, which corresponds to a 19.60% hit rate were identified: AlkG4-2, AlkD7-2, AlkC2-2, AlkB3-2, AlkD4-2, AlkC6-2, AlkF1-2, AlkA6-2, AlkG5-2, AlkB4-2 (Fig. 86). The % triazole formation of the hits identified in the presence of GlxI and the controls is presented in Fig. 85. 0.001.002.003.004.00AlkG4-2 AlkC2-2 AlkA6-2 AlkF5-2 AlkB3-2 AlkC1-2 AlkD4-2 AlkC6-2Protein 0.97 2.64 0.17 3.90 3.19 0.94 0.16 1.89BSA 0.10 0.61 0.06 0.00 1.26 0.40 0.00 0.50Buffer 0.16 0.00 0.00 0.00 1.90 0.35 0.00 0.00% Triazole formation134 Fig. 85. Bar graphs showing % triazole formation observed in a biased kTGS experiment with GlxI, GAz (32) and alkyne library 2 in HEPES buffer. Fig. 86. Compounds identified in biased kTGS with GlxI, GAz (32) and Alkyne library 2 in HEPES. MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. Comparing the two experiments once again demonstrated there was significant overlap between the hits identified, with six compounds in common. Two unique hits were identified in the experiment in PBS, AlkF5-2 and AlkC1-2, and four in the HEPES experiment, AlkF1-2, AlkD7-2, AlkG5-2 and AlkB4-2. In total 12 potential triazoles hits were identified. Overlaying the LC-MS/MS traces of the hit compound with the 1,4-triazole synthetic standards indicated that the following alkynes were indicated to form 1,4-triazoles: AlkB3-2, AlkB4-2, AlkC2-2, AlkC6-2, AlkD7-2, AlkC1-2, AlkD7-2, AlkF1-2, 0.0000.5001.0001.5002.0002.5003.000AlkG4-2AlkF1-2AlkD7-2AlkC2-2AlkA6-2AlkG5-2AlkB3-2AlkB4-2AlkD4-2AlkC6-2Protein 0.012 2.84 0.13 0.18 0.17 0.12 0.15 0.32 0.03 0.05BSA 0.007 0.39 0.04 0.01 0.06 0.01 0.00 0.00 0.01 0.02Buffer 0.00 0.48 0.03 0.04 0.08 0.01 0.00 0.00 0.00 0.02% Triazole formation135 AlkG5-2 and AlkF5-2. Two alkynes, AlkA6-2 and AlkD4-2, were indicated to form 1,5-triazoles and AlkG4-2 appeared to result in a mixture of 1,4 and 1,5-regioisomers. In summary, a total of 18 triazoles were identified over the four kTGS experiments. Taking synthetic tractability and cost into consideration hits were selected for synthesis in order to evaluate the identified hit compounds as GlxI inhibitors. Although the LC/MS overlays suggested that a number of the hits could be the 1,5-regioisomer, due to the challenges described in section 3.2.3 synthetic efforts focused on the 1,4-regioisomer. 3.3.2 Synthesis of alkyne building blocks and triazoles The 1,4-triazoles that were shortlisted for synthesis from alkyne library 1 were: AlkD5, AlkE3, Alk4, AlkA3, AlkC1 and AlkG4. Those from library 2 were: AlkD7-2, AlkF1-2, AlkB4-2, AlkF5-2 and AlkB3-2. Even though AlkD5 and AlkC1 were proposed to result in 1,5 regioisomers, the 1,4-triazoles were synthesised due to the challenges encountered with the synthesis of the 1,5-regioisomers described previously. Synthesis of the required alkyne fragments was carried out in one step using alkylation conditions as summarised in Tables 25. The alkyne fragments that were purchased from chemical suppliers are shown in Fig. 87. Table 25. Summary of alkyne synthesis conditions. SM = starting material. SM1 SM2 Conditions, % yield Alkyne fragment a, 30% AlkD5 (68) a, 53% AlkA3 (42) b, 22% AlkD7-2 (69) Reagents and conditions: (a) K2CO3 (2 eq), DMF, RT, 16 h. (b) NaH (1.1 eq), DMF, 0 °C-RT, 16 h. AlkE3 AlkC1 AlkF1-2 AlkF5-2 Alk4 AlkG4 AlkB4-2 AlkB3-2 Fig. 87. Summary of alkynes from chemical suppliers. 136 All target 1,4-triazoles were successfully synthesised in low to moderate yield using previously optimised conditions as summarised in Scheme 11 and Table 26. Scheme 11. Synthesis of the biased 1,4-triazole hits identified with GlxI. Reagents and conditions: (h) GAz (1 eq), alkyne (1 eq), H2O:t-BuOH (2:1), sodium ascorbate (15 mol%), Cu(II)SO4.5H2O (2 mol%), TBTA (2 mol%), RT, 16-20 h. Table 26. Summary of the 1,4-triazole synthesis for the GlxI hits. Alkyne fragment Structure Yield, ID Alkyne fragment Structure Yield, ID AlkA3 (42) 23% 54 AlkC1 28% 74 AlkE3 40% 55 AlkG4 45% 75 AlkF5-2 15% 70 AlkF1-2 16% 76 AlkB3-2 58% 71 AlkB4-2 18% 77 AlkD5 (68) 32% 72 AlkD7-2 (69) 22% 78 137 Alkyne fragment Structure Yield, ID Alkyne fragment Structure Yield, ID Alk4 37% 73 3.3.3 Biophysical evaluation of the GlxI hits With the synthesised 1,4-triazoles in hand evaluation as inhibitors of GlxI was performed using DSF and enzyme kinetics. Firstly, for the DSF analysis the 1,4-triazoles (1 mM) were incubated with GlxI (10 µM) which resulted in a ΔTm in the range of -6.02 – 4.49 °C (Fig. 88, Table 27). The triazole which had the highest ΔTm for GlxI was 74, which had been predicted to be a 1,5-triazole in the LC-MS trace overlays, however the 1,4- triazole appears to be well tolerated, which could be due to the flexible sp2 nature of the tetrahydrothiophene ring interacting with residues in the H-site. 74 was also the only triazole to show an improved ΔTm compared to GAz, 4.49 °C and 3.66 °C respectively. The other 1,4-triazoles that showed an increase in the ΔTm, were 70 (1.48°C), 71 (1.28 °C), 72 (1.20 °C) and 73 (2.43 °C). Of note is 70, which has a carboxylic acid moiety which would be a useful synthetic handle to grow off to develop this compound further. A small change in ΔTm was observed with 76 (0.43 °C) and 55 (0.25 °C), neither of which was significant. 75 showed a small decrease of the GlxI melt temperature by a thermal shift of -0.57 °C. 77 (-6.02 °C), 54 (-5.91 °C) and 78 (-4.76 °C) showed a decrease in the GlxI Tm indicating binding to the denatured GlxI. In summary, only 1 hit (74), appeared to be an improvement over the starting biased scaffold GAz. In addition, as observed with GlxII, the % triazole formation in the kTGS assay was not a prediction of the relative binding affinities for GlxI rather a reflection on the reactivity of the building blocks within that kTGS environment. 138 Fig. 88. Bar graphs of thermal shift (ΔTm, °C) in the presence of the ligand with respect to GlxI alone. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 27. Summary of melt temperature and thermal shift of GlxI in the presence of the biased triazoles. Compound Tm (°C) Δ Tm ± GAz (32) 66.28 3.66 1.74 54 56.71 -5.91 0.71 55 62.87 0.25 0.10 70 64.09 1.48 1.52 71 63.90 1.28 0.26 72 63.82 1.20 1.64 73 65.05 2.43 3.10 74 67.10 4.49 0.16 75 62.05 -0.57 0.21 76 63.01 0.40 0.39 77 56.60 -6.02 0.92 78 57.86 -4.76 0.44 To evaluate the effect of the synthesised 1,4-triazoles on the rate of production of SLG by GlxI an established enzyme kinetics assay153 was employed, in which the increase in SLG concentration was monitored by a corresponding increase in absorbance at 240nm over 20 min. The protocol developed by Thornalley et al153 was applied: where the HTA was pre-formed by incubation of MG (20 mM) and GSH (20 mM) in buffer (PBS pH 6.6) 139 and water at 37 °C, followed by addition of GlxI (0.2 µM), with or without ligand, and the increase in SLG concentration measured over time at an absorbance of 240 nm (Fig. 89). The change in absorbance over time was used to calculate the specific activity (U/mg) of GlxI using Beers law where the change in molar absorbance coefficient was Δε 240 nm = 2.86 mM-1.cm-1, and was subsequently used to calculate the % inhibition in the presence of the synthesised triazoles. Fig. 89. Schematic of the glyoxalase pathway under analysis. To determine the optimal ligand concentration, GAz was co-incubated with GlxI at a range of concentrations (0 – 1000 µM). The data for GAz is shown in Table 28 and Fig. 90, where unexpectedly a decrease in inhibition was observed increasing from 1 to 100 to 200 µM, however 89% inhibition was observed at 1000 µM. The data, as previously observed with the GlxII enzymatic assays, resulted in a wide spread of data points and a lack of statistical significance led to the removal of the data at 10 and 50 µM. However, to be consistent with the GlxII data reported in section 3.2.4, 10 and 100 µM were chosen to evaluate the synthesised triazole hits as well as to evaluate the applicability of this assay. Table 28. Specific activity data for GlxI and % inhibition by a range of GAz concentrations. ǂ Data not included in the data set. Compound Specific Activity % Inhibition from specific activity GlxI 12.87 0.00 GlxI_2%DMSO 13.63 0.00 GAz 1uM 10.22 25.00 GAz 5uM 13.93 -2.18 GAz 10uM ǂ ǂ GAz 50uM ǂ ǂ GAz 100uM 11.53 15.38 GAz 200uM 12.32 9.65 GAz 500uM 16.72 -22.66 GAz 1000uM 1.46 89.26 140 Fig. 90. % Inhibition of GlxI by GAz (32) over 0 – 1000 µM. The 1,4-triazoles identified as hits in the kTGS were evaluated as inhibitors of GlxI at 10 and 100 µM using the conditions described above. The data is summarised in Fig. 91 and Table 29. 71 demonstrated a concentration dependent % inhibition, increasing from 41 to 79% at 10 and 100 µM respectively. Three compounds demonstrated inhibition at 10 µM, 72 (44%), 73 (87%), 75 (88%). Two compounds demonstrated inhibition at 100 µM, 75 (60%) and 78 (50%). However, as observed with GAz, large errors were associated with the data indicating they were not significant. This assay was challenging for a number of reasons: it was not possible to ensure the MG and GSH had pre-formed HTA prior to the addition of GlxI and the ligand; the OD vs time curves appeared as expected however upon analysis and normalisation a few outliers were observed. These outliers were identified as producing no SLG as no change in absorbance was detected at 240nm (activity = 0) or a decreased SLG (activity = negative value) and were therefore removed from the data set as they were not useful for ligand evaluation. The rationale for these observations, was suggestive of a lack of SLG production, with this assay could include the following: 1) MG and GSH not forming HTA within the preincubation time of the assay due to the reversible nature of this reaction; 2) interaction of nucleophilic ligands with the electrophilic MG thereby preventing HTA formation; 3) interference of the ligands with HTA either through side reactions or stabilisation, thereby inhibiting the formation of SLG, suggesting the ligands were very good inhibitors or interfering with the assay; 4) chelation or binding of the ligands to the Ni2+ metal thereby inactivating GlxI, suggesting the ligands were once again either very good inhibitors or interfering with the assay; 5) a moiety on the ligands causing an optical interference in absorbance at 240nm; 6) the ligands possibly having fast on rates and slow off rates within the time frame of the assay once again indicating either very good inhibitors or interfering with the assay. 141 Future work is therefore suggested to optimise this assay and could include: evaluating the relationship between the pKa of the ligands and the pH of the assay; evaluating the ligands at a lower concentration or synthesising isosteres of the triazoles to remove nucleophilic moieties; investigating the UV/Vis absorbance of the ligands alone. Exploring the premise that the triazoles were interfering in the assay the structures were assessed for nucleophilicity and metal chelation potential. 76 and 77 had primary amines which could initiate nucleophilic reactions as well as acting as metal chelators to inactivate GlxI. The secondary amine in 74 could potentially cause interference in the assay. The pyrrole in 78 and the N,S in 55 could potentially chelate the Ni2+ metal.156 The isatin in 54 could undergo a reaction with a primary amine to form a Schiff base which in turn can chelate to a Ni2+ metal, such compounds are known to have antimicrobial activity i.e. the penicillin class of antibacterials. These compounds could either be blocking the active site, thereby preventing the activity of GlxI which is metal dependent, as very good inhibitors or as interferences to the assay. Therefore, future work should involve further optimisation of the enzyme kinetics assay and evaluation of these compounds using an alternative assay to evaluate inhibition, such as fluorescence polarisation (FP), which would require the development of an FP probe, or SPR. Several compounds (71, 72, 73, 75 and 78) did show inhibition of the enzyme activity of GlxI and a number appeared to activate the enzyme (70, 76 and 77) as shown in Fig. 91 and Table 29. These compounds were therefore evaluated however, due to the error bars and removal of outlier these results were used for compound ranking only and an alternative assay should be investigated to evaluate GlxI in the future. Fig. 91. % Inhibition bar graphs of triazoles at 10 and 100 µM with GlxI. 142 Table 29. Summary of specific activity and % inhibition of GlxI. ǂ Data not included as specific activity <0 and inhibition calculated to be > 100%. Specific Activity 240 nm % Inhibition Structure 10 µM 100 µM 10 µM 100 µM GlxI 2% DMSO 3.70 3.70 0.00 0.00 54 ǂ ǂ ǂ ǂ 55 ǂ ǂ ǂ ǂ 70 7.67 ǂ -107.50 ǂ 71 2.18 0.77 41.13 79.26 72 2.05 3.41 44.49 7.74 73 0.45 ǂ 87.69 ǂ 74 ǂ 4.86 ǂ -31.54 143 Specific Activity 240 nm % Inhibition Structure 10 µM 100 µM 10 µM 100 µM 75 0.42 1.46 88.73 60.62 76 14.44 8.13 -290.82 -120.03 77 13.23 6.84 -258.14 -85.23 78 ǂ 1.84 ǂ 50.12 To get a better understanding of the effect triazoles had on GlxI all the data including in situ % triazole formation, DSF and specific activity data is summarised in Table 30. Unlike with GlxII, no clear differences could be observed between using HEPES or PBS buffer therefore for future KTGS experiments PBS should be used for consistency. This data indicates that 71, 72 and 73 have a good balance of binding, through the GlxI ΔTm, and indicated inhibition of the specific GlxI activity (Fig. 92). However, as the DSF data was very promising for 74 it should also be evaluated further using another biophysical technique, such as ITC, SPR or FP. Of note, is the size of the groups on the triazoles, which are relatively small and flat compared to the hits identified for GlxII, this is due to the smaller size of the hydrophobic binding pocket of GlxI relative to the more open hydrophobic binding site found in GlxII. Overall, from 97 alkynes hit rates of 15.4 – 17%, excluding regioisomers, were identified in the kTGS screens with GlxI and 4 compounds have been identified to evaluate further as inhibitors of GlxI. 144 Fig. 92. Hit compounds for further development targeting GlxI. Table 30. Summary of % triazole formation in kTGS, DSF, specific activity and % inhibition of GlxI. ǂ Data not shown. % triazole formation DSF data Specific Activity of GlxI (U/mg) % Inhibition Compound HEPES PBS ΔTm (°C) 10 µM 100 µM 10 µM 100 µM GlxI 2% DMSO 0.00 3.70 3.70 0.00 0.00 GAz (32) 3.66 ǂ 11.53 ǂ 15.38 54 0.20 0.58 -5.91 ǂ ǂ ǂ ǂ 55 0.06 0.06 0.25 ǂ ǂ ǂ ǂ 70 3.90 - 1.48 7.67 ǂ -107.50 ǂ 71 3.19 0.15 1.28 2.18 0.77 41.13 79.26 72 0.25 - 1.20 2.05 3.41 44.49 7.74 73 0.93 0.55 2.43 0.45 ǂ 87.69 ǂ 74 0.93 0.74 4.49 ǂ 4.86 ǂ -31.54 75 0.23 0.13 -0.57 0.42 1.46 88.73 60.62 76 - 2.84 0.40 14.44 8.13 -290.82 -120.03 77 - 0.32 -6.02 13.23 6.84 -258.14 -85.23 78 - 0.13 -4.76 ǂ 1.84 ǂ 50.12 To obtain additional insight into how the synthesised triazoles were predicted to bind in GlxI, docking studies were carried out. As an X-ray crystal structure of GlxI from E. coli 145 with a bound ligand was unavailable the crystal structure of GlxI from humans (PDB: 1FRO) with S-benzyl glutathione (79) was overlaid with the apo crystal structure from E. coli (PDB: 1F9Z) and the residues important for binding were identified (Fig. 93 and 94A). Conserved amino acid residues within the active site were observed. This was carried out in collaboration with Dr Geoff Wells (UCL, School of Pharmacy). Fig. 93. Overlay of GlxI from E. coli 1F9Z (purple) and from human 1FR0 (grey). As discussed in greater detail in section 1.3.1, the important residues around the Ni2+ metal in the active site were His133, Glu184, His74 and Glu122 with 2 water molecules (Fig. 94B). The ligand S-benzyl-glutathione (79) was held in the glutathione binding site by intermolecular hydrogen bonds to Asn103, Arg37 and 2 water molecules (Fig. 94C and D). 146 A. B. C. D. Fig. 94. (A) E. coli GlxI 1F9Z (purple) and human GlxI 1FR0 (grey) overlaid with a bound benzyl-glutathione conjugate (79) (cyan blue) and (B) the active site of E. coli 1F9Z. (C) H-bonds in glutathione binding site of E. coli 1F9Z. (D) Bound benzyl-glutathione ligand (79) in E. coli 1F9Z. Visualised in Chimera. As with GlxII, AutoClickChem149 was used to generate an in silico library of all possible triazoles, from GAz and the alkyne fragments. Docking of the triazoles was performed using LeDock and Vina.150 On closer analysis of the docking, it was observed that Vina gave a better prediction of docking for the triazoles compared to LeDock with respect to the position of the triazoles compared to the S-benzyl glutathione. The Vina docking scores ranged from -8.9 to 5.5 kcal/mol, where the lower the docking score the better the fit of the docked structure. The hits identified in the kTGS assay are highlighted in red and are scattered throughout the docking score graph (Fig. 95). 147 Fig. 95. Vina docking scores for 1,4 and 1,5 triazoles in GlxI (PDB: 1F9Z). Looking first at 74 which had the highest binding affinity for GlxI with a ΔTm of 4.49 °C and a docking score of -7.3 kcal/mol. The Glu amino acid in the tripeptide backbone was predicated to form a hydrogen bonding interaction with Glu122 and Glu184 which surrounds the Ni2+ metal in the active site, instead of the residues Asn103 and Arg37 in the G-site as anticipated (see discussions in section 1.3.1), and the functional group on the triazole was predicted to have a hydrogen bond interaction with Lys104 on the loop surrounding the binding site (Fig. 96). The predicted orientation of the tripeptide backbone correlates with the anticipated binding through the Glu-amine as the dipeptide fragments demonstrated that the Glu residue was important than Gly for binding to GlxI (section 2.3.4). The slight twisting of the tripeptide backbone within the binding site results in interactions between the Glu-amine and the metal active site, instead of Asn103 and Arg37 in the G-site, could be due to either the size of the H-site and the functionalised triazole not fitting or the modelling requiring optimisation so the tripeptide backbone binds as expected in the G-site. Fig. 96. Predicted docking of 74 in GlxI (19FZ). Visualised in Chimera. -9.5-9-8.5-8-7.5-7-6.5-6-5.5-5Docking score (kcal/mol)Score kTGS hitsG-site H-site 148 70, which had a docking score of -7.4 kcal/mol and ΔTm of 1.48 °C, was also predicted to bind in the same orientation with the tripeptide backbone forming hydrogen bonding interactions to Glu122 and Glu184 and with the carboxylic moiety on the functional group on the triazole forming a hydrogen bond interaction with Lys104. 54, which had a docking score of -7.6 kcal/mol and ΔTm of -5.91 °C, was also predicted to bind in the same orientation with the tripeptide backbone forming hydrogen bond interactions to Glu122 and Glu184 and with the ketone on the isatin as the functional group on the triazole forming a hydrogen bond interaction with Lys104. 71 (docking score -7.6 kcal/mol and ΔTm of 1.28 °C), 72 (docking score -7.3 kcal/mol and ΔTm of 1.20 °C), 73 (docking score -7.0 kcal/mol and ΔTm of 2.43 °C), 75 (docking score -7.5 kcal/mol and ΔTm of -0.57 °C), 76 (docking score -6.7 kcal/mol and ΔTm of 0.40 °C) and 55 (docking score -8.0 kcal/mol and ΔTm of 0.25 °C) were also predicted to bind with the tripeptide backbone forming hydrogen bond interactions Glu122 and Glu184, however the functional groups at the 4 position on the triazole appeared to fill the left side of the binding pocket, as shown in Fig. 97A for 71. 77 had a docking score of -6.1 kcal/mol and ΔTm of -6.02 °C, as shown in Fig. 97B, however was predicted to have the tripeptide backbone twisted out of the G-site and the ethoxy-amine functional group on the triazole hydrogen bonding to the Glu residues surrounding the active site. A B Fig. 97. Predicted docking of A) 71 and B) 77 in GlxI (19FZ). Visualised in Chimera. These docking studies provide an insight as to the binding mode of the triazoles however the docking orientations and predicted interactions did not correlate with the binding data measured in the DSF studies, therefore as with GlxII, either this model needs refining to allow a better prediction of binding alignment or it highlights the ability of kTGS to identify serendipitous hits. G-site H-site H-site G-site 149 3.4 Competition kTGS As discussed in section 3.2 during the pilot kTGS experiment with GlxII, 6 hits (AlkH2, Alk6, AlkE1, AlkB1, AlkB1, AlkA3, AlkE3) were identified that were not identified during the repeat kTGS with both MOPS and PBS buffer. The compounds did inhibit GlxII in the enzyme kinetics assay but two compounds, 57 (derived from AlkH2) and 55 (derived from AlkE3), were shown to stabilise the GlxII melt temperature, by a ΔTm of 0.15 and 1.03 °C respectively, in the DSF assay indicating binding. These results raised some questions around the misidentification of hits i.e. false negatives. Also, considering that none of these compounds have been identified as hits to develop further a question remains over whether there is a way to remove them from the synthesis list by addition of a secondary validation step to the proposed screening cascade prior to synthesis. This would allow the kTGS hit identification process to be streamlined and the time to synthesis and biophysical evaluation decreased. A competition assay was designed to answer two questions regarding whether a change in chemical environment could affect the hit identification: 1) did a change in chemical environment of the fragments reduce the chance of false negatives and 2) could the best hits be identified so as to reduce the number of compounds requiring synthesis and biophysical evaluation. One pitfall for consideration when carrying out kTGS experiments is the specific grouping of building blocks and one should be mindful of the potential for the reagents to be in competition with themselves to give false negatives.74 In order to probe this with the GlxI and II hits, a competition experiment was carried out where all the alkynes that were identified in the biased kTGS were incubated with GAz (32) and GlxI/II together in one pot to see if there was a difference in hit identification due to a different chemical environment. The 5 groups analysed are shown in Table 31. Two alkyne fragments, which were not anticipated to be hits, were included in GlxI-Alk2 and GlxII-Alk2 to explore the possibility of false positives. The alkyne fragments from GlxII alkyne library 1 were split into 2 groups due to the number of hits identified in the pilot experiment. The LC-MS/MS traces were analysed and % triazole formation observed in the protein, BSA and PBS buffer were plotted in bar graphs as previously described. 150 Table 31. Groups of alkynes for competition kTGS with GAz in GlxI or GlxII. Group number Protein Alkyne fragments GlxI-Alk1 GlxI Alk4 AlkD5 AlkE3 AlkA3 AlkC1 AlkG4 GlxI-Alk2 GlxI AlkF1-2 AlkD7-2 AlkB4-2 AlkD2-2 AlkF6-2 GlxII-Alk1-1 GlxII Alk6 AlkD3 AlkB3 AlkB1 AlkE1 AlkH4 GlxII-Alk1-2 GlxII AlkE5 AlkE3 AlkB2 AlkH2 AlkC2 AlkF3 GlxII-Alk2 GlxII AlkB3-2 AlkG6-2 AlkH6-2 AlkC1-2 AlkF5-2 Within the group GlxI-Alk1, all the triazoles were detected in the LC/MS in the presence of the protein. However, applying the criteria set out in section 3.2.1 that for a compound to be considered a hit the % triazole formation must be at least twice in the protein compared to controls, only AlkC1 is considered a hit (Fig. 98) Of note, is the triazole formed with alkyne fragment AlkC1 is 74, which had the highest binding affinity to GlxI (as determined by DSF) and was considered a promising compound to investigate further. The greatest % triazole formation in this experiment was observed with 72 (derived from AlkD5), which also demonstrated good binding affinity and inhibition of GlxI. This experiment demonstrated that: 1) a change in chemical environment affected hit identification and resulted in false negatives suggesting that the fragment grouping of alkynes in the kTGS in section 3.3 was required to identify all the hit compounds; 2) the best hits could be identified through a competition assay thereby refining the synthesis list. Fig. 98. Bar graphs of % triazole formation in the competition assay group GlxI-Alk1. Within the group GlxI-Alk2, again all the triazoles were detected in the LC/MS in the presence of the protein. However, only AlkD7-2 was considered a hit over the controls (Fig. 99). The different chemical environment did appear to affect the hit identification in this case. Fragment AlkD7-2 which formed triazole 78 had a negative thermal shift in the 0.000.501.00Alk4 AlkD5 AlkE3 AlkA3 AlkC1 AlkG4GlxI 0.58 0.95 0.27 0.23 0.45 0.12BSA 0.34 0.48 0.14 0.12 0.18 0.07Buffer 0.28 0.27 0.09 0.07 0.12 0.04% Triazole formation151 DSF studies. Fragment AlkF1-2 which formed triazole 76 demonstrated binding affinity to GlxI in the DSF studies suggesting the fragment grouping of alkynes in the kTGS in section 3.3 was required to identify all the hits and reduce false negatives. AlkD2-2 and AlkF6-2 have formed the corresponding triazoles but were not classed as hits in this competition assay, which was expected as they were hits observed with GlxII and were used as controls in this assay. This observation suggests that if an alkyne is not a hit compound irrespective of what chemical environment it is grouped in, it won’t produce a hit, i.e. a false positive, due to the lack of target guided effect. Fig. 99. Bar graphs of % triazole formation in the competition assay group GlxI-Alk2. Within the group GlxII-Alk1-1, all the triazoles were detected in the LC/MS in the presence of the protein although only AlkE1 was considered a hit over the controls (Fig. 100). Triazole 60, which was derived from AlkE1, appeared to act as an inhibitor but binding was not detected using DSF. The different chemical environment did appear to affect the hits identified in the in situ experiment. Interestingly, all the triazoles formed from these alkyne fragments demonstrated negative ΔTm of GlxII in the DSF screen although inhibition was observed. This experiment demonstrates that a competition kTGS could refine the synthesis list, due to the lack of binding data for these triazole examples. 0.000.20AlkF1-2 AlkD7-2 AlkB4-2 AlkD2-2 AlkF6-2GlxI 0.09 0.17 0.17 0.15 0.05BSA 0.09 0.05 0.12 0.13 0.03Buffer 0.07 0.05 0.10 0.11 0.03% Triazole formation152 Fig. 100. Bar graphs of % triazole formation in the competition assay group GlxII-Alk1-1. Within the group GlxII-Alk1-2, all the triazoles were detected in the LC-MS in the presence of the protein although only AlkH2 and AlkF3 were considered hits over the controls (Fig. 101). In this competition kTGS AlkF3 appears to be a 1.5-triazole where in the kTGS in section 3.2 AlkF3 was indicated to be a 1,4-triazole. The different chemical environment did appear to affect the hit identification in this case leading to a refinement of the synthesis list. Fragment AlkF3 formed triazole 56, which was the compound with the highest binding affinity and % inhibition of GlxII. AlkH2 produced 57 which exhibited binding affinity and % inhibition of GlxII. Therefore, this change in chemical environment allowed confirmation of the best hit compounds through competition. However, 55 (derived from AlkE3) demonstrated good binding affinity and % inhibition of GlxII but on the change of chemical environment this compound was no longer considered a hit, suggesting that the original grouping of alkynes in section 3.2 was required to identify all the hit compounds, i.e. reduce false negatives. The remaining compounds appeared to act as inhibitors but binding was not detected using DSF, suggesting that the change in chemical environment could have possibly reduced the number of compounds requiring synthesis. This data suggests the chemical environment did affect the % triazole formation and hit identification. 0.000.501.00Alk6 AlkD3 AlkB3 AlkB1 AlkE1 AlkH4GlxII 0.84 0.03 0.07 0.07 0.43 0.03BSA 0.70 0.02 0.04 0.10 0.05 0.04Buffer 0.82 0.01 0.05 0.15 0.05 0.04% Triazole formation153 Fig. 101. Bar graphs of % triazole formation in the competition assay group GlxII-Alk1-2. Within the group GlxII-Alk2, all the triazoles were detected in the LC-MS in the presence of the protein although only AlkH6-2, possibly as the 1,5-triazole due to a shift in retention time of the peak compared to the synthetic standard, was considered a hit over the controls (Fig. 102). The different chemical environment again did appear to affect the hits identified. AlkB3-2 and AlkF5-2 were not expected to be hits as they were hits identified with GlxI and were used as negative controls in this competition kTGS to confirm that the change in chemical environment did not encourage false positives. 64 (derived from AlkG6-2), 66 (derived from AlkC1-2) and 63 (derived from AlkH6-2) all showed binding in the DSF assay as well as % inhibition of GlxII suggesting that the original grouping of alkynes in the kTGS in 3.2 is required to identify all the hit compounds, i.e. reduce false negatives. Fig. 102. Bar graphs of % triazole formation in the competition assay of group GlxII-Alk2. In summary, the 5 competition assays that were carried out demonstrated that a change in the chemical environment does affect the % triazole formation and the hits identified. The % triazole formation did change in both the protein and control wells compared to 0.000.100.200.300.40AlkE5 AlkE3 AlkB2 AlkH2 AlkC2 AlkF3GlxII 0.18 0.08 0.16 0.08 0.06 0.33BSA 0.09 0.06 0.13 0.03 0.06 0.08Buffer 0.24 0.08 0.15 0.04 0.07 0.08% Triazole formation0.001.002.00AlkB3-2 AlkG6-2 AlkH6-21.5AlkH6-2 AlkC1-2 AlkF5-2GlxII 0.09 0.11 1.23 0.65 0.18 0.02BSA 0.06 0.07 0.40 0.21 0.01Buffer 0.06 0.06 0.27 0.20 0.01% Triazole formation154 the kTGS experiments in 3.2 and 3.3, however as was demonstrated the amount of triazole formed in situ does not appear to be an indication of the compounds ability to be an inhibitor so the change in triazole formation was not compared in these experiments only whether compound was a hit or not. In some cases, the most promising hits, 74 (AlkC1) and 56 (AlkF3) were identified over the other potential compounds but in other cases compounds that had a good combination of binding affinity and inhibition would have been missed, 76 (AlkF1-2), 55 (AlkE3), 64 (AlkG6-2), 63 (AlkH6-2) and 66 (AlkC1-2). Encouragingly, these competition kTGS suggested that a change in chemical environment did not lead to identification of false positives. Given this information it seems appropriate to use this competition assay between all identified hits more as a ‘hit-ranking’ assay and should be compared and confirmed with any biophysical data. One way to overcome the issue of false negatives in the future is to carry out orthogonal kTGS where each building block is represented in 2 wells in 2 different chemical environments. This will allow each of the building blocks the chance to interact with the binding site as it is incubated with the enzyme in 2 different competitive environments which could reduce the number of false negatives and still maintain the efficiency of synthesis and fragment screening in a one pot procedure.74 3.5 Copper addition to the kTGS assay As it was observed that a different chemical environment could affect the outcome of identified hits and that the hits were usually observed in trace quantities. It was hypothesised that the addition of copper could improve the kinetics of the building blocks within the enzyme and therefore ease of identification, due to increase % triazole formation, but the question remained would the target guided effect of hit formation relative to controls be maintained. There is some precedence in the literature for the addition of copper to kTGS experiments, for example, Balasubramanian et al.157 carried out an in situ click chemistry experiment with a biased scaffold in the presence of excess copper (10 mM) with a DNA G-quadruplex target and observed that the copper improved the kinetics of the reaction but importantly the hit identification was still target guided i.e. controlled by the DNA-template. To determine how much copper to add to the kTGS experiments a few factors were taken into consideration. During the synthesis of the triazoles 2 mol% of copper is used which corresponds to 0.2 mM and as there are 5 potential triazoles per well, the concentration of copper would increase to 1.0 mM. Balasubramanian et al.157 used excess copper. 155 Therefore, two kTGS experiments with both GlxI and GlxII were planned with a range of Cu(II)SO4.5H2O concentrations; 0, 0.2, 0.5, 1, 2 and 4 mM to select the optimal copper concentration for any future kTGS experiments. The ultimate aim of these experiments was to determine whether the addition of copper could aid in the confirmation of hits identified in the kTGS experiments prior to compound synthesis. Two kTGS experiments were carried out and each alkyne was individually investigated to determine the effect of the copper addition. In the first experiment five alkynes; AlkE3, AlkF3, AlkC1, AlkB4 and Alk5 were grouped together. AlkE3 and AlkF3 were hits from the kTGS with GlxII and AlkE3 and AlkC1 were hits with GlxI. AlkB4 and Akk5 had not previously been identified as hits. In the second experiment, AlkD2-2, AlkC1-2, AlkB1-2, AlkA7-2 and AlkA5-2 were grouped together. AlkD2-2 and AlkC1-2 were hits from a kTGS screen with GlxII. AlkB1-2, AlkA7-2, AlkA5-2 were not previously identified as hits. The two sets of alkynes were incubated with GAz (32) in the presence of GlxI, GlxII or BSA and buffer (50 mM PBS, pH 7.4, 0.15 M NaCl) and increasing concentrations of copper (under the conditions previously described). A few trends were observed for all the alkyne fragments: GlxII was more affected by the addition of copper that GlxI and increased % triazole formation was observed; the addition of ≤1 mM appeared to increase the % triazole formation without affecting the results of hit identification compared to no copper; % triazole formation in the controls generally increased above 1 mM. An example demonstrating the effect of copper addition is shown in Fig. 103. The triazole derived from AlkE3 was a hit for both GlxI and GlxII. This triazole (55) demonstrated both binding and inhibition to GlxII and binding to GlxI. This alkyne again identified as a hit under the standard kTGS conditions, i.e. 0 mM copper. Upon the addition of copper the % triazole formation increased in all cases. The protein conditions were affected the greatest on addition of Cu and in all cases the % triazole in the presence of protein was at least double to the controls resulting in the compound being a hit. With 0.5 mM being the optimal copper concentration for GlxII and 2 mM for GlxI. 156 Fig. 103. Bar graphs of % triazole formation between AlkE3 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. A second examples is AlkA7-2, which was not a hit for GlxI or GlxII during the initial kTGS studies. Under standard kTGS conditions, i.e. 0 mM copper, this alkyne was identified as a hit for both GlxII and GlxI, this result could be explained by the change in chemical environment within this group of alkynes, as discussed in section 3.4, (Fig. 104). Upon addition of copper the % triazole formation increased under protein conditions at 0.2, 0.5, 1 and 2 mM with no effect on the controls. At 4 mM, the % triazole formation increased under all conditions, with 1 mM being the optimal copper concentration for GlxII and 2 mM for GlxI for hit confirmation due to increase in the signal in the LC-MS trace. Fig. 104. Bar graphs of % 1,4-triazole formation between AlkA7-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Of interest, is that this triazole, derived from AlkA7-2, was indicated to be a mixture of a 1,4 and 1,5-regioisomers through the LC-MS trace overlays with the synthetic samples for GlxII (Fig. 105). Encouragingly, the addition of copper did not affect or decrease the % 1,5-triazole formation in this example. 0.002.004.006.008.0010.00AlkE3 AlkE3 AlkE3 AlkE3 AlkE3 AlkE3GlxI 0.12 0.19 0.58 2.05 4.95 4.07GlxII 0.12 0.14 9.41 5.27 3.86 8.64BSA 0.06 0.07 0.22 1.08 1.06 2.23Buffer 0.05 0.05 0.11 0.27 0.96 1.60% Triazole formationGlxIGlxIIBSABuffer0.000.100.200.300.400.500.600.70AlkA7-2 AlkA7-2 AlkA7-2 AlkA7-2 AlkA7-2 AlkA7-2GlxI 0.05 0.02 0.08 0.12 0.18 0.13GlxII 0.07 0.03 0.17 0.68 0.18 0.15BSA 0.02 0.01 0.01 0.02 0.02 0.06Buffer 0.02 0.01 0.01 0.01 0.01 0.04% Triazole formation157 Fig. 105. Bar graphs of % 1,5-triazole formation between AlkA7-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. The addition on copper appears to affect different alkynes in different ways. All the individual bar graphs are shown in the Appendix. The % triazole formation is affected more under protein conditions (increases of 0.02 to 72 times) than controls (increases of 0.05 to 10 times). It appears that if a compound is a hit observed in previous kTGS assays then any amount of copper addition does not affect the positive hit outcome, but rather increases the % triazole formation which makes for greater ease of identification through a bigger peak in the LC-MS trace. For alkynes that previously were not hits the addition of copper has varying effects, ranging from greater % triazole formation in protein compared to controls resulting in positive hit identification, to overall increase in % triazole formation under all conditions resulting in no hit identification. These positive hit identifications upon copper addition should be assessed carefully so that false positives are not selected. To summarise, in general the addition of 0.2 mM copper resulted in a decrease or no change in % triazole formation in 9 of the examples, only the triazole derived from AlkB4 had a 3 fold increase in GlxI and 4 fold in GlxII as well as a 5 fold increase in the BSA and buffer controls suggesting that the % triazole formation with the addition of 0.2 mM copper was dependent on the intrinsic reactivity of the fragments and did not ease the identification of target guided hits. Addition of 0.5 mM appears to increase the % triazole formation without effecting the overall result of hit identification compared to the 0 mM copper conditions. Addition of 1 mM copper affected the outcome in a number of ways, from increased % triazole formation without effecting target guided result, to increasing % triazole formation under all conditions with a greater effect on the controls resulting in no hit identification. Addition of 2 and 4 mM copper appears to increase % triazole formation under all conditions. GlxIGlxIIBSABuffer0.000.200.400.600.80AlkA7-21.5AlkA7-21.5AlkA7-21.5AlkA7-21.5AlkA7-21.5AlkA7-21.5GlxI 0.02 0.00 0.01 0.02 0.03 0.05GlxII 0.58 0.41 0.72 0.60 0.73 0.66BSA 0.01 0.02 0.01 0.02 0.01 0.03Buffer 0.02 0.01 0.01 0.01 0.00 0.04% Triazole formation158 In conclusion, 0.5 mM copper addition improves the % triazole formation under protein template conditions but not in the controls allowing for greater ease in identification of triazoles synthesised in the LC-MS traces. Therefore, the inclusion of a secondary copper kTGS assay could enable confirmation of hits identified and be used as an additional screening experiment to confirm hits identified under standard kTGS conditions, i.e. 0 mM copper, prior to individual triazole synthesis. 3.6 Summary and future work suggestions Using a biased kTGS approach to identify hit compounds targeting GlxI and GlxII hit rates of 17% and 19% respectively were obtained. Biophysical evaluation using DSF and enzyme kinetics further validated three compounds targeting GlxII and four targeting GlxI for optimisation using traditional medicinal chemistry approaches, into novel antibacterials. The three GlxII inhibitors were: 56 (ΔTm 2.28 °C, 32 – 55% inhibition), 55 (ΔTm 1.03 °C, 21 – 40% inhibition) and 64 (ΔTm 1.55 °C, 15 – 31% inhibition) as shown in Fig. 106. Fig. 106. Hit compounds for further development targeting GlxII identified in the biased kTGS. The four GlxI inhibitors were: 71 (ΔTm 1.28 °C, 41 – 79% inhibition), 72 (ΔTm 1.20 °C, 44% inhibition), 73 (ΔTm 2.43 °C, 87% inhibition) and 74 (ΔTm 4.49 °C) as shown in Fig. 107. 159 Fig. 107. Hit compounds for further development targeting GlxI identified in the biased kTGS. The biased kTGS together with the design and synthesis of GAz (32) enabled the exploration of the GlxI and GlxII binding sites. These hit compounds allowed the optimisation of the DSF and enzyme kinetics biophysical techniques. The range of % inhibition observed in the enzyme kinetics assays were of concern and a number of attempts to reduce the data spread in the GlxII assay were attempted, including end-point measurements and an alternative assay, however the standard deviation did not improve. Whether this is due to the instruments sensitivity or optical interference by the compounds, future work should include attempts to improve this assay. The lack of statistical significance, size of error bars and outliers with GlxI were also of concern and future work should include optimisation of this assay in order for it to be used as a reliable readout for evaluation of synthesised compounds. Additional methods to evaluate the inhibition of GlxI would be beneficial to confirm the observed data. A complimentary screening method using 19F-NMR to demonstrate ligand binding was evaluated and progress towards this goal is described in Chapter 5. This study has demonstrated the usefulness of a kTGS approach through the identification of hits, targeting GlxI and II, and the exploration of the binding site using a diverse alkyne fragment library. These biased kTGS experiments have allowed the opportunity for kTGS optimisation so as to develop this in situ tool to be used more efficiently in the future and with application to other biological targets. A number of valuable insights have been identified and will be applied to future kTGS experiments including; 1) the competitive environment does affect the hits identified, going forward to 160 minimise false negatives fragments should be represented in a number of competitive environments or orthogonal kTGS to allow all building blocks the chance to interact with the enzyme;74 2) choice of buffer can affect the hit identification, but the additional orthogonal kTGS should remove the chance of false negatives, therefore PBS buffer should be used for GlxI and GlxII to allow for continuity within the assay; 3) competition kTGS experiments can be used to identify the ‘best’ inhibitors, such as AlkC1 / 74 with GlxI and AlkF3 / 56 with GlxII, to refine the synthesis list; 4) addition of 0.5 mM copper allows for more robust confirmation due to a higher % triazole formation in the protein wells and correspondingly greater ease of identification in the LC-MS/MS. Therefore, for future biased kTGS experiments the amended screening cascade outlined in Fig. 108, is proposed. Stage 1, includes two orthogonal kTGS experiments so that each building block is represented in a different chemical environment, thus minimising false negatives. Stage 2, includes a competition kTGS to allow compound ranking and addition of 0.5 mM copper to allow for hit confirmation. Stage 3, involves the synthesis of the hit compounds. Stage 4, biophysical evaluation of all hits synthesised. 161 Fig. 108. Proposed screening cascade for future biased kTGS hit identification experiments. To develop kTGS further as an efficient drug discovery approach, new ligation reactions are required to increase the methods utility. This would also allow more access to chemical libraries and the additional probing of chemical space within a receptors binding site. Oueis et al.77,78 have started to evaluate reductive aminations between amines and aldehydes, amide bond formations between carboxylic acids and amides and alkylations between amines and alkyl bromides. For use in the kTGS, in order to form C-S bonds, S-alkylation reactions between thiol compounds and alkyl halides, nucleophilic substitutions using thiol compounds to open epoxide rings and between thiol containing St e i se T S o tho on ss ys St e Com etition n co e T S it n in n confi m tion St e 3 Synthesis of hits St e io hysic e tion it i entific tion 162 compound and vinyl containing compounds to form thio-ethers are most common. Kolb80, Sharpless80 and Finn81 have explored other click reactions, such as diels-alder, ring openings, oxime ether formations and oxidation reactions thereby increasing the methodology to a wide range of drug discovery opportunities such as lead discovery as well as the tagging and labelling of biological targets. Other options to consider is the inclusion of a 1,5-synthesis plate in addition to the 1,4-synthesis plate to allow for the confirmation of the triazole regioisomers. In summary, the successful identification of peptide inhibitors of GlxI and GlxII using a biased kTGS has been achieved and allowed the optimisation of kTGS conditions, synthesis conditions, DSF and enzyme kinetics assays. 163 Chapter 4: Unbiased kinetic target guided synthesis 4.1 Unbiased approach The unbiased kTGS approach has the potential to significantly impact drug discovery through the identification of novel lead compounds from diverse fragment libraries against biological targets where no inhibitors are known. Unlike the biased kTGS approach, where a known binding moiety was used as a handle in the binding site and kTGS was used to explore the remainder of the pocket and identify lead compounds, the unbiased approach was used to expand the methodology to establish if kTGS can be used in an unbiased fragment based approach to identify small, novel and non-peptidic inhibitors. This unbiased kTGS approach could also be used to build upon the peptidic inhibitors identified in the biased approach, as described in Chapter 3, and to explore the limits of kTGS as a fragment screening and in situ fragment linking tool. The hypothesis was to utilise two fragments that could link in situ to identify novel lead compounds. Potential limits of this approach include fragments typically having lower affinity due to the relatively small fragment size resulting in fewer interactions within the binding site,99,95 as well as the requirement for the both fragments to interact with the receptor and be in close enough proximity to each other within the enzyme to enable the fragments to link together. To investigate the feasibility of this approach azide and alkyne fragment libraries were compiled from in-house reagent stocks and as libraries from chemical suppliers. The proposed screening cascade for this unbiased kTGS approach is outlined in Fig. 109; Stage 1, DSF to pre-select fragments with a favourable stabilising effect on the enzyme, followed by STD-NMR to confirm the binding of the selected fragments and a solubility test to ensure the fragments were soluble in <5% DMSO and water to select fragments to test the unbiased kTGS approach; Stage 2, kTGS optimisation of fragment hits identified using the appropriate partnering fragment library; Stage 3, synthesis of identified triazole linked fragments and confirmation as inhibitors using a number of biophysical methods, including DSF, enzyme kinetics, STD NMR and 19F-NMR. 164 Fig. 109. Proposed screening cascade for unbiased kTGS. 4.1.1 Fragment based drug discovery As discussed in Chapter 1, fragment based drug discovery (FBDD) is an approach to drug discovery that identifies fragments that have a weak binding affinity to a desired biological target, these fragments are then grown or linked and optimised into lead compounds with high affinity and selectivity through an additive effect.92,100,101 An advantage of FBDD is it allows a greater proportion of chemical space to be investigated using a relatively small library due to the small size, efficiency of ligand binding and diversity of fragments. Once fragments have been identified, it is then necessary to design linkers to join two or more fragments or to grow the fragments using traditional structure guided medicinal chemistry into drug-like molecules.92,100,101 The optimal design of linkers is an appealing approach, however can prove challenging and time consuming. Our unbiased kTGS approach could have a significant impact on FBDD as it would provide a route to in situ fragment linking, thereby decreasing the time from identification of a fragment hit to the development of a lead compound. 4.1.2 Fragment library design The aim of this chapter was to explore the feasibility of using kTGS for in situ fragment linking and to discover non-peptidic inhibitors of GlxII, thus building on results in the Stage 1: Preliminary fragment screening Stage 2: kTGS Stage 3: Synthesis and Biophysical evaluation Fragment Library 00’s Compounds selected for Development and SAR 165 biased kTGS described in Chapter 3. To enable this, bespoke fragment libraries containing alkyne and azide functional groups were required. These libraries were assembled from commercial sources and their physicochemical properties evaluated using KNIME® Analytics Platform158,159 to ensure a diverse chemical space was utilised. The compiled libraries contained 102 azides and 98 alkynes respectively and a summary of the physicochemical properties is shown in Table 32 to highlight the diversity of chemical space. The Rule of 399 for fragment like compounds requires <300 Dalton, xlopP ≤ 3, hydrogen bond donors ≤ 3 and hydrogen bond acceptors ≤ 3. Using KNIME® the predicted properties for the azide library showed a range of molecular weights from 82 – 285 Dalton, which fits the Rule of 3 criteria. The xlogP range is -1 to 3.6, which means 5 of the azide fragments fall outside the criteria of ≤ 3. The hydrogen bond donors range from 0 to 2, and the hydrogen bond acceptors from 1 – 6. The H-bond donors fall within the criteria of ≤ 3 although 9 of the H-bond acceptors fall outside the criteria. The azide fragments also have a sp3 range of 0 – 0.3, rotatable bonds from 0 to 12 and total polar surface area from 12 to 104 to explore the space within the binding pocket. The alkyne library contains fragments with a molecular weight range from 56 to 273 Dalton, which fits the fragment like criteria of <300 Dalton. The xlogP range is -1.2 to 3.4, where 3 fragments in the library fall outside the criteria of ≤ 3. The range of H-bond donors is 1 to 2 and acceptors 0 – 6. The H-bond donors fall within the criteria of ≤ 3 although 1 of the H-bond acceptors fall outside the criteria. The alkyne fragments also have a sp3 range of 0 – 0.3, rotatable bonds from 0 to 8 and total polar surface area from 0 to 100 to explore the space within the binding pocket. Even though the a few of the fragments from the azide and alkyne libraries fall outside the Rule of 3 criteria for fragment-like compounds they were not removed from the kTGS assays for several reasons including: the biased kTGS showed bigger scaffolds were tolerated; the physicochemical properties of antibiotic drugs in general are an exception to Lipinski’s rules due to higher molecular weight and polarity;160 the desire not to limit the fragment diversity during the first iteration of the unbiased kTGS; the novel nature of the unbiased methodology, where currently the rules are unknown, and our interest in pushing kTGS to its limits and exploring this approach of in situ covalent fragment linking to identify novel fragment hits for drug discovery. 166 Table 32. Summary of physicochemical properties of azide and alkyne fragment libraries generated by KNIME®. Library MW xlogP TPSA HBD HBA sp3 Rotatable bonds Azide 82 - 285 -1 – 3.6 12 - 104 0 - 2 1 - 6 0 – 0.3 0 - 12 Alkyne 56 - 273 -1.2 – 3.4 0 - 100 0 - 2 0 - 6 0 – 0.3 0 - 8 The azide and alkyne libraries were split into 2 subsets. Alkyne library 1 and 2 consisted of 45 and 52 fragments respectively. (Tables 16 and 17, Chapter 3.2). Azide library 1 and 2 consisted of 50 and 51 fragments respectively. The fragments were named according to their position in the 96-well plate and stored as 100 mM stocks in DMSO at -20 °C. Tables 33 and 34 below show the structures of the azide fragments. Table 33. Structure and identifying codes of azide fragments in library 1 as positioned in the 96-well plates. AzA1 AzA2 AzA3 AzA4 Az1 Az2 Az3 AzB1 AzB2 AzB3 AzB4 Az4 Az5 Az6 AzC1 AzC2 AzC3 AzC4 Az7 Az8 Az9 AzD1 AzD2 AzD3 AzD4 Az10 Az11 Az12 AzE1 AzE2 AzE3 Az13 Az14 Az15 AzF1 AzF2 AzF3 Az16 Az17 Az18 AzG1 AzG2 AzG3 Az19 Az20 Az21 AzH1 AzH2 AzH3 Az22 167 Table 34. Structure and identifying codes of azide fragments in library 2 as positioned in the 96-well plates. AzA1-2 AzA2-2 AzA3-2 AzA4-2 AzA5-2 AzA6-2 AzA7-2 AzB1-2 AzB2-2 AzB3-2 AzB4-2 AzB5-2 AzB6-2 AzB7-2 AzC1-2 AzC2-2 AzC3-2 AzC4-2 AzC5-2 AzC6-2 AzC7-2 AzD1-2 AzD2-2 AzD3-2 AzD4-2 AzD5-2 AzD6-2 AzD7-2 AzE1-2 AzE2-2 AzE3-2 AzE4-2 AzE5-2 AzE6-2 AzF1-2 AzF2-2 AzF3-2 BLANK AzF5-2 AzF6-2 AzG1-2 AzG2-2 AzG3-2 AzG4-2 AzG5-2 AzG6-2 AzH1-2 AzH2-2 AzH3-2 AzH4-2 AzH5-2 AzH6-2 4.2 Towards the development unbiased kTGS methodology In order to investigate the feasibility of using kTGS for in situ fragment linking with low molecular weight and low affinity fragments a proof of principle experiment was carried out with GlxII, alkyne library 1 and azide library 1. GlxII was chosen as a model system due to ease of protein expression and purification. 4.2.1 Stage 1: Preliminary fragment screening To select fragments to test the unbiased kTGS methodology, a fragment screen was performed using DSF. DSF is a widely used technique for fragment screening and is as popular as NMR and SPR techniques.161,162 The fragments had a molecular weight range of <300 Da therefore large thermal shifts were not anticipated due to the small size of the fragments. 168 The alkyne fragments from library 1 (1 mM concentration) were incubated with GlxII (10 µM concentration) using previously optimised conditions. For alkyne library 1 the observed change in melt temperatures ranged from -1.35 – 3.68 °C. Alkyne fragment (AlkF4) was found to precipitate at this concentration with a ΔTm of -24.2 °C and was therefore removed from the reported data. Of the 45 fragments within Alkyne library 1, 28 resulted in a stabilisation of the GlxII melting temperature by an increase in melt temperature with ΔTm’s ranging from 0.05 – 3.68 °C, as summarized in Table 35. Twelve alkyne fragments demonstrated a ΔTm greater than 1 °C, as shown in Fig. 110. The five selected alkyne fragments, highlighted in bold in Table 35, were those that stabilised the melting temperature of the GlxII protein to the greatest extent were: AlkD5, AlkF5, AlkB5, AlkE5 and AlkA4. For the full data set see Table 64 and Fig. 163 in the Appendix. Table 35. Thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from Alkyne library 1. Fragment ΔTm Fragment ΔTm GlxII 0.00 AlkB3 0.61 GSH 2.01 AlkD3 0.14 GAz (32) 1.7 AlkE3 1.19 AlkA1 0.66 AlkF3 1.36 AlkC1 0.32 AlkG3 1.26 AlkD1 0.26 AlkH3 0.50 AlkE1 0.31 AlkA4 1.52 AlkF1 0.43 AlkB4 1.40 AlkG1 0.28 AlkC4 0.31 AlkA2 1.01 AlkE4 0.84 AlkB2 0.46 AlkH4 0.48 AlkC2 -0.10 AlkB5 2.01 AlkD2 0.90 AlkC5 0.05 AlkF2 1.27 AlkD5 3.68 AlkG2 1.31 AlkE5 1.99 AlkH2 0.54 AlkF5 2.99 Fig. 110. Bar graphs of thermal shift (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from Alkyne library 1. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). 169 In a similar manner, the 50 azide fragments from library 1 (1 mM concentration) were incubated with GlxII (10 µM concentration) using DSF using previously optimised conditions. The difference in melt temperatures ranged from -1.94 – 6.96 °C. Of the 50 fragments within azide library 1, 35 showed a stabilisation of the GlxII melting temperature by an increase in ΔTm ranging from 0.06 – 6.96 °C, as summarized in Table 36 and Fig. 111. Eighteen azide fragments demonstrated a ΔTm greater than 1 °C. The azide fragments that showed the highest stabilisation, highlighted in bold in Table 5, for binding to GlxII were: AzG02, AzC03, AzA03, AzA04, AzG01, Az-19, AzA02 and Az-13. For the full data set see Table 65 and Fig. 166 in the Appendix. Table 36. Thermal melt temperatures (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from azide library 1. Fragment ΔTm Fragment ΔTm Fragment ΔTm GlxII 0.00 GSH -2.25 GAz (32) -0.30 AzA1 0.87 AzC3 6.41 Az-11 1.24 AzB1 1.32 AzD3 0.44 Az-12 1.86 AzC1 0.73 AzE3 0.62 Az-13 2.34 AzE1 0.25 AzF3 1.17 Az-14 0.99 AzF1 1.05 AzA4 3.32 Az-15 0.13 AzG1 2.77 AzC4 0.08 Az-16 1.62 AzA2 2.71 Az-1 0.52 Az-17 0.06 AzB2 0.23 Az-2 1.51 Az-18 0.20 AzC2 1.34 Az-3 0.87 Az-19 2.74 AzE2 1.93 Az-5 0.58 Az-20 1.53 AzG2 6.96 Az-8 0.41 Az-22 0.55 AzA3 4.45 Az-9 1.70 Fig. 111. Thermal melt temperatures (ΔTm, °C) in the presence of ligand relative to that of GlxII alone with fragments demonstrating positive ΔTm from azide library 1. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). This fragment hits identified in each screen were ranked in order of ΔTm and the following fragments were selected: AlkD5, AlkF5, AlkB5, AlkE5 and AlkA4; AzG2, AzC3, AzA3, AzA4, AzG1, Az-19, AzA2 and Az-13. Prior to selection as scaffolds for the pilot unbiased 170 kTGS experiments, the fragments were subjected to a solubility test (<5% DMSO in water) as the kTGS assay was carried out in aqueous media. Not all fragments were soluble under the kTGS conditions which resulted in 2 azide fragments, AzG1 and AzG2, and 2 alkyne fragments, AlkD5 and AlkF5, being selected. Table 37 shows the alkyne and azide scaffolds chosen to test the unbiased methodology. Table 37. Selected fragments for the unbiased kTGS due to the combination of ΔTm and aqueous solubility. ID Structure ΔTm ID Structure ΔTm AlkD5 3.68 AzG1 2.77 AlkF5 2.99 AzG2 6.96 4.3 Stage 2: Unbiased kTGS assay With a number of azide and alkyne fragments identified we could begin to explore the potential of the unbiased kTGS for in situ fragment linking. The fragments (AzG1, AzG2, AlkD5 and AlkF5) were chosen to test the unbiased kTGS methodology due to a combination of highest ΔTm with GlxII and aqueous solubility. kTGS experiments were performed as previously described for the pilot biased kTGS, ensuring mixtures of alkynes and azides had distinct molecular weights to enable identification of triazoles formed in situ by LC-MS/MS. Ten mixtures of up to five azides were grouped to screen with the alkyne fragments hits. Nine mixtures of up to five alkynes were grouped to screen with the corresponding azide fragments hits. The mixtures were co-incubated with GlxII, BSA and buffer, respectively. Analysis of the crude mixtures was carried out on an Agilent Triple Quad LC-MS/MS with a MS2 scan and SIM scan, on an AquaSil C18 (100 x 4.6 mm, 5 µm particle size) column. Molecular weights identified to match expected triazoles were considered to be ‘hit’ compounds if observed in the GlxII wells with an area under the curve of at least double compared to the control wells. An example of a hit triazole is shown in Fig. 112, where the triazole derived from the fragments AzG2 and AlkD1 showed a higher area under the curve for the protein sample compared to the controls in BSA and buffer. 171 Fig. 112. LC-MS/MS trace overlays for the synthesis, protein, BSA and Buffer of AzG2 and AlkD1 in the SIM and MS scan. In the unbiased kTGS screens with the two azide fragments (AzG2 and AzG1) and alkyne library 1: five ‘hits’ were observed with AzG2 whereas with AzG1 a single hit was identified. In the corresponding experiments with the two alkynes (AlkF5 and AlkD5) and azide library 1: no ‘hits’ were observed with AlkF5 and with AlkD5 a single hit was identified (Fig. 113). This corresponded to a hit rate of 11% for AzG2, 2% for AzG1, 0% for AlkF5 and 2% for AlkD5, compared to a typical hit rate of 15 – 19% being observed in the biased experiments as described in Chapter 3. Due to a combination of a number of factors the lower hit rate in the unbiased experiments is perhaps to be expected, these include: the lower affinity and small size of the fragments; fragments adopting multiple binding modes and the likelihood of high off rate thus resulting in a lower probability of correctly aligned fragments in close enough proximity within the binding site to react. It could be anticipated that a lower hit rate would be observed in the unbiased approach due to unknown binding modes, whereas in the biased approach having a known handle in the G-site promotes the linking together with fragments in the H-site. One exciting advantage of this unbiased kTGS approach is the potential for the small triazoles to explore multiple binding sites, compared to the controlled biased kTGS, allowing further hit optimisation via traditional medicinal approaches to develop novel non-peptidic inhibitors for GlxII. The percentage triazole formation was calculated from the integration of the area under the curve as before and values ranging 0.7 – 45% were observed as shown in the bar graphs in Fig. 114. A concern from the biased kTGS were compounds that produced large triazole formation were often correlated with masses that were often in the MS trace postulated to be derived from components of the storage buffer and not the kTGS assay. In the unbiased assay the larger triazole formations could be due to the reactivity of the fragments within the binding site, however, as this was a pilot experiment and the overall 172 hit rate was low the compounds were synthesised to undergo biophysical evaluation so that lessons could be identified. Fig. 113. ‘Hit’ compounds identified in the unbiased kTGS. Where MS refers to triazole masses observed in the MS2 scan and SIM refers to triazole masses observed in the SIM scan. Fig. 114. Bar graphs of % triazole formation for the four unbiased kTGS pilot experiments with GlxII. 4.4 Stage 3: Synthesis and biophysical evaluation of the unbiased kTGS hits 4.4.1 Synthesis of the unbiased triazoles The following azide fragments, AzG1 (80) and AzC2 (81), were synthesised using a standard nucleophilic displacement reaction (Scheme 12). The low yield achieved in the 0.0010.0020.0030.0040.0050.00AlkE1 Alk6 AlkG2 AlkC3 AlkD1 Alk7 AzC2GlxII 26.47 0.93 1.70 0.72 23.45 29.18 45.77BSA 1.11 0.00 0.00 0.00 3.47 4.38 10.82Buffer 2.03 0.00 0.72 0.00 3.23 0.00 12.85% Triazole formationGlxII BSA Buffer173 synthesis of AzC2 was due to volatility. Challenges with volatility were also the reason AzG2 was purchased after multiple attempts to isolate the pure compound failed. The following alkyne fragments, AlkD5 (68), AlkE1 (45) and AlkC3 (82) were synthesised under standard alkylation and urea formation conditions as shown in Scheme 13, or purchased from chemical suppliers. Scheme 12. Synthesis of azide fragments. Reagents and conditions: (a) NaN3, DMF, 80 °C, 4 h. AzG1 (80) 80%, AzC2 (81) 20% Scheme 13. Synthesis of alkyne fragments. Reagents and conditions: (b) K2CO3 (2 eq), DMF, RT, 16 h, 81%. (c) DCM, 0 °C-RT, 6 h, quantitative. (d) NaH (1.1 eq), DMF, 0 °C-RT, 16 h, 55%. With the building block quantities of the required fragments in hand, the 1,4-triazoles hits were synthesised using previously optimised conditions (Chapter 3), in good yields. One compound (88) could not be isolated due to decomposition during purification. To observe whether the regioisomers influenced binding and inhibition of GlxII the synthesis of the 1,5-triazoles were also attempted. One 1,5-triazole was successfully synthesised using ruthenium catalysis, however in the three other examples attempted no reaction was observed (Table 38). 174 Table 38. Unbiased triazole synthesis conditions. Triazole 1,4-Triazole Cu 1,5-Triazole Rui g, 96% 83 h, 67% 90 g, 82% 84 g, 87% 85 h, 0% g, 48% 86 h, 0% g, quantitative 87 h, 0% g, 0% 88 g, 53% 89 Reagents and conditions: (g) azide (1 eq), alkyne (1 eq), H2O:t-BuOH (2:1), sodium ascorbate (15 mol%), Cu(II)SO4.5H2O (2 mol%), TBTA (2 mol%), RT, 16-20 h. (h) Ru*Cp(PPh3)2 (2 mol%), dioxane, 90 °C, microwave, 1 hr. To conclusively demonstrate that the 1,5-regioisomer had been synthesised 2D NMR techniques were used to determine the differences between the 1,4 and 1,5 triazoles. In the proton spectrum, the observed δH for the triazole ArCH shifted from 7.54 ppm for the 1,4-triazole to 7.51 for the 1,5-triazole. In the carbon sprectrum, the observed δC for the ArCH shifted from 121.9 ppm for the 1,4-triazole to 131.8 ppm for the 1,5-triazole and the observed δC for the quaternary triazole C shifted from 137.4 ppm for the 1,4-triazole to 143.9 ppm for the 1,5-triazole. The main difference observed in the 13C was an upfield shift of the 1,4-triazole. To further confirm the regioisomers HMQC and HMBC spectra 175 were run. The HMBC spectra for both compounds, 83 and 90 (Fig. 115), are shown in Fig. 116 and 117. In the HMQC spectra for the 1,4-triazole 83, the triazole Ar(C5)-H proton at 7.54 ppm directly couples to the carbon at 121.9 ppm. For the 1,5-triazole 90, the triazole Ar(C4)-H proton at 7.51 ppm directly couples to the carbon at 131.4 ppm in the HMQC. These spectra allowed the assignment of the C4 and C5 signals through direct coupling, the relatively large shift indicated different compounds. HMBC spectra were run to confirm the regiochemistry through differences in longer-range coupling. For the 1,4-triazole 83, the triazole Ar(C5)-H proton at 7.54 ppm couples with the quaternary carbon at 154.5 ppm, as shown by 1, and weakly with the CH2 in the side chain at 51 ppm, as shown by 2. The CH2 at 4.54.ppm in the side chain interacts with, the other CH2 at 35.9 ppm, as shown by 3, and the triazole Ar(C5)-H at 121.9 ppm (as shown by 2) as well as the quaternary ArC on the phenyl ring at 136.5 ppm, as shown by 4. For the 1,5-triazole 90, the triazole Ar(C4)-H proton at 7.51 ppm interacts with the quaternary carbon at 143.9 ppm, as shown by 1. The CH2 in the side chain at 4.71 ppm couples with the other CH2 at 35.9 ppm, as shown by 3, and the quaternary ArC in the phenyl ring at 137.9 ppm, as shown by 4, and the quaternary triazole carbon at 143.9 ppm, as shown by 2. The key difference was the cross over peaks of 1 and 2, where only 1 was observed for the triazole Ar(C4)-H of 90 and no adjacent longer range interactions as observed by 2 for 83, therefore allowing confirmation of the regioisomers. Fig. 115. Structure of 83 and 90 with the interactions under examination numbered. 176 Fig. 116. HMBC of 83. Fig. 117. HMBC of 90. 4.4.2 Biophysical evaluation of the unbiased triazoles The synthesised triazole hits were evaluated using DSF and enzyme kinetics. 87, ΔTm 1.61 °C and 84, ΔTm 1.58 °C demonstrated stabilisation of GlxII comparable to GAz (32), ΔTm 1.57 °C. For ligands 83 (ΔTm 0.49 °C) and 85 (ΔTm 0.65 °C) the observed ΔTm was comparable to that observed with GSH (ΔTm 0.61 °C). Ligands 89 (ΔTm -0.86 °C) and 86 (ΔTm -0.41 °C) appeared to destabilise or bind to the denatured GlxII. Interestingly, a greater thermal shift (2.56 °C) was observed for the 1,5 regioisomer (90) compared with the corresponding 1,4 regioisomer (83). Re-examination of the kTGS LC-MS/MS data revealed two peaks in a 3:1 ratio at 2.15 min and 2.35 min which could indicate that both 177 the 1,4 and 1,5 regioisomers were synthesised in the in situ experiment. This data suggests that the unbiased kTGS has potential as a tool to identify non-peptidic inhibitors of GlxII due to the favourable stabilisation of change in melt temperatures indicating binding to GlxII (Fig. 118, Table 39). Fig. 118. Thermal shift (ΔTm, °C) for unbiased synthesised triazoles and their corresponding fragments with respect to GlxII. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 39. Thermal shift (ΔTm, °C) binding results for synthesised unbiased hits to GlxII. Compound Structure ΔTm Compound Structure ΔTm GlxII 1% DMSO 0.00 84 1.58 GSH 0.61 85 0.65 GAz (32) 1.57 86 -0.41 90 2.56 87 1.61 83 0.49 89 -0.86 Comparing the observed thermal shift in the presence of the triazoles and their corresponding fragments to evaluate any fragment synergy did not identify a trend between the fragments and the triazoles or ‘linked fragments’ as was anticipated due to the theory of additivity proposed by Jencks88 (Table 40). 90 showed a relative increase 178 in binding stabilisation compared to its 2 constituent fragments, Alk7 and AzG1 (80), whereas 83 showed a relative decrease in binding stabilisation compared to Alk7 and AzG1 (80), but still demonstrated a small but significant stabilising effect indicating binding to GlxII. 84 and 87 showed an increase in binding stabilisation compared to their 2 constituent fragments, AlkD5 (68) with AzC2 (81) and Alk6 with AzG2, respectively. 85 showed a slight increase in binding stabilisation compared to its 2 constituent fragments, AlkE1 (45) and AzG2. 86 and 89 showed a decrease in binding stabilisation compared to their 2 constituent fragments, AlkD1 with AzG2 and AlkC3 (82) and AzG2. It is important to note that there was a difference in the preliminary fragment DSF results from library stocks compared to the synthesised fragment results. A thermal shift of 2.77 °C was observed for AzG1 (80) however when the synthetic sample was retested the thermal shift was 1.34 °C. This trend was also observed for AzG2, where the thermal shift dropped from 6.96 °C to 0.58 °C, and AlkD5 (68) where the thermal shift dropped from 3.68 °C to -0.69 °C. AlkF5 had a thermal shift of 2.99 °C but as no hit triazoles were observed in the kTGS this fragment was not synthesised and rescreened. Table 40. Thermal shift (ΔTm, °C) binding results for synthesised triazoles and corresponding alkyne and azide fragments to GlxII. Compound Structure ΔTm Alkyne fragment ΔTm Azide fragment ΔTm 83 0.49 Alk7 1.27 AzG1 (80) 1.34 84 1.58 AlkD5 (68) -0.69 AzC2 (81) -0.60 85 0.65 AlkE1 (45) 0.30 AzG2 0.58 86 -0.41 AlkD1 1.92 AzG2 0.58 87 1.61 Alk6 -0.32 AzG2 0.58 89 -0.86 AlkC3 (82) 2.25 AzG2 0.58 90 2.56 Alk7 1.27 AzG1 1.34 179 In summary, the DSF results indicated that 5 of the 7 unbiased synthesised hits did stabilise the melt temperature of GlxII thus indicating binding. The two 1,4-triazole compounds with the highest thermal shift, 87 and 84, showed an increase in thermal shift compared to their 2 constituent fragments. The 1,5-triazole, 90, showed the greatest increase in thermal shift and an increase from the 2 component fragments. This exciting proof of principle data suggests that unbiased kTGS has potential as a tool for in situ fragment linking. To understand the effect of these synthesised hits on the activity of GlxII and to evaluate these compounds further enzyme kinetics experiments were carried out. The triazoles were evaluated at 10 and 100 µM using the two optimized enzyme kinetics assay conditions described in Chapter 3. As described in Chapter 3, DTNB or Ellman’s reagents can be used to detect GSH which is one of the products of the GlxII catalysed hydrolysis of SLG by monitoring the absorbance at 412nm (Fig. 119). Absorbance was monitored over time, and the slope of the initial rate of reaction was calculated using linear regression in GraphPad Prism. The slopes were normalised and the difference in slopes relative to protein (2% DMSO) in the absence of ligand used to calculate the % inhibition (Fig. 119). A summary of the % inhibition is shown in Table 41. 180 A. B. C. Fig. 119. A) OD (412 nm) vs time for unbiased synthesised triazoles. B) normalised slope and C) % Inhibition of relative GlxII activity for unbiased synthesised triazoles at 10 and 100 µM. Significance of changes evaluated by a Student’s t-test (where ** p ≤ 0.01, * p ≤ 0.05). Table 41: % Inhibition of GlxII by the unbiased triazoles at 10 and 100 μM, measured at 412 nm. Compound 0 μM 00 μM 83 42.94 49.49 84 30.56 41.90 85 -28.02 37.78 86 -1.04 40.44 87 76.28 21.62 89 21.72 45.93 90 17.13 1.67 In summary, the enzyme kinetics data for the unbiased triazoles revealed that 66% showed inhibition at 10 µM and 100% at 100 µM. 84 showed a thermal shift of 1.58 °C indicating binding to GlxII and inhibition of 30 to 41% on increase of 10 to 100 µM of ligand concentration. 83 showed a thermal shift of 0.49 °C indicating binding to GlxII and inhibition 42 to 49% on increase of 10 to 100 µM of ligand. 90 showed a good binding affinity to GlxII and a 17% inhibition at 10 µM, an increase of ligand to 100 µM resulted 181 in the % inhibition decreasing to 1%. 87 showed a good combination of binding affinity (1.61 °C) and 76% inhibition at 10 µM, however a decrease in inhibition (21%) was observed at 100 µM of ligand. Although 89 did not show a stabilization of melt temperature (-0.86 °C) the concentration response in the enzyme kinetics assay showed 21 to 45% inhibition of relative GlxII activity. 85 showed a thermal shift of 0.65 °C and inhibition of 37% at 100 µM of ligand. 86 showed a decrease in the thermal shift of - 0.41 °C and inhibition of 40% at 100 µM of ligand. As discussed in Chapter 3, there was a concern regarding the wide spread of data points that was observed during this DTNB assay which can be due to a number of factors including:154,155 a problem due to interference with the absorbance either by compounds or the biological sample; oxidation of GSH; the instability of DTNB. To address this the assay was repeated with the ligand, GlxII protein and SLG undergoing preincubation for 10 min at room temperature before the addition of the DTNB followed by the OD measurement at 412 nm to allow measurement of the endpoint of the reaction. OD readings at 5.6 min where normalized to calculate the % inhibition (Fig. 120). A summary of the % inhibition is shown in Table 42. All the 1,4-triazoles did inhibit GlxII activity at both 10 and 100 µM ligand concentration, however, 89 was the only compound to show a concentration dependent response. This pre-incubation assay however did not improve the spread of data points and the results were inconsistent compared to the initial assay, therefore the data was not used to evaluate the triazoles. 182 A.B. Fig. 120. A) Normalised curve and B) % Inhibition of relative GlxII activity for unbiased synthesised triazoles. Significance of changes evaluated by a Student’s t-test (where ** p ≤ 0.01, * p ≤ 0.05). Table 42: % Inhibition of GlxII at 10 and 100 μM, measured at 412 nm after preincubation. Compound 0 μM 00 μM 83 42.76 19.91 84 33.44 10.36 85 16.28 8.58 86 30.55 7.42 87 14.07 5.40 89 12.88 21.98 90 14.87 -7.95 An alternative assay to directly measure the hydrolysis of SLG via a decrease in absorbance at 240 nm and calculate the specific activity of GlxII using Beer’s law, where ε = molar extinction coefficient (-3.1 mM-1cm-1) to give the specific activity of GlxII in units/mg, and to calculate the % inhibition in the presence of triazole ligands (10 and 100 µM as before). The protocol to calculate the glyoxalase activities was carried out as described in the publication by Thornalley et al.’153 (Table 43, Fig. 121). 183 Table 43: Activity in U/mg and % Inhibition GlxII at 10 and 100 μM. Unbiased triazoles measure at 240nm. 10µM 100µM Specific Activity % Inhibition Specific Activity % Inhibition GlxII_2%DMSO 5.43 0.00 3.85 0.00 83 3.59 33.96 4.08 -5.90 84 4.26 21.60 0.89 76.99 85 4.24 21.85 1.53 50.71 86 8.19 -50.79 0.78 79.64 87 1.34 75.31 0.93 75.71 89 1.34 74.52 4.00 -3.87 90 2.13 60.79 6.08 -57.78 Fig. 121. Bar graphs showing % inhibition of the unbiased triazoles against specific GlxII activity. Significance of changes evaluated by a Student’s t-test (where * p ≤ 0.05). This activity data suggested that the most promising unbiased inhibitor was 87 (inhibition of 75%). 85 (inhibition of 21 to 50%) and 84 (inhibition of 21 to 76%) showed concentration dependent responses as ligand increased from 10 to 100 µM. At 10 µM 6 of the 7 compounds appeared to inhibit GlxII whilst at 100 µM only 4 showed inhibition. The spread of data and lack of significance was once again a cause for concern. One possible rationale could be the compounds interfering with the absorbance at 240 nm or other side reactions occurring between the ligands and SLG or GSH. A summary of the in situ % triazole formation, DSF and % inhibition data is shown in Table 44. 184 Table 44: Summary of % triazole formation during kTGS, ΔTm, % inhibition and specific activity data for unbiased triazoles. kTGS DSF % Inhibition Specific activity (U/mg) Structure and Compound ID % triazole formation ΔTm 0μM 412nm 00μM 412nm 0μM 240nm 00μM 240nm 0μM 00μM GlxII 0.00 0.00 0.00 0.00 0.00 5.43 3.85 83 29.18% 0.49 42.94 49.49 33.96 -5.90 3.59 4.08 90 29.18% 2.56 17.13 1.67 60.79 -57.78 2.13 6.08 84 45.77% 1.58 30.56 41.90 21.6 76.99 4.26 0.89 85 2.04% 0.65 -28.02 37.78 21.85 50.71 4.24 1.53 86 23.45% -0.41 -1.04 40.44 -50.79 79.64 8.19 0.78 87 0.93% 1.61 76.28 21.62 75.31 75.71 1.34 0.93 89 0.72% -0.86 21.72 45.93 74.52 -3.87 1.38 4.00 Enzyme kinetics at GlxII 0.2 µM. N/A not applicable. Thermal shift values GlxII at 10 µM, ligand at 1 mM. In summary, of the 7 triazoles evaluated using a combination of DSF and enzyme kinetics it appeared that these unbiased ligands showed potential to be used as probes to assess GlxII as an antibacterial target. This exciting proof of principle experiment demonstrated the potential of the unbiased kTGS approach. As with the biased kTGS, % triazole formation was not a predictor of binding affinity or % inhibition. Due to the small size of these triazoles (MW 170 – 380 Da) big thermal shifts and high inhibition results were not anticipated due to the unknown, or possibly multiple, binding modes and potentially low affinity for the binding site. Therefore, thermal shifts of -0.86 to 2.56 °C and inhibition at 10 µM of -50 – 76% and at 100 µM of -57 – 185 79% shows tangible potential for these novel non-peptidic probes to be further developed into GlxII inhibitors. To determine the ranking of the triazoles for development the data from the DSF screen and both enzyme kinetics assays were compared. 87 and 84 were both indicated to bind to GlxII with thermal shifts of 1.61 and 1.58 °C, respectively. These compounds also demonstrated inhibition at 10 and 100 µM in both the OD412nm and OD240nm assays. Of note is both 87 and 84 contain a hydroxy moiety and a nitrile group respectively, which could provide useful handles to be functionalised and grow off for compound development. 85, 83 and 90 all indicated to bind to GlxII with thermal shifts of 0.65, 0.49 and 2.56 °C, respectively. 85 only exhibited inhibition at 100 µM in the OD412nm assay of 37% and a concentration response of 21 – 50% in the OD240nm assay upon ligand concentration increasing from 10 to 100 µM. 83 and 90, the 1,4 and 1,5 regioisomer matching pair, demonstrated inhibition at both concentrations in the OD412nm assay but only at 10 µM in the OD240nm assay. 83 demonstrated a % inhibition of 42 – 49% at OD412nm and the GlxII activity was inhibited at 10 µM by 33% at OD240nm. 90 demonstrated a decrease in inhibition of 17 – 1% at OD412nm and the GlxII activity was inhibited at 10 µM by 60% at OD240nm. 89 did not appear to stabilise GlxII during the DSF screen however the inhibition, at OD412nm, was 21 – 45% and 74% at 10 µM during the OD240nm assay. This compound contains a carboxylic acid moiety on the aromatic ring which could provide a useful handle to develop this compound, resulting in more interactions with the binding site to improve the binding affinity. 86 did not appear to stabilise GlxII during the DSF screen and only demonstrated inhibition at 100 µM in both enzymatic assays, 40% in the OD412nm assay and 79% in the OD240nm assay. This compound also contained a carboxylic acid moiety providing a useful handle for further functionalisation. In terms of using this data to rank the compounds for further development into GlxII antibacterials, the compounds that demonstrated both binding affinity and inhibition in all enzyme kinetics assays at 10 and 100 µM, 87 and 84, were ranked as a higher priority. The second group (85, 83 and 90) demonstrated binding and inhibition in at least one of the assays. The third group (89 and 86) where binding was not detected by DSF, however inhibition was observed at 100 µM and both compounds had good handles which could be used to grow and develop into lead compounds. 186 Therefore, these triazoles exhibited the potential to be developed into chemical probes. Given the fragment based nature of the unbiased kTGS approach to identify and explore the GlxII binding site and the ability of kTGS to link the fragments in situ, these 4 experiments with a total of 190 fragments have not only identified 7 triazoles, which could be ranked for further research, but have demonstrated the exciting potential of this in situ fragment screening and linking approach. 4.4.3 Docking studies To evaluate how the triazoles were predicted to bind, docking studies were carried in collaboration with Dr Geoff Wells (UCL School of Pharmacy). The 1,4-triazoles were docked using Le Dock and Vina150 into the GlxII crystal structure obtained from the PDB. Chimera was used to visualise the predicted Vina dockings into Glyoxalase II from Salmonella thyphimurium (PDB: 2QED).59 87 was predicted to bind in the hydrophobic pocket through hydrogen bond interactions with Tyr168 and a water molecule around the metal in the active site (Fig. 122). This predicted orientation of the triazole with the hydroxyl group pointing towards Tyr168 and the potential for optimization allows future development of 87 to focus on growing the structure off the hydroxy group to pick up additional interactions in the G-site. Fig. 122. 89 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. 84 was predicted to bind with the aromatic ring occupying the hydrophobic site and the triazole and cyano group partially occupying the glutathione binding pocket (Fig. 123). This predicted orientation allows future development of 84 to focus on growing the structure from the nitrile group to pick up additional interactions in the G-site. 187 Fig. 123. 84 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. 83 was predicted to bind with the aromatic ring occupying the hydrophobic pocket forming a π stacking interaction with Tyr168 and the triazole and dimethyl-hydroxy group pointing towards the glutathione binding site (Fig. 124). Fig. 124. 83 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. 89 was predicted to bind with the aromatic ring occupying the hydrophobic pocket through π stacking interactions with Tyr168 and the triazole and ethyl-N-dimethyl group occupying the glutathione binding site (Fig. 125). Fig. 125. 89 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. 188 85 was predicted to bind with the ethyl-N-dimethyl group in close proximity to the Tyr168, the triazole and urea in the right hand side of the glutathione binding site and the difluorophenyl protruding out of the glutathione binding pocket (Fig. 126). Fig. 126. 85 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. 86 was predicted to bind with the aromatic group occupying the hydrophobic pocket through π stacking interactions with Tyr168, the triazole and ethyl-N-dimethyl amino group occupying the middle of the glutathione binding site with hydrogen bonds interaction between the amino group, Asp58 and a water molecule (Fig. 127). Fig. 127. 86 predicted docking in GlxII from S. thyphimurium (PDB: 2QED), visualised in Chimera. In summary, the unbiased triazoles were predicted to fill different parts of the glutathione and hydrophobic pockets indicating further possibilities to develop and grow these hits into lead compounds. The Vina docking scores are shown below (Fig. 128) where the lowest predicated score are compounds with the best fit in the GlxII binding site. Interestingly, these predicted scores didn’t match the experimental ranking, therefore, future work to refine the docking model based on experimental data would be beneficial. 189 Due to the unbiased nature of these triazoles and without the biased peptide handle competition experiments to confirm whether the triazoles are occupying the G- or H-sites would allow confirmation of the docking model. Fig. 128. Predicted Vina docking scores (kcal/mol) for the unbiased 1,4-triazoles. 4.4.4 Summary To test the unbiased methodology using GlxII as a model system 4 fragments (2 azides and 2 alkynes) were chosen through a preliminary DSF screen to identify fragments that indicated binding to GlxII. The 2 azide fragments were incubated with a diverse set of alkynes, in 10 groups of 5, and incubated with GlxII for 3 days at 37 °C. The 2 alkyne fragments were incubated with a diverse set of azides, in 9 groups of 5, and incubated with GlxII for 3 days at 37 °C. Two controls were used where the azides and alkynes were incubated with BSA and buffer. The 4 unbiased kTGS assays were analysed using LC-MS/MS and triazole masses found in the protein samples with at least double the area under to curve to the controls were identified as ‘hit’ compounds. From a total of 190 potential products, excluding regioisomers, 7 triazole hits were identified, which corresponded to approximately 3.6% hit rate over the 4 kTGS experiments. Six of the 1,4-triazoles were successfully synthesised, the remaining triazole decomposed so was therefore not evaluated further. One 1,5-triazole was synthesised, however, to date attempts to synthesise further 1,5-analogues have been unsuccessful. With synthetic samples of the hits in hand, binding and inhibition of these six 1,4-triazoles and one 1,5-triazoles were evaluated using DSF and enzyme kinetics. Evidence indicating binding of five of the triazoles (83, 84, 85, 87 and 90) was obtained using DSF where the compounds exhibited an increase in melt temperature. Two of the compounds (86 and 89) tested appeared to decrease the melt temperature. Subsequently enzyme kinetics experiments were used to evaluate inhibition of the relative GlxII activity, through 86, -6.7 85, -6.683, -689, -5.684, -5.487, -4.1-7-6.5-6-5.5-5-4.5-4-3.5Docking score (kcal/mol)190 monitoring the formation of GSH at 412 nm, and inhibition of the specific activity of GlxII, through monitoring the decrease in SLG at 240 nm. The assay at 412 nm revealed a concentration response for 5 of the triazoles and 2 (including the 1,5-triazole) showed a decrease in inhibition when ligand concentration increased from 10 to 100 µM. The 240 nm assay revealed a concentration response for 3 triazoles. A comparison of all the data indicated that the 7 synthesised triazoles could be clustered into 3 groups determining the order of ranking for further development. Priority group 1 consisted of 87 and 84, due to combination of binding affinity and inhibition in both assays at 10 and 100 µM. Priority group 2 consisted of 85, 83 and 90, due to binding affinity and partial inhibition being observed. Priority group 3 consisted of 89 and 86 where binding was not detected by DSF, however inhibition was observed at 100 µM. In terms of future development of these hits into lead compounds for GlxII inhibitors, 87, 84, 89 and 86 had had good handles which could be used to grow and develop these hits into lead compounds. To gain further insight the 1,4-triazoles were docked into GlxII (PDB:2QED) using Vina and were predicated to occupy different parts of the glutathione and hydrophobic sites highlighting the potential to develop these compounds into leads is possible. Given the challenges encountered with the enzyme kinetics assay it would be beneficial to incorporate another biophysical technique (e.g. STD-NMR, SPR, ITC or FP) into the screening cascade. This would provide increased confidence in the hits identified. However, the initial indications are extremely promising for an unbiased approach to identify novel non-peptidic inhibitors and the usefulness of kTGS as a tool for fragment linking. 4.5 Fragment screening for future unbiased kTGS 4.5.1 DSF screening with GlxII For future fragment selection for additional unbiased kTGS experiments, alkyne library 2 and azide library 2 were screened against GlxII using DSF as before. The alkyne fragments (1 mM concentration) from alkyne library 2 were incubated with GlxII (10 µM). The observed change in melt temperature ranged from -6.68 to 1.43 °C. Out of the 52 fragments within alkyne library 2, 35 demonstrated a stabilisation of the GlxII melt temperature by an increase in thermal shift ranging from 0.13 – 1.43 °C, as summarised in Table 45. The 5 alkyne fragments that stabilised the melting temperature of protein to the greatest extent indicating binding to GlxII were: AlkA2-2, AlkB3-2, AlkC2-2, AlkH2-2 191 and AlkD6-2. However, only 2 alkyne fragments demonstrated a thermal shift greater than 1 °C relative to standard deviation, as highlighted in bold in Table 45 and Fig. 129, and were therefore chosen as alkyne fragments to use in future unbiased kTGS experiments. For a full summary of the data see Table 66 and Fig. 167 in the Appendix. Table 45: Thermal shift data of fragments from alkyne library 2 with a positive ΔTm in the presence of GlxII. Fragment ΔTm Fragment ΔTm Fragment ΔTm Fragment ΔTm GSH -0.72 AlkD2-2 0.64 AlkH2-2 1.16 AlkD5-2 0.16 AlkA1-2 0.67 AlkE1-2 0.29 AlkH3-2 0.40 AlkD6-2 1.43 AlkA2-2 0.90 AlkE2-2 0.13 AlkA4-2 0.43 AlkE4-2 0.33 AlkA3-2 0.17 AlkE3-2 0.34 AlkA5-2 0.69 AlkF4-2 0.44 AlkB1-2 0.19 AlkF1-2 0.24 AlkA6-2 0.67 AlkF5-2 0.54 AlkB3-2 0.84 AlkF2-2 0.78 AlkB4-2 0.66 AlkF6-2 0.21 AlkC1-2 0.29 AlkG1-2 0.15 AlkB6-2 0.76 AlkH4-2 0.29 AlkC2-2 1.27 AlkG3-2 0.18 AlkC4-2 0.35 AlkH5-2 0.21 AlkC3-2 0.38 AlkH1-2 0.21 AlkC6-2 0.84 AlkH6-2 0.46 Fig. 129. Bar graphs of thermal shift (ΔTm, °C) for fragments from alkyne library 2 with greater that 1 °C ΔTm in the presence of GlxII. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). The azides fragments (1 mM concentration) from azide library 2 were incubated with GlxII (10 µM) (Fig. 130, Table 46). The observed change in melt temperature ranged from -1.12 to 2.38 °C. Out of the 51 fragments within azide library 2, 39 demonstrated a stabilisation of the GlxII melt temperature by an increase in thermal shift ranging from 0.06 to 2.38 °C, as summarised in Table 46. However, only 7 azide fragments demonstrated a thermal shift greater than 1 °C and 3 fragments, AzA2-2 (ΔTm 1.33 °C), AzH2-2 (ΔTm 1.59 °C) and AzH6-2 (ΔTm 2.38 °C), greater than 1 °C relative to standard deviation, as highlighted in Table 46 and Fig. 130, and were therefore chosen as azide fragments to use in future unbiased kTGS experiments. For the full summary see Table 67 and Fig. 168 in the Appendix. 192 Table 46: Thermal shift data of azide library 2 in the presence of GlxII. Fragment ΔTm Fragment ΔTm Fragment ΔTm Fragment ΔTm GlxII 0.00 AzD2-2 0.43 AzH3-2 0.26 AzD7-2 0.15 GSH -1.27 AzD3-2 0.46 AzA5-2 0.28 AzE4-2 0.81 GAz (32) 0.26 AzE1-2 0.36 AzA6-2 0.43 AzE5-2 0.69 AzA2-2 1.33 AzE2-2 1.16 AzA7-2 0.39 AzF5-2 1.28 AzA3-2 0.78 AzE3-2 0.54 AzB4-2 0.50 AzF6-2 0.69 AzB1-2 0.78 AzF1-2 0.53 AzB5-2 0.28 AzG4-2 0.93 AzB2-2 0.06 AzF2-2 0.48 AzB6-2 1.05 AzG5-2 0.19 AzB3-2 0.30 AzF3-2 0.50 AzC4-2 0.62 AzH4-2 0.78 AzC1-2 0.68 AzG3-2 0.89 AzC5-2 0.02 AzH5-2 1.33 AzC2-2 0.45 AzH1-2 0.77 AzD4-2 0.18 AzH6-2 2.38 AzC3-2 0.36 AzH2-2 1.59 AzD6-2 0.37 Fig. 130. Bar graphs of thermal shift (ΔTm, °C) of fragments from azide library 2 in the presence of GlxII with a ΔTm greater that 1 °C. Significance of changes evaluated by a Student’s t-test (where ** p ≤ 0.01, * p ≤ 0.05). In summary, the fragments identified in subsequent DSF screens with GlxII to carry out further unbiased kTGS assays were: AlkC2-2, AlkH2-2, AzA2-2, AzH2-2 and AzH6-2 (Table 47). Table 47: Thermal melt temperatures (Tm, °C) with respect to GlxII of selected fragments. Alkyne fragments Structure ΔTm Azide fragments Structure ΔTm AlkC2-2 1.27 AzA2-2 1.33 AlkH2-2 1.16 AzH2-2 1.59 AzH6-2 2.38 4.5.2 DSF screening with GlxI For the identification of fragments for future rounds of in situ linking against GlxI, all azide and alkyne libraries were screened using DSF as previously described. 193 In the screen with alkyne library 1 and GlxI, the observed ΔTm ranged from -6.25 to 1.55 °C. Out of the 45 fragments within alkyne library 1, 33 demonstrated a stabilisation of the GlxI melt temperature by an increase in thermal shift ranging from 0.02 to 1.55 °C, as summarised in Table 48. Of those, 5 alkyne fragments with the largest ΔTm indicating binding to GlxI were: AlkA2, AlkC1, AlkE2, AlkG1 and AlkD4, as shown in Fig. 131. Fragments that had a higher than 1 °C thermal shift relative to standard deviation were AlkG1 (ΔTm 1.45 °C) and AlkD4 (ΔTm 1.55 °C), as highlighted in Table 48, and were therefore selected as fragments to initially test the unbiased kTGS with GlxI. For the full data set see Table 68 and Fig. 169 in the Appendix. Table 48: Positive ΔTm data of fragments from alkyne library 1 in the presence of GlxI. Fragment ΔTm Fragment ΔTm Fragment ΔTm Fragment ΔTm GSH 0.93 AlkD2 0.04 AlkA5 0.02 AlkF5 0.70 GAz (32) 0.96 AlkD3 0.18 AlkB4 0.28 AlkG4 0.10 AlkA1 0.02 AlkE1 0.41 AlkB5 0.37 AlkH4 0.65 AlkA2 1.13 AlkF1 0.49 AlkC4 0.37 Alk-1 0.08 AlkB2 0.55 AlkF2 0.97 AlkC5 0.25 Alk-2 0.25 AlkB3 0.24 AlkG1 1.45 AlkD4 1.55 Alk-5 0.48 AlkC1 1.15 AlkH2 0.58 AlkD5 0.57 Alk-6 0.43 AlkC2 0.54 AlkH3 0.63 AlkE4 0.58 Alk-7 0.15 AlkC3 0.18 AlkA4 0.06 AlkE5 0.26 Alk-8 -0.01 Fig. 131. Bar graphs of thermal shift (ΔTm, °C) of fragments from alkyne library 1 with a greater that 1 °C ΔTm against GlxI. Significance of changes evaluated by a Student’s t-test (where ** p ≤ 0.01, * p ≤ 0.05). In the DSF screen with alkyne library 2, the observed ΔTm ranged from -11.40 to 1.71 °C. Out of the 52 fragments within alkyne library 2, 36 demonstrated a stabilisation of the GlxI melt temperature by an increase in thermal shift ranging from 0.07 to 1.71°C, as summarised in Table 49. The 5 alkyne fragments with the highest ΔTm indicating binding to GlxI were: AlkB2-2, AlkF5-2, AlkF6-2, AlkH4-2 and AlkH6-2 (Fig. 132). Fragments that had a higher than 1 °C thermal shift relative to standard deviation were AlkB2-2 (ΔTm 1.71 °C), AlkF6-2 (ΔTm 1.39 °C) and AlkH4-2 (ΔTm 1.42 °C), as highlighted in bold in Table 49, and were therefore selected as fragments to initially take into the unbiased 194 kTGS with GlxI, as shown in Fig. 132. For the full data set see Table 69 and Fig. 170 in the Appendix. Table 49: Positive thermal shift data of fragments from alkyne library 2 against GlxI. Fragment ΔTm Fragment ΔTm Fragment ΔTm Fragment ΔTm GlxI 0.00 AlkE2-2 0.71 AlkB5-2 0.05 AlkE6-2 0.90 GSH 0.63 AlkE3-2 0.09 AlkB6-2 0.15 AlkF4-2 0.52 GAz (32) 1.50 AlkF1-2 0.56 AlkB7-2 0.14 AlkF5-2 1.09 AlkA1-2 0.09 AlkF3-2 0.46 AlkC4-2 0.05 AlkF6-2 1.39 AlkA2-2 0.96 AlkG1-2 0.46 AlkC6-2 0.69 AlkG5-2 0.78 AlkB2-2 1.71 AlkG3-2 0.91 AlkD4-2 0.25 AlkG6-2 0.59 AlkC1-2 0.12 AlkH1-2 0.18 AlkD6-2 0.43 AlkH4-2 1.42 AlkC3-2 0.90 AlkH3-2 0.43 AlkD7-2 0.50 AlkH5-2 0.55 AlkD1-2 0.07 AlkA4-2 0.60 AlkE4-2 0.86 AlkH6-2 1.01 AlkE1-2 0.42 AlkB4-2 0.04 AlkE5-2 0.84 Fig. 132. Bar graphs of thermal shift (ΔTm, °C) for selected fragments with ΔTm > 1 °C from alkyne library 2 in the presence of GlxI. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). In the DSF screen with azide library 1, the observed ΔTm ranged from -1.88 to 4.87 °C. Out of the 50 fragments within azide library 1, 19 demonstrated a stabilisation of the GlxI melt temperature by an increase in thermal shift ranging from 0.03 - 4.87 °C, as summarised in Table 50. The 5 alkyne fragments with the highest ΔTm indicating binding to GlxI were: AzB1, AzC3, AzD3, AzE2 and AzG2. The only fragment that had a higher than 1 °C thermal shift relative to standard deviation was AzB1 (ΔTm 4.87 °C), as highlighted in bold in Table 50, and was therefore selected as fragment to take into the unbiased kTGS with GlxI, as shown in Fig. 133. For the full data set see Table 70 and Fig. 171 in the Appendix. 195 Table 50: Positive thermal shift data of fragments from azide library 1 against GlxI. Fragment ΔTm Fragment ΔTm Fragment ΔTm GSH 0.92 AzC2 0.21 AzE3 0.25 GAz (32) 1.59 AzC3 0.96 AzF2 0.46 AzA1 0.02 AzD1 0.53 AzF3 0.05 AzA3 0.03 AzD2 0.57 AzG1 0.52 AzB1 4.87 AzD3 0.93 AzG2 0.96 AzB3 0.39 AzE1 0.17 AzG3 0.11 AzC1 0.60 AzE2 0.91 AzH1 0.11 Fig. 133. Bar graphs of thermal shift (ΔTm, °C) for fragments from azide library 1 with ΔTm > 1 °C in the presence of GlxI. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). In the DSF screen with azide library 2, the observed ΔTm ranged from -1.13 to 1.39 °C. Out of the 51 fragments within azide library 2, 16 demonstrated a stabilisation of the GlxI melt temperature by an increase in thermal shift ranging from 0.04 – 1. 39 °C, as summarised in Table 51. The 5 alkyne fragments with the highest ΔTm indicating binding to GlxI were: AzA2-2, AzC1-2, AzC3-2, AzH2-2 and AzD6-2. The only fragment that had a higher than 1 °C thermal shift relative to standard deviation was AzD6-2 (ΔTm 1.39 °C), as highlighted in bold in Table 51, and was therefore selected as fragment to take into the unbiased kTGS with GlxI, as shown in Fig. 134. For the full data set see Table 71 and Fig. 172 in the Appendix. Table 51: Positive thermal shift data of fragments from azide library 2 against GlxI. Fragment ΔTm Fragment ΔTm Fragment ΔTm GSH 0.16 AzC1-2 0.27 AzC6-2 0.12 GAz (32) 1.84 AzC2-2 0.26 AzD4-2 0.06 AzA2-2 0.52 AzC3-2 0.85 AzD6-2 1.39 AzA3-2 0.16 AzD2-2 0.10 AzE5-2 0.18 AzB2-2 0.26 AzE2-2 0.03 AzE6-2 0.12 AzB3-2 0.04 AzH2-2 0.58 AzF5-2 0.20 196 Fig. 134. Bar graphs of thermal shift (ΔTm, °C) for fragments from azide library 2 with ΔTm > 1 °C in the presence of GlxI. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). In summary, the DSF screens with GlxI identified AlkG1, AlkD4, AlkB2-2, AlkF6-2, AlkH4-2, AzB1 and AzD6-2 as fragments to undergo solubility testing to carry out further unbiased kTGS assays. The summary Table 52, shows the chemical diversity of the selected fragments where 3 compounds with flexible chains and a terminal carboxylic acid, which would be useful handles to develop any hits compounds further. Table 52: Thermal shift (ΔTm, °C) in the presence of the selected fragments relative to that of GlxI alone. Alkyne fragment Structure ΔTm Azide fragment Structure ΔTm AlkD4 1.55 AzB1 4.87 AlkG1 1.45 AzD6-2 1.39 AlkB2-2 1.71 AlkF6-2 1.39 AlkH4-2 1.42 4.5.3 Summary The DSF screens with GlxII and GlxI demonstrated the order of ranking showing the highest ΔTm and fragments with a higher than 1 °C thermal shift with respect to standard deviation identified several fragments to undergo solubility testing within an aqueous media to be subsequently used in the unbiased kTGS assay. 197 4.5.4 STD NMR As outlined in the proposed screening cascade for fragment identification prior to undertaking the unbiased kTGS, DSF analysis was carried out of all fragment libraries to ascertain the binding of the fragments to GlxI and II. A small selection of fragments with preferential stabilisation to GlxI and GlxII were selected to undergo further analysis, using STD-NMR, to confirm binding and allow the selection of preferable fragments to test the unbiased kTGS methodology.163 The STD-NMR experiments that were designed and setup in collaboration with the UCL School of Pharmacy Structural Chemistry Services did not provide any additional binding confirmation. This was due to the difficulty in setting up the STD-NMR parameters resulting in low signal to noise ratios. The assay conditions were as follows: ligand (1 mM concentration), protein (10 µM concentration), deuterated buffer 75 mM PBS pH 7.5, 0.15 M NaCl, 400 µL total volume. To reduce the signals observed from the protein storage buffers the protein was buffer exchanged into the deuterated buffer although this did not result in a STD-NMR trace that could confirm binding. Future work should include increasing the saturation time and number of scans to reduce the signal to noise ratio so as to use the STD NMR assay to confirm fragment hit selection. Therefore, at this time this assay was not used to preselect fragments to test the unbiased kTGS methodology. 4.5.5 Further investigation into the DSF conditions with Alkyne library 1 and GlxII To observe if a higher concentration of fragments would show greater differences in melt temperature a second screen was carried out with alkyne fragments at 5 mM final concentration with GlxII (10 µM concentration) (Fig. 135, Table 53, Appendix Fig. 163). Ligand concentration and thermal stability are usually proportional, so an increase in fragment concentration was anticipated to give an increased thermal shift.129 Alkyne fragments AlkD2, AlkC3, Alk-1, Alk-2, Alk-3 and Alk-4 demonstrated poor DSF melt curves. A negative thermal shift was observed for a number of fragments, indicating that the fragments were binding to the unfolded protein in preference to the folded protein.164 Solubility issues were observed for AlkF4, AlkC1, AlkE1, AlkF3, AlkG3 and AlkG4 by precipitation in the wells. ΔTm ranged from –31.1 – 3.11 °C. The 5 alkyne fragments that showed the greatest stabilisation of GlxII were: AlkE1, AlkG1, AlkH2, AlkB4 and AlkA2, highlighted in Fig. 135. The AlkA4 alkyne fragment also showed stabilisation and was the only alkyne that showed consistent preferential stabilisation with GlxII at both 1 mM and 5 mM. 198 Fig. 135. Bar graphs of thermal shift (ΔTm, °C) in the presence of fragments (5 mM) from alkyne library 1 with ΔTm >1 °C relative to that of GlxII alone. Significance of changes evaluated by a Student’s t-test (where ** p ≤ 0.01, * p ≤ 0.05). To exclude solubility issues as a problem for identifying fragments in an aqueous environment, the experiment was repeated in the presence of the detergent TWEEN20 at 0.01% v/v. In a screen with fragments at 1 mM final concentration both AlkF4 and AlkG4 were observed to have poor melt DSF curves and high initial background fluorescence resulting in only n = 1 to obtain the melt temperature. Alk-3 gave no result due to high initial background fluorescence and did not produce a sigmoidal DSF curve. ΔTm melt temperatures ranged from –21.8 to 3.99 °C. Of the 45 fragments within Alkyne library 1, 24 showed a stabilization of GlxII melt temperature by an increase in thermal shift. The 5 alkyne fragments that showed the highest stabilisation to GlxII were: AlkG4, AlkF3, AlkH2, AlkE3 and AlkC1 (Fig. 136, Table 53, Appendix Fig. 165). In a screen with fragments at 5 mM final concentration with 0.01% TWEEN20 only three fragments, AlkE3, AlkC5 and AlkE5, produced a sigmoidal DSF curve to enable the melt temperatures to be measured. AlkC5 and AlkE5 only gave n = 1 data due to high initial fluorescence and not demonstrating a sigmoidal DSF curve for the other repeats. ΔTm melt temperatures ranged from –9.3 to -0.8 °C (Table 53). Fig. 136. Bar graphs of thermal shift (ΔTm, °C) in the presence of fragments (1 mM + 0.01% TWEEN20) from alkyne library 1 with ΔTm >1 °C relative to that of GlxII alone. Significance of changes evaluated by a Student’s t-test (where * p ≤ 0.05). A comparison between GlxII and GlxII co-incubated with 0.01% TWEEN20 showed a higher melt temperature in the presence of TWEEN20, suggesting that 0.01% TWEEN20 199 has a stabilising effect on GlxII (Fig. 137). However, the addition of TWEEN20 was not used in subsequent DSF screens due to the problematic initial high fluorescence creating poor sigmoidal curves and therefore difficulty in measuring the melt temperatures. Fig. 137. Bar graphs of thermal melt temperature (Tm, °C) of GlxII with (green) and without 0.01%TWEEEN20 (olive). These additional 3 DSF screens of alkyne library 1 where the concentration of fragments was increased from 1 to 5 mM concentration, with and without 0.01% TWEEN20, demonstrated that the ideal conditions to screen all the fragment libraries was at 1 mM. This was due to the fragments demonstrating the highest stabilisation for binding to GlxII and demonstrating problematic sigmoidal DSF curves. Table 53 shows a summary of all the DSF results for alkyne library 1 with GlxII. However, it is important to note that the fragments identified in the 1 mM with 0.01% TWEEN20 DSF screen identified alkyne fragments that had also been identified as fragments hits in the biased kTGS due to the hit triazole formation with GAz (32) in GlxII. Therefore, it would be interesting for future work to include TWEEN20 in the DSF screen to observe if other fragments matched those identified in the biased kTGS experiments. Table 53. Thermal shift (ΔTm, °C) in the presence of fragments from alkyne library 1 at 1mM and 5 mM with and without 0.01% TWEEN20 with respect to GlxII alone.NR = no result, N/A = not tested. Fragment concentration 1 mM 1 mM + 0.01% TWEEN20 5 mM 5 mM + 0.01% TWEEN20 Fragment concentration 1 mM 1 mM + 0.01% TWEEN20 5 mM 5 mM + 0.01% TWEEN20 Fragment ΔTm Fragment ΔTm GlxII 0.00 0.00 0.00 0.00 AlkH3 0.50 0.40 0.93 NR GSH 2.01 1.36 -14.9 3.37 AlkA4 1.52 -0.02 1.52 NR AlkA1 0.66 -0.5 2.99 NR AlkB4 1.40 0.36 2.08 NR AlkB1 -0.72 -1.43 -5.31 NR AlkC4 0.31 0.80 1.47 NR AlkC1 0.32 1.02 -25.4 NR AlkD4 -0.91 0.80 -10.3 NR TmGlxIIGlxII_Tween206 76 86 97 0200 Fragment concentration 1 mM 1 mM + 0.01% TWEEN20 5 mM 5 mM + 0.01% TWEEN20 Fragment concentration 1 mM 1 mM + 0.01% TWEEN20 5 mM 5 mM + 0.01% TWEEN20 Fragment ΔTm Fragment ΔTm AlkD1 0.26 0.17 -18 NR AlkE4 0.84 -0.13 0.42 NR AlkE1 0.31 -0.19 19.43 NR AlkF4 -24.2 0.93 -25.1 NR AlkF1 0.43 0.27 0.50 NR AlkG4 -1.35 3.99 -26.1 NR AlkG1 0.28 -0.37 3.11 NR AlkH4 0.48 0.76 -6.63 NR AlkH1 -1.30 -0.17 -0.07 NR AlkA5 -0.76 -0.92 -0.91 NR AlkA2 1.01 0.58 1.61 NR AlkB5 2.01 0.01 0.73 NR AlkB2 0.46 0.24 0.27 NR AlkC5 0.05 -0.48 -2.86 -2.30 AlkC2 -0.1 0.73 -4.18 NR AlkD5 3.68 -0.36 1.29 NR AlkD2 0.90 0.42 NR NR AlkE5 1.99 -0.13 -0.08 -0.8 AlkF2 1.27 -0.01 -24.5 NR AlkF5 2.99 -0.10 -0.17 NR AlkG2 1.31 -1.05 -3.27 NR Alk-1 -0.42 -0.87 NR NR AlkH2 0.54 1.48 2.74 NR Alk-2 -0.35 -1.72 NR NR AlkA3 -0.17 -0.12 -2.47 NR Alk-3 -1.21 -21.8 NR NR AlkB3 0.61 0.05 -1.52 NR Alk-4 -0.08 -11.0 NR NR AlkC3 -0.83 -0.91 NR NR Alk-5 -0.03 0.27 -5.63 NR AlkD3 0.14 0.88 0.46 NR Alk-6 -0.22 1.34 -1.39 NR AlkE3 1.19 1.20 -3.5 -9.3 Alk-7 -0.19 0.39 -0.77 NR AlkF3 1.36 2.58 -31.1 NR Alk-8 -0.32 0.37 -1.17 NR AlkG3 1.26 0.55 -24.4 NR GAz (32) 1.7 N/A N/A N/A 4.6 Conclusions and future work Herein a promising proof of principle experiment has been presented with the initial indications showing potential for an unbiased approach to identify novel non-peptidic inhibitors of GlxII. From 46 alkynes and 50 azides giving a total of 190 potential triazoles (excluding regioisomers) over 4 unbiased kTGS, 7 hits were identified. Through biophysical evaluation these compounds have been clustered into priority groups, based on binding and inhibition data, for further development. This pilot unbiased kTGS has shown the impact this approach could have on FBDD in regards to one-pot fragment screening and in situ linking, thereby significantly reducing the time from identification of a fragment hits to development of a lead compound, as well as an approach towards overcoming the challenges associated with fragment 201 linking. The small nature of the unbiased hits allows for further optimisation using traditional medicinal chemistry approaches to develop a lead candidate. A comparison between the fragment hits identified in the biased and the unbiased kTGS identified further alkyne fragments to carry out additional unbiased experiments. Of note, a number of the alkyne fragments identified in the biased kTGS were identified in the pre-screening assay with DSF and with the addition of TWEEN20. For example, Alk6 and AlkE1 were identified as hit compounds in the biased kTGS (61 and 60) and as hit compounds in the unbiased (85 and 87) and all 4 compounds have shown good binding and inhibition data with the potential to develop into novel GlxII inhibitors. AlkE1 and Alk6 were also identified in the DSF pre-screening assay under the 5 mM and 1 mM+TWEEN20 conditions respectively. Therefore, future work could include using Alk6 and AlkE1 in unbiased kTGS experiments with the azide fragment library. Future work should include using AlkF3 to carry out an unbiased kTGS with the azide libraries as the biased compound, 56, showed the best combination of binding affinity and inhibition and was identified during the 1 mM+TWEEN20 DSF pre-screening assay. Other alkyne fragments identified in the DSF pre-screening that were also observed as hit alkynes in the biased kTGS were AlkE5 (1 mM and 5 mM+TWEEN20), AlkH2 (5 mM and 1 mM+TWEEN20) and AlkE3 (5 mM+TWEEN20). Combining the results from the biased kTGS could enhance the hits identified in the unbiased kTGS, in terms of using the alkyne fragments from the biased results to carry out further unbiased screens. Interestingly, no cross over alkynes were observed between the DSF pre-screens, at 1 mM fragment concentration, and the biased assay with GlxI. However, it would be interesting to observe if this cross over of fragments changes upon the addition of TWEEN20 to the DSF screen. Future work should involve carrying out the unbiased kTGS with the additional fragments identified in the preliminary DSF screens for both GlxI and GlxII as well as repeating the DSF screens with TWEEN20. Using the unbiased kTGS approach should enable the identification of novel small inhibitors for biological targets for which there are no known chemical probes would be a significant addition to the drug discovery tool box. One advantage of the unbiased kTGS is the potential for these small hits to have multiple binding sites thereby allowing further optimisation using traditional medicinal chemistry approaches to develop novel inhibitors. Another advantage is the ability of kTGS to link fragments in situ allowing the approach to be an asset to FBDD and fragment screening. A potential disadvantage is due to the small size of the fragments, and the resulting triazoles, they have a reduced 202 chance for binding interactions within the active site or may not be orientated in such a way as to be in close enough proximity to each other to react forming the linked triazole, resulting in a lower hit rate compared to the biased methodology. In order for this unbiased approach to become a valuable tool in the chemical biology and drug discovery tool box a few additions to the kTGS assay would be beneficial as part of any future work. As described in Chapter 3 and Fig. 108, orthogonal groupings of fragments would reduce false negatives as each fragment would have the opportunity to interact with the enzyme in different chemical environments. The addition of a synthesis plate enables the confirmation of hits as well as providing an indication if the 1,4 or 1,5-regioisomer was formed in the in situ experiments. For any future additions to the fragment libraries it should be remembered that the fragment libraries need to have sufficient diversity so that the same libraries can be used with a wide variety of biological targets. To enhance the fragment linking abilities and to widen the scope of fragment screening and hit identification using kTGS alternative chemistries should be explored. 203 Chapter 5: Fluorine NMR assay development 5.1 19F-NMR for ligand binding analysis In order to have a complimentary screening method for the confirmation of hits identified in the kTGS experiments, the development of a 19F-NMR assay was undertaken. This would allow evaluation of the binding of novel compounds to the glyoxalase proteins, and other glutathione binding proteins, in a straight forward assay as well as to confirm binding affinities of ligands observed in thermal shift and enzyme kinetics screens. Due to the challenges encountered in the enzyme kinetics assay, an additional confirmatory binding assay would give increased confidence in the hits. Competition binding experiments with a known glutathione inhibitor would give an indication of whether the novel inhibitors occupied the G or H sites. 5.1.1 Introduction to NMR for ligand binding Nuclear magnetic resonance (NMR) spectroscopy is a common used technique to detect the binding of compounds to protein targets. There are a variety of ligand or protein observed NMR methods, which are used to monitor ligand-protein interactions including: Saturation Transfer Difference (STD), which relies on NOE to measure magnetisation transfers from the protein to the bound ligand, and allows the measurement of interactions within large proteins; Water-Ligand Observed by Gradient SpectroscopY (WaterLOGSY), which uses the water molecules in the ligand-protein binding site to transfer the magnetisation from the waters to the protein and then to the ligand resulting in stronger intensity signals for the bound ligand signals than the non-bound ligand signals. Competition experiments can be carried out in the presence of a known inhibitor of the target protein, which allows confirmation of the binding site of the ligand.163 An advantage of NMR based methods for the measurement of ligand-protein interactions is the wide affinity range of which the interactions can be measured, from µM to mM concentrations.165 Protein observed NMR methods also allow the observation of fully bound states where there is a ratio of 1:1 ligand to protein. Ligand observed NMR methods have a limit of detection where the dissociation constant ranges from 10 to 100 µM, however this can be overcome by an increase in the sensitivity of the NMR machines. Disadvantages include the need for a second experiment to be included as a reference which increases the time required to carry out the experiments and an increase in difficulty when moving from a one ligand-one protein system.163 204 19F-NMR spectroscopy is a well-established ligand observed technique and is used to detect binding of small molecule ligands containing fluorine atoms to protein targets. 19F-NMR spectroscopy has several advantages over 1H-NMR spectroscopy namely: 19F has a natural abundance of 100% and a wide chemical shift range (-250 - 550 ppm) compared to 1H which has an abundance of 99.9% and a narrower chemical shift range (-4 - 14 ppm);166-167 19F does not occur naturally in proteins and so the method is very sensitive and easier to interpret as there is no background noise associated with components of the buffers or protein;167,168 due to the high sensitivity of this screening method very low protein concentrations can be used even for low affinity ligands.169 The incorporation of a fluorine atom into a ligand is relatively easy due to a large number of commercially available building blocks containing fluorine. Fluorine is an established bioisosteric replacement for hydrogen, due to its similar size, which can mimic the sterics of a hydrogen atom in the binding site of the protein. Fluorine can also have a beneficial influence on the pharmacokinetic properties, such as solubility, which influences absorption rates and transport properties of the compound as well as blocking sites of metabolism. The lipophilicity of a fluorine atom is slightly higher compared to that of a hydrogen atom, and the lipophilicity of a trifluoromethyl group is greater than that of a bromine atom. Fluorine is more electronegative than hydrogen, which in turn affects the physical properties of the ligand as well as the chemical reactivity.170 As the binding of a ligand to a protein is an equilibrium process the association and dissociation rates can be calculated.167 The dissociation constant (Kd) is used to describe the affinity between the ligand and a protein, the higher the affinity the tighter the binding, which results in a more stable complex between ligand and protein. Measured Kd values for the ligands would allow ranking in terms of affinity to the protein and can be carried out by running the NMR without the presence of the protein and at different concentrations in the presence of the protein. When a fluorinated ligand binds to a protein the chemical environment, including electric fields and hydrogen bonds surrounding the 19F nucleus, changes compared to its unbound state. NMR monitoring of this environmental change results in a shift in shielding and relaxation rates in regards to the bound fluorine ligand moving at a slower rate and chemical exchange. This reduction of the relaxation rate (T2) results in linewidth broadening of the NMR resonance compared to the unbound fluorine ligand167 and/or a change in chemical shift (Fig. 138).166 205 Fig. 138. Schematic demonstrating a non-bound and bound F-analogue resulting in peak broadening, loss of signal or peak shift. 5.1.2 Glutathione S-transferase as a model system To develop the 19F-NMR assay glutathione S-transferase (GST) from Schistosoma japonicum was used as a model system due to the ease of protein production and the relatively conserved binding properties amongst GSH binding proteins. GST contributes to phase II metabolism, by enzymatically conjugating glutathione (GSH) to a wide range of metabolites and toxins increasing their water solubility and enabling excretion, in a process known as xenobiotic metabolism.30, 171 GST from Schistosoma japonicum (Sj26GST) is a fusion tag commonly used in recombinant protein production. The pGEX-2T plasmid was developed as a gene fusion expression system that could be readily purified by glutathione affinity columns, and contains the 217 residues and an additional 9-residue linker at the C-terminus containing a thrombin cleavage site. The pGEX gene fusion expression systems were developed to aid the over-expression, solubility and purification of Sj26GST in E.coli systems.65, 172, 173 GST’s exist as both homodimers and heterodimers with an active site in each monomer. The N-terminal consists of an α/β domain whilst the C-terminal is primarily a α-helical. The G-site, in the N-terminal, and the H-site, in the C-terminal, are both required for GSH conjugation with electrophilic species (Fig. 139).171 The GSH tripeptide backbone is aligned in the G-site by the interactions between the γ-glutamate and Gln67 and the glycine carboxylate to Trp8. The G-site is found to be conserved throughout the GST class but the H site varies and provides an opportunity for substrate specificity.174 The interactions of the cysteine sulfhydryl moiety with Tyr111 and Phe7 are essential for enzymatic activity.175 The residue at position 7, whether it is a tyrosine or phenylalanine 206 plays an important role in GSH binding. The hydroxyl group on Tyr111 forms a conserved hydrogen bond between the residue and the thiol of GSH. Fig. 139. Pymol representation of GST from Schistosoma Japonicum complexed with glutathione (PDB: 1U87) bound in the G-site.174 The purpose of this study was to investigate if the binding of 19F containing glutathione S-conjugates to GST could be detected using 19F-NMR and to carry out competition experiments with novel inhibitors to provide a qualitative and quantitative measure of binding. The ultimate aim was to apply these experiments to GlxI and GlxII to gain confidence in the hits identified in the kTGS assays described in Chapters 3 and 4. 5.1.3 Protein expression and purification optimisation GST The E. coli expression strain BL21(DE3) was transformed with pGEX-2T. Transformed cells were grown (LB supplemented with 50 μg/L ampicillin) to an optical density (OD600nm) of 0.70 at 37 °C and induced with ITPG (1.0 mM) for 4 hours at 25 °C. Cell lysate was prepared by sonication and the addition of 1% Triton-X100. The GST protein was purified by a glutathione sepharose 4B column (2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, 140 mM NaCl, pH 7.3) and eluted with a gradient of glutathione (50 mM Tris, pH 8, 120 mM NaCl, 20 – 100 mM reduced L-glutathione), followed by buffer exchange (10 mM PBS, pH 7.4, 200 mM NaCl).176,106 Under these conditions purified GST was obtained in yields of 30.27 mg/ pellet batch. Fractions were analysed with SDS-PAGE were GST was identified as a band migrating approximately the same distance as the 28 kDa marker (Carbonic Anhydrase) in the protein standard ladder (Fig. 140). 207 Fig. 140. SDS-PAGE Gel showing fractions containing glutathione S-transferase after purification and buffer exchange. The purification protocol underwent optimisation by the addition of a gradient elution buffer. In the initial purification, the protein was eluted very slowly from the column, with 20 mM reduced-L-glutathione, over 25 factions. In the second attempt, the elution buffers were optimised by increasing the gradient concentration of reduced-L-glutathione from 20 mM to 100 mM, which reduced the protein elution from 25 fractions to 7. This optimisation improved the time required for purification whilst maintaining a comparable yield, 34.5 mg/ pellet to 30.27 mg/ pellet. A sample was submitted for MALDI-TOF analysis where the mass expected was calculated using UniProt and ExPASy123 to be 25498.6 Dalton. The observed mass was found to be 26795.5 – 28659.6 Da with a centroid mass of 27826.2 Da. Machine error is 0.05%, 26782.1 – 28673.9 Da. In order to carry out a sequence comparison the plasmid (pGEX-2T) was transformed into competent E. coli DH5α, primers pGEX forward and reverse were used. The sequence was carried out by the DNA sequencing and services at University of Dundee. The full DNA sequence is in the Appendix, Table 72. Sequence comparisons and alignments were produced by 4peaks, CLC sequence viewer and BLAST NCBI database. The DNA sequence was 99% (forward) and 100% (reverse) identical at the nucleotide level to pGEX-2T cloning vector. 5.1.4 Ligand design and synthesis Glutathione S-conjugates were identified as a starting point for ligand design due to the conserved GSH backbone to ensure recognition by the GST and glyoxalase enzymes.171 Modification of the Cys thiol with fluorine containing substrates was an essential design element for the 19F-NMR experiments to observe the 19F peak in the NMR spectra and to detect binding. In order to explore size, potential interactions and SAR around the H-site a F atom and CF3 group were used to replace the proton in all positions around the aromatic ring, as well as one example where the sulfur was removed (Fig. 141). To 208 establish the competition assay with a known ligand, the glutathione S-conjugate, S-(4-bromo-benzyl) glutathione (3), developed by Thornalley et al.114 was selected. Fig. 141. Example of fluorine analogues (4-10, 17). The desired chemical tools were synthesised as described in section 2.1.3. 5.1.5 Preliminary binding experiments 5.1.5.1 DSF screening results To get an initial evaluation of the binding of the fluorine analogues to GST DSF experiments were carried out. The DSF results with GlxI and GlxII are discussed in section 2.3.4. Based on the DSF conditions for GlxI and II the following conditions were used for GST. 10 μM GST protein concentration, 5 x SYPRO orange dye, 1 mM ligand, 40 μL total volume and the buffer used was 10 mM PBS, pH 7.4, 0.2 M NaCl for glutathione S-transferase. The assay was run at 25 °C for 30 sec, step and hold with a 1 °C increase to 95 °C, then held for 30 sec. In the presence of 1 mM 19F-ligands and GST (10 μM) a ΔTm of between 0.67 to 11.60 °C was observed, which indicated binding. No clear SAR could be discerned for the glutathione S-conjugates tested where a ΔTm of between 8.88 and 11.60 °C was observed. However, for the cysteine replacement analogue (17) a ΔTm of 0.67 °C was measured, suggesting that the SH is required for binding (Fig. 142, Table 54). The data for 17 is consistent with the function of GST as the thiol is required for conjugation and therefore recognition and selectivity. The competitive inhibitor (3) showed equivalent stabilization to the fluorine analogues, suggesting similar affinity. 209 Fig. 142. Bar graphs of thermal shift (ΔTm, °C) in the presence of fluorine containing ligands, 3 and GSH relative to GST alone. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001). Table 54. Thermal shift in the presence of fluorine containing ligands, 3 and GSH relative to GST alone (ΔTm, °C). GST GST ID Structure ΔTm (°C) ID Structure ΔTm (°C) GSH 1.77 7 11.60 3 11.84 8 8.88 4 9.09 9 9.33 5 10.76 10 11.39 6 11.10 17 0.67 210 5.1.5.2 Preliminary 19F NMR experiments Preliminary 19F-NMR binding experiments167,168 were completed for a small library of fluorinated ligands with GST to evaluate if this assay could be used as a complimentary screening method, and these results are described below. To determine the limit of detection, 19F-NMR experiments were run at different concentrations of 19F chemical probes (5, 10, 20, 50 & 100 μM) with TFA as an internal standard. These studies determined that in order to collect data of sufficient quality the minimum concentration for trifluoromethyl analogues and the mono- or di- fluoro analogues is 10 μM and 50 μM respectively, where all experiments were run in a 1:1 ratio of ligand:protein. With limits of detection established for the CF3 analogues, S-(2-trifluoromethybenzyl)-glutathione (7), S-(3-trifluoromethybenzyl)-glutathione (8), S-(4-trifluoromethybenzyl)-glutathione (9) the 19F NMR’s were run in PBS buffer at 10 μM of the ligand. The resulting CF3 peak was observed at a chemical shift of -58.72 ppm, -61.01 ppm and -60.79 ppm respectively (Fig. 143). When the 19F NMR’s were run in the presence of the protein GST at a 1:1 ratio (10 μM:10 μM), the CF3 signal was not observed for any of the ligands tested. These results suggest that each of the ligands tested bound to the protein and the difference in chemical environment for the 19F nucleus in the ligands upon protein binding resulted in a loss or broadening of CF3 peak. To explore the usefulness of these results further 19F NMR’s were run with a 10-fold excess (100 μM) of a competitive inhibitor, S-(4-bromobenzyl)-glutathione (3), which resulted in the return of the CF3 peak in the spectrum, suggesting that the known inhibitor, S-(4-bromobenzyl)-glutathione (3), could displace the fluorinated ligands and therefore bound to the protein in preference over the fluorinated ligands when present at a 10-fold excess. This data also confirmed the CF3 analogues were bound in the G and H-sites as anticipated. The NMR data correlated with the DSF data confirming the CF3 analogues did bind to GST and that 3 could compete out the ligands from the GST binding site. Moving onto the F analogues, S-(2-fluorobenzyl)-glutathione (4), S-(3-fluorobenzyl)-glutathione (5), S-(4-fluorobenzyl)-glutathione (6, provided by Dr Jess Healy), the 19F NMR’s were run in PBS buffer at 50 μM ligand concentration. The resulting F peaks were observed at a chemical shift of -116.43 ppm, -113.40 ppm and -115.69 ppm respectively (Fig. 143). When the 19F NMR of the fluorinated ligands were run in the presence of the protein GST in a 1:1 ratio (50 μM:50 μM), the F signal was not observed for S-(2-211 fluorobenzyl)-glutathione (4) and S-(3-fluorobenzyl)-glutathione (5) and a shift of -3.98 to –119.67 ppm was observed for S-(4-fluorobenzyl)-glutathione (6). These results suggest that each of the ligands tested bound to the protein and due to the difference in chemical environment for the 19F nucleus in the ligands upon protein binding loss or broadening or a shift of F peak was observed. Further 19F NMR’s were run with a 10-fold excess (500 μM) of a competitive inhibitor, S-(4-bromobenzyl)-glutathione (3), which resulted in the return of the F peak being observed, suggesting that the known inhibitor, S-(4-bromobenzyl)-glutathione, could displace the fluorinated ligands and therefore bound to the protein in preference over the fluorinated ligands at a 10-fold excess. This data once again confirmed the F analogues were bound in the G and H-sites as anticipated and that the NMR data correlated with the DSF data confirming the F analogues did bind to GST and that 3 could compete the ligands out of the GST binding site. Fig. 143. 19F NMR for fluorine ligands (4, 5, 6, 7, 8, 9) showing (A) F-ligand, protein and competitive inhibitor (3) (1:1:10) (B) F-ligand and protein (1:1) and (C) the F-ligand in buffer. 212 Finally, for the fluorinated ligands, S-(3,5-di-fluorobenzyl)-glutathione (10) and γ-glutamyl-4-fluoro-L-phenylalaninylglycine (17) the 19F NMR were run in PBS buffer at 50 μM ligand concentration. The resulting F peak was observed at a chemical shift of -110.29 ppm and -116.40 ppm respectively (Fig. 144). When the 19F NMR of the fluorinated ligands were run in the presence of the protein GST in a 1:1 ratio (50 μM:50 μM), the F signal for S-(3,5-di-fluorobenzyl)-glutathione (10) was still observed, however, the integration had decreased. This NMR was repeated and the same result observed. This suggests that the S-(3,5-di-fluorobenzyl)-glutathione (10), does bind to GST but has a weaker affinity than previous ligands. Of note is that this is the only analogue that does not correlate with the DSF data. When the 19F NMR of the fluorinated ligand, γ-glutamyl-4-fluoro-L-phenylalaninylglycine (17), was run in the presence of the protein GST in a 1:1 ratio (50 μM:50 μM), no signal was observed suggesting that the ligand had bound to the protein. Further 19F NMR’s were run with a 10-fold excess (500 μM) of a competitive inhibitor, S-(4-bromobenzyl)-glutathione (3), which resulted in the return of the F peak being observed, suggesting that the known inhibitor, S-(4-bromobenzyl)-glutathione (3), could displace the fluorinated ligands and was therefore bound to the protein in preference over the fluorinated ligands at a 10-fold excess. This NMR data correlated with the DSF data where, γ-glutamyl-4-fluoro-L-phenylalaninylglycine (17) was bound to GST in the G and H-site. Fig. 144. 19F NMR for fluorine ligands (10, 17) showing (A) F-ligand, protein and competitive inhibitor (3) (1:1:10) (B) F-ligand and protein (1:1) and (C) the F-ligand in buffer. 5.1.6 Quantitative Determinations To determine the affinity of the ligands for GST, dissociation constants (Kd) were measured using titration experiments. The ligand concentration [L] was varied whilst the 213 protein concentration [P] kept constant to determine the integration of bound and unbound ligand to allow quantification of the ligand and protein-ligand concentrations [PL]. The mole fraction of the bound ligand was expected to range from 0 to 1 during the titration. GraphPad Prism was used to calculate the dissociation constant from a plot of percentage ligand bound to ratio of [L]/[P].163 The assay conditions were as follows: GST concentration was kept constant at 100 µM and ligand concentration ranged from 10 to 1000 µM. As the limit of detection was lower for the CF3 probes, the 2-CF3 chemical probe (7) was used to run the initial titration experiments. A few difficulties were encountered during the titration NMR’s specifically: the volume of solvent required per NMR tube was a minimum of 0.4 mL, therefore a large quantity of protein was required; the lower ligand concentration samples did not result in the same sensitivity as had been expected after the qualitative experiments as a result it was not possible to obtain three replicates for all concentration points as the NMR’s did not provide sufficient quality readouts to calculate the bound fraction. Further optimisation is required in order to measure the bound mole fraction more accurately either by improving the sensitivity of the spectra or by using competition ligands. The data obtained with the 2CF3 ligand (7) resulted in a Kd of 2.2 µM being calculated (Fig. 145 and 146, Table 55), however this should be viewed as preliminary data and future work should include repeating these titration experiments on the 700 MHz NMR which is currently being installed at UCL, Department of Chemistry. 214 Fig. 145. 19F NMR for the titration spectra for S-(2-trifluoromethybenzyl)-glutathione (7). Table 55. Average mole fraction bound % for S-(2-trifluoromethybenzyl)-glutathione (7) [P] µM [L] µM Average mole fraction bound n 100 10 100.00 1 100 20 100.00 1 100 50 86.28 3 100 75 78.92 1 100 100 62.40 2 100 150 2.45 1 100 200 15.29 1 100 400 0.00 1 100 500 7.45 1 100 1000 0.00 1 Fig. 146. [L]/[P] vs % bound, where [P] is 100 μM. Kd 2.2 μM for 2-CF3. 5.1.7 Discussion All the synthesised 19F glutathione conjugates were screened against purified proteins GST, GlxI and GlxII. The synthesised peptide analogues screened against GST showed no clear SAR was observed with the different positions of the F or CF3 around the benzyl 215 ring, but removal of the cysteine thiol showed decreased stabilization of the GST in the DSF assay. No clear SAR was observed with the peptide analogues against GlxII, suggesting the tripeptide backbone has more of an effect on the binding of the analogues in the G-site than the substitutions on the cysteine sulfur in the H-site. The 4-CF3 (9) and 3,5-diF (10) synthesised peptides provided the highest stabilisation of the GlxI melt temperature, with the remaining glutathione S-conjugates also binding to GlxI, and the removal the thiol functionality with (17) showed a destabilisation, indicating the thiol is essential for binding. The 19F-NMR’s of the fluorine ligands and GST have demonstrated that this methodology could be employed to detect ligands binding via competition 19F probes and have application as a complimentary screening tool. Due to the relatively flat structure activity relationships with the F and CF3 functional groups and the limit of detection being lower for the CF3 groups, this resulted in the CF3 ligands being utilised to optimise the titration experiment in order to calculate the Kd’s in GST. Initial titration 19F-NMR experiments could provide a binding curve and a Kd calculation however, further optimisation is required in order to measure the unbound mole fraction more accurately either by improving the sensitivity of the spectra or by using competition ligands. 5.2 Application of the 19F NMR to the kTGS hits The qualitative and competition 19F NMR experiments were applied to the kTGS hits containing a fluorine moiety in order to gain confidence in these hits identified in the kTGS assays described in Chapters 3 and 4. 60 contained a 3,5-difluorophenyl moiety and was identified in the biased kTGS screen with GlxII. 72 contained a 3-fluorophenyl and was identified in the biased kTGS screen with GlxI. Of the six synthesised compounds identified in the unbiased kTGS in GlxII two included F atoms, 84 contained a 3-fluorphenyl moiety and 85 a 3,5-difluoro moiety. A summary of the DSF and % inhibition data is shown in Table 56, where 60 was observed to have a ΔTm of -3.61 °C. As 72 and 84 had been identified as hits to be developed further as GlxI and GlxII inhibitors, respectively in Chapter 3 and 4, this 19F NMR experiment anticipated to be used as a confirmation of binding and ultimately proof of the usefulness of 19F NMR as a complimentary screening method. 216 Table 56. Overall summary of data for F-analogues including: DSF, % inhibition at 412 nm and from specific activity at 240nm. DSF % Inhibition 412 nm Specific Activity U/mg % Inhibition 240 nm Compound ID Protein ΔTm 10 µM 100 µM 10 µM 100 µM 10 µM 100 µM 60 GlxII -3.61 50.66 57.33 1.10 0.78 5.48 21.30 72 GlxI 1.20 N/A N/A 2.05 3.41 44.49 7.74 84 GlxII 1.58 30.56 41.90 4.26 0.89 21.6 76.99 85 GlxII 0.65 -28.0 37.78 4.24 1.53 21.85 50.71 The 19F-NMR’s were set up as previously described. However, due to initial complications involving TFA, which was used for reference purposes, binding to GlxII the concentration of ligand to protein was increased from a 1:1 ratio to 6.8:1 and the F ligand to competitive inhibitor (3) ratio was 1:10 to evaluate 60 with GlxII. It was observed that the integration of the F peaks doubled in the presence of GlxII indicating that TFA bound to GlxII in preference to 60. The known inhibitor, 3, displaced the TFA returning the integration of both TFA and the F peaks on 60 to the starting integration as observed in the buffer NMR. This 19F-NMR result did confirm the DSF result where GlxII showed a negative thermal shift of -3.61 °C in the presence of 60. Of note is that the triazole derived from AlkE1 had been predicted to be a 1,5-triazole in the kTGS LCMS overlay suggesting the usefulness of the synthesis kTGS in hit identification (Fig. 147). 217 Fig. 147. 19F NMR for 60 showing (A) the di-F-ligand in buffer (100 μM) (B) di-F-ligand, GlxII and competitive inhibitor (100 μM: 14.5 μM: 1000 μM) (C) di-F-ligand and GlxII (10 μM: 14.5 μM). To evaluate 72 with GlxI in the 19F-NMR, the ligand to protein ratio was 2:1 and the F ligand to competitive inhibitor ratio was 1:10. Through the integration of the F peak, with reference to TFA, it was calculated that 72 was 76% bound to GlxI. The known inhibitor, 3, could partially displace the ligand with 26% remaining bound, thus indicating the triazole was bound in the G- and H-site. This 19F-NMR result confirmed the DSF results, where a thermal shift of 1.20 °C was observed for GlxI upon co-incubation with 72, the inhibition studies, where 72 inhibited the GlxI activity by 44% at 10 µM, and the docking studies, where the tripeptide backbone was predicted to bind in the G-site (Fig. 148). Fig. 148. 19F NMR for 72 showing (A) the F-ligand in buffer (50 μM) (B) F-ligand, GlxI and competitive inhibitor (50 μM: 25 μM: 500 μM) (C) F-ligand and GlxI (50 μM: 25 μM). 218 To evaluate the unbiased fluorinated analogues (84 and 85), the ligands were prepared in 0.1 M PBS, pH 7.6 with 5% DMSO-d6 at 50 and 100 µM with TFA as a reference. The NMR’s were run with and without GlxII protein at 10 µM (Fig. 149 and 150). The ligand to protein ratio was adjusted in this assay due to preliminary NMR’s indicating TFA was binding to the GlxII protein in preference to the ligands. To evaluate the binding of 84 to GlxII, the ligand to protein ratio was 5:1 and the F ligand to competitive inhibitor ratio was 1:10. Through the integration of the F peak, with reference to TFA, it was calculated that 84 was bound to GlxII. However, due to the small nature of this ligand and the excess ligand to protein ratio, it was possible the ligand had multiple binding sites, therefore through the integration of the fluorine signal it is was calculated that 325% of 84 was bound to GlxII. The known inhibitor, 3, could displace 200% of the ligand with 125% remaining bound, thus indicating the majority of the triazole was bound in the G- and H-site. The change in integration of the 19F signal does confirm the ligand was bound to GlxII and the competitive inhibitor did displace the ligand confirming it had been bound in the G- and H-site. This 19F-NMR result confirmed the DSF results, where a thermal shift of 1.58 °C was observed for GlxII upon co-incubation with 84, and the inhibition studies, where 84 inhibited the GlxII activity by 20 – 76% (Fig. 149). Fig. 149. 19F NMR for 84 showing (A) F-ligand and GlxI (50 μM: 10 μM) (B) F-ligand, GlxI and competitive inhibitor (50 μM: 10 μM: 500 μM) (C) the F-ligand in buffer (50 μM). To evaluate the binding of 85 to GlxII, the ligand to protein ratio was 10:1 and the F ligand to competitive inhibitor ratio was 1:10. Through the integration of the F peak, with 219 reference to TFA, it was observed that TFA bound in preference to 85, and that 85 was not bound to GlxII. The 19F-NMR in the presence of the known inhibitor, 3, showed that 85 was bound in preference to GlxII. This 19F-NMR result confirmed previous NMR’s that the binding of 85 to GlxII could not be evaluated with this assay (Fig. 150). The DSF data for 85 was 0.65 °C indicating it was bound with weak affinity to GlxII, therefore the 19F-NMR could either not be sensitive enough to confirm binding for this ligand or the di-fluorophenyl moiety was not tolerated by GlxII as observed with 60. Fig. 150. 19F NMR for 85 showing (A) F-ligand and GlxI (100 μM: 10 μM) (B) F-ligand, GlxI and competitive inhibitor (100 μM: 10 μM: 1000 μM) (C) the F-ligand in buffer (100 μM). In conclusion, the binding of ligands containing a 3-fluorophenyl moiety (72 and 84) was evaluated with GlxI and GlxII, respectively. The 3,5-difluorophenyl moiety (60 and 85) however could not be evaluated with GlxII, either due to the ligands lack of binding to the protein or an unexpected problem with the 3,5-difluorophenyl group. These results have indicated the usefulness of 19F-NMR as a complimentary screening assay to confirm binding observed in the DSF screens as well as confirmation of the synthesis kTGS as a positive control in the kTGS experiments in predicting regioisomers. The addition of a competitive inhibitor allowed the confirmation of the site of binding for the triazoles within the G- and H-sites. Future work should include further optimisation of the titration experiments to allow for the calculation of dissociation constants. 220 Chapter 6: Conclusions The main aim of this project was to identify novel, potent and selective inhibitors of GlxI and GlxII using kinetic target guided synthesis. A scaffold for kTGS optimisation, γ-glutamyl-L-azidoalaninylglycine (GAz, 32), was designed and synthesised based on the glutathione tripeptide backbone. The incorporation of an azide functionality allowed GAz to act as a handle in the glutathione binding site whilst using a diverse set of alkyne fragments to rapidly screen the hydrophobic binding site in a biased kTGS. Through a series of biased kTGS experiments a robust and reliable methodology for sample handling and analysis was developed. Overall from 97 alkynes, an average hit rate of 17% was observed with GlxII. Three compounds were identified, 56, 55 and 64, as having potential to be developed further due to a combination of binding affinity as observed in the DSF screens and two enzyme kinetics assays providing an indication of % inhibition of GlxII (Fig. 151). 56 55 64 DSF ΔTm, °C 2.28 1.03 1.55 % Inhibition 412 nm 10 µM 52.31 21.62 31.28 100 µM 55.04 40.98 30.07 % Inhibition 240 nm 10 µM 33.73 29.15 15.87 100 µM 32.45 30.86 19.45 Fig. 151. Structure and data summary of hit compounds selected for further development targeting GlxII. Overall from 97 alkynes an average hit rate of 16% was observed with GlxI. Four compounds, 71, 72, 73 and 74, were identified as having potential to be developed further due to binding affinities as observed in the DSF screens and an enzyme kinetics assay which provided an indication of % inhibition (Fig. 152). 221 71 72 73 74 DSF ΔTm, °C 1.28 1.20 2.43 4.49 % Inhibition 240 nm 10 µM 41.13 44.49 87.69 ǂ 100 µM 79.26 7.74 ǂ -31.54 Fig. 152. Structure and data of hit compounds selected for further development targeting GlxI. ǂ Data not shown. The biased kTGS allowed for the identification of novel triazole containing mimetics based on glutathione S-conjugates which showed binding to and inhibition of the bacterial glyoxalases. To build upon the biased approach and to explore the limits of kTGS, the methodology was expanded to establish if kTGS can be used in an unbiased fragment based approach to identify small, novel and non-peptidic inhibitors. The hypothesis was to utilise two fragments that could link in situ to identify novel lead compounds. Preliminary fragment screening using DSF allowed for the identification of a number of alkyne and azide fragments which bound to GlxII. Two azide and two alkyne fragments were selected for use in the proof of principle experiment to explore the unbiased kTGS approach. Four pilot experiments were undertaken using 45 alkyne fragments and 50 azide fragments. From a total of 190 potential products, excluding regioisomers, 7 triazoles were identified which corresponded to a 3% hit rate over the 4 unbiased kTGS experiments. Six of the 1,4-triazoles were successfully synthesised alongside one 1,5-triazole analogue. A comparison of all the biophysical data allowed the 7 synthesised triazoles to be clustered into three groups determining the order of ranking for further development (Fig. 153). Priority group 1 consisted of 87 and 84, due to the combination of binding affinity and inhibition in both assays at 10 and 100 µM. Priority group 2 222 consisted of 85, 83 and 90, due to binding affinity and some inhibition being observed. Priority group 3 consisted of 89 and 86 where binding was not detected using DSF, however inhibition was observed at 100 µM. In terms of future development of these scaffolds into lead compounds for GlxII inhibitors, 87, 84, 89 and 86 contain handles which could be used to grow and develop the ligands to improve affinity and selectivity. Given the small size of these hits they are a good start for a traditional medicinal chemistry program for the development of GlxII inhibitors into novel antibacterials. 87 84 DSF ΔTm, °C 1.61 1.58 % Inhibition 412 nm 10 µM 76.28 30.56 100 µM 21.62 41.90 % Inhibition 240 nm 10 µM 75.31 21.6 100 µM 75.71 76.99 85 83 90 DSF ΔTm, °C 0.65 0.49 2.56 % Inhibition 412 nm 10 µM -28.02 42.94 17.13 100 µM 37.78 49.49 1.67 % Inhibition 240 nm 10 µM 21.85 33.96 60.79 100 µM 50.71 -5.90 -57.78 86 89 DSF ΔTm, °C -0.41 -0.86 % Inhibition 412 nm 10 µM -1.04 21.72 100 µM 40.44 45.93 % Inhibition 240 nm 10 µM -50.79 74.52 100 µM 79.64 -3.87 Fig. 153. Unbiased hit compounds prioritised for further development targeting GlxII. 223 To support compound profiling and provide an assay which can qualitatively and quantitatively confirm the hit compounds a 19F-NMR assay was developed using synthesised 19F containing glutathione S-conjugates. S-(4-bromobenzyl)-glutathione, was used as a competitive inhibitor to determine if the ligands were bound in the G-site. Evaluation of the 19F containing triazoles demonstrated that 72 was 76% bound to GlxI and could be partially displaced by the competitive inhibitor indicating it was bound in the G-site. 84 was bound to GlxII and could be partially displaced by the competitive inhibitor indicating it was bound in the G-site. However, binding of 85 and 60 to GlxII was not observed in the 19F-NMR assay, and in the case of 60 was a confirmation of the DSF results. In conclusion, the successful identification of peptidic and non-peptidic inhibitors of GlxI and GlxII using a biased and unbiased kTGS approach has been achieved. The biased kTGS allowed for the optimisation of a screening cascade for future experiments and the unbiased kTGS experiments have shown promising proof of principle indications as a fragment screening and in situ linking method to allow for the identification of novel inhibitors. This study allowed the optimisation of kTGS conditions, synthesis conditions, DSF and enzyme kinetics assays. Future work will include the medicinal chemistry optimisation of the identified lead compounds as inhibitors of GlxI and GlxII and evaluation of their potential as novel antibacterials agents. 224 Chapter 7: Experimental 7.1 Biological materials and methods 7.1.1 Chemicals and Buffers All chemical reagents were purchased from Fisher Scientific, Sigma Aldrich UK, VWR or Life Sciences unless otherwise stated. SDS PAGE gel electrophoresis was carried out using NOVEX™ precast gels, buffers and reagents from Invitrogen Ltd. Instant Blue was obtained from Expedeon through Fisher Scientific. All solutions for growth were autoclaved (Priorclave Front Loading Autoclave machine) and stored at room temperature unless otherwise stated. A Beckman Coulter Avanti J-E Centrifuge was used. 7.1.2 Growth media All growth media were prepared using d-water and autoclaved to sterilize. LB Media was supplied by Sigma Aldrich UK and contained: 10 g/L Tryptone 5 g/L Yeast extract 5 g/L NaCl LB media: LB (10 g) in water (500 mL). LB-agar: as above, with the addition of agar (7.5 g) per 500 mL of solution LB-agar Ampicillin: as for LB-agar above. Once cooled, ampicillin (50 mg/mL) added per 500 mL, to a final concentration of 100 μg/mL. TB Media was supplied by Merck and contained: 12.0 g/L peptone from casein 24.0 g/L Yeast extract 9.4 g/L di-potassium hydrogen phosphate 2.2 g/L potassium dihydrogen phosphate TB media: TB (23.8 g) in water (500 mL) containing 2 mL glycerol as per manufacturer’s instructions. 225 Ampicillin: Stock solutions (100 mg/mL) were prepared as follows: Ampicillin (0.1 g) was dissolved in d-water (1 mL), filter sterilized (Whatman 0.45 µm) and stored in 1 mL aliquots at -20 °C. Stock solutions (50 mg/mL) were prepared as follows: Ampicillin (0.05 g) was dissolved in d-water (1 mL), filter sterilized (Whatman 0.45 µm) and stored in 1 mL aliquots at -20 °C. Chloramphenicol Stock solutions (25 mg/mL) were prepared as follows: Chloramphenicol (0.025 g) was dissolved in d-water (1 mL), filter sterilized (Whatman 0.45 µm) and stored in 1 mL aliquots at -20 °C. Nickel chloride Stock solutions (100 mM) were prepared as follows: nickel chloride (0.129 g) was dissolved in d-water (10 mL) and filter sterilised (Whatman 0.45 μm), transferred to 1.5 mL Eppendorf tubes in 10 x 1 mL aliquots and stored at -20 °C. IPTG: Stock solutions (100 mM) were prepared as follows: Isopropyl-β-ᴅ-1-thiogalactopyranoside (IPTG) (0.238 g) was dissolved in d-water (10 mL) and filter sterilised (Whatman 0.45 μm), transferred to 1.5 mL Eppendorf tubes in 10 x 1 mL aliquots and stored at -20 °C. 7.1.3 Plasmids The plasmids used are listed in Table 57. Plasmid DNA was stored at -20 °C or -80 °C as required by the supplier. Table 57. Plasmid details. Plasmid Description of protein(s) encoded Source pET3a-gloA Synthetic clone; optimised for E. coli expression; N-terminal T7 tag; C-terminal His tag; amp resistance Dr Samantha Miller, University of Aberdeen pTrcgloBH6 Synthetic clone; optimised for E. coli expression in place of MscS (pMscS used for cloning, MscS replaced by cloning into NcoI and XhoI cut pMscSH6); Insert is N-terminal His-tagged; amp resistance Dr Samantha Miller, University of Aberdeen pGEX-2T GST fusion vector with Ptac promoter, lacI repressor, thrombin cleavage and kinase site ORFs; amp resistance VWR 226 7.1.4 Strains The strains listed below in Table 58 were stored at -80 °C as required by the supplier. Table 58. Bacterial strain details. Strain Description Source BL21(DE3) E. coli strain F-, ompT, hsdS(r-,m), gal, dcm(DE3) Invitrogen BL21(DE3)pLysS E. coli strain F-, ompT, hsdSB (rB- mB-), gal, dcm(DE3), pLysS(CamR) Invitrogen DH5α E. coli strain F-, φ80lacZ∆M15 ∆(lacZYA-argF)U169 recA1 endA1 hsdR17(rk-, mk+)phoA supE44 thi-1 gyrA96 relA1 tonA (confers resistance to phage T1) Prof Frank Kozielski, UCL School of Pharmacy BW25113, E. coli K-12 Host Keio collection, E. coli strain rrnB3 ΔlacZ4787 hsdR514 Δ(araBAD)567 Δ(rhaBAD)568 rph-1, Kan resistance Dharmacon JW1643, gloA knockout Keio collection. E. coli strain National BioResource Project (NIG, Japan): E. coli JW0202, gloB knock out Keio collection. E. coli strain National BioResource Project (NIG, Japan): E. coli 7.1.5 Primers The primers listed below in Table 59 were stored at -20 °C as required by the supplier. Table 59. Primer details. Primer Desc i tion (5’ to 3’) Source pGEX forward GGGCTGGCAAGCCACGTTTGGTG, 3' end of glutathione-S-transferase, forward primer Sigma Aldrich pGEX reverse CCGGGAGCTGCATGTGTCAGAGG 3' of MCS in pGEX vectors, reverse primer Sigma Aldrich T7 TAATACGACTCACTATAGGG T7 promoter, forward primer Sigma Aldrich pET TAATACGACTCACTATAGG pET expression vetor primer, T7 promoter Sigma Aldrich pTrcHis Forward GAGGTATATATTAATGTATCG (Invitrogen) 5’ of MCS in pTrcHis vector, forward primer Sigma Aldrich pTrcHis Reverse GATTTAATCTGTATCAGG (Invitrogen) 3’ of MCS in pTrcHis vector, same as pBAD-R, reverse primer Sigma Aldrich 227 7.2 Plasmid techniques 7.2.1 Transformation of competent cells BL21(DE3), BL21(DE3)pLysS and DH5α competent cells were thawed on ice. The cells were mixed to ensure an even suspension prior to pipetting. 50 μL was aliquoted into sterile Eppendorf tubes (1 tube/transformation and 1 negative control). The Eppendorf tubes were left on ice for 10 min. 1 μL of the desired plasmid was added and the tubes left on ice for 30 min, followed by heat shock for 20 sec at 42 °C. 150 μL of prewarmed filter sterilized LB was added to each Eppendorf tube. The tubes were incubated with shaking (200 rpm) at 37 °C for 1 h. The cells were mixed carefully and 200 μL pipetted onto LB/Agar plates containing the appropriate antibiotic. 100 μL of the negative control containing BL21(DE3), BL21(DE3)pLysS or DH5α and LB was pipetted onto LB/Agar plates containing the appropriate antibiotic and 100 μL onto LB/Agar plate with no antibiotic. The transformed cells were spread onto the agar plates and incubated at 37 °C for 16 h. To purify the transformants a single colony was picked and spread onto an LB/Agar plate containing the appropriate antibiotic and incubated at 37 °C for 16 h. The agar plates were sealed with parafilm and stored at 4 °C for 2 - 3 weeks. 7.2.2 Mini plasmid purification Mini plasmid purification was carried out using the Illustra Plasmid Prep Mini Spin kit (GE Healthcare). Overnight cultures (5 mL LB containing the appropriate antibiotic) of transformed strains containing the expression plasmids of interest were incubated with shaking (200 rpm) at 37 °C for 16 h. 1 mL from each falcon tube was pipetted into a 1.5 mL Eppendorf tube and centrifuged at 11000 rpm for 30 sec. The supernatant was removed taking care not to disturb the pellet. This process was repeated until a total of 3 mL of the overnight culture had been centrifuged. Lysis buffer type 7 (175 μL) was added and the pellet re-suspended. Lysis buffer type 8 (175 μL) was added and mixed gently by inversion (repeated 5 times). Lysis buffer type 9 (350 μL) was added and mixed gently by inversion. Tubes were centrifuged at 11000 rpm for 4 min (or until pellet had formed). The supernatant was loaded onto a spin column in a micro-centrifuge tube and centrifuged for 30 sec. The flow through was discarded by emptying the collection tube, whilst the plasmid remained bound to the column. Lysis buffer type 9 (400 μL) was added and centrifuged at 11000 rpm for 30 sec. The flow through was discarded. Wash buffer type 1 (400 μL) was added to the column and centrifuged for 1 min. The flow through was discarded. The tube was re-centrifuged for 30 sec and the flow through discarded. The mini column was transferred to a new micro-centrifuge tube and elution buffer type 228 4 (50 μL) added. The micro-centrifuge tubes were incubated for 60 sec then centrifuged for 30 sec. The flow through contained the plasmid of interest. The elution process was repeated. Isolated plasmid was stored in Eppendorf tubes in 10 μL aliquots at -80 °C. The concentration of the isolated plasmids of interest were measured by UV Spectroscopy at an absorbance of 260 nm, where concentration (µg/mL) = A260nm x dilution factor x 50 µg/mL. Dilution factor is 100 if the absorbance of 10 µL purified plasmid in 990 µL water is measured in a quartz cuvette. The purity of the isolated plasmid DNA should be in the range of 1.8 – 2.0 and is calculated from A260nm/A280nm. 7.2.3 Midi plasmid purification Midi plasmid purification was carried out using the Qiagen Midi Prep kit (Qiagen). Overnight cultures of transformed strains containing the expression plasmids of interest were incubated with shaking (200 rpm) at 37 °C for 16 h in 200 mL LB containing the appropriate antibiotic. Cells were harvested by centrifugation (3000 rpm, 4 °C, 10 min) and the supernatant was discarded. The pellet was resuspended in buffer P1 (6 mL). Buffer P2 (6 mL) was added and mixed by inversion 6 times followed by incubation for 15 min at room temperature. Chilled buffer P3 (6 mL) was added to the lysate and mixed by inversion 6 times. The lysate was poured into the barrel of the QIAfilter cartridge and incubated for 10 min. To equilibrate a HiSpeed midi Tip, Buffer QBT (4 mL) was added and allowed to empty under gravity. The cap was removed from the QIAfilter outlet nozzle and the precipitate was scraped out. The plunge was inserted into the filter and the lysate filtered into the HiSpeed Tip where the cleared lysate was allowed to enter the resin by gravity. The HiSpeed Midi Tip was washed by Buffer QC (20 mL). Into a sterile falcon 3.5 mL isopropanol was added and the DNA was eluted with Buffer QF (5 mL) and mixed gently followed by incubation for 5 min at room temperature. A QIA precipitator was placed over a wash bottle and the DNA/isopropanol mixture transferred into a 20 mL syringe and filtered under constant pressure. The precipitator was removed and the plunger pulled out, the precipitator was reattached and 70% ethanol / water (2 mL) added. The DNA was washed by reinserting the plunger and pressing with constant pressure. The precipitator was removed and the plunger pulled out. The precipitator was reattached, followed by inserting the plunger and pressing through the air. This was repeated twice to dry the outlet nozzle and to prevent ethanol carry over. The plunger was removed from a new syringe and the precipitator attached. TE buffer (1.5 mL) was added and eluted into a sterile Eppendorf. The eluent was transferred into the same syringe and eluted a second time. An agarose gel was carried out to confirm the identity 229 and purity of the DNA. The concentration of the purified plasmids were measured by UV Spectroscopy (260 nm), where concentration (µg/mL) = A260nm x dilution factor x 50 µg/mL. Dilution factor is 100 if the absorbance of 10 µL of purified plasmid in 990 µL water is measured in a quartz cuvette. The purity should be in a range of 1.8 – 2.0 and is calculated from A260nm/A280nm.The DNA was stored at -20 °C. 7.2.4 Agarose gel To prepare the agarose gel, agarose (2 g) was added to water (200 mL). The DNA sample was prepared as follows: DNA (30 µL) and binding buffer (15 µL) (TAE buffer, 20 µL EDTA and 980 µL water). The gel was run by electrophoresis at 175 V, 400 mAmp and 100 W. A DNA ladder (10 µL) was used to monitor the gel. The gel was stained for 15 min using TAE buffer and ethidium bromide followed by destaining for 10 min using water. The gel was visualised by placing on an ultraviolet transilluminator. 7.2.5 Glycerol stocks Overnight cultures (5 mL LB containing the appropriate antibiotic) of transformed strains containing the expression plasmids of interest were incubated with shaking (200 rpm) at 37 °C for 16 h. Overnight cultures of LB/BL21(DE3)/plasmid mixture (700 μL) or LB/BL21(DE3)pLysS /plasmid mixture (700 μL) and sterile 80% glycerol in water (300 μL) were pipetted into sterile 1.5 mL Eppendorf tubes, flash frozen and stored at -80 °C. 7.2.6 Long-term storage Glycerol stocks transformed with the plasmid of interest in an appropriate E. coli strain were flash frozen before storage at -80 °C. The cells were recovered by streaking onto LB agar (containing the appropriate antibiotic) and incubated at 37 °C for 16 h. 7.2.7 Short-term storage The strains recovered from -80 °C and grown on LB agar (containing the appropriate antibiotic) were stored at 4 °C for 2 - 3 weeks. 7.3 Protein Purification Techniques 7.3.1 Growth and over expression of proteins Overnight cultures (5 mL LB containing the appropriate antibiotic) of transformed strains containing the expression plasmids of interest were incubated with shaking (200 rpm) at 37 °C for 16 h. The overnight culture was added to LB (500 mL or 1 L), containing the appropriate antibiotics and additive, and grown to an OD600 0.6 – 0.7. IPTG (final 230 concentration 0.5 – 1.0 mM) was added and the cells were incubated with shaking at 25 °C for 4 h for GlxII and GST and 18 °C for 17 h for GlxI. The cells were harvested by centrifugation at 5000 rpm for 15 min at 4 °C. The supernatant was discarded and the pellet suspended in Lysis Buffer and stored in Falcon tubes at -80 °C or re-centrifuged at 4000 rpm for 10 min and the supernatant discarded and the pellet stored at -80 °C. A summary of the growth and expression conditions for GlxI, GlxII and GST are shown in Table 60. Table 60. Summary of expression conditions. Protein name GlxI GlxII GST Media LB LB LB Antibiotic Amp 50 μg/L Chloramp 25 μg/L Amp 50 μg/L Amp 50 μg/L Additives Ni(II)Cl2 1 mM N/A N/A Cell growth temperature 37 °C 37 °C 37 °C OD600 nm 0.5 – 0.7 0.7 0.7 Cell growth time ~ 5 h ~ 4 h ~ 4 h IPTG concentration 0.4 mM 0.5 mM 1.0 mM Expression temperature 18 °C 25 °C 25 °C Expression time 17 h 4 h 4 h 7.3.2 Protein Purification 7.3.2.1 Purification Buffers Sodium chloride Stock solutions (3 M) were prepared by dissolving sodium chloride (8.77 g) in d-water (500 mL) and stored at 4 °C. Calcium chloride Stock solutions (50 mM) were prepared by dissolving calcium chloride (1.1 g) in d-water (200 mL) and stored at 4 °C. Potassium chloride Stock solutions (100 mM and 5 M) were prepared by dissolving potassium chloride (1.1 g) in d-water (200 mL) and potassium chloride (74.55 g) in d-water (200 mL) and stored at 4 °C. Imidazole Stock solutions, 0.2 M and 1 M respectively, were prepared by dissolving imidazole 231 (2.72 g) in d-water (200 mL) and imidazole (13.61 g) in d-water (200 mL) and stored at 4 °C. Sodium phosphate monobasic dihydrate Stock solutions (0.2 M) were prepared by dissolving sodium phosphate monobasic dihydrate (15.60 g) in d-water (500 mL) and stored at 4 °C. Sodium phosphate dibasic heptahydrate Stock solutions (0.2 M) were prepared by dissolving sodium phosphate dibasic heptahydrate (53.65 g) in d-water (1 L) and stored at 4 °C. Tris buffer pH 8 Stock solutions (0.5 M) were prepared by dissolving Tris (24.23 g) in d-water (400 mL). HCl (conc.) was added until a pH of 8 was reached and stored at 4 °C. PMSF Stock solutions (100 mM) were prepared in ethanol by dissolving phenylmethane sulfonyl fluoride (0.174 g) in ethanol (10 mL). 1mL aliquots were stored at -20 °C. Hen egg lysozyme: Stock solutions (10 mM) were prepared by dissolving lysozyme (0.14 g) in d-water (1 mL) and stored at -20 °C. GlxI Purification Buffers Tris Lysis or Binding Buffer 50 mL volumes of this solution were prepared from stock solutions on the day of use. 2 mL 0.5 M Tris (pH 8) to give a final concentration of 20 mM 8.3 mL 3 M NaCl to give a final concentration of 500 mM The volume was adjusted to 50 mL with d-water. Tris Wash Buffer 50 mL volumes of this solution were prepared from stock solutions on the day of use. 2 mL 0.5 M Tris (pH 8) to give a final concentration of 20 mM 8.3 mL 3 M NaCl to give a final concentration of 500 mM 1 mL of 0.5 M Imidazole to give a final concentration of 10 mM The volume was adjusted to 50 mL with d-water. Tris Elution Buffer 1 50 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 232 0.4 mL 0.5 M Tris (pH 8) to give a final concentration of 20 mM 1.6 mL 3 M NaCl to give a final concentration of 500 mM 1 mL of 0.5 M Imidazole to give a final concentration of 50 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 2 100 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 0.4 mL 0.5 M Tris (pH 8) to give a final concentration of 20 mM 1.6 mL 3 M NaCl to give a final concentration of 500 mM 2 mL of 0.5 M Imidazole to give a final concentration of 100 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 3 200 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 0.4 mL 0.5 M Tris (pH 8) to give a final concentration of 20 mM 1.6 mL 3 M NaCl to give a final concentration of 500 mM 4 mL of 0.5 M Imidazole to give a final concentration of 200 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 4 300 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 0.4 mL 0.5 M Tris (pH 8) to give a final concentration of 20 mM 1.6 mL 3 M NaCl to give a final concentration of 500 mM 6 mL of 0.5 M Imidazole to give a final concentration of 300 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 5 400 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 0.4 mL 0.5 M Tris (pH 8) to give a final concentration of 20 mM 1.6 mL 3 M NaCl to give a final concentration of 500 mM 8 mL of 0.5 M Imidazole to give a final concentration of 400 mM The volume was adjusted to 10 mL with d-water. Buffer Exchange 50 mL volumes of this solution were prepared from stock solutions on the day of use. 5 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 1.6 mL 3 M NaCl to give a final concentration of 100 mM The volume was adjusted to 50 mL with d-water. 233 GlxII Purification Buffers Tris Lysis Buffer 50 mL volumes of this solution were prepared from stock solutions on the day of use. 5 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 5 mL 3 M NaCl to give a final concentration of 300 mM The volume was adjusted to 50 mL with d-water. Tris Wash Buffer 50 mL volumes of this solution were prepared from stock solutions on the day of use. 5 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 5 mL 3 M NaCl to give a final concentration of 300 mM 0.5 mL 1 M Imidazole to give a final concentration of 10 mM The volume was adjusted to 50 mL with d-water. Tris Elution Buffer 1 25 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 1 mL 3 M NaCl to give a final concentration of 300 mM 0.25 mL 1 M Imidazole to give a final concentration of 25 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 2 50 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 1 mL 3 M NaCl to give a final concentration of 300 mM 0.5 mL 1 M Imidazole to give a final concentration of 50 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 3 75 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 1 mL 3 M NaCl to give a final concentration of 300 mM 0.75 mL 1 M Imidazole to give a final concentration of 75 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 4 100 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 234 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 1 mL 3 M NaCl to give a final concentration of 300 mM 1 mL 1 M Imidazole to give a final concentration of 100 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 5 150 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 1 mL 3 M NaCl to give a final concentration of 300 mM 1.5 mL 1 M Imidazole to give a final concentration of 150 mM The volume was adjusted to 10 mL with d-water. Tris Elution Buffer 6 300 mM Imidazole 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 1 mL 3 M NaCl to give a final concentration of 300 mM 3 mL 1 M Imidazole to give a final concentration of 300 mM The volume was adjusted to 10 mL with d-water. Buffer Exchange 50 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M MOPS to give a final concentration of 10 mM 3.3 mL 3 M NaCl to give a final concentration of 200 mM The volume was adjusted to 50 mL with d-water and pH adjusted to 7.2 with 1 M NaOH. GST Purification Buffers Binding Buffer 50 mL volumes of this solution were prepared from stock solutions on the day of use. 2.3 mL 3 M NaCl to give a final concentration of 140 mM 1.3 mL 0.1 M KCl to give a final concentration of 2.7 mM 2.5 mL 0.2 M Na2HPO4 to give a final concentration of 10 mM 0.09 mL 1 M KH2PO4 to give a final concentration of 1.8 mM The volume was adjusted to 50 mL with d-water and pH adjusted to 7.3. Elution Buffer 1 20 mM GSH 10 mL volumes of this solution were prepared from stock solutions on the day of use. 235 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 0.4 mL 3 M NaCl to give a final concentration of 120 mM 0.4 mL 0.5 M reduced L-glutathione to give a final concentration of 20 mM The volume was adjusted to 10 mL with d-water. Elution Buffer 2 40 mM GSH 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 0.4 mL 3 M NaCl to give a final concentration of 120 mM 0.8 mL 0.5 M reduced L-glutathione to give a final concentration of 40 mM The volume was adjusted to 10 mL with d-water. Elution Buffer 3 60 mM GSH 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 0.4 mL 3 M NaCl to give a final concentration of 120 mM 1.2 mL 0.5 M reduced L-glutathione to give a final concentration of 60 mM The volume was adjusted to 10 mL with d-water. Elution Buffer 4 100 mM GSH 10 mL volumes of this solution were prepared from stock solutions on the day of use. 1 mL 0.5 M Tris (pH 8) to give a final concentration of 50 mM 0.4 mL 3 M NaCl to give a final concentration of 120 mM 2 mL 0.5 M reduced L-glutathione to give a final concentration of 100 mM The volume was adjusted to 10 mL with d-water. Buffer Exchange 50 mL volumes of this solution were prepared from stock solutions on the day of use. 5 mL 0.1 M PBS, pH 7.4 to give a final concentration of 10 mM 3.3 mL 3 M NaCl to give a final concentration of 200 mM The volume was adjusted to 50 mL with d-water. 7.3.2.2 Protein Purification Glyoxalase I Frozen cells (stored at -80 °C) were thawed on ice. Lysis buffer (60 mL) was added and the pellet resuspended. The cells were lysed using sonication (10 x 1 min on 30 sec off at 16 Amp) followed by centrifugation at 18000 rpm for 30 min at 4 °C. The resulting 236 supernatant was collected and filtered (Whatman filter, 0.45 μm), loaded onto Ni-NTA agarose gel (1 mL) in a 10 mL polypropylene column. The flow through was collected and the column washed with binding buffer (25 mL) then wash buffer (25 mL). To elute the protein, the column was treated with 4 mL of each of the outlined elution buffers and 1 mL fractions collected. SDS PAGE gel electrophoresis was employed to determine the purity of each fraction and the Bradford assay was used to calculate the concentration of the fractions containing protein. For protein fractions the buffer was exchanged for protein fractions using PD-10 columns eluting with the buffer exchange buffer and 1 mL fractions were collected. SDS PAGE gel and Bradford assay were used to identify the purity and concentration of fractions containing protein. The sample was aliquoted (100 μL) into 0.5 mL Eppendorf’s with 10% glycerol and stored at -80 °C. Glyoxalase II Frozen cells (stored at -80 °C) were thawed on ice. Lysis buffer (60 mL) and PMSF (final concentration 1 mM) were added and the pellet resuspended. The cells were lysed using sonication (10 x1 min on 30 sec off at 16 Amp) followed by centrifugation at 18000 rpm for 30 min at 4 °C. The resulting supernatant was collected and filtered (Whatman filter, 0.45 μm), loaded onto Ni-NTA agarose gel (2 mL) in a 10 mL polypropylene column. The flow through was collected and the column washed with wash buffer (30 mL). To elute the protein, the column was treated with 4 mL of each of the outlined elution buffers and 1 mL fractions collected. SDS PAGE gel electrophoresis was employed to determine the purity of each fraction and the Bradford assay was used to calculate the concentration of the fractions containing protein. The buffer was exchanged using PD-10 columns eluting with the buffer exchange buffer and 1 mL fractions were collected. SDS PAGE gel and the Bradford assay were used to identify the purity and concentration of the fractions containing protein. The sample was aliquoted (100 μL) into 0.5 mL Eppendorf’s with 10% glycerol and stored at -80 °C. Glutathione S-transferase Frozen cells (stored at -80 °C) were thawed on ice. Lysis Buffer (60 mL) and 1% Triton-X100 were added and the pellet resuspended. The cells were lysed using sonication (10 x 1 min on 30 sec off at 16 Amp) followed by centrifugation at 5000 rpm for 20 min at 4 °C. The resulting supernatant was collected and filtered (Whatman filter, 0.45 μm). A glutathione sepharose 4B column (2 mL bed volume), was prepared using a 10 mL polypropylene column, followed by washing with binding buffer (5 x column volume). The 237 supernatant was loaded onto the column and flow through was collected. The gel was washed with binding buffer (5 x column volume), this step was repeated three times. To elute the protein, the column was treated with elution buffer (0.5 mL per mL of bed volume) and incubated at room temperature for 10 min prior to collection, this process was repeated three times for each elution buffer until all the protein was eluted. SDS PAGE gel electrophoresis was employed to determine the purity of each fraction and the Bradford assay was used to calculate the concentration of the fractions containing protein. The buffer was exchanged using PD-10 columns eluting with the buffer exchange buffer and 1 mL fractions collected. SDS PAGE gel and the Bradford assay were used to identify the purity and concentration of the fractions containing protein. The sample was stored in 100 μL aliquots with 10% glycerol at -80 °C. 7.3.2.3 Analytical Protein Techniques 7.3.2.3.1 SDS-PAGE Gel SDS PAGE Running Buffer: 30 mL NuPAGE® Novex ® MOPS SDS Running Buffer (20 X) (Invitrogen). The final volume was adjusted to 600 mL with d-water. SDS loading Buffer: NuPAGE® Novex ® Sample Buffer(4 X) NuPAGE® Reducing Agent (10 X) SDS gel electrophoresis was carried out using NuPAGE® Novex ® Bis-Tris gradient gels (4-12%) (Invitrogen). The sample to be analysed (7 μL) was mixed with NuPAGE® Sample Buffer (5 μL) and NuPAGE® Reducing Agent (2 μL) and heated at 72 °C for 10 min. The chambers of the gel tank were filled with the prepared NuPAGE Running Buffer. After heating, the samples were loaded onto the gel along with a protein marker (SeeBlue Plus 2 prestained standard, Invitrogen). The gel was run for 30-45 min at 200 V and 150 mA. After the run was completed the gel was opened with a metal spatula, placed in a plastic tray and washed with water (x 3). The protein was visualised using Instant Blue (Expedeon) and destained with water. 7.3.2.4 Spectroscopic Techniques 7.3.2.4.1 UV/Visible Absorbance Spectroscopy Absorbance spectroscopy of samples in the UV/Vis range (260 - 650 nm) were carried out using a Jenway 6305 Spectrophotometer. The scans were performed in 3 mL or 1 mL plastic cuvettes. 238 7.3.2.4.2 Bradford Assay to calculate protein concentration From the absorbance of a sample in Quickstart Bradford 1 X Dye (BioRad) the protein concentration could be determined using the following calculation: A*16/V = mg/mL where A = Absorbance at 595 nm V = 2 for the 2 μL sample in 1 mL of 1 X Bradford Dye Using the molecular weight of the protein obtained from the sequence using ExPASy ProtParam tool the concentration of protein (molar) could be calculated. 7.3.2.4.3 MALDI-TOF The molecular mass of the protein was determined using an Applied Biosystems Voyager-DE Pro (MALDI-TOF) machine at UCL, SOP. 10% aqueous solution of TFA was prepared. The following solutions were prepared in 1 mL Eppendorf’s. Solution Composition Acetonitrile Vo me (μL) Water o me (μL) 10% TFA o me (μL) Total volume (μL) 1 Hydration 50:50 ACN:H2O:0.1% TFA 500 490 10 1000 2 Wash 0.1% TFA in water 0 990 10 1000 3 Protein Elution 75:25 ACN:H2O:0.1% TFA 750 240 10 1000 Prepare 3 Eppendorf tubes for: Tube Waste Wash Protein sample Contents Empty ~ 60 μL 1.3 μL A ZipTip (C18) was placed on a P10 pipette. For hydration, 10 μL of 1 was aspirated, then discarded to waste (x 2) followed by 2 repetitions of 10 μL of 2 being aspirated then discarded. The sample was loaded by aspirating 10 μL of protein sample and expelling back into the liquid tube, repeating 6 times. The tip was then washed by aspirating 10 μL of solution 2 and expelling into the waste, repeating 10 times. The sample was eluted by aspiration and expelling 1.3 μL of 3 twice into the protein sample tube. The molecular weight was calculated using UniProt and ExPASy and compared with the analytical results observed. 239 7.3.2.4.4 Trypsin Digest Solution preparation: Trypsin working solution: 5 µL trypsin stock was added to 45 µL d-water. Destaining solution: 80 mg ammonium bicarbonate, 20 mL acetonitrile, 20 mL d-water Digestion buffer: 10 mg ammonium bicarbonate, 5 mL water. Final concentration 25 mM. Reducing Buffer: 3.3 µL TCEP, 30 µL digestion buffer. Alkylation buffer: 7 mg IAA, 70 µL water gives a 500 mM stock solution. 7 µL 500 mM stock added to 28 µL digestion buffer. Activated trypsin: 1 µL trypsin working solution added to 9 µL digestion buffer. Band Preparation: A scalpel was used to cut out the bands of interest, from the SDS-PAGE gel. Digestion solution (200 µL) was added and incubated at 37 °C with shaking for 30 min. The solution was decanted and the process repeated. Reduction and alkylation: Reducing buffer (30 µL) was added to the sample and incubated at 60 °C for 10 min. The solution was cooled then discarded. Alkylation buffer (30 µL) was added and incubated at RT for 1 hour in the dark. The solution was discarded and the sample washed with destaining solution (200 µL). The sample was incubated at 37 °C for 15 min with shaking after which time the solution was decanted the step repeated. Digestion: Acetonitrile (50 µL) was added and the samples incubated for 15 min at room temperature. The solution was removed and the gel air-dried for 10 min. Activated trypsin (10 µL) was added and incubated at room temperature for 15 min. Finally, digestion buffer (25 µL) was added and the sample incubated at 37 °C for 4 h with shaking. The digestion mixture was removed and transferred to a new Eppendorf for submission for analysis by MALDI-TOF. 7.4 Biophysical and biochemical assays 7.4.1 Differential scanning fluorimetry (Thermal shift) All analyses were performed on a 7500 Real time PCR machine. Life Technologies software was used to set up the run on 7500 Software v.2.0.0 and analysis was carried 240 out on Protein thermal shift solution software. The mid temperature of the melt curve was determined using excel calculations and fitted using the Boltzman equation using GraphPad Prism.129 The thermal shift was calculated as the difference between protein control melt temperatures and protein containing ligand samples. Each run was performed as follows: samples were held at 25 °C for 30 sec, step and hold with a 1 °C increase to 95 °C, then held for 30 sec. The optimal assay conditions for each protein was optimised and summarised below. Glyoxalase II: 10 μM protein, 5 x SYPRO orange dye, 1 mM ligand, 40 μL total volume and the buffer used was 10 mM MOPS, pH 7.2, 0.2 M NaCl. Glyoxalase I: 10 μM GlxI protein concentration, 5 x SYPRO orange dye, 1 mM ligand, 40 μL total volume and the buffer used was 10 mM HEPES, pH 7.5, 0.15 M NaCl. Glutathione S-transferase: 10 μM GST protein concentration, 5 x SYPRO orange dye, 1 mM ligand, 40 μL total volume and the buffer used was 10 mM PBS, pH 7.4, 0.2 M NaCl. 7.4.2 Enzyme kinetics All analyses were performed on a SPECTROstar Omega UV/Vis spectrum absorbance spectrometer (BMG labtech) at wavelength 412 or 240 nm at 37 °C in 96-well half area microclear plates (Greiner BIO-One Ltd). Data analysis was performed using MARS data analysis software. The slope, Km and Vmax were determined using GraphPad Prism and used to evaluate the effect of ligand on the enzyme activity. Inhibition or effect on relative GlxI/II activity was calculated as: slope (ligand) / slope (GlxI/II + 2% DMSO). Plots of OD (240nm) vs time graphs were used to calculate the specific activity of GlxI/II using Beer’s law. Ligands were evaluated at two fixed concentrations, 10 and 100 µM, respectively. Glyoxalase II: Elman’s reagent (5,5'-dithio-bis-(2-nitrobenzoic acid) (DTNB)) was employed to measure the production of GSH when GlxII hydrolyses SLG to GSH. The relative activity of GlxII was calculated from the slope of the plot of absorbance (OD) at 412 nm vs time. The optimal conditions were: 0.2 μM Glyoxalase II protein, 0.4 mM S-lactoyl glutathione, 0.2 mM 5,5'-dithio-bis-(2-nitrobenzoic acid) (DTNB), 100 μL total volume and the buffer used was 40 mM PBS, pH 7.5, 0.26 M NaCl and 2% DMSO. OD412nm was measured for 40 min. 241 An alternative assay in which the decrease in SLG concentration as GlxII hydrolyses the SLG to GSH by a decrease in OD (240 nm) was employed. The relative activity of GlxII was calculated from the slope of the plot of absorbance (OD) at 240 nm vs time. The OD vs time graphs at 240 nm were also used to calculate the specific activity of GlxII using Beer’s law. The optimal conditions were: 2 µL Glyoxalase II protein (final concentration 0.2 μM), 10 µL 3 mM S-lactoyl glutathione, 38 µL d-water, 50 µL 100 mM Tris pH 7.4 buffer preheated at 37 °C and 2 µL ligand with 2% DMSO (final volume %) to give a final concentration of 10 or 100 µM. OD240nm was measured for 20 min. Glyoxalase I: The enzyme kinetics assay was employed to monitor the increase in SLG concentration as GlxI catalyses the formation of SLG from MG and GSH by an increase in OD (240 nm). The relative activity of GlxI was calculated from the slope of the plot of absorbance (OD) at 240 nm vs time. The OD vs time graphs at 240 nm were also used to calculate the specific activity of GlxI using Beer’s law. The optimal conditions were: 50 µL 100 mM Tris pH 6.6 buffer, 10 µL of 20 mM glutathione, 10 µL of 20 mM methylglyoxal and 28 µL water. The mixture was preheated for 10 min at 37 °C. 2 µL ligand, with 2% DMSO (final volume %), to give a final concentration of 10 or 100 µM and 2 µL Glyoxalase I protein (final concentration 0.2 μM) were added and OD240nm was measured for 20 min. The OD vs time graphs at 240 nm were used to calculate the specific activity of GlxII and GlxI using Beer’s law. Beers law = A = (ε)(l)(c), where A = absorbance (OD240 nm), ε = molar extinction coefficient (GlxII = -3.1 mM-1.cm-1, GlxI = 2.86 mM-1.cm-1), l = path length (1 cm) and c = concentration. Therefore, ΔA240 = (ΔC)(ε) Activity = ΔC = ΔA240/ε = μmol/min/mL or unit/mL = μmol of SLG hydrolysed/min/mL Taking enzyme dilutions into account: Activity = (ΔC)(total volume)(dilution factor)/(volume of enzyme) Protein = Mass Protein/Volume extract = mg/mL 242 Total Protein = (Protein)(Total volume) = (mg/mL)(mL) = mg Specific Activity = (Total Activity)/(Total Protein) = Units/mg Alternatively Activity = (ΔC)(total volume) = μmol/min Total Protein = (Mass Protein/Volume)(Volume extract) = (mg/mL)(mL) = mg Specific Activity = (Total Activity)/(Total Protein) = Units/mg Alternatively153 Activity (GloA) = (ΔA240nm)/ (Δt)0/2.86 and Activity (GloB) = -(ΔA240nm)/ (Δt)0/3.10 7.4.3 Kinetic target guided synthesis (kTGS) Biased assay: Stock solutions of GAz at 10 mM were prepared in < 5% DMSO and water. Mixture of up to 5 alkynes were prepared at 10 mM in DMSO from the 100 mM stock solutions. Alkynes were pooled into groups based on final triazole molecular weights to ensure that all possible triazole products had a distinct molecular weight, enabling hit identification using LC-MS/MS. A stock solution of protein at a concentration of 12.5 µM was prepared in buffer. The GAz and alkyne mixtures were incubated with the protein to give final concentrations as follows: GAz 1 mM, alkynes 1 mM, protein (GlxI/GlxII) 10 µM. The total volume was 100 µL. Two control samples were prepared where the fragments were incubated with 10 µM BSA and in buffer alone. The 96-well plates were sealed and incubated at 37 °C in a water bath for 3 days. Analysis of the crude mixtures was carried by LC-MS/MS out on an Aquasil C18 (100 x 4.6 mm, 5 µm particle size) column or Aquasil C18 (50 x 4.6 mm, 5 µm particle size) column. Masses were identified to match the molecular weight of the expected triazole and were considered a ‘hit’ compound if observed in the protein samples and not in the control samples. Synthesis assay: The following stock solutions were prepared: 100 mM of GAz in DMSO, 25.5 µL of which was added to 74.5 µL water to give a 25.5 mM stock solution. Mixtures of up to 5 alkyne fragments at 20 mM in DMSO from the 100 mM stock solutions, 5.1 µL of which was added to 14.9 µL DMSO to give a 5.1 mM stock solution. 100 mM of copper (II) sulfate pentahydrate in water, 27.4 µL of which was added to 72.6 µL of water. 100 mM of (+)-sodium L-ascorbate in water, 25.5 µL of which was added to 74.5 µL water. The volumes per well for incubation were as follows: 3 µL of GAz, 3 µL of alkyne fragments, 2 µL of copper (II) sulfate pentahydrate, 3 µL of (+)-sodium L-ascorbate, 293 µL water:butanol (2:1). The 96-well plates were sealed and gently shaken at RT for 3 243 days. Analysis of the crude mixtures was carried by LC-MS/MS out on an Aquasil C18 (50 x 4.6 mm, 5 µm particle size) column. Unbiased assay: 10 mM stock solutions of selected azide or alkyne hits were prepared in < 5% DMSO and water. Mixtures of up to 5 alkyne or azide fragments were prepared at 10 mM in DMSO from the 100 mM stock solutions. Alkynes or azides fragments were pooled into groups based on final triazole molecular weights to ensure distinct molecular weight differences. A stock solution of protein was prepared at a concentration of 12.5 µM in buffer. The fragment mixtures were incubated with the protein to give final concentrations as follows: azide 1 mM, alkyne 1 mM, protein 10 µM. The total volume was 100 µL. Two control samples were prepared where the fragments were incubated with 10 µM BSA and in buffer alone. The 96-well plates were sealed and incubated at 37 °C in a water bath for 3 days. Analysis of the crude mixtures was carried out by LC-MS/MS on an Aquasil C18 (100 x 4.6 mm, 5 µm particle size) column or Aquasil C18 (50 x 4.6 mm, 5 µm particle size) column. Masses were identified to match the molecular weight of expected triazoles and were considered to be a ‘hit’ compound if observed in the protein samples with at least twice the area under the curve, which indicated % triazole formation, compared to the control samples. The LC-MS/MS used was an Agilent triple quad 6460. The data acquisition and analysis Mass Hunter software was used. The optimised run methods for the MS scan and SIM scan on the Aquasil C18 (50 x 4.6 mm, 5 µm particle size) was a gradient of 0 – 95% acetonitrile (0.1% formic acid) in water (0.1% formic acid) over 4 min, held at 95% acetonitrile (0.1% formic acid) in water (0.1% formic acid) for 1 min followed by re-equilibration to 0% acetonitrile (0.1% formic acid) in water (0.1% formic acid) over 2 min. 7.4.4 19F-NMR binding assay The 19F-NMR samples were prepared as follows: fluorinated ligands (10 – 50 µM final concentration) and protein (10 – 50 µM final concentration) in <5% DMSO-d6 and PBS (pH 7.6) or fluorinated ligands (10 – 50 µM final concentration) in <5% DMSO-d6 and PBS (pH 7.6). The samples were analysed by 19F NMR using trifluoroacetic acid as an internal standard (-73 – -74 ppm). No interaction between TFA and the protein was observed for GST and GlxI but some interaction was observed with GlxII. All proton decoupled 19F NMR spectra were run at ambient room temperature (25 °C) on a Bruker Avance 300 MHz NMR with 4K scans. Carried out with the help of Dr Abil Aliev at UCL’s NMR department. 244 7.4.5 STD-NMR The assay conditions were as follows: ligand (1 mM final concentration), protein (10 µM final concentration), deuterated buffer 75 mM PBS pH 7.5, 0.15 M NaCl, 400 µL total volume. The STD-NMR’s were run with UCL School of Pharmacy Structural Chemistry Services on a Bruker Avance 500 MHz NMR. 7.4.6 Disk diffusion assay Sterile antimicrobial susceptibility discs containing the antibiotic or toxins of interest were prepared by pipetting a solution onto the discs and allowing them to dry at room temperature. To prepare the inoculum, 2 mL of autoclaved 0.9% saline solution was pipetted into a 15 mL Falcon tube. Using an inoculating loop, 4-5 colonies of bacterial strain was added to the solution and vortexed to suspend the colonies. The turbidity was matched to a 0.5 McFarland standard, adjusting with more colonies or by dilution as required. Agar plates were prepared using Müller-Hinton agar. Each plate was streaked using a sterile swab with the bacterial strain (x 3), rotating the plate 60 °C to ensure full coverage. One containing solvent and antibiotic and another containing solvent alone (negative control) were placed on the plate with sterilised tweezers. The plates were incubated at 37 °C for 16 h. The diameter of the zone of growth inhibition surrounding the disc was measured (mm). 7.4.7 Molecular modelling The human glyoxalase II protein structure in complex with glutathione was obtained from the Protein Data Bank (PDB: 1QH5) and the Salmonella enterica glyoxalase II protein structure was obtained from the Protein Data Bank (PDB: 2QED). To date there is no known crystal structure of glyoxalase II in E. coli. The human glyoxalase I protein structure in complex with S-benzyl glutathione was obtained from the Protein Data Bank (PDB: 1FRO) and the E. coli glyoxalase I protein structure was obtained from the Protein Data Bank (PDB: 1F9Z). The protein was processed using a text editor to prepare pdb files for the protein, ligand and protein-ligand complex for use in subsequent analyses. The protein pdb file was opened in Autodock Tools (ADT),177 the solvent and ions were removed and the resulting structure was saved as a pdbqt file for use in Autodock Vina.150 The small molecule ligands were prepared by constructing the 2D structures in ChemBioDraw Ultra 14.0 which were saved in sdf format. The 2D representations were converted into 3D structures using an in-house script that used Balloon 1.6.1.1220178 to generate an initial 3D conformer, followed by Openbabel 2.3.2179 to add hydrogens 245 appropriate for pH 7.4 and to perform a weighted conformer search (nconf = 1000, weighted, gasteiger partial charges). Energy minimisation was performed on the lowest energy conformer using obmimimize (n = 10000, c = 1e-30) followed by MOPAC2012180 using the semi-empirical PM7 parameters (keywords: PM7, MMOK, T=30000, PDBOUT). The compounds were saved in a mol2 file format and converted to pdb, pdbqt and mol2 files formats (the latter two containing partial charge information) using Openbabel 2.3.2. The properties of the compounds were determined using Openbabel 2.3.2 (molecular weight, number of non-hydrogen atoms, cLogP). The alkyne and azide fragment libraries were parameterised similarly from the corresponding sd files, with visual inspection of the output to check for anomalies. Autodock Vina was run using an Exhaustiveness setting of 9 and a binding site box centred on the ligand binding pocket that was 24 x 20 x 20 Å in size. All other parameters were set to their default values. The lowest energy conformation of each docked ligand was extracted from the results file using a custom script and analysed in Chimera 1.11.2.181 7.5 General Chemistry Experimental 1H-NMR spectra were recorded on a Bruker Avance 500 MHz NMR spectrometer (cryoprobe) or a Bruker Avance 400 MHz NMR spectrometer using deuterated solvents. The chemical shift data for each signal are given as δH in units of parts per million (ppm), referenced against deuterated solvent and the NMR was calibrated relative to tetramethylsilane. The number of protons (n) for each signal is indicated by nH. The multiplicity of each signal is indicated by s (singlet), br s (broad singlet), d (doublet), t (triplet), q (quartet), dd (doublet of doublets), ddd (doublet of doublets of doublets), dt (doublet of triplets) or m (multiplet). Coupling constants (J) are quoted in Hz and recorded to the nearest 0.1 Hz. The coupling constants were determined by analysis using ACD/NMR Processor Academic Edition software or Bruker TopSpin 3.2 and 3.5pl6. 13C-NMR spectra were recorded on a Bruker Avance 125 MHz NMR spectrometer (cryoprobe) or a Bruker Avance 100 MHz NMR spectrometer using deuterated solvents. The chemical shift data for each signal are given as δC in units of parts per million (ppm), referenced against deuterated solvent and the NMR was calibrated relative to tetramethylsilane. The coupling constants, where necessary, were determined by analysis using ACD/NMR Processor Academic Edition software or Bruker TopSpin 3.2 and 3.5pl6. 246 19F-NMR spectra were recorded on a Bruker Avance 300 MHz NMR spectrometer using deuterated solvents or aqueous buffers at UCL, NMR Department. The chemical shift data for each signal are given as δF in units of parts per million (ppm) and referenced against trifluoroacetic acid solvent. COSY, HMBC, HMQC spectra were recorded on a Bruker Avance 500 MHz NMR spectrometer (cryoprobe) spectrometer using deuterated solvents. The chemical shift data for each signal are in units of parts per million (ppm) and referenced against deuterated solvent. The multiplicity of each signal is indicated by s (singlet), br s (broad singlet), d (doublet), t (triplet), q (quartet), dd (doublet of doublets), ddd (doublet of doublets of doublets), dt (doublet of triplets) or m (multiplet). Coupling constants (J) are quoted in Hz and recorded to the nearest 0.1 Hz. The coupling constants were determined by analysis using ACD/NMR Processor Academic Edition software or Bruker TopSpin 3.2 and 3.5pl6. Mass spectra. High resolution mass spectrometry was obtained from the UCL Chemistry Mass Spectrometry Facility or from the EPSRC National Mass Spectrometry Facility Swansea, College of Medicine, Swansea University, Swansea, operating in positive or negative mode. The Orbitrap has a precision of 5 ppm. m/z values are reported in Daltons and followed by their percentage. Low resolution mass spectrometry was obtained from UCL School of Pharmacy, Structural Chemistry, operating in positive or negative mode on a ThermoLCQ Duo mass spectrometer coupled with LC system to 10 femtomoles sensitivity. Ionisation mode was ESI for all samples. Melting Points were determined using a Stuart SMP10 in duplicate and are uncorrected. With a temperature accuracy of ±1 °C at 20 °C and ±2.5 °C at 300 °C. Infrared Spectra were obtained from neat samples, either as solids or oils, using a Bruker Platinum ATR Alpha spectrometer or a Perkin Elmer Spectrum 100 FT-IT Spectrophotometer. The spectra were measured on an Opus lab manager or Perkin Elmer Spectrum express software. Absorption maxima are reported in wavenumbers (cm-1) and reported as br (broad), s (strong), m (medium) or w (weak). Specific Optical Rotations were measured on an Optical Activity PolAAR 21 polarimeter, in a cell with a path length of 1 dm. The light source was maintained at 546 nm. The concentration (c) is expressed in mg/mL. Specific rotations are denoted [𝛼]54620 and are given in implied units of 10-1 deg cm2 g-1 (T = ambient temperature at 20 °C). The cell was a stainless steel unjacketed flow tube (2.5 x 100 mm). 247 HPLC were performed on an Agilent technologies 1200 series HPLC system using a Waters XSelect® CSH C18 Column 50 x 6 mm (particle size: 2.5 μm) at a flow rate of 1 mL/min. The eluent system and method were as follow; Eluent A H2O + 0.1% formic acid; Eluent B MeCN + 0.1% formic acid; initial fixed composition 5% B to 95% B over 20 min, held at 95% B for 4 min then returned to 5% B over 1 min, the total run time was 25 min. Analytical thin layer chromatography (TLC) was carried out on normal phase Merck silica gel 60 F254 aluminium-supported thin layer chromatography sheets or Merck silica gel 60 RP-18 F254 aluminium-supported thin layer chromatography sheets. Visualisation was by absorption of UV light (λ 254 or 365 nm) or thermal development after staining with ninhydrin or KMnO4. Flash column chromatography was carried out manually on Merck silica gel 60 (230-400 mesh ASTM) or Sigma Aldrich silica gel 60 Å (230-400 mesh particle size, 40-63 µm particle size) eluting with indicated solvents under positive pressure of compressed air. Reverse phase column chromatography was carried out on Fluka Silica gel 100 C18 reversed phase under a positive pressure of compressed air or on the Biotage SP1 or SP4 using a Reveleris Grace reverse phase cartridge. Anhydrous solvents were obtained from Sigma Aldrich UK in SureSeal® bottles or Fisher Scientific or Acros in AcroSeal® bottles. All other solvents were used as supplied (analytical or HPLC grade). All non-aqueous reactions were performed in heat-dried flasks and under an inert nitrogen atmosphere. Chemicals were purchased from Sigma Aldrich UK, Alfa Aesar, Acros, VWR, Fisher UK or Enamine and were used without further purification unless otherwise stated. In Vacuo refers to solvents being removed under reduced pressure on a Heidolph rotary evaporator or Buchi rotary evaporator. Brine refers to a saturated aqueous solution of sodium chloride in distilled water. 10% Citric Acid refers to a 10% w/v citric acid/distilled water. Bicarb refers to a saturated aqueous solution of sodium hydrogen carbonate in distilled water. DP refers to desired product. 248 7.6 Experimental data of synthesised compounds 7.6.1 Synthesis of Glutathione S-conjugates The naming of the glutathione S-conjugates has been simplified, due to the conserved nature of the glutathione tripeptide backbone for all synthesised analogues, i.e. from N5-((R)-3-((4-bromobenzyl)thio)-1-((carboxymethyl)amino)-1-oxopropan-2-yl)-L-glutamine to S-(4-bromobenzyl)-glutathione. S-(4-Bromobenzyl)-glutathione (3) Reduced L-glutathione (1.00 g, 3.26 mmol, 1.00 eq) was dissolved in 1 M NaOH (aq) (6.52 mL). Ethanol (16 mL) was added until the solution turned cloudy. 4-Bromobenzyl bromide (0.81 g, 3.26 mmol, 1.00 eq) was added portion wise in ethanol (4 mL) over 10 min. The reaction mixture was stirred vigorously for 3 h at room temperature. The reaction mixture was acidified with conc. HCl, cooled and the precipitate filtered to afford S-(4-bromobenzyl)-glutathione 3 as a colourless solid (1.26 g, 82%): Rf 0.21 (methanol: water: acetic acid 1:1:0.1, RP C-18 TLC); mp 186 – 189 °C [Lit182 206 – 207 °C]; [𝛼]54620 -36.42 (c 0.96 DMSO); vmax (neat)/cm-1 3373.44 (w), 3354.73 (w), 3044.27 (br), 1648.30 (s), 1513.33 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.67 (1H, t, J = 5.8 Hz, Gly-NH); 8.38 (1H, d, J = 8.8 Hz, Cys-NH); 7.50 (2H, d, J = 8.2 Hz, Ar(2,6)-H); 7.30 (2H, d, J = 8.2 Hz, Ar(3,5)-H); 4.60 (1H, ddd, J = 13.7, 8.8, 4.7 Hz, Cys-α-CH); 3.77 – 3.73 (4H, m, Gly-CH2, Ar-CH2); 3.48 (1H, t, J = 6.4 Hz, Glu-α-CH); 2.78 (1H, dd, J = 13.7, 4.7 Hz, Cys-β-CH2); 2.55 (1H, dd, J = 13.7, 8.8 Hz, Cys-β-CH2); 2.34 (2H, t, J = 7.4 Hz, Gly-ϒ-CH2); 1.97 – 1.90 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ 171.5 (C=O), 170.8 (C=O), 170.6 (C=O), 170.2 (C=O), 138.0 (Ar(C1)-C), 131.2 (Ar(C2,6)-H), 131.1 (Ar(C3,5)-H), 119.8 (Ar(C4)-Br), 52.7 (Gly-α-CH), 51.9 (Cys-α-CH), 40.9 (Gly-CH2), 34.3 (Ar-CH2), 32.9 (Cys-CH2), 31.2 (Glu-ϒ-CH2), 26.6 (Glu-β-CH2); HRMS m/z (ES+) [Found [M79Br+H]+ / [M81Br+H]+ 476.0481 / 478.0458 C17H22BrN3O6S requires [M79Br+H]+ / [M81 Br+H]+ 476.0485 / 478.0466]; m/z (ES-) 474.0 ([M-H]-, 100%); HPLC RT 9.34 min; >95% content. The data are in agreement with the literature.108, 117, 182,121, 183 249 S-(2-Fluorobenzyl)-glutathione (4) Reduced L-glutathione (1.00 g, 3.25 mmol, 1.00 eq) was dissolved in 1 M NaOH (aq) (6.5 mL). Ethanol (16 mL) was added until the solution turned cloudy. 2-Fluorobenzyl bromide (0.62 g, 0.39 mL, 3.25 mmol, 1.00 eq) was added portion wise in ethanol (4 mL) over 10 min. The reaction mixture was stirred vigorously for 4.5 h at room temperature, after which time the reaction mixture was acidified with conc. HCl, concentrated in vacuo. The resulting solid was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/methanol in 0.1% formic acid/water) to afford a colourless solid which was further purified by recrystallisation from ethanol and water. The solid was filtered and dried under high vacuum to afford S-(2-fluorobenzyl)-glutathione 4 as a colourless solid (0.88 g, 65%): Rf 0.37 (methanol: water: acetic acid, 1:1:0.1, RP C-18 TLC); mp 200 °C (dec.) (Ethanol/Water); [𝛼]54620 -34.96 (c 0.94 DMSO); vmax (neat)/cm-1 3345.33 (br), 3035.63 (br), 1641.15 (s), 1511.39 (s), 1229.64 (s), 756.95 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.72 (1H, t, J = 5.8 Hz, Gly-NH); 8.42 (1H, d, J = 9.0 Hz, Cys-NH); 7.44 – 7.40 (1H, m, Ar(4)-H); 7.41 – 7.31 (1H, m, Ar(5)-H); 7.21 – 7.13 (2H, m, Ar(3,6)-H); 4.52 (1H, ddd, J = 13.5, 9.0, 4.4 Hz, Cys-α-CH); 3.77 (2H, s, Ar-CH2); 3.70 – 3.74 (2H, m, Gly-CH2); 3.33 (1H, t, J = 6.6 Hz, Glu-α-CH); 2.85 (1H, dd, J = 13.5, 4.4 Hz, Cys-β-CH2); 2.62 (1H, dd, J = 13.5, 9.0 Hz, Cys-β-CH2); 2.39 – 2.27 (2H, m, Glu-γ-CH2); 1.98 – 1.83 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ ppm 171.7 (C=O), 170.8 (C=O), 170.6 (C=O), 170.1 (C=O), 159.2 (d, J = 245.6 Hz, Ar-(C2)-F), 131.2 (d, J = 3.6 Hz, Ar-(C6)-H), 129.0 (d, J = 8.2 Hz, Ar-(C4)-H), 125.4 (d, J = 14.7 Hz, Ar-(C1)-C), 124.2 (d, J = 3.7 Hz, Ar-(C5)-H), 115.3 (d, J = 22.0 Hz, Ar-(C3)-H), 53.0 (Gly-α-CH), 52.1 (Cys-α-CH), 41.1 (Gly-CH2), 33.1 (Ar-CH2), 31.4 (Cys-β-CH2), 28.3 (Glu-γ-CH2), 26.8 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -117.80 (s); HRMS m/z (ES+) [Found [M+H]+ 416.1286; C17H22FN3O6S requires [M+H]+ 416.1286]; m/z (ES+) 416.1 ([M+H]+, 100%). 250 S-(3-Fluorobenzyl)-glutathione (5) Reduced L-glutathione (1.00 g, 3.25 mmol, 1.00 eq) was dissolved in 1 M NaOH (aq) (6.5 mL). Ethanol (16 mL) was added until the solution turned cloudy. 3-Fluoro-bromobenzyl bromide (0.62 g, 0.39 mL, 3.25 mmol, 1.00 eq) was added portion wise in ethanol (4 mL) over 10 min. The reaction mixture was stirred vigorously for 4.5 h at room temperature, after which time the reaction mixture was acidified with conc. HCl and concentrated in vacuo. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/methanol in 0.1% formic acid/water) to afford a colourless solid, which was further purified by recrystallization from ethanol and water. The solid was filtered and dried under high vacuum to afford S-(3-fluorobenzyl)-glutathione 5 as a colourless solid (0.24 g, 18%): Rf 0.43 (methanol: water: acetic acid, 1:1:0.1, RP C-18 TLC); mp 205 °C (dec.) (Ethanol/Water); [𝛼]54620 -36.61 (c 1.04 DMSO);vmax (neat)/cm-1 3345.68 (br), 3030.89 (br), 1640.16 (s), 1509.73 (s), 1307.27 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.70 (1H, t, J = 5.6 Hz, Gly-NH); 8.44 (1H, d, J = 8.8 Hz, Cys-NH); 7.39 – 7.32 (1H, m, Ar(6)-H); 7.18 (1H, s, Ar(2)-H); 7.16 – 7.15 (1H, m, Ar(4)-H); 7.07 (1H, dt, J = 8.9, 1.8 Hz, Ar(5)-H); 4.51 (1H, ddd, J = 13.7, 8.8, 4.6 Hz, Cys-α-CH); 3.77 (2H, s, Ar-CH2); 3.73 – 3.70 (2H, m, Gly-CH2); 3.34 (2H, t, J = 6.6 Hz Glu-α-CH); 2.79 (1H, dd, J = 13.7, 4.6 Hz Cys-β-CH2); 2.58 (1H, dd, J = 13.7, 8.8 Hz Cys-β-CH2); 2.38 – 2.27 (2H, m, Glu-γ-CH2); 1.99 – 1.84 (2H, m Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ ppm 171.7 (C=O), 170.9 (C=O), 170.6 (C=O), 170.2 (C=O), 161.1 (d, J = 243.8 Hz, Ar-(C3)-F), 141.4 (d, J = 8.2 Hz, Ar-(C1)-C), 130.1 (d, J = 8.2 Hz, Ar-(C5)-H), 125.0 (d, J = 2.7 Hz Ar-(C6)-H), 115.4 (d, J = 21.0 Hz, Ar-(C2)-H), 113.5 (d, J = 21.0 Hz, Ar-(C4)-H), 53.0 (Gly-α-CH), 52.0 (Cys-α-CH), 41.1 (Gly-CH2), 34.5 (Ar-CH2), 32.8 (Cys-β-CH2), 31.4 (Glu-γ-CH2), 26.8 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -113.37 (s); HRMS m/z (ES+) [Found [M+H]+ 416.1288; C17H22FN3O6S requires [M+H]+ 416.1286]; m/z (ES+) 416.1 ([M+H]+, 100%); HPLC RT 6.83 min; >95% content. 251 S-(4-Fluorobenzyl)-glutathione (6) S-(4-fluorobenzyl)-glutathione (6) was synthesised using the conditions described for S-(2-trifluormethylbenzyl)-glutathione (7) by Dr Jess Healy. Rf 0.4 (methanol: water: acetic acid, 1:1:0.1, ninhydrin stain, RP C-18 TLC); mp 198 °C (dec.) [Lit108 196 – 197 °C]; [𝛼]54620 -37.12 (c 0.91 DMSO);vmax (neat)/cm-1 3346.02 (br), 1673.58 (s), 1640.46 (s), 1506.95 (s), 1218.42 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.71 (1H, t, J = 5.5 Hz, Gly-NH); 8.44 (1H, d, J = 9.1 Hz, Cys-NH); 7.36 (2H, dd, J = 7.8, 5.6 Hz, Ar(3,5)-H); 7.14 – 7.11 (2H, m, Ar(2,6)-H); 4.51 (1H, ddd, J = 13.6, 9.1, 4.5 Hz, Cys-α-CH); 3.75 (2H, s, Ar-CH2); 3.73 – 3.70 (2H, m, Gly-CH2); 3.34 (1H, t, J = 6.6 Hz, Glu-α-CH); 2.79 (1H, dd, J = 13.6, 4.5 Hz, Cys-β-CH2); 2.56 (1H, dd, J = 13.6, 9.1 Hz, Cys-β-CH2); 2.26 – 2.40 (2H, m, Glu-ϒ-CH2); 1.98 – 1.84 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ171.7 (C=O), 170.9 (C=O), 170.6 (C=O), 170.1 (C=O), 160.1 (d, J = 241.9 Hz, Ar(C4)-F), 134.6 (d, J = 3.7 Hz, Ar(C1))-C), 130.7 (d, J = 8.2 Hz, Ar(C2,6)-H), 115.1 (d, J = 21.1 Hz, Ar(C3,5)-H), 53.0 (Gly-α-CH), 52.0 (Cys-α-CH), 41.1 (Gly-CH2), 34.2 (Ar-CH2), 32.8 (Cys-β-CH2), 31.4 (Glu-ϒ-CH2), 26.8 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -115.898 (s); HRMS m/z (ES+) [Found [M+H]+ 416.1285; C17H22FN3O6S requires [M+H]+ 416.1286]; m/z (ES+) 416.1 ([M+H]+, 100%); HPLC RT 6.90 min; >95% content. The data are in agreement with the literature.108 S-(2-Trifluormethylbenzyl)-glutathione (7) Reduced L-glutathione (0.20 g, 0.65 mmol, 1.00 eq) was dissolved in 1 M NaOH (aq) (1.3 mL). Ethanol (4 mL) was added until the solution turned cloudy. To this 2-trifluoro-bromobenzyl bromide (0.16 g, 0.65 mmol, 1.00 eq) was added portion wise in ethanol (4 252 mL) over 10 min. The reaction mixture was stirred vigorously for 3 h at room temperature, after which time the reaction mixture was acidified with conc. HCl, and concentrated in vacuo. Water and diethyl ether were added and the mixture cooled over ice, the resulting precipitate was isolated by filtration and recrystallized from ethanol and water. The solid was filtered and dried under high vacuum to afford S-(2-trifluormethylbenzyl)-glutathione 7 as a colourless solid (0.16 g, 52%): Rf 0.17 (methanol: water: acetic acid, 1:1:0.1, RP C-18 TLC); mp 201 °C (dec.); [𝛼]54620 -33.33 (c 0.93 DMSO);vmax (neat)/cm-1 3380.89 (m), 3349.45 (m), 3033.73 (br), 2959.02 (br), 1673.15 (s), 1642.95 (s), 1511.07 (s), 1306.15 (s), 1232.60 (s), 1119.23 (s), 767.99 (m); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.74 (1H, t, J = 5.8 Hz, Gly-NH); 8.46 (1H, d, J = 8.9 Hz, Cys-NH); 7.70 (1H, d, J = 7.8 Hz, Ar(3)-H); 7.66 – 7.62 (2H, m, Ar(4,5)-H); 7.50 – 7.45 (1H, m, Ar(6)-H); 4.55 (1H, ddd, J = 13.7, 8.9, 4.2 Hz, Cys-α-CH); 3.92 (2H, s, Ar-CH2); 3.75 – 3.71 (2H, m, Gly-CH2); 3.35 (1H, t, J = 6.6 Hz, Glu-α-CH); 2.94 (1H, dd, J = 13.7, 4.2 Hz, Cys-β-CH2); 2.69 (1H, dd, J = 13.7, 8.9 Hz, Cys-β-CH2); 2.41 – 2.28 (2H, m, Gly-ϒ-CH2); 2.00 – 1.84 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ 171.8 (C=O), 170.9 (C=O), 170.6 (C=O), 170.2 (C=O), 136.8 (Ar(C5)-H), 132.5 (Ar(C6)-H), 131.9 (Ar(C4)-H), 127.5 (Ar(C3)-H), 126.7 (q, J = 31.2 Hz, Ar(C2)-C), 125.9 (q, J = 5.5 Hz, Ar(C1)-C), 123.2 (q, J = 275.9 Hz, CF3), 53.0 (Gly-α-CH), 52.1 (Cys-α-CH), 41.1 (Gly-CH2), 33.8 (Ar-CH2), 32.1 (Cys-β-CH2), 31.4 (Glu-ϒ-CH2), 26.8 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -58.73 (s); HRMS m/z (ES+) [Found [M+H]+ 466.1246; C18H22F3N3O6S requires [M+H]+ 466.1254]; m/z (ES- ) 464.1 ([M-H]-, 100%); HPLC RT 9.21 min; >95% content. S-(3-Trifluoromethylbenzyl)-glutathione (8) Reduced L-glutathione (1.00 g, 3.25 mmol, 1.00 eq) was dissolved in 1 M NaOH (aq) (6.5 mL). Ethanol (16 mL) was added until the solution turned cloudy. 3-Trifluoromethyl-benzyl bromide (0.77 g, 0.49 mL, 3.25 mmol, 1.00 eq) was added portion wise in ethanol (4 mL) over 10 min. The reaction mixture was stirred vigorously for 4.5 h at room temperature, after which time the reaction mixture was acidified with conc. HCl, concentrated in vacuo. The crude material was purified by reverse phase column 253 chromatography (10 – 90% 0.1% formic acid/methanol in 0.1% formic acid/water). The resulting solid was recrystallised from ethanol and water, filtered and dried under high vacuum to afford S-(3-trifluoromethylbenzyl)-glutathione as a colourless solid (0.17 g, 13%): Rf 0.24 (methanol: water: Acetic acid, 1:1:0.1, RP C-18 TLC); mp 193 °C (dec.) (Ethanol/Water); [𝛼]54620 -23.14 (c 0.99 DMSO); vmax (neat)/cm-1 3345.87 (br), 3038.92 (br), 1640.74 (s), 1510.09 (s), 1327.05 (s), 1112.43 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.70 (1H, t, J = 5.6 Hz, Gly-NH); 8.40 (1H, d, J = 8.9 Hz, Cys-NH); 7.70 (1H, s, Ar(2)-H); 7.65 (1H, d, J = 7.5 Hz, Ar(6)-H); 7.59 (1H, d, J = 7.5 Hz, Ar(4)-H); 7.54 (1H, t, J = 7.2 Hz, Ar(5)-H); 4.54 (1H, ddd, J = 13.7, 8.9, 4.8 Hz, Cys-α-CH); 3.87 (2H, s, Ar-CH2); 3.74 – 3.71 (2H, m, Gly-CH2); 3.42 (1H, t, J = 6.3 Hz, Glu-α-CH); 2.79 (1H, dd, J = 13.7, 4.8 Hz, Cys-β-CH2); 2.58 (1H, dd, J = 13.7, 8.9 Hz, Cys-β-CH2); 2.38 – 2.31 (2H, m, Glu-γ-CH2); 2.01 – 1.86 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ ppm 171.6 (C=O), 170.8 (C=O), 170.6 (C=O), 170.2 (C=O), 140.1 (Ar(C1)-C), 133.0 (Ar(C6)-C), 129.3 (Ar(C5)-H), 128.7 (q, J = 32.0 Hz, Ar(C3)-CF3), 120.8 (q, J = 272.4 Hz, CF3),125.2 (q, J = 3.7 Hz, Ar(C2)-H), 123.4 (q, J = 3.7 Hz, Ar(C4)-H), 52.7 (Gly-α-CH), 52.0 (Cys-α-CH), 40.9 (Gly-CH2), 34.4 (Ar-CH2), 32.9 (Cys-β-CH2), 31.2 (Glu-γ-CH2), 26.6 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -61.01 (s, CF3); HRMS m/z (ES+) [Found [M+H]+ 466.1249; C18H22F3N3O6S requires [M+H]+ 466.1254]; m/z (ES+) 466.1 ([M+H]+, 100%); HPLC RT 9.78 min; >90% content. S-(4-Trifluoromethylbenzyl)-glutathione (9) Reduced L-glutathione (0.40 g, 1.30 mmol, 1.00 eq) was dissolved in 1 M NaOH (aq) (2.6 mL). Ethanol (6.5 mL) was added until the solution turned cloudy. 4-Trifluoromethyl-benzyl bromide (0.31 g, 0.20 mL, 1.30 mmol, 1.00 eq) was added portion wise in ethanol (1.6 mL) over 10 min. The reaction mixture was stirred vigorously for 4.5 h at room temperature, after which time the reaction mixture was acidified with conc. HCl, concentrated in vacuo. The crude material was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/methanol in 0.1% formic acid/water). The resulting solid was recrystallised from ethanol and water, the solid filtered off and dried under high vacuum to afford S-(4-trifluoromethylbenzyl)-glutathione 9 as a colourless 254 solid (0.17 g, 28%): Rf 0.24 (methanol: water: acetic acid, 1:1:0.1, ninhydrin stain, RP C-18 TLC); mp 218 °C (dec.) (Ethanol/Water); [𝛼]54620 -37.50 (c 0.80 DMSO); vmax (neat)/cm-1 3348.08 (br), 2934.00 (br), 1646.47 (s), 1512.62 (s), 1321.74 (s), 1118.20 (s); 1H NMR (500MHz, DMSO-d6) δ ppm 8.74 (1H, t, J = 5.6 Hz, Gly-NH); 8.44 (1H, d, J = 8.9 Hz, Cys-NH); 7.68 (2H, d, J = 8.0 Hz, Ar(3,5)-H); 7.56 (2H, d, J = 8.0 Hz, Ar(2,6)-H); 4.53 (1H, ddd, J = 13.6, 8.9, 4.5 Hz, Cys-α-CH); 3.85 (2H, s, Ar-CH2); 3.72 (2H, dd, J = 5.6, 3.1 Hz, Gly-CH2); 3.33 (1H, t, J = 6.7 Hz, Glu-α-CH); 2.79 (1H, dd, J = 13.6, 4.5 Hz, Cys-β-CH2); 2.58 (1H, dd, J = 13.6, 8.9 Hz, Cys-β-CH2); 2.42 – 2.25 (2H, m, Glu-γ-CH2); 2.00 – 1.83 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ ppm 171.8 (C=O), 170.9 (C=O), 170.6 (C=O), 170.0 (C=O), 143.5 (Ar(C1)-C), 129.6 (Ar (C2,6)-H), 128.9 (q, J = 31.4 Hz, Ar(C4)-CF3), 123.1 (q, J = 268.4 Hz, CF3), 125.3 (q, J = 3.7 Hz, Ar(C3,5)-H), 53.0 (Gly-α-CH), 52.0 (Cys-α-CH), 41.1 (Gly-CH2), 34.4 (Ar-CH2), 32.8 (Cys-β-CH2), 31.4 (Glu-γ-CH2), 26.8 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -60.79 (s); HRMS m/z (ES+) [Found [M+H]+ 466.1250; C18H22F3N3O6S requires [M+H]+ 466.1254]; m/z (ES+) 466.1 ([M+H]+, 100%); HPLC RT 10.07 min; >95% content. S-(3,5-Difluorobenzyl)-glutathione (10) Reduced L-glutathione (0.20 g, 0.65 mmol, 1.00 eq) was dissolved in 1 M NaOH (aq) (1.3 mL). Ethanol (4 mL) was added until the solution turned cloudy. 3,5-Difluoro-bromobenzyl bromide (0.14 g, 0.65 mmol, 1.00 eq) was added portion wise in ethanol (4 mL) over 10 min. The reaction mixture was stirred vigorously for 3 h at room temperature, after which time the reaction mixture was acidified with conc. HCl, concentrated in vacuo and purified by reverse phase column chromatography (5 – 95% 0.1% formic acid/methanol in 0.1% formic acid/water). The compound was further purified by recrystallization from ethanol and water, filtered and dried under high vacuum to afford S-(3,5-difluorobenzyl)-glutathione 10 as a colourless solid (0.03 g, 12%): Rf 0.30 (methanol: water: acetic acid, 1:1:0.1, RP C-18 TLC); mp 194 °C (dec.); [𝛼]54620 -37.04 (c 0.08 DMSO); vmax (neat)/cm-1 3379.92 (m), 3348.39 (m), 3032.17 (br), 2953.09 (br), 1640.83 (s), 1509.10 (s), 1235.17 (s), 971.86 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 255 8.74 (1H, t, J = 5.6 Hz, Gly-NH); 8.45 (1H, d, J = 8.8 Hz, Cys-NH); 7.04 – 7.13 (3H, m, ArH); 4.50 (1H, ddd, J = 13.5, 8.8, 4.7 Hz, Cys-α-CH); 3.78 (2H, s, Ar-CH2); 3.73 – 3.69 (2H, m, Gly-CH2); 3.34 (1H, t, J = 6.6 Hz, Glu-α-CH); 2.80 (1H, dd, J = 13.5, 4.7 Hz, Cys-β-CH2); 2.59 (1H, dd, J = 13.5, 8.8 Hz Cys-β-CH2); 2.39 – 2.37 (2H, m, Glu-ϒ-CH2); 2.00 – 1.82 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ ppm 171.7 (C=O), 170.8 (C=O), 170.6 (C=O), 170.1 (C=O), 161.2 (d, J = 245.6 Hz, Ar(C3)-F), 161.1 (d, J = 245.6 Hz, Ar(C5)-F), 143.4 (t, J = 9.2 Hz, Ar(C1)-C), 112.0 (dd, J = 19.2, 6.4 Hz, Ar(C2,6)-H), 102.2 (t, J = 25.6 Hz, Ar(C4)-H), 53.0 (Gly-α-CH), 52.1 (Cys-α-CH), 41.1 (Gly-CH2), 34.2 (Ar-CH2), 32.7 (Cys-β-CH2), 31.4 (Glu-ϒ-CH2), 26.8 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -110.05 (s); HRMS m/z (ES+) [Found [M+H]+ 434.1186; C17H21F2N3O6S requires [M+H]+ 434.1192]; m/z (ES-) 432.1 ([M-H]-, 100%); HPLC RT 7.57 min; >95% content. N-Fluorenylmethoxycarbonyl-4-fluoro-L-phenylalanine-glycine tert-butyl ester (13) To N-fluorenylmethyloxycarbonyl-4-fluoro-L-phenylalanine (0.4 g, 0.99 mmol, 1.00 eq), glycine-tert-butyl ester hydrochloride (0.17 g, 1.04 mmol, 1.05 eq) and HBTU (0.39 g, 1.04 mmol, 1.05 eq) in anhydrous N,N-dimethylformamide (10 mL) was added diisopropylethylamine (0.52 mL, 2.96 mmol, 3.00 eq) dropwise. The reaction stirred at room temperature for 3.5 h, after which time the reaction was complete by TLC analysis. Ethyl acetate (40 mL) was added and washed with water (40 mL), 10% citric acid (40 mL), saturated bicarb (40 mL) and brine (2 x 20 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was purified by silica gel column chromatography (0 – 100% ethyl acetate/hexane) to afford N-fluorenylmethoxycarbonyl-4-fluoro-L-phenylalanine-glycine tert-butyl ester 13 as a colourless solid (0.54 g, 84% yield): Rf 0.81 (Hexane: EtOAc, 1:1); mp 162 – 163 °C; vmax (neat)/cm-1 3320.65 (br), 1741.24 (m), 1693.24 (m), 1653.76 (s), 1524.07 (s), 1298.78 (s), 733.86 (s); 1H NMR (500 MHz, CDCl3-d) δ ppm 7.70 (2H, d, J = 7.5 Hz, Fmoc-ArH); 7.46 (2H, t, J = 7.5 Hz, Fmoc-ArH); 7.34 (2H, t, J = 7.5 Hz, Fmoc-ArCH); 7.22 (2H, t, J = 7.5 Hz, Fmoc-ArCH); 7.07 (2H, m, Ar(3,5)-H); 6.90 (2H, t, J = 8.3 Hz, Ar(2,6)-H); 6.20 – 6.12 (1H, br. s, Gly-256 NH); 5.27 – 5.16 (1H, m, Phe(4F)-NH); 4.40 - 4.33 (2H, m, Fmoc-CH2); 4.33 – 4.27 (1H, m, Fmoc-CH); 4.15 – 4.09 (1H, m, Phe(4F)-α-CH); 3.88 – 3.72 (2H, m, Gly-CH2); 3.05 – 2.94 (2H, m, Phe(4F)-β-CH2); 1.38 (9H, s, CH3); 13C NMR (125 MHz, CDCl3) δ 170.5 (C=O), 168.3 (C=O), 161.0 (d, J = 244.6 Hz, Ar(C4)-F), 155.8 (C=O), 143.7 (d, J = 6.4 Hz, Ar-(C1)-C), 141.3 (Fmoc-ArC-C), 131.9 (Fmoc-ArC-C), 130.8 (d, J = 7.3 Hz, Ar(C2,6)-H), 127.7 (Fmoc-ArC-H), 127.1 (Fmoc-ArC-H), 120.0 (Fmoc-ArC-H), 125.0 (Fmoc-ArC-H), 115.6 (d, J = 22.0 Hz, Ar(C3,5)-H), 82.6 (C), 67.0 (Fmoc-CH), 56.0 (CH), 47.1 (Fmoc-CH2), 41.9 (Gly-CH2), 37.6 (Ar-CH2), 28.0 (CH3); 19F NMR (300 MHz, DMSO-d6) δ ppm -116.94 (s); HRMS m/z (ES+) [Found [M+H]+ 519.2295; C30H31FN2O5 requires [M+H]+ 519.2290]; m/z (ES+) 541.2 ([M+Na]+, 100%). 4-Fluoro-L-phenylalanine-glycine tert-butyl ester (14) To N-fluorenylmethoxycarbonyl-4-fluoro-L-phenylalanine-glycine tert-butyl ester (13) (0.49 g, 1.94 mmol, 1.00 eq) was added 10% v/v piperidine/N,N-dimethylformamide (10 mL). The reaction mixture was stirred at room temperature for 35 min, after which time TLC analysis showed consumption of starting material. The reaction was concentrated in vacuo to give a colourless solid. The crude material was purified by silica gel column chromatography (50 – 100% ethyl acetate in hexane to 20% methanol/ethyl acetate) to afford 4-Fluoro-L-phenylalanine-glycine tert-butyl ester 14 as a colourless gum (0.27 g, 95%): Rf 0.16 (Hexane: Ethyl acetate, KMnO4 stain, 1:1); mp 63 – 64 °C; vmax (neat)/cm-1 3351.36 (br), 2980.97 (br), 1734.15 (s), 1645.71 (s), 1509.08 (s), 1149.79 (s), 840.08 (s), 524.14 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.58 – 7.73 (1H, br. s, Gly-NH); 7.12 (2H, dd, J = 8.0, 5.5 Hz, Ar(3,5)-H); 6.94 (2H, m, Ar(2,6)-H); 3.88 (2H, dd, J = 7.2, 5.6 Hz, Gly-CH2); 3.54 (1H, dd, J = 9.4, 3.7 Hz, Phe(4F)-α-CH); 3.18 (1H, dd, J = 13.8, 3.7 Hz, Phe(4F)-β-CH2); 2.64 (1H, dd, J = 13.8, 9.4 Hz, Phe(4F)-β-CH2); 1.41 (9H, s, CH3); 1.34 – 1.19 (2H, br. s, NH2); 13C NMR (125 MHz, CDCl3) δ 174.2 (C=O), 169.0 (C=O), 160.9 (d, J = 246.5 Hz, Ar(C4)-F), 133.5 (d, J = 4.6 Hz, Ar(C1)-C), 130.7 (d, J = 8.2 Hz, Ar(C2,6)-H), 115.6 (d, J = 21.1 Hz, Ar(C3,5)-H), 82.2 (C), 56.4 (CH), 41.6 (Gly-CH2), 40.0 (Ar-CH2), 28.0 (CH3); 19F NMR (300 MHz, DMSO-d6) δ ppm -115.82 (s); HRMS m/z (ES+) [Found [M+H]+ 297.1608; C15H21FN2O3 requires [M+H]+ 297.1609]; m/z (ES+) 296.9 ([M+H]+, 100%). 257 N-Boc-α-tert-butyl-ester-γ-glutamyl-4-fluoro-L-phenylalaninylglycine tert-butyl ester (16) To 4-fluoro-L-phenylalanine-glycine tert-butyl ester (14) (0.21 g, 0.73 mmol, 1.00 eq), N-Boc-L-glutamic acid-α-tert butyl ester (0.22 g, 0.73 mmol, 1.00 eq) and HBTU (0.29 g, 0.77 mmol, 1.05 eq) in N,N-dimethyformamide (7 mL) was added diisopropylethylamine (0.38 mL, 2.19 mmol, 3.00 eq) dropwise. The reaction was allowed to stir at room temperature for 4 h, after which time TLC analysis showed consumption of starting material. Ethyl acetate (40 mL) was added and the organic layer washed with water (40 mL), 10% citric acid (40 mL), saturated bicarb (40 mL) and brine (2 x 20 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo. The resulting material was further purified by silica gel column chromatography (0 – 100% ethyl acetate in hexane) to afford N-Boc-α-tert-butyl-ester-γ-glutamyl-4-fluoro-L-phenylalaninylglycine tert-butyl ester 16 as a glassy colourless solid (0.26 g, 60%): Rf 0.65 (Hexane: Ethyl acetate, 1:1); mp 69 – 71 °C; vmax (neat)/cm-1 3294.64 (br), 2977.72 (br), 1736.96 (m), 1715.32 (m), 1641.77 (m), 1509.42 (m), 1366.34 (m), 1220.75 (m), 1150.13 (s); 1H NMR (500 MHz, CDCl3-d) δ ppm 7.23 (2H, dd J = 8.0, 5.5 Hz, Ar(3.5)-H); 7.09 – 7.05 (2H, m, Ar(2,6)-H); 6.93 (1H, d, J = 7.1 Hz, Phe(4F)CH2CH-NH); 6.57 – 6.51 (1H, br. s, Gly-NH); 5.26 (1H, d, J = 7.2 Hz, Phe(4F)-NH); 4.74 – 4.67 (1H, m, Phe(4F)-α-CH); 3.97 – 3.82 (3H, m, Glu-α-CH, Gly-CH2); 3.20 (1H, dd, J = 14.0, 7.1 Hz, Phe(4F)-β-CH2); 3.07 (1H, dd, J = 14.0, 7.1 Hz, Phe(4F)-β-CH2); 2.34 – 2.22 (2H, m, Glu-ϒ-CH2); 2.13 – 2.05 (1H, m, Glu-β-CH2); 1.81 – 1.72 (1H, m, Glu-β-CH2); 1.48 (27H, s, CH3); 13C NMR (125 MHz, CDCl3) δ 172.4 (C=O), 171.2 (C=O), 170.8 (C=O), 168.4 (C=O), 160.8 (d, J = 245.6 Hz, Ar(C4)-F), 155.9 (C=O), 132.5 (d, J = 2.7 Hz, Ar(C1)-C), 130.7 (d, J = 8.2 Hz, Ar(C2,6)-H), 115.4 (d, J = 21.1 Hz, Ar(C3,5)-H), 82.5 (C), 82.3 (C), 80.1 (C), 54.5 (Gly-α-CH), 52.9 (Cys-α-CH), 42.0 (Gly-CH2), 36.6 (Ar-CH2), 32.4 (Glu-ϒ-CH2), 29.8 (Glu-β-CH2), 28.3 (CH3), 28.0 (CH3), 27.9 (CH3); 19F NMR (300 MHz, DMSO-d6) δ ppm -117.04 (s); HRMS m/z (ES+) [Found [M+H]+ 582.3193; C29H44FN3O8 requires [M+H]+ 582.3185]; m/z (ES+) 582.1 ([M+H]+, 29%), 604.3 ([M+Na]+, 100%). 258 γ-Glutamyl-4-fluoro-L-phenylalaninylglycine (17) N-Boc-α-tert-butyl-ester-γ-glutamyl-4-fluoro-L-phenylalaninylglycine tert-butyl ester (16) (0.21 g, 0.36 mmol, 1.00 eq) and triethylsilane (0.35 mL, 2.17 mmol, 6.00 eq) in 80% v/v trifluoroacetic acid/dichloromethane (4 mL) was stirred at room temperature for 4 h, after which time TLC analysis showed consumption of the starting material. The crude reaction mixture was concentrated in vacuo and purified by reverse phase column chromatography (5 – 95% 0.1% formic acid/methanol in 0.1% formic acid/water). The compound was further purified by recrystallization from ethanol and water to afford γ-Glutamyl-4-fluoro-L-phenylalaninylglycine 17 as a colourless solid (0.02 g, 21%): Rf 0.52 (methanol: water: acetic acid, ninhydrin stain, 1:1:0.1, RP C-18 TLC); mp 218 °C (dec) (Ethanol/Water); vmax (neat)/cm-1 3343.06 (br), 2940.11 (br), 1725.15 (m), 1621.02 (s), 1509.91 (s), 1219.72 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.75 (1H, t, J = 5.5 Hz, Gly-NH); 8.50 (1H, d, J = 9.7 Hz, Phe(4F)-NH); 7.30 (2H, dd, J = 8.2, 5.6 Hz Ar(3,5)-H); 7.08 (2H, t, J = 8.8 Hz, Ar(2,6)-H); 4.46 – 4.40 (1H, m, Phe(4F)-α-CH); 3.77 – 3.67 (2H, m, Gly-CH2); 3.27 – 3.23 (1H, m, Glu-α-CH); 3.07 (1H, dd, J = 13.6, 3.6 Hz, Phe(4F)-β-CH2); 2.72 (1H, dd, J = 13.6, 9.7 Hz, Phe(4F)-β-CH2); 2.33 – 2.23 (1H, m, Glu-γ-CH2); 2.16 – 2.09 (1H, m, Glu-γ-CH2); 1.88 – 1.79 (1H, m, Glu-β-CH2); 1.77 – 1.69 (1H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ ppm 171.6 (C=O), 171.5 (C=O), 171.0 (C=O), 169.9 (C=O),159.9 (d, J = 242.8 Hz Ar(C4)-F), 134.5 (d, J = 3.2 Hz, Ar(C1)-C), 130.9 (d, J = 8.2 Hz, Ar(C2,6)-H), 114.7 (d, J = 22.0 Hz, Ar(C3,5)-H), 54.1 (Gly-α-CH), 52.9 (Phe(4F)-α-CH), 41.0 (Gly-CH2), 36.2 (Phe(4F)-β-CH2), 31.3 (Glu-γ-CH2), 26.7 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -116.90 (s); HRMS m/z (ES+) [Found [M+H]+ 370.1413; C16H20FN3O6 requires [M+H]+ 370.1409]; m/z (ES-) 368.2 ([M-H]-, 100%); HPLC RT 4.23 min; >95% content. 259 7.6.2 Synthesis of Glutathione S-conjugate esters S-(4-Bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione (26) To S-(4-bromobenzyl)-glutathione (3) (0.20 g, 0.42 mmol, 1.00 eq) in 10% sodium carbonate (aq) (1.3 mL) at 0 °C was added N-(9-fluorenylmethoxycarbonyloxy)-succinimide (0.14 g, 0.42 mmol, 1.00 eq) in acetone (1 mL). The reaction was stirred at 0 °C for 2.5 h and warmed to room temperature for 16 h. The acetone was removed in vacuo and the aqueous layer acidified to pH 2 with conc. HCl.183 The resulting white precipitate was filtered and washed with ethyl acetate and diethyl ether and dried under high vacuum to afford S-(4-bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione 26 as a colourless solid (0.26 g, 90%): Rf 0.88 (MeOH: H2O: Acetic Acid, 1:1:0.1, RP C-18 TLC), 0.04 (CHCl3:MeOH, 9:1); mp 220 °C (dec.); vmax (neat)/cm-1 3307.49 (br), 3260.99 (br), 1714.90 (s), 1686.87 (s), 1643.69 (s), 1533.43 (s), 758.61 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 12.64 (2H, br. s, OH); 8.42 (1H, t, J = 5.8 Hz, Gly-NH); 8.18 (1H, d, J = 8.5 Hz, Cys-NH); 7.90 (2H, d, J = 7.3 Hz, Fmoc-ArH); 7.73 (2H, d, J = 7.3 Hz, Fmoc-ArH); 7.70 (1H, d, J = 7.3 Hz, Fmoc-NH); 7.49 (2H, d, J = 8.2 Hz, Ar(3,5)-H); 7.43 (2H, t, J = 7.5 Hz, Fmoc-ArH); 7.34 (2H, t, J = 7.3 Hz, Fmoc-ArH); 7.29 (2H, d, J = 8.2 Hz, Ar(2,6)-H); 4.57 (1H, ddd, J = 14.1, 8.5, 5.1 Hz, Cys-α-CH); 4.31 – 4.15 (3H, m, Fmoc-CH2, Fmoc-CH); 4.14 – 3.96 (1H, m, Gly-CH); 3.77 (1H, d, J = 5.6 Hz, Glu-α-CH); 3.73 (2H, s, Ar-CH2); 2.74 (1H, dd, J = 14.1, 5.1 Hz, Cys-β-CH2); 2.61 – 2.54 (1H, m, Cys-β-CH2); 2.27 (2H, t, J = 7.8 Hz, Glu-ϒ-CH2); 2.07 – 1.98 (1H, m, Glu-β-CH); 1.88 – 1.75 (1H, m, Glu-β-CH); 13C NMR (125 MHz, DMSO-d6) δ 173.6 (C=O), 171.5 (C=O), 170.9 (C=O), 170.6 (C=O), 156.1 (C=O), 143.8 (Fmoc-ArC-C), 140.6 (Fmoc-ArC-C), 138.0 (Ar-(C4)-Br), 131.1 (Ar(C3,5)-H), 131.1 (Ar(C2,6)-H), 127.6 (Fmoc-ArC-H), 127.0 (Fmoc-ArC-H), 125.2 (Fmoc-ArC-H), 120.1 (Fmoc-ArC-H), 119.8 (Ar(C1)-C), 65.6 (Fmoc-CH), 53.4 (Gly-α-CH), 51.6 (Cys-α-CH), 46.5 (Fmoc-CH2), 40.6 (Gly-CH2), 34.2 (Ar-CH2), 33.1 (Cys-β-CH2), 31.6 (Glu-ϒ-CH2), 26.7 (Glu-β-CH2); HRMS m/z (ES+) [Found 260 ([M79Br+NH4]+ / [M81Br+NH4]+ 100%, 715.1432 / 717.1411); C32H32BrN3O8S.NH4 requires [M79Br+NH4]+ / [M81Br+NH4]+ 715.1432 / 717.1416]. S-(4-Bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione diethyl ester (27) To S-(4-bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione (26) (0.3 g, 0.42 mmol, 1.00 eq) in ethanol (18 mL) was added thionyl chloride (0.15 g, 0.09 mL, 1.28 mmol, 3.00 eq) dropwise. The reaction was allowed to stir at room temperature for 23 h.121 The crude reaction mixture was filtered, the precipitate collected and washed with diethyl ether to afford S-(4-bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione diethyl ester 27 as an off white solid (0.23 g, 70%): Rf 0.75 (9:1 DCM:MeOH); mp 184 – 185 °C; vmax (neat)/cm-1 3297.33 (br), 1692.78 (s), 1637.79 (s), 1533.10 (s), 1298.05 (s), 738.91 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.55 (0.7H, t, J = 5.6 Hz, Gly-NH); 8.39 (0.3H, t, J = 5.6 Hz, Gly-NH); 8.19 (1H, d, J = 8.1 Hz, Cys-NH); 7.90 (2H, d, J = 7.5 Hz, Fmoc-ArH); 7.82 (1H, d, J = 7.5 Hz, Fmoc-NH); 7.75 – 7.71 (2H, m, Fmoc-ArH); 7.49 (2H, d, J = 7.8 Hz, Ar(3,5)-H); 7.43 (2H, t, J = 7.5 Hz, Fmoc-ArH); 7.37 – 7.32 (2H, m, Fmoc-ArH); 7.29 (2H, d, J = 7.8 Hz, Ar(2,6)-H); 4.57 (1H, dd, J = 13.8, 8.1 Hz, Cys-α-CH); 4.34 – 4.20 (3H, m, Fmoc-CH2, CH); 4.13 – 4.02 (4H, m, O-CH2); 3.84 (2H, d, J = 5.9 Hz, Gly-CH2); 3.73 (2H, s, Ar-CH2); 3.63 (1H, d, J = 7.5 Hz, Glu-α-CH); 2.72 (1H, dd, J = 13.8, 5.0 Hz, Cys-β-CH2); 2.56 – 2.53 (1H, m, Cys-β-CH2); 2.26 (2H, t, J = 7.5 Hz, Glu-ϒ-CH2); 2.05 – 1.96 (1H, m, Glu-β-CH2); 1.87 – 1.78 (1H, m, Glu-β-CH2); 1.21 – 1.15 (6H, m, CH3); 13C NMR (125 MHz, DMSO-d6) δ 172.1 (C=O), 171.3 (C=O), 170.7 (C=O), 169.4 (C=O), 156.0 (C=O), 143.7 (Fmoc-ArC-C), 140.6 (Fmoc-ArC-C), 137.9 (Ar-(C4)-Br), 131.1 (Ar(C3,5)-H), 131.1 (Ar(C2,6)-H), 127.6 (Fmoc-ArC-H), 127.0 (Fmoc-ArC-H), 125.2 (Fmoc-ArC-H), 120.1 (Ar(C1))-C), 119.8 (Fmoc-ArC-H), 65.6 (Fmoc-CH), 60.4 (O-CH2-CH3), 60.4 (O-CH2), 53.5 (Gly-α-CH), 51.6 (Cys-α-CH), 46.5 (Fmoc-CH2), 40.7 (Gly-CH2), 34.2 (Ar-CH2), 33.0 (Cys-CH2), 31.4 (Glu-ϒ-CH2), 26.6 (Glu-β-CH2), 14.0 (CH2-CH3), 14.0 (CH2-CH3); HRMS m/z (ES+) [Found [M79Br+H]+ / [M81Br+H]+ 754.1795 / 261 756.1774; requires C36H40BrN3O8S [M79Br+H]+ / [M81Br+H]+ 754.1792 / 756.1778]; m/z (ES+) 776.4 / 778.4 ([M79Br+H]+ / [M81Br+H]+ , 93 / 100%). S-(4-Bromobenzyl)-glutathione diethyl ester (29) S-(4-Bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione diethyl ester (27) (0.08 g, 0.10 mmol, 1.00 eq) in 10% v/v piperidine/DMF (1 mL) was stirred at room temperature for 1.5 h after which time the TLC showed consumption of starting material. The crude reaction mixture was concentrated in vacuo at room temperature to afford a colourless solid. The solid was triturated with diethyl ether, filtered and dried under vacuum to afford S-(4-bromobenzyl)-glutathione diethyl ester 29 as a colourless solid (0.04 g, 86%): Rf 0.17 (MeOH:water: acetic acid, ninhydrin, 1:1:0.1, RP C-18 TLC); mp 114 – 119 °C [Lit184,185 107 – 109 °C]; [𝛼]54620 -29.41 (c 0.06 DMSO); vmax (neat)/cm-1 3274.32 (br), 1740.36 (s), 1635.47 (s), 1535.21 (s), 1210.24 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.54 (1H, t, J = 5.6 Hz, Gly-NH); 8.20 (1H, d, J = 8.6 Hz, Cys-NH); 7.50 (2H, d, J = 8.2 Hz, Ar(3,5)-H); 7.30 (2H, d, J = 8.2 Hz, Ar(2,6)-H); 4.55 (1H, ddd, J = 13.6, 8.6, 5.2 Hz, Cys-α-CH); 4.14 – 4.04 (4H, m, OCH2); 3.84 (2H, d, J = 5.6 Hz, Gly-CH2); 3.74 (2H, s, Ar-CH2); 3.63 (1H, d, J = 7.8 Hz, Glu-α-CH); 3.03 – 2.98 (1H, m, Cys-β-CH2); 2.74 (1H, dd, J = 13.6, 5.2 Hz, Cys-β-CH2); 2.29 – 2.24 (2H, m, Glu-ϒ-CH2); 1.93 – 1.83 (1H, m, Glu-β-CH2); 1.72 – 1.52 (1H, m, Glu-β-CH2); 1.23 – 1.15 (6H, dt, J = 15.2, 7.2 Hz, CH3); 13C NMR (125 MHz, DMSO-d6) δ 171.9 (C=O), 170.8 (C=O), 170.3 (C=O), 169.5 (C=O), 137.9 (Ar-(C4)-Br), 131.1 (Ar(C3,5)-H), 131.1 (Ar(C2,6)-H), 119.8 (Ar(C1)-C), 60.4 (O-CH2CH3), 60.2 (O-CH2CH3), 53.3 (Gly-α-CH), 51.6 (Cys-α-CH), 39.9 (Gly-CH2), 34.2 (Ar-CH2), 32.9 (Cys-β-CH2), 31.5 (Glu-ϒ-CH2), 22.2 (Glu-β-CH2), 14.0 (CH2-CH3), 14.0 (CH2-CH3); HRMS m/z (ES+) [Found [M79Br+H]+ / [M81Br+H]+ 532.1105 / 534.1082; C21H30BrN3O6S requires [M79Br+H]+ / [M81Br+H]+ 532.1111 / 534.1093]; m/z (ES+) 532.2 / 534.2 ([M79Br+H]+ / [M81Br+H]+ , 98 / 100%). The data are in agreement with the literature.185 262 S-(4-Bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione dipropyl ester (28) To S-(4-bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione (26) (0.3 g, 0.42 mmol, 1.00 eq) in n-propanol (18 mL) was added thionyl chloride (0.15 g, 0.09 mL, 1.28 mmol, 3.00 eq) dropwise. The reaction was allowed to stir at room temperature for 23 h.108 The crude reaction mixture was filtered and washed with diethyl ether to afford S-(4-bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione dipropyl ester 28 as a colourless solid (0.28 g, 83%): Rf 0.92 (CHCl3: MeOH, 9:1); mp 179 – 180 °C; vmax (neat)/cm-1 3292.19 (br), 1746.68 (m), 1728.08 (m), 1694.10 (s), 1641.33 (s), 1538.35 (s), 1038.05 (s), 755.98 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.55 (1H, t, J = 5.3 Hz, Gly-NH); 8.20 (1H, d, J = 8.5 Hz, Cys-NH); 7.90 (2H, d, J = 7.5 Hz, Fmoc-ArH); 7.84 (1H, d, J = 7.5 Hz, Fmoc-NH); 7.72 (2H, d, J = 7.5 Hz, Fmoc-ArH); 7.49 (2H, d, J = 8.2 Hz, Ar(3,5)-H); 7.43 (2H, t, J = 7.5 Hz, Fmoc-ArH); 7.34 (2H, t, J = 7.4 Hz, Fmoc-ArH); 7.29 (2H, d, J = 8.4 Hz, Ar(2,6)-H); 4.57 (1H, ddd, J = 13.7, 8.5, 5.3 Hz, Cys-α-CH); 4.33 – 4.21 (3H, m, Fmoc-CH2, CH); 4.10 – 4.04 (1H, m, Cys-β-CH2); 4.03 – 3.97 (4H, m, O-CH2); 3.85 (2H, d, J = 5.9 Hz, Gly-CH2); 3.73 (2H, s, Ar-CH2); 2.71 (1H, dd, J = 13.7, 5.3 Hz, Cys-β-CH2); 2.53 – 2.51 (1H, m, Cys-β-CH2); 2.27 (2H, t, J = 7.7 Hz, Glu-ϒ-CH2); 2.05 – 1.97 (1H, m, Glu-β-CH2); 1.88 - 1.80 (1H, m, Glu-β-CH2); 1.60 – 1.53 (4H, m, O-CH2CH2CH3); 0.86 (6H, t, J = 7.2 Hz, O-CH2CH2CH3); 13C NMR (125 MHz, DMSO-d6) δ 172.2 (C=O), 171.3 (C=O), 170.7 (C=O), 169.5 (C=O), 156.0 (C=O), 143.7 (Fmoc-ArC-C), 140.6 (Fmoc-ArC-C), 137.9 (Ar-(C4)-Br), 131.1 (Ar(C3,5)-H), 131.1 (Ar(C2,6)-H), 127.6 (Fmoc-ArC-H), 127.0 (Fmoc-ArC-H), 125.2 (Fmoc-ArC-H), 120.0 (Fmoc-ArC-H), 119.8 (Ar(C1)-C), 65.8 (Fmoc-CH), 65.8 (O-CH2CH2CH3), 65.6 (O-CH2CH2CH3), 53.5 (Gly-α-CH), 51.6 (Cys-α-CH), 46.5 (Fmoc-CH2), 40.7 (Gly-CH2), 34.2 (Ar-CH2), 33.0 (Cys-β-CH2), 31.4 (Glu-ϒ-CH2), 26.6 (Glu-β-CH2), 21.4 (CH2), 21.4 (CH2), 10.1 (CH2CH2CH3); HRMS m/z (ES+) [Found [M79Br+H]+ / [M81Br+H]+ 782.2112 / 784.2092; C38H44BrN3O8S requires [M79Br+H]+ / [M81Br+H]+ 782.2105 / 784.2092]; m/z (ES+) 804.5 / 806.4 ([M79Br+Na]+ / [M81Br+Na]+ , 95 / 100%). 263 S-(4-Bromobenzyl)-glutathione dipropyl ester (30) S-(4-Bromobenzyl)-N-fluorenylmethyloxycarbonyl-glutathione dipropyl ester (28) (0.15 g, 0.19 mmol, 1.00 eq) in 10% v/v piperidine/DMF (2 mL) was stirred at room temperature for 2 h after which time the TLC showed consumption of starting material. The crude reaction mixture was concentrated in vacuo at room temperature to afford a colourless solid. The solid was triturated with diethyl ether, filtered and dried under vacuum to afford S-(4-bromobenzyl)-glutathione dipropyl ester 30 as a cream solid (0.057 g, 53%): Rf 0.21 (MeOH: water: Acetic acid, ninhydrin, 1:1:0.1, RP C-18 TLC); mp 119 – 124 °C [Lit185 118 – 120 °C]; [𝛼]54620 -24.65 (c 0.56 DMSO); vmax (neat)/cm-1 3278.41 (br), 1742.95 (m), 1719.40 (m), 1636.20 (s), 1537.94 (m), 1205.32 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.56 (1H, t, J = 5.9 Hz, Gly-NH); 8.21 (1H, d, J = 8.8 Hz, Cys-NH); 7.50 (2, d, J = 8.2 Hz, Ar(3,5)-H); 7.29 (2H, d, J = 8.2 Hz, Ar(2,6)-H); 4.55 (1H, ddd, J = 13.6, 8.8, 5.0 Hz, Cys-α-CH); 4.05 – 3.96 (4H, m, O-CH2CH2); 3.84 (2H, d, J = 5.9 Hz, Gly-CH2); 3.74 (2H, s, Ar-CH2); 2.74 (1H, dd, J = 13.6, 5.0 Hz, Cys-β-CH2); 2.53 – 2.51 (1H, m, Cys-β-CH2); 2.29 – 2.24 (2H, m, Glu-ϒ-CH2) 1.92 – 1.84 (1H, m, Glu-β-CH); 1.69 – 1.62 (1H, m, Glu-β-CH); 1.62 – 1.53 (4H, m, OCH2CH2); 0.88 (6H, dt, J = 11.6, 7.4 Hz, CH3); 13C NMR (125 MHz, DMSO-d6) δ 175.3 (C=O), 171.9 (C=O), 170.8 (C=O), 169.5 (C=O), 137.9 (Ar-(C4)-Br), 131.1 (Ar(C3,5)-H), 131.1 (Ar(C2,6)-H), 119.8 (Ar(C1)-C), 65.8 (O-CH2CH2CH3), 65.5 (O- CH2CH2CH3), 53.5 (Glu-α-CH), 51.6 (Cys-α-CH), 40.7 (Gly-CH2), 34.2 (Ar-CH2), 32.9 (Cys-β-CH2), 31.6 (Glu-ϒ-CH2), 30.2 (Glu-β-CH2), 21.5 (CH2), 21.4 (CH2), 10.2 (O-CH2CH2CH3), 10.1 (O-CH2CH2CH3); HRMS m/z (ES+) [Found [M79Br+H]+ / [M81Br+H]+ 560.1420 / 562.1399; C23H34BrN3O6S requires [M79Br+H]+ / [M81Br+H]+ 560.1424 / 562.1407]; m/z (ES+) 560.3 / 562.2 ([M79Br+H]+ / [M81Br+H]+ 98 / 100%). The data are in agreement with the literature.185 264 7.6.3 Synthesis of γ-Glutamyl-L-azidoalaninylglycine (GAz) Imidazole-1-sulfonyl azide hydrochloride (33) Sodium azide (5.00 g, 76.91 mmol, 1.00 eq) was suspended in anhydrous acetonitrile (76 mL) and cooled to 0 °C. Sulfuryl chloride (6.23 mL, 10.38 g, 76.91 mmol, 1.00 eq,) was added dropwise over 20 min. The reaction was stirred at room temperature over 19 h. After which time the reaction was cooled to 0 °C and imidazole (10.47 g, 153.82 mmol, 2.00 eq) was added portion wise over 20 min. After stirring at room temperature for 3 h, the reaction mixture was diluted with ethyl acetate (200 mL) and the organic layer washed with water (2 x 200 mL) and saturated bicarb (2 x 200 mL), dried over anhydrous MgSO4, filtered and cooled to 0 °C. Acetyl chloride (8.20 mL, 9.04 g, 115.37 mmol, 1.50 eq) was added dropwise to ice cold anhydrous ethanol (35 mL) over 15 min, before stirring for 10 min at 0 °C. This solution was then added to the ice-cold ethyl acetate solution over 20 min. The resulting precipitate was filtered, washed with ethyl acetate (2 x 100 mL) and dried under suction for 30 min to afford imidazole-1-sulfonyl azide hydrochloride 33 as a colourless solid (11.33 g, 70%): Rf 0.44 (50:50 MeOH: water, RP C-18 TLC); mp 90 – 92 °C [Lit186 96 – 98 °C]; vmax (neat)/cm-1 3101.83 (br), 2491.36 (m), 2167.94 (s), 1581.08 (m), 1423.31 (s), 1157.47 (s), 581.03 (s); 1H NMR (500 MHz, D2O) δ ppm 7.45 (1H, dd, J = 1.8, 0.9 Hz, Ar(2)-H); 7.88 (1H, t, J = 1.8 Hz, Ar(5)-H); 9.10 – 9.08 (1H, m, Ar(4)-H); 13C NMR (125 MHz, D2O) δ 137.7 (Ar(C4)-H), 125.0 (Ar(C5)-H), 119.6 (Ar(C2)-H); HRMS m/z (ES+) [Found [M+H]+ 174.0080; C3H4ClN5O2S requires [M+H]+ 174.0080]; m/z (ES+) 173.95 ([M+H]+; 91%). The data are in agreement with the literature.186-187,138 N-Fluorenylmethyloxycarbonyl-L-2, 3-diaminopropionic acid (35) To [bis(trifluoroacetoxy)iodo]benzene (9.10 g, 21.16 mmol, 1.50 eq) in N,N-dimethylformamide (66 mL) and water (33 mL), N-fluorenylmethyloxycarbonyl-L-asparagine (5.00 g, 14.11 mmol, 1.00 eq) was added. After stirring at room temperature 265 for 15 min, pyridine (2.39 mL, 29.60 mmol, 2.10 eq) was added and the reaction stirred at room temperature for a further 21 h. The reaction mixture went from a pale suspension to a clear yellow solution. After which time TLC showed consumption of the starting material. The reaction mixture was concentrated in vacuo. The resulting oily residue was diluted with water (76 mL), acidified with conc. HCl (1.5 mL) and washed with diethyl ether (3 x 70 mL). The aqueous layer was adjusted to pH 6 with 2M NaOH, and the resulting precipitate was filtered, washed with water (3 x 45 mL), ice cold ethanol (15 mL) and diethyl ether (5 x 45 mL) and dried under vacuum to afford N-fluorenylmethyloxycarbonyl-L-2, 3-diaminopropionic acid 35 as a beige solid (2.96 g, 64%): Rf 0.12 (50:50 MeOH:H2O, RP C-18 TLC); Rf 0.04 (EtOAc); mp 138 °C (dec.); [𝛼]54620 +23.33 (c 0.6 DMSO) (Lit141 [𝛼]𝐷27 +27.7 (c 0.41 DMSO); vmax (neat)/cm-1 3291.17 (br), 2960.55 (br), 1685.11 (m), 1407.71 (m), 735.98 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 7.90 (2H, d, J = 7.4 Hz, FmocAr-CH); 7.72 (2H, dd, J = 7.4, 3.1 Hz, FmocAr-CH); 7.43 (2H, t, J = 7.4 Hz, FmocAr-CH); 7.37 – 7.32 (2H, m, FmocAr-CH); 6.89 (1H, d, J = 5.9 Hz, NH); 4.20 – 4.35 (3H, m, Fmoc-CH, Fmoc-CH2); 3.70 – 3.66 (1H, m, α-CH); 3.06 (1H, dd, J = 10.7, 4.4 Hz, β-CH2); 2.87 – 2.75 (1H, m, β-CH2); 13C NMR (125MHz, DMSO-d6) δ 155.6 (OC=ONH), 143.8 (OC=O), 142.5 (FmocArC-C), 140.6 (FmocArC-C), 139.3 (FmocArC-C), 137.3 (FmocArC-C), 128.9 (FmocArC-H), 127.6 (FmocArC-H), 127.2 (FmocArC-H), 127.0 (FmocArC-H), 125.1 (FmocArC-H), 121.3 (FmocArC-H), 120.0 (FmocArC-H), 109.7 (FmocArC-H), 65.6 (Fmoc-CH, α-CH), 46.6 (Fmoc-CH2, β-CH2); HRMS m/z (ES+) [Found [M+H]+ 327.1342; C18H18N2O4 requires [M+H]+ 327.1339]; m/z (ES+) 327.25 ([M+H]+; 81.8%). The data are in agreement with the literature.141 N-Fluorenylmethoxycarbonyl-L-azidoalanine (36) N-Fluorenylmethyloxycarbonyl-L-2, 3-diaminopropionic acid (35) (1.5 g, 4.59 mmol, 1.00 eq) was dissolved in a biphasic mixture of water (23 mL), methanol (45 mL) and dichloromethane (36 mL). CuSO4.5H2O (6.90 mg, 0.03 mmol, 0.006 eq) and imidazole-1-sulfonyl azide hydrochloride (33) (2.98 g, 14.25 mmol, 3.10 eq) were added. The reaction mixture was adjusted to pH 9 with a saturated aqueous potassium carbonate 266 solution and stirred vigorously at room temperature for 23 h. TLC analysis showed the complete consumption of starting material. The crude reaction mixture was diluted with dichloromethane (45 mL) and the aqueous phase isolated. The organic layer was washed with saturated bicarb (2 x 70 mL). The combined aqueous layers were washed with diethyl ether (2 x 70 mL) and acidified to pH 2 with conc. HCl. The aqueous layer was extracted with diethyl ether (3 x 90 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo to obtain the N-fluorenylmethoxycarbonyl-L-azidoalanine 36 as a yellow gum (1.25 g, 77%): Rf 0.36 (50:50 MeOH:H2O, RP C-18 TLC); Rf 0.35 (EtOAc); mp 117 – 120 °C; [𝛼]54620 -4.23 (c 0.47 DMF) (Lit141 [𝛼]𝐷27 -10.3 (c 1.0 DMF); vmax (neat)/cm-1 3306.48 (m), 2103.96 (m), 1687.22 (s), 1253.05 (s), 733.66 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.80 (2H, d, J = 7.0 Hz, FmocAr-CH); 7.63 (2H, d, J = 7.0 Hz, FmocAr-CH); 7.44 (2H, t, J = 7.0 Hz, FmocAr-CH); 7.35 (2H, t, J = 7.0 Hz, FmocAr-CH); 5.66 (1H, d, J = 7.2 Hz, NH); 4.65 – 4.56 (1H, m, Aza-α-CH); 4.50 – 4.43 (2H, m, Fmoc-CH2); 4.27 (1H, t, J = 6.6 Hz, Fmoc-CH); 3.88 – 3.81 (2H, m, Aza-β-CH2); 13C NMR (125 MHz, CDCl3) δ 173.2 (HOC=O), 155.8 (NHC=O), 143.6 (FmocArC-C), 143.5 (FmocArC-C), 141.3 (FmocArC-C), 127.8 (FmocArC-H), 127.1 (FmocArC), 125.0 (FmocArC-H), 124.5 (FmocArC-H), 120.2 (FmocArC-H), 120.0 (FmocArC-H), 67.5 (Fmoc-CH2), 53.9 (Aza-α-CH), 51.9 (Aza-β-CH2), 47.0 (Fmoc-CH); HRMS m/z (ES+) [Found [M+H]+ 353.1249 C18H16N4O4 requires [M+H]+ 353.1244]; m/z (ES+) 353 ([M+H]+; 91%). The data are in agreement with the literature.141 N-Fluorenylmethoxycarbonyl-L-azidoalaninylglycine tert-butyl ester (37) To N-fluorenylmethyloxycarbonyl-L-azidoalanine (36) (1.20 g, 3.41 mmol, 1.00 eq), glycine-tert-butyl ester hydrochloride (0.59 g, 3.58 mmol, 1.05 eq) and HBTU (1.36 g, 3.58 mmol, 1.05 eq) in anhydrous N,N-dimethylformamide (17 mL) was added diisopropylethylamine (1.78 mL, 10.22 mmol, 3.00 eq) dropwise. The reaction was stirred at room temperature for 3.5 h, after which time the reaction was complete by TLC analysis. Ethyl acetate (70 mL) was added and then washed with water (70 mL), 10% citric acid (70 mL), saturated bicarb (70 mL) and brine (2 x 50 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was purified by silica gel 267 column chromatography (0 – 100% Ethyl acetate/Hexane) to afford N-fluorenylmethoxycarbonyl-L-azidoalaninylglycine tert-butyl ester 37 as a yellow solid (0.88 g, 55%): Rf 0.73 (EtOAc); mp 84 – 86 °C; [𝛼]54620 +28.11 (c 0.67 CHCl3); vmax (neat)/cm-1 3293.34 (m), 2977.84 (br), 2095.72 (m), 1649.45 (s), 1229.95 (s), 1155.34 (s), 737.43 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.80 (2H, d, J = 7.4 Hz, FmocAr-H); 7.62 (2H, d, J = 7.4 Hz, FmocAr-H); 7.44 (2H, t, J = 7.4 Hz, FmocAr-H); 7.35 (2H, t, J = 7.4 Hz, FmocAr-H); 6.68 (1H, br. s, Gly-NH); 5.61 (1H, d, J = 6.6 Hz, Aza-NH); 4.49 (2H, d, J = 5.6 Hz, Fmoc-CH2); 4.44 – 4.38 (1H, m, Fmoc-CH); 4.05 – 3.84 (4H, m, Gly-CH2, Aza-β-CH2); 3.63 – 3.56 (1H, m, Aza-α-CH); 1.50 (9H, s, CH3); 13C NMR (125 MHz, CDCl3) δ 171.1 (C=O), 168.8 (C=O), 155.9 (C=O), 143.5 (FmocArC-C), 141.3 (FmocArC-C), 127.8 (FmocArC-H), 127.1 (FmocArC-H), 125.0 (FmocArC-H), 120.0 (FmocArC-H), 82.7 (C), 67.3 (Fmoc-CH2), 53.9 (Aza-α-CH), 52.0 (Aza-β-CH2), 47.1 (Fmoc-CH), 42.2 (Gly-CH2), 28.0 (CH3); HRMS m/z (ES+) [Found [M+H]+ 466.2078 C24H27N5O5 requires [M+H]+ 466.2085; [M+Na]+ 488.1889 C24H27N5O5Na requires M+ 488.19]; m/z (ES+) 488.3 ([M+Na]+; 88%). L-Azidoalaninylglycine tert-butyl ester (38) To N-fluorenylmethoxycarbonyl-L-azidoalaninylglycine tert-butyl ester (37) (0.88 g, 1.89 mmol, 1.00 eq) was added 10% v/v piperidine/dimethylformamide (19 mL). The reaction mixture was stirred at room temperature for 30 min, after which time TLC analysis showed complete consumption of the starting material. The crude reaction mixture was concentrated under high vacuum to give a yellow solid, purified by silica gel column chromatography (0 – 100% Ethyl acetate in hexane to 10% methanol in ethyl acetate) to afford L-azidoalaninylglycine tert-butyl ester 38 as a yellow gum (0.32 g, 70% yield): Rf 0.53 (1:9 methanol:ethyl acetate); 1H NMR (400 MHz, CDCl3) δ ppm 7.82 (1H, br. s, Gly-NH); 3.98 – 3.95 (2H, m, Aza-β-CH2); 3.70 (2H, d, J = 5.2 Hz, Gly-CH2); 3.57 (1H, t, J = 5.3 Hz, Aza-α-CH); 1.50 (9H, s, CH3); 13C NMR (100 MHz, CDCl3) δ 171.9 (C=O), 168.8 (C=O), 82.3 (C), 55.2 (Aza-β-CH2), 54.5 (Aza-α-CH), 41.8 (Gly-CH2), 28.0 (CH3); HRMS m/z (ES+) [Found [M+H]+ 244.1406 C9H17N5O3 requires [M+H]+ 244.1404]; m/z (ES+) 273.0 ([M+CH3OH-2H] +; 100%). 268 N-Boc-α-tert-butyl-ester-γ-glutamyl-L-azidoalaninylglycine tert-butyl ester (39) To L-azidoalaninylglycine tert-butyl ester (38) (0.32 g, 1.32 mmol, 1.00 eq), N-Boc-L-glutamic acid-α-tert butyl ester (0.39 g, 1.32 mmol, 1.00 eq) and HBTU (0.52 g, 1.38 mmol, 1.05 eq) in N,N-dimethyformamide (7 mL), N,N-diisopropylethylamine (0.69 mL, 3.95 mmol, 3.00 eq) was added dropwise. The reaction was allowed to stir at room temperature for 3 h, after which time TLC analysis showed consumption of starting material. Ethyl acetate (80 mL) was added and the organic layer washed with water (80 mL), 10% citric acid (80 mL), saturated bicarb (80 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was further purified by silica gel column chromatography (0 – 100% ethyl acetate in 40-60 petroleum ether) to afford N-Boc-α-tert-butyl-ester-γ-glutamyl-L-azidoalaninylglycine tert-butyl ester 39 as a yellow gum (0.59 g, 84%): Rf 0.75 (1:1 ethyl acetate:petroleum ether 40–60); [𝛼]54620 +38.46 (c 0.05 CHCl3); vmax (neat)/cm-1 3306.11 (br), 2978.52 (w), 2104.29 (m), 1651.43 (br), 1366.15 (m), 1149.20 (s), 844.36 (m); 1H NMR (500 MHz, CDCl3) δ ppm 6.94 (1H, d, J = 7.5 Hz, Aza-NH); 6.84 (1H, t, J = 4.1 Hz, Gly-NH); 5.22 (1H, d, J = 7.6 Hz, Aza-NH); 4.56 (1H, dd, J = 12.1, 7.6 Hz, Aza-α-CH); 4.18 – 4.10 (1H, m, Glu-α-CH); 3.96 – 3.75 (3H, m, Gly-CH2, Aza-β-CH2); 3.58 (1H, dd, J = 12.1, 5.3 Hz, Aza-β-CH2); 2.30 (2H, t, J = 7.0 Hz, Glu-ϒ-CH2); 2.16 – 2.08 (1H, m, Glu-β-CH); 1.85 – 1.76 (1H, m, Glu-β-CH); 1.40 (18H, s, CH3); 1.38 (9H, s, CH3); 13C NMR (125 MHz, CDCl3) δ 172.7 (C=O), 169.1 (C=O), 168.4 (C=O), 156.0 (C=O), 82.5 (C), 82.5 (C), 80.2 (C), 53.1 (Glu-α-CH), 52.4 (Aza-α-CH), 51.4 (Aza-β-CH2), 42.1 (Gly-CH2), 38.6 (Glu-ϒ-CH2), 32.4 (Glu-β-CH2), 28.3 (CH3), 28.0 (CH3), 27.9 (CH3); HRMS m/z (ES+) [Found [M+H]+ 529.2972; C23H40N6O8 requires [M+H]+ 529.2980]; m/z (ES+) 551.3 ([M+Na]+, 100%). 269 γ-Glutamyl-L-azidoalaninylglycine (32, GAz) N-Boc-α-tert-butyl-ester-γ-glutamyl-L-azidoalaninylglycine tert-butyl ester (39) (0.58 g, 1.09 mmol, 1.00 eq) and triethylsilane (1.05 mL, 6.58 mmol, 6.00 eq) in 80% v/v trifluoroacetic acid/dichloromethane (10.9 mL) was stirred at room temperature for 4 h, after which time TLC analysis showed complete consumption of the starting material. The reaction mixture was cooled to 0 °C and cold diethyl ether (50 mL) added. The resulting colourless precipitate was isolated by filtration. The white precipitate turned to a yellow gum, this was dissolved in water, concentrated in vacuo and purified by reverse phase chromatography (5 – 95% 0.1% TFA/Acetonitrile in 0.1% TFA/water) to give γ-glutamyl-L-azidoalaninylglycine 32 as an off white solid (0.26 g, 74%): Rf 0.13 (EtOAc:Petroleum ether 40–60 1:1); Rf 0.84 (MeOH:water 1:1, ninhydrin); [𝛼]54620 +12.22 (c 0.9 H2O); vmax (neat)/cm-1 3269.98 (br), 3043.73 (br), 2943.52 (br), 2111.34 (m), 1646.13 (s), 1523.78 (s), 1183.96 (s), 1133.57 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.43 – 8.49 (2H, m, Aza-NH, Gly-NH); 8.27 (2H, br. s, NH2); 4.60 (1H, ddd, J = 12.7, 8.0, 4.5 Hz, Aza-α-CH); 3.92 – 3.98 (1H, m, Glu-α-CH); 3.79 (1H, d, J = 5.6 Hz, Gly-CH); 3.78 (1H, d, J = 5.6 Hz, Gly-CH); 3.55 (1H, dd, J = 12.7, 4.5 Hz, Aza-β-CH2); 3.46 (1H, dd, J = 12.7, 8.0 Hz, Aza-β-CH2); 2.45 – 2.31 (2H, m, Glu-ϒ-CH2); 2.10 – 1.95 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, DMSO-d6) δ 171.1 (C=O), 170.8 (C=O), 170.8 (C=O), 169.1 (C=O), 52.0 (Glu-CH), 51.6 (Aza-α-CH), 51.2 (Aza-β-CH2), 40.8 (Gly-CH2), 30.5 (Glu-ϒ-CH2), 25.7 (Glu-β-CH2); HRMS m/z (ES+) [Found [M+H]+ 317.1208; C10H16N6O6 requires [M+H]+ 317.1204]; m/z (ES+) 317.00 ([M+H]+; 83.2%). 270 7.6.4 Synthesis of Alkynes identified in the biased kTGS Alkynes were either synthesised as described or purchased from chemical suppliers: Enamine, Sigma, VWR or Alfa Aesar. 2-Phenyl-1-(prop-2-yn-1-yl)-1H-imidazole (40, AlkC2) To 2-phenylimidazole (0.5 g, 3.46 mmol, 1.00 eq) in N,N-dimethylformamide (10 mL) was added potassium carbonate (0.96 g, 6.93 mmol, 2.00 eq) and propargyl bromide 80% in toluene (0.46 mL, 5.20 mmol, 1.50 eq). The reaction was stirred at room temperature for 16 h, after which time the reaction mixture was quenched with water (1 mL), extracted with diethyl ether three times, washed with brine, dried over anhydrous magnesium sulphate, filtered and concentrated in vacuo. The crude material was purified by silica gel column chromatography eluting with 0 – 100% ethyl acetate in hexane to afford 2-phenyl-1-(prop-2-yn-1-yl)-1H-imidazole 40 as a yellow oil (0.25 g, 40%): Rf 0.37 (hexane: ethyl acetate, 1:1) 0.62 (5% methanol in DCM); vmax (neat)/cm-1 3286.50 (br), 3217.60 (br), 2120.25 (w), 1271.17 (s), 696.25 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.67 (2H, dd, J = 7.8, 2.0 Hz, Ar(7,11)-H); 7.54 – 7.43 (3H, m, Ar(8,9,10)-H); 7.21 (2H, dd, J = 10.0, 1.2 Hz, Ar(4,5)-H); 4.78 (2H, d, J = 2.5 Hz, CH2); 2.52 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 147.6 (Ar-(C2)-C), 130.0 (Ar-(C6)-C), 129.1 (Ar-(C5)-H), 129.0 (Ar-(C4)-H), 128.8 (Ar-(C7,11)-H), 128.7 (Ar-(C8,10)-H), 120.6 (Ar-(C9)-H), 77.6 (CH), 74.2 (C≡C-CH2), 36.6 (CH2); HRMS m/z (ES+) [Found [M+Na]+ 205.0732; C12H10N2Na requires [M+Na]+ 205.0736]; m/z (ES+) 183.2 ([M+H]+, 100%). 1-(4-(Prop-2-yn-1-yloxy)phenyl)ethan-1-one (41, AlkB2) To 4-hydroxyacetophenone (0.1 g, 0.73 mmol, 1.00 eq) in N,N-dimethylformamide (1 mL) was added potassium carbonate (0.2 g, 1.46 mmol, 2.00 eq) and propargyl bromide 80% in toluene (0.09 mL, 1.1 mmol, 1.50 eq). The reaction was stirred at room temperature for 16 h, after which time the reaction mixture was quenched with water (1 mL), extracted with diethyl ether three times, washed with brine, dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was triturated with 271 hexane, filtered and dried to afford 1-(4-(prop-2-yn-1-yloxy)phenyl)ethan-1-one 41 as an off white solid (0.075 g, 59%): Rf 0.62 (hexane: ethyl acetate, 1:1); mp 78 – 79 °C (dec.) (Lit188: 77.2 – 78 °C; Lit189: 76.3 – 76.7 (ethanol)); vmax (neat)/cm-1 3218.97 (m), 2119.88 (w), 1656.49 (s), 1238.44 (s), 823.57 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.89 (2H, d, J = 4.8 Hz, Ar(2,6)-H); 6.95 (2H, d, J = 4.8 Hz, Ar(3,5)-H); 4.69 (2H, d, J = 2.2 Hz, CH2); 2.50 (3H, s, CH3); 2.48 (1H, t, J = 2.2 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 196.7 (C=O), 161.2 (Ar-(C1)-C), 131.0 (Ar-(C5)-C), 130.5 (Ar-(C2,6)-H), 114.5 (Ar-(C3,5)-H), 77.7 (CH), 76.1 (C≡C-CH2), 55.8 (CH2), 26.4 (CH3); HRMS m/z (ES+) [Found [M+Na]+ 197.0570; C11H10O2Na requires [M+Na]+ 197.0573]; m/z (ES+) 349.0 ([2M+H]+, 100%). Data are in agreement with the literature.188,189 1-(Prop-2-yn-1-yl)indoline-2,3-dione (42, AlkA3) To isatin (0.5 g, 3.39 mmol, 1.00 eq) in N,N-dimethylformamide (10 mL) was added potassium carbonate (0.94 g, 6.79 mmol, 2.00 eq) and propargyl bromide 80% in toluene (0.45 mL, 5.09 mmol, 1.50 eq) and stirred at room temperature for 16 h, after which time the reaction mixture was quenched with water (1 mL), extracted with diethyl ether three times, washed with brine, dried over anhydrous MgSO4, filtered and concentrated. The crude material was purified by silica gel column chromatography eluting with 0 – 100% ethyl acetate in hexane to afford 1-(prop-2-yn-1-yl)indoline-2,3-dione 42 as a red solid (0.33 g, 53%): Rf 0.62 (hexane: ethyl acetate, 1:1) 0.70 (5% methanol in DCM); mp 146 – 153 °C (dec.) (Lit: 156 – 158 °C190, 158 – 160 °C191); vmax (neat)/cm-1 3262.04 (br), 2120.21 (w), 1734.21 (s), 1611.71 (s), 1371.48 (s), 674.08 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.70 – 7.65 (2H, m, Ar(6,7)-H); 7.21 (1H, t, J = 8.0 Hz, Ar(5)-H); 7.16 (1H, d, J = 8.0 Hz, Ar(4)-H); 4.57 (2H, d, J = 2.5 Hz, CH2); 2.34 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 182.5 (C=O), 157.1 (C=O), 149.6 (Ar-(C7’)-C), 138.4 (Ar-(C6)-H), 125.5 (Ar-(C5)-H), 124.2 (Ar-(C4’)-C), 117.7 (Ar-(C4)-H), 111.0 (Ar-(C7)-H), 75.6 (CH), 73.3 (C≡C-CH2), 29.4 (CH2); HRMS m/z (ES+) [Found [M+Na]+ 208.0369; C11H7NO2Na requires [M+Na]+ 208.0369]; m/z (ES+) 431.1 ([2M+ACN+Na]+, 97%). Data are in agreement with the literature.190,191 272 6-Bromo-4-(prop-2-yn-1-yl)-2H-benzo[b][1,4]oxazin-3(4H)-one (43, AlkF3) To 6-bromo-2H,1,4-benzoxazine-3-(4H)-one (0.4 g, 1.75 mmol, 1.00 eq) in N,N-dimethylformamide (5 mL) was added potassium carbonate (0.48 g, 3.50 mmol, 2.00 eq) and propargyl bromide 80% in toluene (0.23 mL, 2.63 mmol, 1.50 eq) and stirred at room temperature for 16 h, after which time the reaction mixture was quenched with water (1 mL), extracted with diethyl ether three times, washed with brine, dried over anhydrous MgSO4, filtered and concentrated to afford 6-bromo-4-(prop-2-yn-1-yl)-2H-benzo[b][1,4]oxazin-3(4H)-one 43 as a colourless solid (0.4 g, 85%): Rf 0.7 (hexane: ethyl acetate, 1:1); Rf 0.81 (methanol:DCM 0.5:9.5); mp 119 – 121 °C; vmax (neat)/cm-1 3249.57 (m), 2113.95 (w), 1686.46 (s), 1373.49 (s), 1270.05 (s), 675.49 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.29 – 7.23 (1H, m, Ar(9)-H); 7.11 – 7.05 (1H, m, Ar(6)-H); 6.82 (1H, dd, J = 8.5, 1.5 Hz, Ar(7)-H); 4.63 – 4.54 (4H, m, CH2, C(3)H2); 2.24 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 163.6 (C=O), 144.3 (Ar-(C8’)-C), 129.2 (Ar-(C5’)-C), 126.9 (Ar-(C7)-H), 118.4 (Ar-(C5)-H), 118.2 (Ar-(C8)-H), 115.0 (Ar-(C6)-C), 76.8 (CH), 73.0 (C≡C-CH2), 67.5 (C(2)H2), 30.7 (CH3); HRMS m/z (ES+) [Found [M79Br+H]+ / [M81Br+H]+ 265.9815 / 267.9792; C11H8BrNO2 requires Found [M79Br+H]+ / [M81Br+H]+ 265.9811 / 267.9791]; m/z (ES+) 349.1 / 351.1 ([M79Br+IsoProp+Na+H]+ / [M81Br+IsoProp+Na+H]+, 51 / 43%). 1-(Prop-2-yn-1-yl)pyrrolidin-2-one (44, AlkE5) To 2-pyrolidinone (0.5 g, 5.87 mmol, 1.00 eq) in N,N-dimethylformamide (30 mL) at 0 °C was added sodium hydride 60% dispersion in oil (0.15 g, 6.46 mmol, 1.10 eq). The reaction was stirred for 30 min. After which time, propargyl bromide 80% in toluene (0.57 mL, 6.46 mmol, 1.10 eq) was added at 0 °C. The reaction was warmed to room temperature over 24 h, quenched with ammonium chloride (aqueous), partitioned between water and ethyl acetate, washed with brine, dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was purified by silica gel column chromatography eluting with 0 – 100% ethyl acetate in hexane and 10% methanol in ethyl acetate to afford 1-(prop-2-yn-1-yl)pyrrolidin-2-one 44 as a yellow oil (0.23 g, 33%): 273 Rf 0.37 (hexane: ethyl acetate, 1:1); Rf 0.6 (methanol:DCM 0.5:9.5); vmax (neat)/cm-1 3249.13 (br), 2951.17 (br), 2114.76 (w), 1677.73 (s), 1266.64 (s), 673.94 (s); 1H NMR (500 MHz, CDCl3) δ ppm 4.04 (2H, d, J = 2.5 Hz, CH2); 3.40 (2H, t, J = 6.9 Hz, C(5)H2); 2.33 (2H, t, J = 8.0 Hz, C(3)H2); 2.16 (1H, t, J = 2.5 Hz, CH); 2.06 – 1.95 (2H, m, C(4)H2); 13C NMR (125 MHz, CDCl3) δ ppm 173.5 (C=O), 76.7 (C≡C-CH2), 71.1 (CH), 45.3 (C(5)-C), 30.8 (C(3)-C), 29.6 (CH2), 16.5 (C(4)-C); HRMS m/z (ES+) [Found [M+H]+ 124.0754; C7H9NO requires [M+H]+ 124.0757]; m/z (ES+) 265.2 ([2M+NH4]+, 97%). 1-(3,4-Difluorophenyl)-3-(prop-2-yn-1-yl)urea (45, AlkE1) To propargyl amine (0.19 g, 3.54 mmol, 1.1 eq) in dichloromethane (6 mL) at 0 °C was added 3,4-difluorophenyl isocyanate (0.5 mL, 3.22 mmol, 1.00 eq) in dichloromethane (6 mL). The reaction was warmed to room temperature over 3.5 h. The resulting precipitate was filtered and washed with DCM. The crude solid was recrystallized from DCM to afford 1-(3,4-difluorophenyl)-3-(prop-2-yn-1-yl)urea 45 as an off white solid (0.8 g, quantitative): Rf 0.46 (methanol:DCM 0.5:9.5); mp 174 – 175 °C (Lit192: 173 – 174 °C); vmax (neat)/cm-1 3331.07 (br), 3295.60 (br), 1639.99 (s), 1505.43 (s), 650.87 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.83 (1H, s, N(1)H); 7.62 (1H, ddd, J = 13.7, 7.4, 2.5 Hz, Ar(2)H); 7.29 (1H, ddd, J = 18.2, 10.6, 9.1, Ar(5)H); 7.10 – 7. 01 (1H, m, Ar(6)H); 6.57 (1H, t, J = 5.6 Hz, N(3)H); 3.88 (2H, dd, J = 5.6, 2.5 Hz, CH2); 3.11 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, DMSO-d6) δ ppm 154.5 (C=O), 147.9 (dd, J = 241.3, 12.8 Hz, ArC(3)-C); 143.0 (dd, J = 239.2, 12.8 Hz, ArC(4)-C); 137.4 (dd, J = 9.1, 1.8 Hz, ArC(1)-C); 117.2 (d, J = 17.5 Hz, ArC(5)-H); 113.8 (dd, J = 5.5, 2.7 Hz, ArC(6)-H); 106.6 (d, J = 22.1 Hz, ArC(2)-H), 81.9 (C≡C-CH2), 72.82 (CH), 28.7 (CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -137.65 (1F, d, J = 23.3 Hz, Ar-F); -147.86 (1F, d, J = 23.3 Hz, Ar-F); HRMS m/z (ES+) [Found [M+Na]+ 233.0491; C10H8F2N2ONa requires [M+Na]+ 233.0497]; m/z (ES+) 211.2 ([M+H]+, 68%). Data are in agreement with the literature.192 N-(Prop-2-yn-1-yl)acetamide (46, AlkD3) To propargyl amine (0.5 g, 9.07 mmol, 1.00 eq) in dichloromethane (6 mL) at 0 °C was added acetyl chloride (0.32 mL, 4.53 mmol, 0.50 eq). The reaction was warmed to room 274 temperature over 6 h. The resulting precipitate was filtered and washed with DCM. The filtrate was concentrated and purified by silica gel column chromatography eluting with 0 – 10% methanol in dichloromethane to afford N-(prop-2-yn-1-yl)acetamide 46 as a yellow oil (0.39 g, 45%): Rf 0.4 (methanol:DCM 0.5:9.5); vmax (neat)/cm-1 3222.88 (br), 3073.93 (br), 2114.66 (w), 1631.99 (m), 711.66 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 8.28 (1H, br. s, NH); 3.83 (2H, dd, J = 5.6, 2.5 Hz, CH2); 3.10 (1H, t, J = 2.5 Hz, CH); 1.82 (3H, s, CH3); 13C NMR (125 MHz, DMSO-d6) δ ppm 168.8 (C=O), 81.2 (C≡C-CH2), 72.8 (CH), 27.7 (CH2), 22.2 (CH3); HRMS m/z (ES+) [Found [M+H]+ 98.0600; C5H7NO requires [M+H]+ 98.0600]; m/z (ES+) 98.1 ([M+H]+, 13%); Data are in agreement with the literature.193 6-Chloro-N-(prop-2-yn-1-yl)nicotinamide (47, AlkB1) To propargyl amine (0.5 g, 9.07 mmol, 1.00 eq) in dichloromethane (6 mL) at 0 °C was added 6-chloronictonoyl chloride (0.79 mL, 4.53 mmol, 0.50 eq). The reaction was warmed to room temperature over 6 h. The resulting precipitate was filtered and washed with DCM. The filtrate was concentrated in vacuo and purified by silica gel column chromatography eluting with 0 – 10% methanol in DCM to afford N-(prop-2-yn-1-yl)acetamide 47 as a cream solid (0.65 g, 37%): Rf 0.5 (hexane: ethyl acetate 1:1, KMnO4 stain); mp 142 – 143 °C; vmax (neat)/cm-1 3272.94 (br), 3047.80 (br), 1643.74 (s), 1537.04 (s), 643.99 (s); 1H NMR (500 MHz, DMSO-d6) δ ppm 9.23 (1H, t, J = 5.3 Hz, NH); 8.85 (1H, d, J = 2.5 Hz, Ar(2)H); 8.25 (1H, dd, J = 8.5, 2.5 Hz, Ar(4)H); 7.64 (1H, d, J = 8.5 Hz, Ar(5)H); 4.09 (2H, dd, J = 5.3, 2.5 Hz, CH2); 3.19 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, DMSO-d6) δ ppm 163.4 (C=O), 152.7 (ArC(2)-C), 148.9 (ArC(6)-H), 138.7 (ArC(4)-H), 128.7 (ArC(3)-C), 124.1 (ArC(5)-H), 80.7 (C≡C-CH2), 73.3 (CH), 28.5 (CH2); HRMS m/z (ES+) [Found [M35Cl+H]+ / [M37Cl+H]+ 195.0319 / 197.0287; C9H7ClN2O requires [M35Cl+H]+ / [M37Cl+H]+ 195.0320 / 197.0290]; m/z (ES-) 306.7 / 308.7 ([M35Cl+TFA-H]- / ([M37Cl+TFA-H]-, 100 / 25%). Data are in agreement with the literature.194 275 N-(Furan-2-ylmethyl)prop-2-yn-1-amine (48, AlkH6-2) To 2-fufurylamine (1.00 g, 0.9 mL, 10.30 mmol, 1.20 eq) and triethylamine (1.04 g, 1.43 mL, 10.30 mmol, 1.20 eq) in tetrahydrofuran (12 mL) at 0 °C was added propargyl bromide 80% in toluene (1.28 g, 0.95 mL, 8.58 mmol, 1.00 eq) dropwise over 30 min. The reaction was warmed to room temperature over 4 h. The resulting precipitate was filtered and washed with THF. The crude filtrate was purified by silica gel column chromatography eluting with 0 – 100% DCM in ethyl acetate to afford N-(furan-2-ylmethyl)prop-2-yn-1-amine 48 as a yellow oil (0.10 g, 7% yield): Rf 0.46 (5:1 DCM:EtOAc, KMnO4); vmax (neat)/cm-1 1509.7 (w), 1146.10 (w), 1010.36 (w), 620.69 (m)); 1H NMR (500 MHz, CDCl3) δ ppm 7.30 (1H, dd, J = 1.9, 0.8 Hz, Ar(C5)-H); 6.25 (1H, dd, J = 2.9, 1.9 Hz, Ar(C4)-H); 6.16 (1H, dd, J = 2.9, 0.8 Hz, Ar(C3)-H); 3.82 (2H, s, Ar-CH2); 3.37 (2H, d, J = 2.4 Hz, HC≡C-CH2); 2.18 (1H, t, J = 2.4 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 151.9 (ArC(2)-C), 141.0 (ArC(5)-H), 109.1 (ArC(4)-H), 106.5 (ArC(3)-H), 80.6 (C≡C-CH2), 70.6 (CH), 43.5 (Ar-CH2), 36.1 (C-CH2); HRMS m/z (ES+) [Found [M+H]+ 136.0753; C8H9NO requires [M+H]+ 136.0757]; m/z (ES+) 307.2 ([2M+K]+, 100%). Data are in agreement with the literature.195 1-(Prop-2-yn-1-yl)piperidin-4-one (49, AlkD2-2) To piperidinone (0.25 g, 2.52 mmol, 1.00 eq) in chloroform (2 mL) and water (2 mL) at 0 °C was added potassium carbonate (1.04 g, 7.56 mmol, 3.00 eq) and propargyl bromide 80% in toluene (0.34 mL, 3.78 mmol, 1.50 eq). The reaction was warmed to room temperature over 17 h. The resulting crude material was extracted with chloroform, dried over sodium sulfate (anhydrous), filtered and concentrated in vacuo. The crude material was purified twice by silica gel column chromatography eluting with 0 – 100% ethyl acetate in hexane to afford 1-(prop-2-yn-1-yl)piperidin-4-one 49 as a colourless oil (0.06 g, 18%): Rf 0.41 (1:1 hexane: ethyl acetate); vmax (neat)/cm-1 1711.55 (w), 1325.50 (w), 1191.32 (w), 1083.73 (w), 720.51 (m)); 1H NMR (500 MHz, CDCl3) δ ppm 3.38 (2H, d, J = 2.5 Hz, CH2); 2.80 (4H, t, J = 6.2 Hz, (C2)H2); 2.44 (4H, t, J = 6.2 Hz, (C3)H2); 2.22 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 207.2 (C=O), 77.2 (C≡C-CH2), 72.5 (CH), 50.8 (C(2)H2), 45.4 (CH2), 40.1 (C(3)H2); HRMS m/z (ES+) [Found [M+H]+ 276 138.0916; C8H11NO requires [M+H]+ 138.0913]; m/z (ES+) 169.9 ([M+CH3OH+H]+,100%). Data are in agreement with the literature.196 1-(Prop-2-yn-1-yl)-1H-pyrrole (69, AlkD7-2) To pyrrole (0.5 g, 0.51 mL, 7.45 mmol, 1.00 eq) in anhydrous N,N’-dimethylformamide (40 mL) at 0 °C was added sodium hydride (0.26 g, 11.18 mmol, 1.50 eq) portion wise and stirred for 30 min. After which time, propargyl bromide (1.32 mL, 14.91 mmol, 2.00 eq) was added dropwise. The reaction was warmed to room temperature over 16 h. The resulting mixture was filtered through celite, ammonum chloride (aq) (1 mL) added and partitioned between water and ethyl acetate. The organic layer was washed with brine, dried over sodium sulfate (anhydrous), filtered and concentrated. The crude material was purified by silica gel column chromatography eluting with 0 – 100% ethyl acetate in hexane to afford 1-(prop-2-yn-1-yl)-1H-pyrrole 69 as a yellow oil (0.18 g, 22%): Rf 0.86 (1:1 hexane: ethyl acetate); vmax (neat)/cm-1 1491.42 (w), 1278.31 (w), 1087.07 (w), 617.99 (m)); 1H NMR (500 MHz, CDCl3) δ ppm 6.70 (2H, t, J = 2.3 Hz, Ar-(C2)-H); 6.12 (2H, t, J = 2.3 Hz, Ar-(C3)-H); 4.61 (2H, d, J = 2.5 Hz, CH2); 2.35 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 120.4 (Ar(C2)-H), 108.9 (Ar(C3)-H), 78.0 (C≡C-CH2), 73.6 (CH), 38.6 (CH2); HRMS m/z (ES+) [Found [M+H]+ 106.0651; C7H7N requires [M+H]+106.0651]; m/z (ES+/-) no mass ion observed in LRMS. 7.6.5 Synthesis of triazoles hits from the biased kTGS assay Triazoles were synthesised as described. The naming of the biased triazoles has been changed for all synthesised analogues, due to the conserved Glu-Ala-Gly backbone, i.e. from N5- ((S)-1-((carboxymethyl)amino)-1-oxo-3-(4-((2-phenyl-1H-imidazol-1-yl)methyl)-1H-1,2,3-triazol-1-yl)propan-2-yl)- L -glutamine formic acid salt to γ-Glutamyl-L-(4-((2-phenyl-1H-imidazol-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt. 277 γ-Glutamyl-L-(4-((2-phenyl-1H-imidazol-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt (50) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 2-phenyl-1-(prop-2-yn-1-yl)-1H-imidazole (40) (0.003 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h. After which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-(acetamidomethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt 50 as a pale oil (0.026 g, 33%): Rf 0.68 (methanol: water 1:1 RP C-18 TLC); vmax (neat)/cm-1 3262.79 (br), 1658.10 (m), 1591.76 (m), 1130.10 (m), 696.52 (s); 1H NMR (500 MHz, D2O) δ ppm 8.38 (2H, br s, formic acid); 7.91 (1H, s, Ar(C5)-H); 7.70 – 7.64 (2H, m, Ar(C4’, 5’)H); 7.63 – 7.58 (3H, m, Ar(C8’, 9’, 10’)H); 7.54 – 7.45 (2H, m, Ar(C7’, 11’)H); 5.49 (1H, s, (C6)H2); 4.90 (dd, J = 8.2, 4.5 Hz, Ala-α-CH); 4.85 (1H, dd, J = 14.5, 4.5 Hz, Ala-β-CH2); 4.65 (1H, dd, J = 14.5, 8.2 Hz, Ala-β-CH2); 3.71 – 3.63 (2H, m, Gly-CH2); 3.62 (1H, t, J = 6.3 Hz, Glu-α-CH); 2.34 – 2.29 (2H, m, Glu-Ɣ-CH2); 1.96 – 1.89 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 174.4 (C=O), 169.3 (C=O), 141.2 (Ar(C4)-C), 132.4 (Ar(C1’)-C), 132.4 (Ar-(C8)-H), 129.5 (Ar-(C10)-H), 129.4 (Ar-(C4’)-H), 129.2 (Ar-(C6’)-C), 129.1 (Ar-(C7,11)-H), 125.5 (Ar(C5)-H), 119.6 (Ar(C5’)-H), 52.9 (Ala-α-CH), 50.6 (Ala-β-CH2), 43.3 (Gly-CH2), 42.8 (Glu-α-CH), 37.9 ((C6)H2), 31.3 (Glu-Ɣ-CH2), 26.0 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 497.1899; C22H26N8O6 requires [M-H]- 497.1903]; m/z (ES-) 497.2; ([M-H]-, 100%); 278 γ-Glutamyl-L-(4-((6-chloronicotinamido)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt (51) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and 6-chloro-N-(prop-2-yn-1-yl)nicotinamide (47) (0.03 g, 0.15 mmol, 1.00 eq) in water (0.6 mL) and t-butanol (0.3 mL), (+)-sodium L-ascorbate (0.03 g, 0.15 mmol, 1.00 eq) and copper sulfate pentahydrate (0.02 g, 0.11 mmol, 0.75 eq) were added and the reaction stirred at room temperature for 20 h. The crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-((6-chloronicotinamido)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt 51 as a light brown solid (0.037 g, 45%): Rf 0.8 (methanol: water 1:1 RP C-18 TLC); mp 213 °C (dec.); [𝛼]54620 -3.33 (c 0.3 H2O); vmax (neat)/cm-1 3398.44 (br), 2827.69 (w), 1583.02 (s), 1353.59 (s), 768.04 (m); 1H NMR (500 MHz, D2O) δ ppm 8.65 (1H, d, J = 2.2 H, Ar(C7)-H); 8.26 (0.8H, s, formic acid); 8.10 (1H, dd, J = 8.4, 2.2 Hz, Ar(C9)-H); 7.92 (1 H, s, Ar(C5)-H); 7.54 (1H, d, J = 8.4 Hz, Ar(C10)-H); 4.95 (1H, dd, J = 8.3, 4.3 Hz, Ala-α-CH); 4.85 (1H, dd, J = 14.1, 4.3 Hz, Ala-β-CH2); 4.65 (1 H, dd, J = 14.1, 8.3 Hz, Ala-β-CH2); 4.59 (2H, s, C(6)H2); 3.77 (2H, d, J = 2.2 Hz, Gly-CH2); 3.62 (1H, t, J = 5.9 Hz, Glu-α-CH); 2.34 – 2.31 (2H, m, Glu-Ɣ-CH2); 1.98 – 1.89 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 174.7 (C=O), 174.4 (C=O), 173.6 (C=O), 169.8 (C=O), 168.2 (Ar(C7)-H), 167.4 (Ar(C11)-C), 153.6 (Ar(C9)-H), 148.1 (quaternary Ar(C4)-C), 138.8 (quaternary Ar(C8)-C), 128.8 (Ar(C10)-H), 124.8 (Ar(C5)-H), 53.0 (Ala-β-CH2), 50.4 (C(6)-H2), 48.8 (Ala-α-CH), 42.5 (Glu-α-CH), 34.9 (Gly-CH2), 31.1 (Glu-Ɣ-CH2), 25.9 (Glu-β-CH2); HRMS m/z (ES-) [Found [M35Cl-H]- /[M37Cl-H]- 509.1303 / 511.1274; C19H23ClN8O7 requires [M35Cl-H]- /[M37Cl-H]- 509.1305 / 511.1277]; m/z (ES-) 509.2 / 511.2 ([M35Cl-H]- /[M37Cl-H]-, 100 / 16%); HPLC RT 4.01 min; >95% content. 279 γ-Glutamyl-L-(4-((4-acetylphenoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (52) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and 1-(4-(prop-2-yn-1-yloxy)phenyl)ethan-1-one (41) (0.03 g, 0.15 mmol, 1.00 eq) in water (0.6 mL) and t-butanol (0.3 mL), (+)-sodium L-ascorbate (0.03 g, 0.15 mmol, 1.00 eq) and copper sulfate pentahydrate (0.02 g, 0.11 mmol, 0.75 eq) were added and the reaction stirred at room temperature for 20 h. The crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-((4-acetylphenoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 52 as an off white solid (0.012 g, 16%): Rf 0.65 (methanol: water 1:1 RP C-18 TLC); mp 193 °C (dec.); vmax (neat)/cm-1 2944.93 (br), 1660.20 (m), 1641.50 (s), 1512.17 (s), 1223.95 (s), 832.02 (m); 1H NMR (500 MHz, D2O) δ ppm 8.05 (1H, s, Ar(5)-H); 7.93 (2H, d, J = 8.8 Hz, Ar(9,11)-H); 7.06 (2H, d, J = 8.8 Hz, Ar(8, 12)-H); 5.28 (2H, s, C(6)-H2); 4.94 (1H, dd, J = 8.2, 4.8 Hz, Ala-α-CH); 4.86 (1H, dd, J = 14.4, 4.8 Hz, Ala-β–CH2); 3.77 (2H, d, J = 5.9 Hz, Gly-CH2); 3.63 (1H, t, J = 5.9 Hz, Glu-α-CH); 2.54 (3H, s, CH3); 2.34 – 2.22 (2H, m, Glu-Ɣ-CH2); 1.95 – 1.88 (2H, m, Glu-β-CH2); (1H, Ala-β–CH2) under D2O peak; 13C NMR (125 MHz, D2O) δ ppm 202.5 (C=O), 174.4 (C=O), 174.0 (C=O), 173.5 (C=O), 170.0 (C=O), 161.9 (Ar(C7)-C), 131.1 (Ar(C9,11)-H), 130.1 ((Ar(C4)-C), 125.9 (Ar(C5)-H), 114.9 (Ar(C8,12)-H), 77.7 ((C6)H2), 60.9 (Ala-β-CH2), 52.9 (Ala-α-CH), 50.4 (Gly-CH2), 41.8 (Glu-α-CH), 31.0 (Glu-Ɣ-CH2), 25.9 (Glu-β-CH2), 25.8 (CH3); HRMS m/z (ES-) [Found [M-H]- 489.1733; C21H26N6O8 requires [M-H]- 489.1739]; m/z (ES-) 489.2 ([M-H]-, 100%); HPLC RT 5.82 min; >95% content. 280 γ-Glutamyl-L-(4-(2-hydroxy-3-methylbutyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (53) To γ-glutamyl-L-azidoalaninylglycine (32) (0.056 g, 0.17 mmol, 1.00 eq) and 2-methylhex-5-yn-3-ol (EN300-97981) (0.019 g, 0.17 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0015 g, 0.003 mmol, 2 mol%) were added and the reaction stirred at room temperature for 20 h. After which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-(2-hydroxy-3-methylbutyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 53 as an off white gum (0.020 g, 26%): Rf 0.68 (methanol: water 1:1 RP C-18 TLC); [𝛼]54620 -1.82 (c 1.1 H2O); vmax (neat)/cm-1 3386.88 (br), 3311.49 (br), 1630.23 (s), 1564.07 (s), 1388.19 (m), 646.72 (w); 1H NMR (500 MHz, D2O) δ ppm 7.76 (1H, s, Ar(C5)-H); 4.94 (1H, dd, J = 8.4, 4.8 Hz, Ala-α-CH); 4.82 (1H, dd, J = 14.5, 4.8 Hz, Ala-β-CH2); 4.64 (1H, dd, J = 14.5, 8.4 Hz, Ala-β-CH2); 3.85 (2H, s, Gly-CH2); 3.66 (1H, t, J = 6.3 Hz, Glu-α-CH); 3.64 – 3.57 (1H, m, (C7)H); 2.86 (1H, dd, J = 15.0, 4.0 Hz, (C6)-H); 2.72 (1H, dd, J = 15.0, 8.5 Hz, (C6)-H); 2.34 (2H, t, J = 8.5 Hz, Glu-Ɣ-CH2); 2.02 – 1.90 (1H, m, Glu-β-CH2); 1.64 – 1.53 (1H, m (C8)-H); 0.87 (3H, d, J = 8.5 Hz, CH3); 0.86 (3H, d, J = 8.5 Hz, CH3); 13C NMR (125 MHz, D2O) δ ppm 174.4 (C=O), 173.9 (C=O), 173.5 (C=O), 170.2 (C=O), 145.7 (Ar(C4)-C), 124.7 (Ar(C5)-H), 75.7 ((C7)-H), 53.0 (Ala-β-CH2), 50.2 (Ala-α-CH), 48.8 (Glu-α-CH), 41.8 (Gly-CH2), 32.3 (C(6)H2), 31.1 (C(8)H), 25.9 (Glu-Ɣ-CH2), 18.0 (Glu-β-CH2), 16.5 (CH3), 16.5 (CH3); HRMS m/z (ES-) [Found [M-H]- 427.1941; C17H28N6O7 requires [M-H]- 427.1947]; m/z (ES-) 474.1 ([M+2Na]-, 100%); 281 γ-Glutamyl-L-(4-((2,3-dioxoindolin-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (54) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 1-(prop-2-yn-1-yl)indoline-2,3-dione (42) (0.03 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h. After which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised in ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-((2,3-dioxoindolin-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 54 as a brown solid (0.02 g, 23%): Rf 0.47 (methanol: water 1:1 RP C-18 TLC); mp 150 °C (dec.) (ethanol:water); [𝛼]54620 -20.0 (c 0.6 H2O); vmax (neat)/cm-1 1656.60 (m), 1611.8 (m), 1375.3 (m), 1218.50 (m), 760.54 (m); 1H NMR (500 MHz, D2O) δ ppm 8.03 (1H, s, Ar(C5)-H); 7.63 – 7.55 (2H, m, Ar(C6’)-H, Ar(C7’)-H); 7.12 (1H, t, J = 7.9 Hz, Ar(C5’)-H); 7.00 (1H, d, J = 7.9 Hz, Ar(C4’)-H); 4.98 (2H, d, J = 1.8 Hz, (C6)H2); 4.83 (1H, dd, J = 12.1, 4.7 Hz, Ala-α-CH); 4.78 – 4.76 (1H, m, Ala-β-CH2); 4.67 – 4.63 (1H, m, Ala-β-CH2); 3.91 (1H, t, J = 6.6 Hz, Glu-α-CH); 3.63 – 3.62 (1H, m, Gly-CH2); 2.26 (2H, t, J = 7.2 Hz, Glu-Ɣ-CH2); 1.97 – 1.92 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 184.6 (C(3’)=O), 169.3 (C=O), 168.8 (C=O), 160.1 (C=O), 149.8 (C(2’)=O), 139.0 (Ar-(C6’)-H), 131.2 (Ar(C4)-C), 125.6 (Ar(C4’)-H), 125.2 (Ar(C5)-H), 124.5 (Ar(C5’)-H), 117.5 (Ar(C4’’)-C), 111.2 (Ar(C7’)-H), 57.3 (Ala-β-CH2), 53.3 (Ala-α-CH), 52.4 ((C6)H2), 50.0 (Glu-α-CH), 30.3 (Glu-Ɣ-CH2), 26.2 (Glu-β-CH2); HRMS m/z (ES+) [Found [M+H]+ 502.1675; C21H23N7O8 requires [M+H]+ 502.1681]; m/z (ES-) 500.1; ([M-H]-, 100%); HPLC RT 2.90 min; >95% content. 282 γ-Glutamyl-L-(4-(benzo[d]thiazol-2-yl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (55) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and 2-ethynylbenzo[d]thiazole (EN300-107722) (0.03 g, 0.15 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.004 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0007 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h. The crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford a brown solid. The compound was further purified by recrystallisation from ethanol water to afford γ-glutamyl-L-(4-(benzo[d]thiazol-2-yl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 55 as a cream solid (0.03 g, 40%): Rf 0.39 (methanol: water 1:1 RP C-18 TLC); mp 215 °C (dec.) (ethanol:water 1:1); vmax (neat)/cm-1 3355.43 (w), 3052.47 (br), 1679.54 (s), 1644.22 (s), 1508.31 (s), 758.83 (m); H NMR (500 MHz, DMSO-d6) δ ppm 9.15 (1H, t, J = 5.5 Hz, Gly-NH); 8.82 (1H, s, Ar(C5)-H); 8.73 (1H, d, J = 8.9 Hz, Ala-NH); 8.17 (1H, d, J = 7.8 Hz, Ar(C10)-H); 8.06 (1H, d, J = 7.8 Hz, Ar(C7)-H); 7.56 (1H, t, J = 7.6, Hz, Ar(C9)-H); 7.48 (1H, t, J = 7.6 Hz, Ar(C8)-H); 4.96 – 4.84 (2H, m, Ala-α-CH, Ala-β-CH2); 4.68 (1H, dd, J = 13.5, 8.9 Hz, Ala-β-CH2); 3.76 (2H, d, J = 5.5 Hz, Gly-CH2); 3.24 (1H, dd, J = 7.8, 6.3 Hz, Glu-α-CH); 2.36 – 2.26 (1H, m, Glu-Ɣ-CH) 2.25 – 2.16 (1H, m Glu-Ɣ-CH); 1.96 – 1.87 (1H, m Glu-β-CH); 1.79 – 1.70 (1 H, m, Glu-β-CH); 13C NMR (125 MHz, DMSO-d6) δ ppm 171.9 (C=O), 170.7 (C=O), 169.9 (C=O), 168.9 (C=O), 159.0 (Ar(C7’)-C), 153.1 (Ar(C6)-C), 141.6 (Ar(C10’)-C), 133.7 (Ar(C4)-C), 126.6 (Ar(C5)-H), 125.5 (Ar(C8)-H), 124.4 (Ar(C9)-H), 122.6 (Ar(C7)-H), 122.4 (Ar(C10)-H), 52.8 (Ala-β-CH2), 52.7 (Ala-α-CH), 50.5 (Glu-α-CH), 41.2 (Gly-CH2), 31.2 (Glu-Ɣ-CH2), 26.4 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 474.1192; C19H21N7O6S requires [M-H]- 474.1201]; m/z (ES-) 474.1 ([M-H]-, 100%); HPLC RT 7.67 min; >94% content 283 γ-Glutamyl-L-(4-((6-bromo-3-oxo-2,3-dihydro-4H-benzo[b][1,4]oxazin-4-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (56) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and 6-bromo-4-(prop-2-yn-1-yl)-2H-benzo[b][1,4]oxazin-3(4H)-one (43) (0.04 g, 0.15 mmol, 1.00 eq) in water (0.6 mL) and t-butanol (0.3 mL), (+)-sodium L-ascorbate (0.02 g, 0.12 mmol, 0.8 eq) and copper sulfate pentahydrate (0.01 g, 0.07 mmol, 0.5 eq) were added and the reaction stirred at room temperature for 19 h. The crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-((6-bromo-3-oxo-2,3-dihydro-4H-benzo[b][1,4]oxazin-4-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 56 as a light brown solid (0.013 g, 15%): Rf 0.45 (methanol: water 1:1 RP C-18 TLC); mp 194 °C (dec.); [𝛼]54620 -6.67 (c 0.3 H2O); vmax (neat)/cm-1 1681.93 (m), 1492.79 (m), 1433.08 (m), 1359.65 (m), 1268.99 (m), 808.07 (m); 1H NMR (500 MHz, D2O) d ppm 7.92 (1H, s, Ar(C5)-H); 7.27 (1H, d, J = 2.2 Hz, Ar(C13)-H); 7.17 (1H, dd, J = 8.5, 2.2 Hz, Ar(C10)-H); 6.90 (1H, d, J = 8.5 Hz, Ar(C11)-H); 5.14 (2H, s, (C6)H2); 4.92 (1H, dd, J = 8.6, 4.7 Hz, Ala-α-CH); 4.81 (1H, dd, J = 14.3, 4.7 Hz, Ala-β-CH2); 4.62 (1H, dd, J = 14.3, 8.6 Hz, Ala-β-CH2); 3.70 (2H, d, J = 7.8 Hz, Gly-CH2); 3.59 (1H, t, J = 6.3 Hz, Glu-α-CH); 2.38 – 2.16 (2H, m, Glu-Ɣ-CH2); 1.90 – 1.83 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 174.3 (C=O), 169.7 (C=O), 166.5 (C=O), 144.2 (Ar(C10’)-C), 128.8 (Ar(C13’)-C), 127.4 (Ar(C11)-H), 124.8 (Ar(C4)-C), 118.5 (Ar(C5)-H), 118.4 (Ar(C13)-H), 114.4 (Ar(C10)-C, Ar(C12)-Br), 67.0 ((C9)H2), 52.9 (Ala-β-CH2), 50.5 ((C6)H2), 42.7 (Ala-α-CH, Glu-CH), 36.5 (Gly-β-CH2), 31.1 (Glu-Ɣ-CH2), 25.8 (Glu-β-CH2); HRMS m/z (ES-) [Found [M79-H]-/ [M81-H]- 580.0791 / 582.0767; C21H24BrN7O8 requires [M79Br-H]-/ [M81Br-H]- 580.0797 / 582.0780]; m/z (ES-) 580.1 / 582.1 ([M79Br-H]-/ [M81Br-H]-, 100 / 98%); HPLC RT 8.11 min; >95% content. 284 γ-Glutamyl-L-(4-((2-(aminomethyl)phenoxy)methyl)-1H-1,2,3-triazol-1-yl)alaninylglycine (57) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and (2-(prop-2-yn-1-yloxy)phenyl)methanamine (EN300-73330.HCl) (0.03 g, 0.15 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.004 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0007 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h. After which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford a colourless solid. The compound was further purified by recrystallisation from ethanol and water to afford γ-glutamyl-L-(4-((2-(aminomethyl)phenoxy)methyl)-1H-1,2,3-triazol-1-yl)alaninylglycine 57 as a yellow solid (0.021 g, 28%): Rf 0.48 (methanol: water 1:1 RP C-18 TLC); mp 201 °C (dec.) (ethanol:water); vmax (neat)/cm-1 3320.04 (w), 3082.56 (br), 1631.84 (s), 1574.00 (s), 1057.16 (w), 595.72 (w); 1H NMR (500 MHz, D2O) δ ppm 8.07 (1H, s, Ar(5)-H); 7.38 (1H, t, J = 7.7 Hz, Ar(C9)-H); 7.29 (1H, d, J = 7.7 Hz, Ar(C11)-H); 7.12 (1H, d, J = 7.7 Hz, Ar(C8)-H); 7.00 (1H, t, J = 7.7 Hz, Ar(10)-H); 5.29 (2H, s, (C6)H2); 4.93 (1H, dd, J = 7.8, 4.6 Hz, Ala-α-CH); 4.85 (1H, dd, J = 14.3, 4.6 Hz, Ala-β-CH2); 4.66 (1H, dd, J = 14.3, 7.8 Hz, Ala-β-CH2); 4.08 (2H, s, (C13)H2); 3.68 – 3.56 (3H, m, Gly-CH2, Glu-α-CH); 2.33 – 2.21 (2H, m Glu-Ɣ-CH2); 1.93 – 1.85 (1 H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 176.1 (C=O), 174.4 (C=O), 173.7 (C=O), 169.4 (C=O), 155.7 (quaternary Ar(C7)-C), 143.4 (quaternary Ar(C12)-C), 131.0 (Ar(C9)-H), 130.9 (Ar(C11)-H), 125.8 (Ar(C5)-H), 121.7 (Ar(C10)-H), 121.2 (Ar(C4)-C), 112.7 (Ar(C8)-H), 60.87 (Ala-β-CH2), 53.9 (Ala-α-CH), 52.9 (Glu-α-CH), 50.5 ((C6)H2), 43.2 (Gly-CH2), 39.2 (Glu-Ɣ-CH2), 26.0 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 476.1892; C20H27N7O7 requires [M-H]- 476.1899]; m/z (ES+) 478.3 ([M+H]+, 100%); HPLC RT 1.01 min; >95% content. 285 γ-Glutamyl-L-(4-((2-oxopyrrolidin-1-yl)methyl)-1H-1,2,3-triazol-1-yl)alaninylglycine (58) To γ-glutamyl-L-azidoalaninylglycine (32) (0.056 g, 0.17 mmol, 1.00 eq) and 1-(prop-2-yn-1-yl)pyrrolidin-2-one (44) (0.021 g, 0.17 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0015 g, 0.003 mmol, 2 mol%) were added. The reaction stirred at room temperature for 20 h after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-((2-oxopyrrolidin-1-yl)methyl)-1H-1,2,3-triazol-1-yl)alaninylglycine 58 as a colourless gum (0.039 g, 50%): Rf 0.83 (methanol: water 1:1 RP C-18 TLC); [𝛼]54620 -2.86 (c 0.35 H2O); vmax (neat)/cm-1 3281.09 (br), 2935.98 (br), 1638.23 (s), 1522.25 (s), 1220.44 (m), 535.61 (w); 1H NMR (500 MHz, D2O) δ ppm 7.89 (1H, s, Ar(C5)-H); 4.94 (1H, dd, J = 8.3, 4.7 Hz, Ala-α-CH); 4.84 (1H, dd, J = 14.3, 4.7 Hz, Ala-β-CH2); 4.65 (1H, dd, J = 14.3, 8.3 Hz, Ala-β-CH2); 4.46 (2H, s, (C6)-H2); 3.88 (2H, s, Gly-CH2); 3.69 (1H, t, J = 6.3 Hz, Glu-α-CH); 3.34 (2H, t, J = 7.2 Hz, (C11)-H2); 2.40 – 2.32 (4H, m, Glu-Ɣ-CH2, (C9)-H2); 2.00 - 1.91 (4H, m, Glu-β-CH2, (C10)-H2); 13C NMR (125 MHz, D2O) δ ppm 178.2 (C=O); 174.3 (C=O), 173.4 (C=O), 170.2 (C=O), 166.2 (C=O), 142.7 (quaternary Ar(C4)-C), 124.9 (Ar(C5)-H), 53.5 (Ala-β-CH2), 52.9 (Ala-α-CH), 50.4 (Glu-CH), 47.7 (C(6)H2), 41.7 ((C11)H2), 37.3 (Gly-CH2), 31.0 (Glu-Ɣ-CH2), 30.7 ((C9)H2), 25.8 (Glu-β-CH2), 17.1 ((C10)H2); HRMS m/z (ES-) [Found [M-H]- 438.1737; C17H25N7O7 requires [M-H]- 438.1743]; m/z (ES-) 438.2 ([M-H]-, 100%). 286 γ-Glutamyl-L-(4-((2-(thiophen-2-yl)-1H-imidazol-1-yl)methyl)-1H-1,2,3-triazol-1-yl)alaninylglycine (59) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and 1-(prop-2-yn-1-yl)-2-(thiophen-2-yl)-1H-imidazole (EN300-73177) (0.03 g, 0.15 mmol, 1.00 eq) in water (0.6 mL) and t-butanol (0.3 mL), (+)-sodium L-ascorbate (0.003 g, 0.015 mmol, 10 mol%), copper sulfate pentahydrate (0.004 g, 0.0015 mmol, 1 mol%) and tris(benzyltriazolylmethyl)amine (0.0006 g, 0.0015 mmol, 1 mol%) were added and the reaction stirred at room temperature for 20 h. After which time a further (+)-sodium L-ascorbate (0.0015 g, 0.007 mmol, 5 mol%), copper sulfate pentahydrate (0.002 g, 0.0007 mmol, 0.5 mol%) and tris(benzyltriazolylmethyl)amine (0.0003 g, 0.0007 mmol, 0.5 mol%) was added and the reaction stirred at room temperature for 20 h. The crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford a yellow gum. The compound was further purified by recrystallisation from ethanol and water to afford γ-glutamyl-L-(4-((2-(thiophen-2-yl)-1H-imidazol-1-yl)methyl)-1H-1,2,3-triazol-1-yl)alaninylglycine 59 as a yellow solid (0.019 g, 24%): Rf 0.58 (methanol: water 1:1 RP C-18 TLC); mp 180 – 182 °C (ethanol:water); [𝛼]54620 -1.18 (c 1.7 H2O); vmax (neat)/cm-1 3089.54 (br), 1612.60 (s), 1535.06 (s), 1391.17 (m), 712.71 (m); 1H NMR (500 MHz, D2O) δ ppm 7.96 (1H, s, Ar(C5)-H); 7.83 (1H, dd, J = 5.0, 1.2 Hz, Ar(C15)-H); 7.58 (1H, dd, J = 3.7, 1.2 Hz, Ar(C13)-H); 7.50 (1H, d, J = 1.7 Hz, Ar(C9)-H); 7.45 (1H, d, J = 1.7 Hz, Ar(8)-H); 7.25 (1H, dd, J = 5.0, 3.7 Hz, Ar(C14)-H); 5.54 (2H, s, (C6)H2); 4.91 (1H, dd, J = 8.1, 4.5 Hz, Ala-α-CH); 4.84 (1H, dd, J = 14.3, 4.5 Hz, Ala-β-CH2); 4.64 (1H, dd, J = 14.3, 8.1 Hz, Ala-β-CH2); 3.68 (1H, d, J = 8.2 Hz, Gly-CH2); 3.60 (1H, t, J = 6.3 Hz, Glu-α-CH); 2.36 – 2.25 (2 H, m, Glu-Ɣ-CH2); 1.94 – 1.88 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 175.9 (C=O), 174.4 (C=O), 173.7 (C=O), 169.3 (C=O), 141.2 (Ar(C11)-C), 139.8 (quaternary Ar(C12)-C), 132.4 (Ar(C9)-H), 132.4 (Ar(C13)-H), 128.6 (Ar(C4)-C), 125.4 (Ar(C14)-H), 123.2 (Ar(C15)-H), 121.3 (Ar(C5)-H), 120.2 287 (Ar(C8)-H), 57.3 (Ala-β-CH2), 53.9 (Ala-α-CH), 53.0 (Glu-α-CH), 50.6 ((C6)H2), 43.1 (Gly-CH2), 31.3 (Glu-Ɣ-CH2), 26.0 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 503.1467; C21H26N6O8 requires [M-H]- 503.1467]; m/z (ES+) 505.3 ([M+H]+, 100%); HPLC RT 0.82 min; >95% content. γ-Glutamyl-L-(4,3-(3,4-difluorophenyl)ureido)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (60) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and 1-(3,4-difluorophenyl)-3-(prop-2-yn-1-yl)urea (45) (0.03 g, 0.15 mmol, 1.00 eq) in water (0.6 mL) and t-butanol (0.3 mL), (+)-sodium L-ascorbate (0.03 g, 0.15 mmol, 1.00 eq) and copper sulfate pentahydrate (0.02 g, 0.11 mmol, 0.75 eq) were added and the reaction stirred at room temperature for 20 h. The crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated. The compound was purified by reverse phase column chromatography (10 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(43-(3,4-difluorophenyl)ureido)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 60 as a brown solid (0.029 g, 34%): Rf 0.52 (methanol: water 1:1 RP C-18 TLC); [𝛼]546 20 -0.72 (c 1.4 H2O); vmax (neat)/cm-1 3693.65 (br), 1651.22 (m),1510.68 (m), 1098.21 (m), 797.50 (m); 1H NMR (500 MHz, D2O) δ ppm 7.85 (1H, s, Ar(C5)-H), 7.29 – 7.20 (1H, m, Ar(C15)-H); 7.14 (1H, dd, J = 18.9, 9.1 Hz, Ar(C12)-H); 6.99 – 6.92 (1H, m, Ar(C11)-H); 4.96 (1H, dd, J = 7.8, 4.1 Hz, Ala-α-CH); 4.85 (1H, dd, J = 14.1, 4.1 Hz, Ala-β-CH2); 4.65 (1H, dd, J = 14.1, 7.8 Hz, Ala-β-CH2); 4.37 (2H, s, CH2), 3.79 (1H, s, Gly-CH2); 3.61 (1H, t, J = 5.9 Hz, Glu-α-CH); 2.33 (2H, t, J = 7.25 Hz, Glu-Ɣ-CH2); 2.0 – 1.89 (2H, m, Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -138.14 (1F, d, J = 23.6 Hz, Ar(C)-F), -149.09 (1F, d, J = 23.6 Hz, Ar(C)-F); HRMS m/z (ES-) [Found [M-H]- 525.1657; C20H24F2N8O7 requires [M-H]- 525.1663]; HPLC RT 6.72 min; >95% content. 288 γ-Glutamyl-L-(4-(hydroxymethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt (61) To γ-glutamyl-L-azidoalaninylglycine (32) (0.056 g, 0.17 mmol, 1.00 eq) and propargyl alcohol (0.009 g, 0.17 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0015 g, 0.003 mmol, 2 mol%) were added. The reaction stirred at room temperature for 20 h. After which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 - 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-(hydroxymethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt 61 as a colourless oil (0.019g, 33%): Rf 0.96 (methanol: water 1:1 RP C-18 TLC); vmax (neat)/cm-1 3271.57 (br), 1581.85 (s), 1387.83 (m), 1345.20 (m), 1132.21 (w), 527.75 (w); 1H NMR (500 MHz, D2O) δ ppm 8.37 (1H, br s, formic acid); 7.93 (1H, s, Ar(C5)-H); 4.99 – 4.94 (1H, m, Ala-α-CH); 4.86 (1H, dd, J = 14.3, 4.1 Hz, Ala-β-CH2); 4.67 – 4.65 (1H, m, Ala-β-CH2); 4.63 (2H, s, (C6)-H2); 3.73 – 3.65 (2H, m, Gly-CH2); 3.62 (1H, t, J = 6.3 Hz, Glu-α-CH); 2.34 (2H, t, J = 7.8 Hz, Glu-Ɣ-CH2); 1.96 – 1.92 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 174.5 (C=O); 173.8 (C=O), 170.8 (C=O), 169.5 (C=O), 147.0 (quaternary Ar(C4)-C), 124.7 (Ar(C5)-H), 54.4 (Ala-β-CH2), 54.0 (Ala-α-CH), 53.0 ((C6)H2), 50.4 (Gly-CH2), 43.2 (Glu-α-CH), 31.2 (Glu-Ɣ-CH2), 26.0 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 371.1318; C13H20N6O7 requires [M-H]- 371.1321]; m/z (ES-) 371.1 ([M-H]-, 100%). 289 γ-Glutamyl-L-(4-(acetamidomethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt (62) To γ-glutamyl-L-azidoalaninylglycine (32) (0.056 g, 0.17 mmol, 1.00 eq) and N-(prop-2-yn-1-yl)acetamide (46) (0.009 g, 0.17 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0015 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 - 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-(acetamidomethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt 62 as an off white solid (0.039g, 60%): Rf 0.95 (methanol: water 1:1 RP C-18 TLC); mp 177 °C (dec.); [𝛼]54620 -2.94 (c 1.7 H2O); vmax (neat)/cm-1 3657.85 (w), 1789.10 (s), 1548.17 (m), 1461.97 (m), 951.53 (m), 869.98 (m); 1H NMR (500 MHz, D2O) δ ppm 8.18 (0.4H, br. s, formic acid); 7.83 (1H, s, Ar(C5)-H); 4.94 (1H, dd, J = 8.2, 4.7 Hz, Ala-α-CH); 4.83 (1H, dd, J = 14.1, 4.7 Hz, Ala-β-CH2); 4.65 (1H, dd, J = 14.1, 8.2 Hz, Ala-β-CH2); 4.35 (2H, s, (C6)H2); 3.88 (2H, s, Gly-CH2); 3.69 (1H, t, J = 6.3 Hz, Glu-α-CH); 2.35 (2H, t, J = 7.5 Hz, Glu-Ɣ-CH2); 2.01 – 1.92 (2H, m, Glu-β-CH2); 1.94 (3H, s, CH3); 13C NMR (125 MHz, D2O) δ ppm 174.4 (C=O), 174.1 (C=O), 173.4 (C=O), 170.2 (C=O), 166.3 (C=O); 144.8 (quaternary Ar(C4)-C), 124.36 (Ar(C5)-H), 53.6 (Ala-α-CH), 52.9 (Ala-β-CH2), 50.3 (Gly-CH2), 48.8 (Glu-α-CH), 41.5 ((C6)H2), 34.4 (Glu-Ɣ-CH2), 25.8 (Glu-β-CH2), 21.7 (CH3); HRMS m/z (ES-) [Found [M-H]- 412.1583; C15H23N7O7 requires [M-H]- 412.1586]; m/z (ES+) 414.1563 ([M+H]+, 100%). 290 γ-Glutamyl-L-(4-(((furan-2-ylmethyl)amino)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (63) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and N-(furan-2-ylmethyl)prop-2-yn-1-amine (48) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), and recrystallised from ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-(((furan-2-ylmethyl)amino)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 63 as a yellow oil (0.02 g, 25.2%): Rf 0.50 (methanol: water 1:1 RP C-18 TLC); [𝛼]54620 -3.61 (c 0.55 H2O); vmax (neat)/cm-1 2999.41 (br), 1661.59 (s), 1428.28 (m), 1129.68 (s), 720.87 (m); 1H NMR (500 MHz, D2O) δ ppm 8.11 (1H, s, Ar(5)-H); 7.55 – 7.49 (1H, m, Ar(3’)-H); 6.55 (1H, dd, J = 10.1, 3.1 Hz, Ar(5’)-H); 6.44 – 6.37 (1H, m, Ar(4’)-H); 4.90 (1H, dd, J = 11.6, 5.3 Hz, Ala-α-CH); 4.84 – 4.65 (2H, m, Ala-β-CH2); 4.30 (2H, s, C(6)H2); 4.24 (2H, s, NH-CH2); 4.01 (1H, t, J = 6.3 Hz, Glu-α-CH); 3.82 (2H, d, J = 2.5 Hz, Gly-CH2); 2.37 (2H, t, J = 7.25 Hz, Glu-Ɣ-CH2); 2.07 – 2.01 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 176.9 (C=O), 169.6 (C=O), 144.9 (Ar(C1’)-C), 144.0 (Ar(C4)-C), 138.0 (Ar(C3’)-H), 127.1 (Ar(C5)-H), 112.9 (Ar(C5’)-H), 111.0 (Ar(C4’)-H), 53.2 (Ala-α-CH), 52.3 (Ala-β-CH2), 50.2 (NH-CH2), 42.2 (Gly-CH2), 41.9 (Glu-α-CH), 35.2 (C(6)H2), 29.5 (Glu-Ɣ-CH2), 25.9 (Glu-β-CH2); HRMS m/z (ES+) [Found [M+H]+ 452.1889; C18H25N7O7 requires [M+H]+ 452.1888]; m/z (ES+) 452.2; ([M+H]+, 98%), 903.3; ([2M+H]+, 100%); HPLC RT 0.51 min; >95% content. 291 γ-Glutamyl-L-(4-((4-hydroxytetrahydro-2H-pyran-4-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (64) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.15 mmol, 1.00 eq) and 4-(prop-2-yn-1-yl)tetrahydro-2H-pyran-4-ol (EN300-68505) (0.02 g, 0.15 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0007 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(4-((4-hydroxytetrahydro-2H-pyran-4-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 64 as a colourless solid (0.008 g, 11.1%): Rf 0.76 (methanol: water 2:8 RP C-18 TLC); mp 170 °C (dec.); [𝛼]54620 0.95 (c 2.1 H2O); vmax (neat)/cm-1 3739.99 (br), 2022.50 (w), 2669.00 (w), 1542.00 (w), 675.92 (w); 1H NMR (500 MHz, D2O) δ ppm 7.81 (1H, s, Ar(C5)-H); 4.87 (1H, dd, J = 7.2, 5.2 Hz, Ala-α-CH); 4.80 (1H, dd, J = 14.5, 5.2 Hz, Ala-β-CH2); 4.74 (1H, dd, J = 14.5, 7.2, Ala-β-CH2); 3.97 (1H, t, J = 6.3 Hz, Glu-α-CH), 3.77 – 3.60 (6H, m, Gly-CH2, (C9)H2); 2.82 (2H, s, (C6)H2); 2.33 (2H t, J = 7.8 Hz, Glu-Ɣ-CH2); 2.02 – 1.98 (2 H, m Glu-β-CH2); 1.69 – 1.60 (2H, m, C(8)H2); 1.41 (2H, d, J = 14.2 Hz, C(8)H2); 13C NMR (125 MHz, D2O) δ ppm 175.3 (C=O), 169.3 (C=O), 168.9 (C=O), 143.1 (Ar(C4)-C), 125.6 (Ar(C5)-H), 68.1 (quaternary C7), 63.4 ((C9)H2), 53.4 (Ala-α-CH), 52.5 (Ala-β-CH2), 50.0 (Glu-α-CH), 42.8 (Gly-CH2), 37.6 (C(6)H2), 35.7 ((C8)H2), 30.6 (Glu-Ɣ-CH2), 26.3 (Glu-β-CH2); HRMS m/z (ES+) [Found [M+H]+ 457.2040; C18H28N6O8 requires [M+H]+ 457.2041]; m/z (ES-) 455.2 ([M-H]-, 100%). 292 γ-Glutamyl-L-(4-(2-((carboxymethyl(methyl)amino)ethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (65) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and N-(but-3-yn-1-yl)-N-methylglycine (EN300-75325) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised from ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-(2-((carboxymethyl(methyl)amino)ethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 65 as a brown hygroscopic solid (0.03 g, 41.5%): Rf 0.90 (methanol: water 1:1 RP C-18 TLC); mp 185 °C (dec.) (ethanol:water); vmax (neat)/cm-1 3715.00 (br), 1661.60 (m), 1524.70 (m), 1390.30 (m), 993.92 (m); 1H NMR (500 MHz, D2O) δ ppm 7.90 (1H, s, Ar(C5)-H); 4.88 (1H, dd, J = 7.2, 4.4 Hz, Ala-α-CH); 4.83 - 4.76 (2H, m, Ala-β-CH2); 3.94 (1H, t, J = 6.6 Hz, Glu-α-CH); 3.72 – 3.63 (4H, m, (C8)H2, Gly-CH2); 3.48 – 3.36 (2H, m, (C7)H2); 3.13 (2H, t, J = 7.2 Hz, (C6)H2 ); 2.89 (3H, s, CH3); 2.27 – 2.18 (2H, m, Glu-Ɣ-CH2); 2.04 – 1.90 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 170.1 (C=O), 169.8 (C=O), 168.8 (C=O), 142.9 (Ar(C4)-C), 124.7 (Ar(C5)-H), 58.5 ((C8)H2), 58.4 (Ala-α-CH), 58.4 ((C7)H2), 55.4 (Ala-β-CH2), 53.4 (Glu-α-CH), 52.8 (Gly-CH2), 50.0 ((C6)H2), 41.2 (Glu-Ɣ-CH2), 27.2 (Glu-β-CH2), 20.2 (CH3); HRMS m/z (ES-) [Found [M-H]- 456.1853; C17H27N7O8 requires [M-H]- 456.1848]; m/z (ES-) 456.2 ([M-H]-, 100%). 293 γ-Glutamyl-L-(4-((3-(1-hydroxyethyl)phenoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (66) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 1-(3-(prop-2-yn-1-yloxy)phenyl)ethan-1-ol (EN300-72077) (0.03 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water) and dried under high vacuum to afford γ-glutamyl-L-(43-(1-hydroxyethyl)phenoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 66 as an off white solid (0.01 g, 12.8%): Rf 0.64 (methanol: water 2:8 RP C-18 TLC); mp 170 °C (dec.); vmax (neat)/cm-1 1987.60 (w), 1664.00 (w), 1537.10 (w), 1228.50 (w), 690.85 (w); 1H NMR (500 MHz, D2O) δ ppm 8.04 (1H, s, Ar(5)-H); 7.27 (1H, t, J = 8.2 Hz, Ar(5’)-H); 7.01 – 6.95 (2H, m, Ar(2’)-H, Ar(6’)-H); 6.92 – 6.86 (1H, m, Ar(4’)-H); 5.18 (2H, s, O-C(6)H2); 4.88 (1H, dd, J = 7.2, 5.0 Hz, Ala-α-CH); 4.84 – 4.70 (3H, m, O-CH-CH3); Ala-β-CH2); 3.93 (1H, t, J = 6.6 Hz, Glu-α-CH); 3.71 – 3.63 (2H, m, Gly-CH2); 2.31 (2H, t, J = 7.2 Hz Glu-Ɣ-CH2); 2.07 – 1.96 (2H, m, Glu-β-CH2); 1.35 (3H, d, J = 6.6 Hz, CH3); 13C NMR (125 MHz, D2O) δ ppm 177.6 (C=O), 175.2 (C=O), 169.3 (C=O), 168.8 (C=O), 157.3 (Ar(C1’)-C), 147.1 (Ar(C4)-C), 143.5 (Ar(C3’)-C), 130.0 (Ar(C5’)-H), 125.9 (Ar(C5)-H), 119.0 (Ar(C4’)-H), 114.1 (Ar(C6’)-H), 112.4 (Ar(C2’)-H), 69.4 (C(6)H2), 60.9 (O-CH-CH3), 53.3 (Ala-α-CH), 52.4 (Ala-β-CH2), 50.0 (Glu-α-CH), 42.6 (Gly-CH2), 30.4 (Glu-Ɣ-CH2), 26.2 (Glu-β-CH2), 23.5 (CH3); HRMS m/z (ES+) [Found [M+H]+ 493.2033; C21H28N6O8 requires [M+H]+ 493.2041]; m/z (ES-) 491.2 ([M-H]-, 100%); HPLC RT 3.91 min; >95% content. 294 γ-Glutamyl-L-(4-((4-oxopiperidin-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (67) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 1-(prop-2-yn-1-yl)piperidin-4-one (49) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised in ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-((4-oxopiperidin-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 67 as a yellow solid (0.05 g, 70%): Rf 0.85 (methanol: water 1:1 RP C-18 TLC); vmax (neat)/cm- 1 1676.50 (m), 1581.45 (m), 1386.86 (m), 1130.94 (m), 799.80 (m); 1H NMR (500 MHz, D2O) δ ppm 8.25 (1H, s, Ar(C5)-H); 4.90 – 4.85 (1H, m, Ala-α-CH); 4.44 – 4.36 (2H, m, (C2’)H2); 4.01 – 3.93 (2H, m, (C6’)H2); 3.75 – 3.61 (2H, m, Gly-CH2); 3.24 (2H, s, (C6)H2); 3.05 – 2.99 (1H, m, Glu-α-CH); 2.31 – 2.23 (2H, m, Glu-Ɣ-CH2); 2.07 – 1.87 (6H, m, (C3’,5’)H2, Glu-β-CH2); Ala-β-CH2 under D2O peak; HRMS m/z (ES-) [Found [M-H]- 452.1900; C18H27N7O7 requires [M-H]- 452.1899]; m/z (ES+/-) mass ion not observed in LRMS; HPLC RT 0.53 min; >95% content. 295 γ-Glutamyl-L-(4-((4-oxopropylbenzoic acid)-1H-1,2,3-triazol-1-yl)-alaninylglycine (70) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 3-(but-3-yn-1-yloxy)benzoic acid (EN300-131583) (0.03g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised in ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-((3-oxopropylbenzoic acid)-1H-1,2,3-triazol-1-yl)-alaninylglycine 70 as a colourless solid (0.012 g, 15.0%): Rf 0.39 (methanol: water 1:1 RP C-18 TLC); mp 165 °C (dec.); [𝛼]54620 -4.00 (c 0.5 H2O); vmax (neat)/cm-1 1703.90 (w), 1644.52 (w), 1251.61 (w), 1172.65 (w), 775.62 (w); 1H NMR (500 MHz, D2O) δ ppm 7.87 (2H, d, J = 9.4 Hz, Ar(C3’)H, Ar(C5’)H); 7.83 (1 H, s, Ar(C5)-H); 6.94 (2H, d, J = 9.4 Hz, Ar(C2’)H, Ar(C6’)H); 4.85 (1H, dd, J = 12.7, 5.9 Hz, Ala-α-CH); 4.78 – 4.75 (1H, m, Ala-β-CH2); 4.67 – 4.65 (1H, m, Ala-β-CH2); 4.30 (1H, t, J = 5.9 Hz, C(7)H2); 3.92 (1H, t, J = 6.3 Hz Glu-α-CH); 3.69 (1H, d, J = 4.1 Hz, Gly-CH2); 3.12 (2H, t, J = 5.9 Hz, C(6)H2); 2.29 (2H, t, J = 7.2 Hz, Glu-Ɣ-CH2); 2.00 – 1.94 (2H, m Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 177.1 (C=O), 174.7 (C=O), 171.0 (C=O), 169.2 (C=O), 169.0 (C=O), 162.0 (Ar(C1’)-C), 145.1 (quaternary Ar(C4)-C), 131.7 (Ar(C3’)-H, (Ar(C5’)-H)), 124.6 (Ar(C5)-H), 123.2 (quaternary Ar(C4’)-C), 114.4 (Ar(C2’)-H), (Ar(C6’)-H)), 66.9 ((C7)H2), 53.4 (Ala-α-CH), 52.4 (Ala-β-CH2), 49.8 (Glu-α-CH), 42.4 (Gly-CH2), 29.9 (Glu-Ɣ-CH2), 26.0 (Glu-β-CH2), 24.8 (C(6)H2); HRMS m/z (ES+) [Found [M+H]+ 507.1827; C21H26N6O9 requires [M+H]+ 507.1834]; m/z (ES+) 507.2 ([M+H]+, 100%); HPLC RT 4.73 min; >94% content. 296 γ-Glutamyl-L-(4-(6-methylpyridin-2-yl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt (71) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 2-ethynyl-6-methylpyridine (EN300-101232) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised in ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-(6-methylpyridin-2-yl)-1H-1,2,3-triazol-1-yl)-alaninylglycine formic acid salt 71 as a yellow solid (0.04 g, 58%): Rf 0.4 (methanol: water 2:8 RP C-18 TLC); mp 165 °C (dec.); [𝛼]54620 -13.01 (c 0.53 H2O); vmax (neat)/cm-1 2942.53 (br), 1586.26 (s), 1385.66 (m), 1197.78 (m), 795.54 (m); 1H NMR (500 MHz, D2O) δ ppm 8.49 (1H, s, formic acid salt); 8.33 (1H, s, Ar(C5)-H); 7.94 – 7.89 (1H, m, Ar(C4’)H); 7.77 (1H, d, J = 7.8 Hz, Ar(C3’)H); 7.35 (1H, d, J = 7.2 Hz, Ar(C5’)H); 4.98 – 4.85 (2H, m, Ala-α-CH, Ala-β-CH2); 4.44 – 4.41 (1H, m, Ala-β-CH2); 3.95 (1H, t, J = 6.1 Hz, Glu-α-CH); 3.74 – 3.66 (2H, m, Gly-CH2); 2.54 (2H, s, CH3); 2.26 – 2.22 (2H, m, Glu-Ɣ-CH2); 2.05 – 1.95 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 175.9 (C=O), 170.6 (C=O), 169.7 (C=O), 168.8 (C=O), 157.9 (Ar(C6’)-C), 145.4 (Ar(C2’)-C), 140.7 (Ar(C4)-C), 125.1 (Ar(C4’)-H), 124.6 (Ar(C5’)-H, Ar(C5)-H), 119.3 (Ar(C3’)-H), 53.4 (Ala-α-CH), 52.8 (Ala-β-CH2), 50.3 (Gly-CH2), 43.2 (Glu-α-CH), 32.2 (Glu-Ɣ-CH2), 27.0 (Glu-β-CH2); 21.8 (CH3); HRMS m/z (ES+) [Found [M+H]+ 434.1773; C18H23N7O6 requires [M+H]+ 434.1783]; m/z (ES-) 432.1; ([M-H]-, 100%). 297 γ-Glutamyl-L-(4-((3-fluorophenoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (72) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 1-fluoro-3-(prop-2-yn-1-yloxy)benzene (68) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised from ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-((3-fluorophenoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 72 as a cream solid (0.02 g, 32.5%): Rf 0.18 (methanol: water 1:1 RP C-18 TLC); mp 142 – 144 °C (ethanol:water); vmax (neat)/cm-1 1653.12 (m), 1592.08 (m), 1488.98 (m), 1133.70 (m), 678.15 (m); 1H NMR (500 MHz, D2O) δ ppm 8.05 (1H, s, Ar(C5)H); 7.25 (1H, dd, J = 15.1, 8.5 Hz, Ar(C5’)H); 6.80 – 6.69 (3H, m, Ar(C2’)H, Ar(4’)H, Ar(C6’)H); 5.17 (2H, s, (C6)H2); 4.88 (1H, dd, J = 7.2, 5.6 Hz, Ala-α-CH); 4.79 (1H, dd, J = 14.1, 5.6 Hz, Ala-β-CH2); 4.67 – 4.64 (1H, m, Ala-β-CH2); 3.93 (1H, t, J = 6.6 Hz, Glu-α-CH); 3.72 – 3.52 (2H, m, Gly-CH2); 2.32 (2H, t, J = 7.5 Hz, Glu-Ɣ-CH2); 2.02 – 1.96 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 169.3 (C=O), 168.8 (C=O), 162.2 (d, J = 244.4 Hz, Ar(C3’)-F), 158.4 (d, J = 15.2 Hz, Ar(C1’)-C), 143.5 (Ar(C4)-C), 130.6 (d, J = 10.1 Hz, Ar(C5’)-H), 125.9 (Ar(C5)-H), 111.0 (d, J = 2.5 Hz, Ar(C6’)-H), 108.4 (d, J = 21.7 Hz, Ar(C4’)-H), 102.7 (d, J = 24.9 Hz, Ar(C2’)-H), 61.1 (Ala-α-CH), 53.3 (Ala-β-CH2), 52.4 (Gly-CH2), 50.0 (Glu-α-CH), 30.2 (Glu-Ɣ-CH2), 26.1 (Glu-β-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -111.48 (1F, s, Ar(C)-F); HRMS m/z (ES+) [Found [M+H]+ 467.1680; C19H23FN6O7 requires [M+H]+ 467.1685]; m/z (ES+) 933.1 ([2M+H]+, 100%); HPLC RT 5.15 min; >95% content. 298 γ-Glutamyl-L-(4-((pyridine-4-yl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (73) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 4-ethynylpyridine hydrochloride (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised in ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-((pyridine-4-yl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 73 as a yellow gummy solid (0.03 g, 37.3%): Rf 0.75 (methanol: water 1:1 RP C-18 TLC); mp 130 – 131 °C (ethanol:water); [𝛼]54620 -3.57 (c 1 H2O); vmax (neat)/cm-1 1655.52 (m), 1534.60 (m), 1188.7 (m), 1130.48 (m), 797.46 (m); 1H NMR (500 MHz, D2O) δ ppm 8.81 (1H, s, Ar(C5)-H); 8.71 (2H, br s, Ar(C3’)-H, Ar(C5’)-H); 8.32 (2H, d, J = 6.3 Hz, Ar(C2’)-H, Ar(C4’)-H ); 5.00 – 4.93 (3H, m, Ala-β-CH2, Ala-α-CH); 4.01 (1H, t, J = 5.3 Hz, Glu-α-CH); 3.83 – 3.74 (2H, m, Gly-CH2); 2.40 – 2.32 (2H, m, Glu-Ɣ-CH2); 2.10 – 1.99 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 176.8 (C=O), 169.3 (C=O), 168.9 (C=O), 146.9 (Ar(C1’)-C), 142.6 (Ar(C4)-C), 141.7 (Ar(C3’)-H), (Ar(C5’)-H), 127.6 (Ar(C2’)-H), (Ar(C6’)-H), 122.6 (Ar(C5)-H), 57.3 (Ala-α-CH), 53.2 (Ala-β-CH2), 52.3 (Gly-CH2), 50.4 (Glu-α-CH), 29.6 (Glu-Ɣ-CH2), 25.9 (Glu-β-CH2); HRMS m/z (ES+) [Found [M+H]+ 420.1619; C17H21N7O6 requires [M+H]+ 420.1626]; m/z (ES+) 420.2; ([M+H]+, 100%); HPLC RT 0.47 min; >95% content. 299 γ-Glutamyl-L-(4-(((1,1-dioxidotetrahydrothiophen-3-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (74) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 3-(prop-2-yn-1-ylamino)tetrahydrothiophene 1,1-dioxide hydrochloride (EN300-50270) (0.03 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised in ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-(((1,1-dioxidotetrahydrothiophen-3-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 74 as a colourless oil (0.02 g, 28%): Rf 0.58 (methanol: water 1:1 RP C-18 TLC); vmax (neat)/cm-1 1660.63 (s), 1310.56 (m), 1184.58 (s), 1129.19 (s), 721.27 (m); 1H NMR (500 MHz, D2O) δ ppm 8.16 (1H, s, Ar(C5)-H); 4.93 – 4.87 (1H, m, Ala-α-CH); 4.86 – 4.80 (2H, m, Ala-β-CH2); 4.32 – 4.25 (1H, m, (C1’)H); 4.02 (1H, t, J = 6.6 Hz, Glu-α-CH); 3.94 – 3.93 (2H, m, (C6)H2); 3.86 – 3.83 (2H, m, Gly-CH2); 3.70 – 3.60 (1H, m, (C2’)H); 3.51 – 3.39 (1H, m, (C4’)H); 3.35 – 3.31 (1H, m, (C2’)H); 3.27 – 3.19 (1H, m, (C4’)H); 2.79 – 2.71 (1H, m, (C5’)H); 2.39 (2H, t, J = 6.9 Hz, Glu-Ɣ-CH2); 2.39 – 2.24 (1H, m, (C5’)H); 2.08 – 2.01 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 169.2 (C=O), 169.0 (C=O), 162.8 (C=O), 136.1 (Ar(C5)-H), 127.3 (Ar(C4)-C), 78.6 (C2’-H2), 53.2 (C1’-H), 52.3 (Ala-β-CH2), 52.0 (Ala-α-CH), 51.2 (C4’-H2), 50.3 (C(6)H2), 49.7 (Gly-CH2), 49.6 (Glu-α-CH), 35.4 (Glu-Ɣ-CH2), 25.9 (Glu-β-CH2), 25.2 (C5’-H2); HRMS m/z (ES+) [Found [M+H]+ 490.1708 (100%); C17H27N7O8S requires [M+H]+ 490.1715]; m/z (ES+/-) no mass ion observed in LRMS. 300 γ-Glutamyl-L-(4-(((7-bromochroman-4-yl)amino)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (75) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 7-bromo-N-(prop-2-yn-1-yl)chroman-4-amine (0.04 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised from ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-(((7-bromochroman-4-yl)amino)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 75 as an off white solid (0.04 g, 45%): Rf 0.26 (methanol: water 1:1 RP C-18 TLC); mp 160 °C (dec.) (ethanol:water); [𝛼]54620 -3.93 (c 1 H2O); vmax (neat)/cm-1 1652.65 (w), 1387.88 (w), 1226.67 (w), 1130.08 (w), 819.21 (w); 1H NMR (500 MHz, D2O) δ ppm 8.15 (1H, s, Ar(C5)-H); 7.46 (1H, t, J = 2.5 Hz Ar(C8’)-H); 7.42 (1H, dd, J = 8.8, 2.5 Hz, Ar(C6’)-H); 6.80 (1H, d, J = 8.8 Hz, Ar(C5’)-H); 4.87 – 4.91 (1H, m, Ala-α-CH); 4.87 – 4.83 (2H, m, Ala-β-CH2); 4.58 – 4.52 (1H, m, (C1’)H); 4.45 (2H, d, J = 4.7 Hz, Gly-CH2); 4.33 – 4.26 (1H, m, (C2’)H2); 4.24 – 4.18 (1H, m, (C2’)H2); 3.98 (1H, dt, J = 6.6, 1.5 Hz, Glu-α-CH); 3.77 – 3.63 (2H, m, (C6)H2); 2.39 – 2.31 (2H, m, (C3’)H2); 2.31 – 2.22 (2H, m, Glu-Ɣ-CH2); 2.04 – 1.95 (2H, m Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 169.5 (C=O), 168.8 (C=O), 153.8 (Ar(C8’’)-C), 138.1 (Ar(C4)-C), 134.2 (Ar(C5’)-H), 132.4 (Ar(C6’)-H), 127.2 (Ar(C4’’)-C), 119.5 (Ar(C5)-H), 117.2 (Ar(C7’)-Br), 112.2 (Ar(C8’)-H), 61.4 (C(2’)H2), 57.3 (C(4’)-H), 53.3 (Ala-β-CH2), 52.6 (Ala-α-CH), 50.8 (Glu-α-CH), 50.2 Gly-CH2), 43.0 (C(6)H2), 39.0 (C(3’)H2), 26.6 (Glu-Ɣ-CH2), 23.5 (Glu-β-CH2); HRMS m/z (ES- ) [Found [M79Br-H]-/ [M81Br-H]- 580.1160 / 582.1138; C22H28BrN7O7 requires [M79Br-H]-/ [M81Br-H]- 580.1161 / 582.1144]; m/z (ES-) 580.1 / 583.1; ([M79Br-H]-/ [M81Br-H]-, 100 / 99%); HPLC RT 2.06 min; >90% content. 301 γ-Glutamyl-L-(4-(ureidomethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (76) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 1-(prop-2-yn-1-yl)urea (EN300-40362) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), and recrystallised from ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-(ureidomethyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 76 as a colourless solid (0.006 g, 9%): Rf 0.86 (methanol: water 1:1 RP C-18 TLC); mp 160 °C (dec.) (ethanol:water); vmax (neat)/cm-1 3077.55 (br), 1651.60 (m), 1544.02 (m), 1395.20 (m), 1226.00 (m); 1H NMR (500 MHz, D2O) δ ppm 7.84 (1H, s, Ar(C5)-H); 4.88 (1H, dd, J = 7.2, 4.9 Hz, Ala-α-CH); 4.80 (1H, dd, J = 14.5, 4.9 Hz, Ala-β-CH2); 4.76 – 4.74 (1H, m, Ala-β-CH2); 4.26 (2H, s, Gly-CH2); 3.95 (1H, t, J = 6.3 Hz, Glu-α-CH); 3.73 (2H, s, (C6)H2); 2.33 (1H, t, J = 6.6 Hz Glu-Ɣ-CH2); 2.04 – 1.98 (1H, m, Glu-β-CH2); No NH2 observed; 13C NMR (125 MHz, D2O) δ ppm 178.6 (C=O), 168.9 (C=O), 137.2 (quaternary Ar(C4)-C), 124.5 (Ar(C5)-H), 57.3 (Ala-β-CH2), 53.3 (Ala-α-CH), 52.4 (Gly-CH2), 50.0 (C(6)H2), 35.0 (Glu-Ɣ-CH2), 26.3 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 413.1538; C14H22N8O7 requires [M-H]- 413.1539]; m/z (ES-) 413.1; ([M-H]-, 100%); HPLC RT 0.56 min; >94% content. γ-Glutamyl-L-(4-((2-aminoethoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (77) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 2-(prop-2-yn-1-yloxy)ethan-1-amine (EN300-134148) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper 302 sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised in ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-((2-aminoethoxy)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 77 as an orange solid (0.01 g, 18%): Rf 0.50 (methanol: water 1:1 RP C-18 TLC); mp 150 °C (dec.) (ethanol:water); vmax (neat)/cm-1 3243.84 (br), 1654.10 (m), 1538.87 (m), 1392.63 (m), 1024.27 (m); 1H NMR (500 MHz, D2O) δ ppm 8.04 (1H, s, Ar(C5)-H); 4.91 – 4.86 (1H, m, Ala-α-CH); 4.86 – 4.80 (2H, m, Ala-β-CH2); 4.61 (2H, s, (C6)H2); 3.96 (1H, t, J = 6.3 Hz, Glu-α-CH); 3.76 – 3.63 (4H, m, (C7)H2, Gly-CH2 ); 3.17 – 3.07 (2H, t, J = 4.7 Hz, (C8)H2); 2.30 – 2.22 (1H, m, Glu-Ɣ-CH2); 2.04 – 1.94 (1H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 169.5 (C=O), 168.8 (C=O), 139.9 (Ar(C4)-C), 127.8 (Ar(C5)-C), 74.2 ((C7)H2), 68.1 ((C6)H2), 65.5 (Ala-β-CH2), 62.9 (Ala-α-CH), 53.4 (Glu-α-CH), 52.7 (Gly-CH2), 50.0 ((C8)H2), 38.9 (Glu-Ɣ-CH2), 26.7 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 414.1743; C15H25N7O7 requires [M-H]- 414.1743]; m/z (ES-) 413.1; ([M-H]-, 100%). γ-Glutamyl-L-(4-((1H-pyrrol-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine (78) To γ-glutamyl-L-azidoalaninylglycine (32) (0.05 g, 0.1 mmol, 1.00 eq) and 1-(prop-2-yn-1-yl)-1H-pyrrole (69) (0.02 g, 0.1 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.005 g, 0.02 mmol, 15 mol%), copper sulfate pentahydrate (0.0008 g, 0.003 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.0013 g, 0.003 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 20 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 – 90% 0.1% formic acid/ACN in 0.1% formic acid/water), recrystallised from ethanol:water and dried under high vacuum to afford γ-glutamyl-L-(4-((1H-pyrrol-1-yl)methyl)-1H-1,2,3-triazol-1-yl)-alaninylglycine 78 as a yellow solid (0.02 303 g, 22.5%): Rf 0.91 (methanol: water 1:1 RP C-18 TLC); mp 140 °C (dec.) (ethanol:water); vmax (neat)/cm-1 1658.76 (m), 1531.94 (m), 1388.56 (m), 1226.80 (m), 1054.95 (m); 1H NMR (500 MHz, D2O) δ ppm 8.10 (1H, br s, (Ar(C1’)-H); 7.89 (1H, s, Ar(C5)-H); 6.79 – 6.75 (1H, m Ar(C2’)-H, Ar(3’)-H); 5.12 (2H, s, C(6)H2); 4.84 (1H, dd, J = 7.5, 5.0 Hz, Ala-α-CH); 4.81 – 4.75 (2H, m, Ala-β-CH2); 3.89 (1H, t, J = 6.6 Hz, Glu-α-CH); 3.69 (2H, d, J = 5.3 Hz, Gly-CH2); 2.30 (2H, t, J = 7.5 Hz, Glu-Ɣ-CH2) 1.98 – 1.92 (2H, m, Glu-β-CH2); 13C NMR (125 MHz, D2O) δ ppm 169.3 (C=O), 168.8 (C=O), 141.0 (Ar(C4)-C), 124.7 (Ar(C5)-H), 108.23 (Ar(C2’)-H), 108.2 (Ar(C5’)-H), 108.12 (Ar(C3’)-H), 108.1 (Ar(C4’)-H), 57.3 (Ala-β-CH2), 53.4 (Ala-α-CH), 52.5 (Gly-CH2), 50.0 (Glu-α-CH), 43.2 (Glu-Ɣ-CH2), 26.2 (Glu-β-CH2); HRMS m/z (ES-) [Found [M-H]- 420.1638; C17H23N7O6 requires [M-H]- 420.1637]; m/z (ES+) 423.2; ([M+H]+, 100%); HPLC RT 1.63 min; >95% content. 7.6.6 Synthesis of Alkynes identified in the unbiased kTGS Alkynes were synthesised as described or purchased from Enamine, Sigma, VWR or Alfa Aesar. 1-Fluoro-3-(prop-2-yn-1-yloxy)benzene (68, AlkD5) To 3-fluorophenol (0.50 g, 4.45 mmol, 1.00 eq) in N,N-dimethylformamide (10 mL) at 0 °C was added potassium carbonate (1.23 g, 8.92 mmol, 2.00 eq). Propargyl bromide 80% in toluene (0.59 mL, 6.6 mmol, 1.50 eq) was added and warmed to room temperature for 16 h, after which time water (10 mL) was added, extracted with ethyl acetate (3 x 10 mL), washed with brine (10 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was purified by silica gel column chromatography (0 – 100% ethyl acetate/ hexane) to afford 1-fluoro-3-(prop-2-yn-1-yloxy)benzene 68 as a colourless oil (0.54 g, 81%): Rf 0.75 (hexane: ethyl acetate, 1:1); vmax (neat)/cm-1 3297.00 (br), 1592.61 (s), 1486.96 (s), 1130.21 (s), 1029.92 (s), 676.02 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.32 (1H, dd, J = 15.4, 8.5 Hz, Ar(C5)-H); 6.83 (1H, dd, J = 8.5, 1.5 Hz, Ar(C2)-H); 6.76 – 6.69 (2H, m, Ar(C4,6)-H); 4.71 (2H, d, J = 2.5 Hz, CH2); 2.57 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 162.5 (d, J = 245.2 Hz, Ar(C3) -F), 158.8 (d, J = 10.9 Hz, Ar(C1)-O), 130.3 (d, J = 10.0 Hz, Ar(C5)-H), 110.6 (d, J = 2.7 Hz, Ar(C6)-H), 108.5 (d, J = 21.1 Hz, Ar(C4)-H), 102.9 (d, J = 24.7 Hz, Ar(C2)-H), 78.0 (C≡C-CH2), 56.0 (CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -111.66 304 (1F, s, Ar(C)-F); HRMS m/z (ES+) [Found [M+H]+ 150.0473; C9H7FO requires [M+H]+ 150.0481]; m/z (ES+) mass ion not observed in LRMS. Data in agreement with the literature.197 3-(Prop-2-yn-1-ylthio)benzoic acid (82, AlkC3) To 3-mercaptobenzoic acid (0.20 g, 1.29 mmol, 1.00 eq) in N,N-dimethylformamide (7 mL) at 0 °C was added sodium hydride (0.03 g, 1.42 mmol, 1.10 eq). Propargyl bromide 80% in toluene (0.13 mL, 1.1 mmol, 1.50 eq) was added and the reaction warmed to room temperature for 16 h, after which time the reaction mixture was quenched with saturated ammonium chloride (1 mL) and water (1 mL), extracted with ethyl acetate (3 x 10 mL), washed with brine (10 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was purified by silica gel column chromatography (0 – 100% ethyl acetate/ hexane) to afford 3-(prop-2-yn-1-ylthio)benzoic acid 82 as a colourless solid (0.12 g, 47%): Rf 0.60 (hexane: ethyl acetate, 1:1); mp 121 – 123 °C; vmax (neat)/cm-1 3286.73 (w), 2849.32 (br), 1671.94 (m), 1294.04 (s), 638.75 (s); 1H NMR (500 MHz, CDCl3) δ ppm 8.12 (1H, dd, J = 1.3, 1.3 Hz, Ar(C2)-H); 7.92 (1H, dd, J = 7.8, 1.3 Hz, Ar(C4)-H); 7.62 (1H, dd, J = 7.8, 1.3 Hz, Ar(6)-H); 7.38 (1H, t, J = 7.8 Hz, Ar(5)-H); 3.60 (2H, d, J = 2.5 Hz, CH2); 2.19 (1H, t, J = 2.5 Hz, CH); 13C NMR (125 MHz, CDCl3) δ ppm 170.1 (C=O), 134.9 (Ar(C1)-C), 133.9 (Ar(C6)-H), 130.2 (Ar(C5)-H), 129.0 (Ar(C3)-C), 128.0 (Ar(C2)-H), 127.5 (Ar(C4)-H), 78.1 (C≡C-CH2), 70.9 (CH), 21.2 (CH2); HRMS m/z (ES+) [Found [M+H]+ 191.0174; C10H8O2S requires [M+H]+ 192.0172]; m/z (ES-) 191.0 ([M-H]-, 100%). 7.6.7 Synthesis of Azides identified in the unbiased kTGS Azides were synthesised as described or purchased from Enamine. (2-Azidoethyl)benzene (80, AzG1) To sodium azide (0.26 g, 4.05 mmol, 3.00 eq) in N,N-dimethylformamide (14 mL) was added 2-phenethyl bromide (0.25 g, 0.18 mL, 1.35 mmol, 1.00 eq) dropwise and stirred at room temperature for 30 min. The reaction was then heated to 80 °C for 3 h, after 305 which time the reaction mixture was cooled, quenched with saturated ammonium chloride (5 mL) then water (5 mL) was added, extracted with ethyl acetate (3 x 10 mL), washed with brine (10 mL), dried over anhydrous MgSO4, filtered and concentrated in vacuo. The crude material was purified by silica gel column chromatography (0 – 100% ethyl acetate/ hexane) to afford (2-azidoethyl)benzene 80 as a colourless oil (0.18 g, 90%): Rf 0.81 (hexane: ethyl acetate, 1:1); vmax (neat)/cm-1 3286.84 (w), 2089.79 (m), 1669.40 (m), 1261.47 (s), 638.10 (s); 1H NMR (500 MHz, CDCl3) δ ppm 7.29 – 7.22 (2H, m, Ar(C2,6)-H); 7.20 – 7.17 (3H, m, Ar(C3,4,5)-H); 3.43 (2H, t, J = 7.2 Hz, N-CH2); 2.83 (2H, t, J = 7.2 Hz, C-CH2); 13C NMR (125 MHz, CDCl3) δ ppm 138.0 (Ar(C1)-C), 128.7 (Ar(C3,5)-H), 128.6 (Ar(C2,6)-H), 126.8 (Ar(C4)-H), 52.4 (N-CH2), 35.3 (C-CH2); m/z (ES+/-) mass ion not observed in LRMS. 2-Azidoacetonitrile (81, AzC2) To sodium azide (0.41 g, 6.25 mmol, 3.00 eq) in N,N-dimethylformamide (20 mL) was added bromoacetonitrile (0.25 g, 0.15 mL, 2.08 mmol, 1.00 eq) dropwise and stirred at room temperature for 30 min. The reaction was then heated 80 °C for 4 h, after which time the reaction mixture was cooled, quenched with saturated ammonium chloride (5 mL) then water (5 mL) was added, extracted with ethyl acetate (3 x 10 mL), washed with brine (10 mL), dried over anhydrous magnesium sulphate, filtered and concentrated in vacuo. The crude material was purified by silica gel column chromatography (0 – 100% ethyl acetate/ hexane) to afford 2-azidoacetonitrile 81 as a colourless oil (0.04 g, 23%): Rf 0.83 (hexane: ethyl acetate, 1:1); vmax (neat)/cm-1 2920.87 (s), 2852.45 (m), 2110.34 (w); 1H NMR (500 MHz, CDCl3) δ ppm 4.08 (2H, s, CH2); 13C NMR (125 MHz, CDCl3) δ ppm 113.6 (C-C≡N), 37.6 (CH2); HRMS m/z (ES+) [Found [M+H]+ 82.0000; C2H2N4 requires [M+H]+ 82.0279]; m/z (ES-) 163.0047 ([2M-H]-, 100%). 2-Azido-N,N-dimethylethan-1-amine (91, AzG2) To 2-chloro-N,N-dimethylethyl hydrochloride (1.00 g, 6.94 mmol, 1.00 eq) in water (7 mL) was added sodium azide (0.90 g, 13.88 mmol, 2.00 eq) and potassium hydroxide (0.50 g, 10.41 mmol, 1.50 eq) and stirred at room temperature for 5 days. The product was obtained by distillation at 50 °C to afford 2-azido-N,N-dimethylethan-1-amine 91 as a colourless oil (0.70 g, 88%). The compound subsequently evaporated under normal 306 fumehood temperatures and extraction conditions due to a low predicted flash point: Rf 0.09 (hexane: ethyl acetate, 1:1); vmax (neat)/cm-1 3377.46 (s), 2133.76 (w); 1H NMR (500 MHz, CDCl3) δ ppm 3.39 (2H, t, J = 6.3 Hz, (C2)H2); 2.54 (2H, t, J = 6.3 Hz, N3-(C3)H2); 2.31 (6H, s, CH3); 13C NMR (125 MHz, CDCl3) δ ppm 58.0 (C-(C2)H2), 49.0 (N3-(C3)H2), 45.4 (CH3); m/z (ES+/-) mass ion not observed in LRMS. 7.6.8 Synthesis of triazoles identified in the unbiased kTGS 2-(1-Phenethyl-1H-1,2,3-triazol-4-yl)propan-2-ol (83) To (2-azidoethyl)benzene (80) (0.05 g, 0.33 mmol, 1.00 eq) and 2-methyl-3-butyn-3-ol (0.02 g, 0.03 mL, 0.33 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.01 g, 0.05 mmol, 15 mol%), copper sulfate pentahydrate (0.001 g, 0.006 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.002 g, 0.006 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 24 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by reverse phase column chromatography (0 –90% 0.1% formic acid/methanol in 0.1% formic acid/water) and dried under high vacuum to afford 2-(1-phenethyl-1H-1,2,3-triazol-4-yl)propan-2-ol 83 as a yellow solid (0.07g, 96%): Rf 0.36 (methanol: water 1:1 RP C-18 TLC), 0.19 (hexane: ethyl acetate 1:1); mp 53 – 56 °C (Methanol / water); vmax (neat)/cm-1 1790.66 (br), 1602.77 (w), 1556.78 (w), 1454.22 (m), 960.59 (m), 730.85 (m), 692.88 (m); 1H NMR (500 MHz, D2O) δ ppm 7.54 (1H, s, Ar(C5)-H); 7.26 – 7.16 (3H, m, Ar(C3’, 4’, 5’)H); 6.95 (2H, d, J = 6.6 Hz, Ar(C2’, 6,)H); 4.58 (2H, t, J = 6.6 Hz, N-CH2); 3.11 (2H, t, J = 6.6 Hz, C-CH2-Ar); 1.45 (6H, s, CH3); 13C NMR (125 MHz, D2O) δ ppm 154.5 (Ar(C1’)-C), 137.4 (Ar(C4)-C), 128.7 (Ar(C3’,5’)-H), 128.6 (Ar(C2’,6’)-H), 126.8 (Ar(C4’)-H), 121.9 (Ar(C5)-H), 68.1 (quaternary C), 51.6 (N-CH2), 35.8 (C-CH2), 28.8 (CH3); HRMS m/z (ES+) [Found [M+H]+ 232.1442; C13H17N3O requires [M+H]+ 232.1444]; m/z (ES+) 485.1 ([2M+Na]+, 100%); HPLC RT 9.62 min; >90% content. 307 2-(1-Phenethyl-1H-1,2,3-triazol-5-yl)propan-2-ol (90) To (2-azidoethyl)benzene (80) (0.05 g, 0.33 mmol, 1.00 eq) and 2-methyl-3-butyn-3-ol (0.02 g, 0.03 mL, 0.33 mmol, 1.00 eq) in dioxane (2 mL) was added Cp*RuCl(PPh3)2 (0.005 g, 0.006 mmol, 2 mol%). The reaction was heated in the microwave at 90 °C for 1 hour. The compound was purified by silica gel column chromatography eluting with 0 – 100% ethyl acetate in hexane to 20% methanol in ethyl acetate and dried under high vacuum to afford 2-(1-phenethyl-1H-1,2,3-triazol-5-yl)propan-2-ol 90 as a cream solid (0.05 g, 67.4%). Rf 0.43 (hexane: ethyl acetate 1:1); mp 101 – 103 °C; vmax (neat)/cm-1 3226.24 (br), 1596.12 (w), 1383.26 (w), 1174.21(w), 736.02 (w); 1H NMR (500 MHz, D2O) δ ppm 7.51 (1H, s, Ar(C4)-H); 7.26 – 7.15 (3H, m, Ar(C3’, 4’,5’)H); 7.07 (2H, d, J = 6.9 Hz, ArC(2’,5’)H); 3.19 (2H, t, J = 6.6 Hz, C-CH2); 1.43 (6H, s, CH3); N-CH2 under D2O peak; 13C NMR (125 MHz, D2O) δ ppm 143.9 (Ar(C5)-C), 137.8 (Ar(C1’)-C), 131.8 (Ar(C4)-H), 131.4 (ArC(2’)-H), 131.0 (ArC(6’)-H), 128.9 (ArC(2’)-H), 128.6 (ArC(5’)-H), 126.8 (ArC(4’)-H), 62.2 (C-CH3), 50.7 (N-CH2), 35.8 (C-CH2), 28.8 (CH3), 28.7 (CH3); HRMS m/z (ES+) [Found [M+H]+ 232.1443; C13H17N3O requires [M+H]+ 232.1444]; m/z (ES+) 463.1 ([2M+H]+, 47%); HPLC RT 11.42 min; >95% content. 2-(4-((3-Fluorophenoxy)methyl)-1H-1,2,3-triazol-1-yl)acetonitrile (84) To 2-azidoacetonitrile (81) (0.04 g, 0.48 mmol, 1.00 eq) and 1-fluoro-3-(prop-2-yn-1-yloxy)benzene (CM-A-124) (0.073 g, 0.48 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.015 g, 0.07 mmol, 15 mol%), copper sulfate pentahydrate (0.002 g, 0.009 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.004 g, 0.009 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 24 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by silica gel column chromatography (0 – 100% ethyl acetate / hexane) and dried under high vacuum to afford 2-(4-((3-fluorophenoxy)methyl)-1H-1,2,3-triazol-1-yl)acetonitrile 84 as an off white solid (0.054 g, 47%). Rf 0.53 (hexane: ethyl acetate 1:1); mp 67 – 68 °C; vmax (neat)/cm-1 3154.07 (w), 1536.55 (m), 1488.88 (m), 1134.93 (m), 784.94 (m); 1H NMR (500 MHz, 308 D2O) δ ppm 8.16 (1H, s, Ar(C5)-H); 7.21 (1H, dd, J = 15.7, 7.5 Hz Ar(C4’)-H); 6.81 – 6.68 (3H, m, Ar(C2’)-H, Ar(C5’)-H, Ar(C6’)-H); 5.19 (2H, s, O-CH2), N-CH2 under D2O peak; 13C NMR (125 MHz, D2O) δ ppm 162.3 (d, J = 244.0 Hz, Ar(3’)C-F), 158.5 (d, J = 10.9 Hz, Ar(C1’)-O), 143.7 (Ar(C4)-C), 130.6 (d, J = 10.1 Hz, Ar(C5’)-H), 125.7 (Ar(C5)-H), 111.0 (d, J = 2.7 Hz, Ar(C6’)-H), 108.5 (d, J = 21.1 Hz, Ar(C4’)-H), 102.7 (d, J = 25.5 Hz, Ar(C2’)-H), 61.1 (O-CH2), 37.8 (N-CH2); 19F NMR (300 MHz, DMSO-d6) δ ppm -111.52 (1F, s, Ar(C)-F); HRMS m/z (ES+) [Found [M+H]+ 233.0833; C11H9FN4O requires [M+H]+ 233.0833]; m/z (ES-) 230.9 ([M-H]-, 100%); HPLC RT 13.3 min; >95% content. 1-(3,4-Difluorophenyl)-3-((1-(2-(dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methyl)urea (85) To 2-azido-N,N-dimethylethan-1-amine (0.05 g, 0.33 mmol, 1.00 eq) and 1-(3,4-difluorophenyl)-3-(prop-2-yn-1-yl)urea (45) (0.07 g, 0.3 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.01 g, 0.05 mmol, 15 mol%), copper sulfate pentahydrate (0.002 g, 0.007 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.003 g, 0.007 mmol, 2 mol%) were added. The reaction stirred at room temperature for 24 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by silica gel chromatography (0 – 100% ethyl acetate / hexane to 20% methanol / ethyl acetate) and dried under high vacuum to afford 1-(3,4-difluorophenyl)-3-((1-(2-(dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methyl)urea 85 as a yellow hygroscopic solid (0.09 g, 87.0%). Rf 0.15 (methanol: Ethyl Acetate 2:8); mp 63 – 65 °C; vmax (neat)/cm-1 3319.75 (br), 2950.02 (br), 2829.38 (br), 1520.24 (s), 1206.48 (s), 761.16 (m); 1H NMR (500 MHz, D2O) δ ppm 7.84 (1H, s, Ar(C5)-H); 7.22 – 7.17 (1H, m, Ar(C2’)-H); 7.11 (1H, dd, J = 19.2, 9.1 Hz, Ar(C5’)-H); 6.93 – 6.88 (1H, m, Ar(C6’)-H); 4.50 (1H, t, J = 6.3 Hz, N(triazole)-CH2-CH2); 4.35 (2H, s, NH-CH2); 3.02 – 2.87 (2H, m, N(CH3)-CH2); 2.23 (6H, s, CH3); 13C NMR (125 MHz, D2O) δ ppm 157.9 (C=O), 157.9 (C=O), 155.3 (dd, J = 189.1, 4.1 Hz, Ar(C3’)-F), 143.9 (dd, J = 209.5, 5.6 Hz, Ar(C4’)-F), 128.3 (Ar(C4)-C), 123.6 (Ar(C5)-H), 123.3 (d, J = 14.2 Hz, Ar(C6’)-H), 117.4 (d, J = 6.9 Hz, Ar(C1’)-C), 117.1 (d, J = 18.0 Hz, Ar(C5’)-C), 110.5 309 (d, J = 21.8 Hz, Ar(C2’)-C), 57.0 (N(CH3)-CH2), 47.1 (C-CH2-triazole), 43.6 (CH3), 34.9 (NH-CH2-triazole); HRMS m/z (ES+) [Found [M+H]+ 325.1583; C14H18F2N6O requires [M+H]+ 325.1583]; m/z (ES+) 671.1 ([2M+Na]+, 100%); HPLC RT 3.65 min; >95% content. 2-((1-(2-(Dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methyl)-3-oxo-2,3-dihydrobenzo[d]isothiazole-6-carboxylic acid 1,1-dioxide (86) To 2-azido-N,N-dimethylethan-1-amine (0.05 g, 0.33 mmol, 1.00 eq) and 3-oxo-2-(prop-2-yn-1-yl)-2,3-dihydrobenzo[d]isothiazole-6-carboxylic acid 1,1-dioxide (0.09 g, 0.3 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.01 g, 0.05 mmol, 15 mol%), copper sulfate pentahydrate (0.002 g, 0.007 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.003 g, 0.007 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 24 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by silica gel chromatography (0 – 100% ethyl acetate / hexane to 20% methanol / ethyl acetate) and dried under high vacuum to afford 2-((1-(2-(dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methyl)-3-oxo-2,3-dihydrobenzo[d]isothiazole-6-carboxylic acid 1,1-dioxide 86 as an off white solid (0.06 g, 48.0%). Rf 0.20 (methanol: Ethyl Acetate 2:8); mp 158 °C (dec.); vmax (neat)/cm-1 1721.3, 1298.20 (w), 1163.80 (w), 1049.30 (w), 768.00 (w); 1H NMR (500 MHz, D2O) δ ppm 7.99 (1H, d, J = 8.5 Hz, Ar(C4’)-H); 7.93 (1H, s, Ar(C5)-H); 7.70 (1H, s, Ar(C7’)-H); 7.68 (1H, d, J = 8.5 Hz, Ar(C5’)-H); 4.54 (2H, t, J = 6.1 Hz, CH2-triazole); 4.36 (2H, s, N-(C6)H2); 3.51 (2H, t, J = 6.1 Hz, N-CH2); 2.82 (6H, s, CH3); 13C NMR (125 MHz, D2O) δ ppm 171.5 (C=O), 169.0 (C=O), 143.3 (Ar(C4)-C), 139.3 (Ar(C7’)-C), 137.5 (Ar(C4’)-C), 132.9 (Ar(C5’)-H), 132.0 (Ar(C4)-C), 130.0 (Ar(C4’)-H), 128.7 (Ar(C7’)-H), 124.8 (Ar(C5)-H), 55.7 (N-(C6)H2), 44.4 (CH2-CH2-triazole), 43.0 (CH3), 37.4 (N(CH3)2-CH2-CH2); HRMS m/z (ES+) [Found [M+H]+ 380.1029; C15H17N5O5S requires [M+H]+ 380.1026]; m/z (ES+) 413.2 ([M+CH3OH+H]+, 83%); HPLC RT 0.78 min; >95% content. 310 (1-(2-(Dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methanol (87) To 2-azido-N,N-dimethylethan-1-amine (0.05 g, 0.33 mmol, 1.00 eq) and propargyl alcohol (0.02 g, 0.33 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.01 g, 0.05 mmol, 15 mol%), copper sulfate pentahydrate (0.002 g, 0.007 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.003 g, 0.007 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 24 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by silica gel chromatography (0 – 100% ethyl acetate / hexane to 20% methanol / ethyl acetate) and dried under high vacuum to afford (1-(2-(dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methanol 87 as a yellow gum (0.056 g, quantitative). Rf 0.12 (methanol: ethyl acetate 2:8); vmax (neat)/cm-1 3358.21 (br), 2922.77 (m), 1632.41 (m), 1467.84 (m), 1039.59 (m); 1H NMR (500 MHz, D2O) δ ppm 7.91 (1H, s, Ar(C5)-H); 4.62 (2H, s, O-CH2); 4.49 (2H, t, J = 6.4 Hz, C-CH2-triazole); 2.88 (2H, t, J = 6.4 Hz, N(CH3)2-CH2); 2.18 (6H, s, CH3); 13C NMR (125 MHz, D2O) δ ppm 146.7 (quaternary Ar(C4)-C), 124.1 (Ar(C5)-H), 57.2 (N(CH3)2-CH2), 54.5 (O-CH2), 47.4 (CH2-CH2-triazole), 43.7 (CH3); HRMS m/z (ES+) [Found [M+H]+ 171.1237; C7H14N4O requires [M+H]+ 171.1240]; m/z (ES+) 363.01 ([2M+Na]+, 100%). 3-(((1-(2-(Dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methyl)thio)benzoic acid (89) To 2-azido-N,N-dimethylethan-1-amine (0.05 g, 0.33 mmol, 1.00 eq) and 3-(prop-2-yn-1-ylthio)benzoic acid (82) (0.06 g, 0.3 mmol, 1.00 eq) in water (0.3 mL) and t-butanol (0.6 mL), (+)-sodium L-ascorbate (0.01 g, 0.05 mmol, 15 mol%), copper sulfate pentahydrate (0.002 g, 0.007 mmol, 2 mol%) and tris(benzyltriazolylmethyl)amine (0.003 g, 0.007 mmol, 2 mol%) were added. The reaction was stirred at room temperature for 24 h, after which time, the crude reaction was stirred gently with Chelex® for 2 h, filtered and concentrated in vacuo. The compound was purified by silica gel chromatography (0 – 100% ethyl acetate / hexane to 20% methanol / ethyl acetate) and dried under high vacuum to afford 3-(((1-(2-(dimethylamino)ethyl)-1H-1,2,3-triazol-4-yl)methyl)thio)benzoic acid 89 as a colourless solid (0.06 g, 53.0%). Rf 0.10 (methanol: 311 Ethyl Acetate 2:8); mp 135 °C (dec.); vmax (neat)/cm-1 1589.33 (m), 1552.95 (m), 1391.64 (w), 1053.62 (w), 763.51 (m); 1H NMR (500 MHz, D2O) δ ppm 7.63 – 7.61 (3H, m, Ar(C2’)-H, Ar(C6’)-H, Ar(C5)-H); 7.39 (1H, d, J = 7.5 Hz, Ar(C4’)-H); 7.27 (1H, t, J = 7.5 Hz, Ar(C5’)-H); 4.54 (2H, t, J = 5.3 Hz, N(triazole)-CH2); 4.16 (2H, s, S-CH2); 3.20 (2H, t, J = 5.3, N(CH3)-CH2); 2.46 (6H, s, CH3); 13C NMR (125 MHz, D2O) δ ppm 174.4 (C=O), 145.0 (Ar(C1’)-C), 137.1 (Ar(C6’)-H), 133.7 (Ar(C3’)-C), 133.2 (Ar(C4)-C), 131.4 (Ar(C5’)-H), 129.0 (Ar(C2’)-H), 127.7 (Ar(C4’)-H), 124.4 (Ar(C5)-H), 56.3 (N-CH2), 45.7 (C-CH2-triazole), 43.1 (CH3), 27.8 (S-CH2); HRMS m/z (ES-) [Found [M-H]- 305.1070; C14H18N4O2S requires [M-H]- 305.1078]; m/z (ES-) 305.1 ([M-H]-, 100%); HPLC RT 4.38 min; >95% content. 312 References 1. Cameron, A. D.; Ridderström, M.; Olin, B.; Mannervik, B., Structure 1999, 7 (9), 1067-1078. 2. World Health Organisation http://www.who.int/drugresistance/use/en/. 3. O’Neill, J. Antimicrobial Resistance:Tackling a crisis for the health and wealth of nations; 2014. 4. O’Neill, J. The Review on Antimicrobial Resistance. Tacking drug-resistant infections globally: Final Report and Recommendations.; 2016. 5. Fair, R. J.; Tor, Y., Perspect. Med. Chem. 2014, 6, 25-64. 6. Gram, C., Fortschritte der Medicin 1884, 2, 185-189 7. Williams, K. J., J. R. Soc. Med. 2009, 102 (8), 343-348. 8. 1st: Goodman, L. S. G., Alfred The Pharmacological Basis of Therapeutics: A Textbook of Pharmacology, Toxicology, and Therapeutics for Physicians and Medical Students. Macmillan Publishing Co.: New York, 1941; p 1387 9. Patrick, G. L., An Introduction to Medicinal Chemistry. John Wiley & Sons Ltd: The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England, Vol. 2nd edition. 10. Waksman, S. A., Proc. Natl. Acad. Sci. 1959, 45 (7), 1043-1047. 11. Walsh, C. T.; Wencewicz, T. A., J. Antibiot. 2014, 67 (1), 7-22. 12. Livermore, D. M.; Mushtaq, S.; Warner, M.; Woodford, N., J. Antimicrob. Chemother. 2009, 63 (4), 713-715. 13. Livermore, D. M., J. Antimicrob. Chemother. 2005, 56 (4), 611-614. 14. Kahne, D.; Leimkuhler, C.; Lu, W.; Walsh, C., Chem. Rev. 2005, 105 (2), 425-448. 15. Ling, L. L.; Schneider, T.; Peoples, A. J.; Spoering, A. L.; Engels, I.; Conlon, B. P.; Mueller, A.; Schaberle, T. F.; Hughes, D. E.; Epstein, S.; Jones, M.; Lazarides, L.; Steadman, V. A.; Cohen, D. R.; Felix, C. R.; Fetterman, K. A.; Millett, W. P.; Nitti, A. G.; Zullo, A. M.; Chen, C.; Lewis, K., Nature 2015, 517 (7535), 455-459. 16. Kohanski, M. A.; Dwyer, D. J.; Hayete, B.; Lawrence, C. A.; Collins, J. J., Cell 2007, 130 (5), 797-810. 17. Albesa, I.; Becerra, M. C.; Battán, P. C.; Páez, P. L., Biochem. Biophys. Res. Commun. 2004, 317 (2), 605-609. 18. Liu, Y.; Imlay, J. A., Science 2013, 339 (6124), 1210-1213. 19. Hassett, D. J.; Imlay, J. A., ACS Chem. Biol. 2007, 2 (11), 708-710. 20. Brynildsen, M. P.; Winkler, J. A.; Spina, C. S.; MacDonald, I. C.; Collins, J. J., Nat. Biotech. 2013, 31 (2), 160-165. 21. Todar, K., The online Textbook of Bacteriology. 22. D'Costa, V. M.; McGrann, K. M.; Hughes, D. W.; Wright, G. D., Science 2006, 311 (5759), 374-377. 23. Perry, J. A.; Westman, E. L.; Wright, G. D., Curr. Opin. Microbiol. 2014, 21, 45-50. 24. Perry, J. A.; Wright, G. D., BioEssays 2014, 36 (12), 1179-1184. 25. Cooper, R. A.; Anderson, A., FEBS Lett. 1970, 11 (4), 273-276. 26. Booth, I. R.; Ferguson, G. P.; Miller, S.; Li, C.; Gunasekera, B.; Kinghorn, S., Biochem. Soc. Trans. 2003, 31 (6), 1406-1408. 27. Ozyamak, E.; Black, S. S.; Walker, C. A.; Maclean, M. J.; Bartlett, W.; Miller, S.; Booth, I. R., Mol. Microbiol. 2010, 78 (6), 1577-1590. 28. Thornalley, P. J., Critical Reviews in Oncology/Hematology 1995, 20 (1–2), 99-128. 29. Rabbani, N.; Thornalley, P. J., Semin. Cell Dev. Biol. 2011, 22 (3), 261-326. 30. Sousa Silva, M.; Gomes, R. A.; Ferreira, A. E.; Ponces Freire, A.; Cordeiro, C., Biochem. J. 2013, 453 (1), 1-15. 313 31. Nagai, R.; Ikeda, K.; Kawasaki, Y.; Sano, H.; Yoshida, M.; Araki, T.; Ueda, S.; Horiuchi, S., FEBS Lett. 1998, 425 (2), 355-360. 32. Armeni, T.; Cianfruglia, L.; Piva, F.; Urbanelli, L.; Luisa Caniglia, M.; Pugnaloni, A.; Principato, G., Free Radical Biol. Med. 2014, 67 (0), 451-459. 33. Honek, J. F., Biochem. Soc. Trans. 2014, 42 (2), 479-484. 34. Thornalley, P. J., Mol. Aspects Med. 1993, 14 (4), 287-371. 35. Honek, J. F., Biomol. Concepts 2015, 6 (5-6), 401-414. 36. Ferguson, G. P.; Munro, A. W.; Douglas, R. M.; McLaggan, D.; Booth, I. R., Mol. Microbiol. 1993, 9 (6), 1297-1303. 37. Deponte, M., Biochem. Soc. Trans. 2014, 42 (2), 473-478. 38. Rabbani, N.; Thornalley, P. J., Amino acids 2012, 42 (4), 1087-1096. 39. Ozyamak, E.; de Almeida, C.; de Moura, A. P. S.; Miller, S.; Booth, I. R., Mol. Microbiol. 2013, 88 (5), 936-950. 40. Sakamoto, H.; Mashima, T.; Sato, S.; Hashimoto, Y.; Yamori, T.; Tsuruo, T., Clinical cancer research : an official journal of the American Association for Cancer Research 2001, 7 (8), 2513-2518. 41. Nass, N.; Sel, S.; Ignatov, A.; Roessner, A.; Kalinski, T., Biochim. Biophys. Acta. 2016, 1860 (6), 1272-1280. 42. Dong, L.; Zhou, Q.; Zhang, Z.; Zhu, Y.; Duan, T.; Feng, Y., The journal of obstetrics and gynaecology research 2012, 38 (8), 1077-1085. 43. Thornalley, P. J.; Rabbani, N., Semin. Cell Dev. Biol. 2011, 22 (3), 318-325. 44. Ferguson, G. P.; Battista, J. R.; Lee, A. T.; Booth, I. R., Mol. Microbiol. 2000, 35 (1), 113-122. 45. Saint-Jean, A. P.; Phillips, K. R.; Creighton, D. J.; Stone, M. J., Biochemistry 1998, 37 (29), 10345-10353. 46. Sukdeo, N.; Clugston, Susan L.; Daub, E.; Honek, John F., Biochem. J. 2004, 384 (Pt 1), 111-117. 47. Kizil, G.; Wilks, K.; Wells, D.; Ala'Aldeen, D. A., J. Med. Microbiol. 2000, 49 (7), 669-673. 48. Vickers, T. J.; Greig, N.; Fairlamb, A. H., Proc. Natl. Acad. Sci. U S A 2004, 101 (36), 13186-13191. 49. Suttisansanee, U.; Honek, J. F., Semin. Cell Dev. Biol. 2011, 22 (3), 285-292. 50. He, P.; Moran, G. R., J. Inorg. Biochem. 2011, 105 (10), 1259-1272. 51. He, M. M.; Clugston, S. L.; Honek, J. F.; Matthews, B. W., Biochemistry 2000, 39 (30), 8719-8727. 52. He, M. M. C., S. L.; Honek, J. F.; Matthews, B. W; Biochemistry 2000, 39 (30), 8719-8727. 53. Cameron, A. D.; Olin, B.; Ridderstrom, M.; Mannervik, B.; Jones, T. A., EMBO J. 1997, 16 (12), 3386-3395. 54. Inoue, Y.; Kimura, A., J. Biol. Chem. 1996, 271 (42), 25958-25965. 55. Kim, N. S.; Umezawa, Y.; Ohmura, S.; Kato, S., J. Biol. Chem. 1993, 268 (15), 11217-11221. 56. Ly, H. D.; Susan, L. C.; Sampson, P. B.; Honek, J. F., Bioorg. Med. Chem. Lett. 1998, 8 (7), 705-710. 57. Cameron, A. D.; Ridderstrom, M.; Olin, B.; Kavarana, M. J.; Creighton, D. J.; Mannervik, B., Biochemistry 1999, 38 (41), 13480-13490. 58. Suttisansanee, U.; Honek, J. F., Semin. Cell. Dev. Biol. 2011, 22 (3), 285-292. 59. Campos-Bermudez, V. A.; Leite, N. R.; Krog, R.; Costa-Filho, A. J.; Soncini, F. C.; Oliva, G.; Vila, A. J., Biochemistry 2007, 46 (39), 11069-11079. 60. Ferguson, G. P.; Booth, I. R., J. Bacteriol. 1998, 180 (16), 4314-4318. 314 61. Pittman, M. S.; Robinson, H. C.; Poole, R. K., J. Biol. Chem. 2005, 280 (37), 32254-32261. 62. Cacciatore, I.; Cornacchia, C.; Pinnen, F.; Mollica, A.; Di Stefano, A., Molecules 2010, 15 (3), 1242-1264. 63. Loi, V. V.; Rossius, M.; Antelmann, H., Frontiers in Microbiology 2015, 6, 187. 64. Gopal, S.; Borovok, I.; Ofer, A.; Yanku, M.; Cohen, G.; Goebel, W.; Kreft, J.; Aharonowitz, Y., J. Bacteriol. 2005, 187 (11), 3839-3847. 65. Smirnova, G. V.; Oktyabrsky, O. N., Biochemistry (Moscow) 2005, 70 (11), 1199-1211. 66. Krezel, A.; Wojcik, J.; Maciejczyk, M.; Bal, W., Inorg. Chem. 2011, 50 (1), 72-85. 67. Dalle-Donne, I.; Rossi, R.; Colombo, G.; Giustarini, D.; Milzani, A., Trends Biochem. Sci. 34 (2), 85-96. 68. Manetsch, R.; Krasinski, A.; Radic, Z.; Raushel, J.; Taylor, P.; Sharpless, K. B.; Kolb, H. C., J. Am. Chem. Soc. 2004, 126 (40), 12809-12818. 69. Rideout, D., Science 1986, 233 (4763), 561-563. 70. Manetsch, R.; Krasiński, A.; Radić, Z.; Raushel, J.; Taylor, P.; Sharpless, K. B.; Kolb, H. C., J. Am. Chem. Soc. 2004, 126 (40), 12809-12818. 71. Oueis, E.; Sabot, C.; Renard, P.-Y., Chem. Commun. 2015, 51 (61), 12158-12169. 72. Huc, I.; Lehn, J.-M., Proc. Natl. Acad. Sci. 1997, 94 (6), 2106-2110. 73. Ramström, O.; Lehn, J.-M., ChemBioChem 2000, 1 (1), 41-48. 74. Bosc, D.; Jakhlal, J.; Deprez, B.; Deprez-Poulain, R., Future Med. Chem. 2016, 8 (4), 381-404. 75. Mamidyala, S. K.; Finn, M. G., Chem. Soc. Rev. 2010, 39 (4), 1252-1261. 76. Krasiński, A.; Radić, Z.; Manetsch, R.; Raushel, J.; Taylor, P.; Sharpless, K. B.; Kolb, H. C., J. Am. Chem. Soc. 2005, 127 (18), 6686-6692. 77. Oueis, E.; Nachon, F.; Sabot, C.; Renard, P.-Y., Chem. Commun. 2014, 50 (16), 2043-2045. 78. Oueis, E.; Sabot, C.; Renard, P.-Y., Chem. Commun. 2015, 51 (61), 12158-12169. 79. Jaegle, M.; Steinmetzer, T.; Rademann, J., Angew. Chem., Int. Ed. 2017, 56 (13), 3718-3722. 80. Kolb, H. C.; Sharpless, K. B., Drug Discovery Today 2003, 8 (24), 1128-1137. 81. Kolb, H. C.; Finn, M. G.; Sharpless, K. B., Angew. Chem. Int. Ed. Engl. 2001, 40 (11), 2004-2021. 82. Lewis, W. G.; Green, L. G.; Grynszpan, F.; Radić, Z.; Carlier, P. R.; Taylor, P.; Finn, M. G.; Sharpless, K. B., Angew. Chem. Int. Ed. 2002, 41 (6), 875-875. 83. Mocharla, V. P.; Colasson, B.; Lee, L. V.; Roper, S.; Sharpless, K. B.; Wong, C. H.; Kolb, H. C., Angew. Chem. Int. Ed. Engl. 2004, 44 (1), 116-120. 84. Whiting, M.; Muldoon, J.; Lin, Y. C.; Silverman, S. M.; Lindstrom, W.; Olson, A. J.; Kolb, H. C.; Finn, M. G.; Sharpless, K. B.; Elder, J. H.; Fokin, V. V., Angew. Chem. Int. Ed. Engl. 2006, 45 (9), 1435-1439. 85. Hirose, T.; Sunazuka, T.; Sugawara, A.; Endo, A.; Iguchi, K.; Yamamoto, T.; Ui, H.; Shiomi, K.; Watanabe, T.; Sharpless, K. B.; Omura, S., The Journal of antibiotics 2009, 62 (5), 277-282. 86. Willand, N.; Desroses, M.; Toto, P.; Dirié, B.; Lens, Z.; Villeret, V.; Rucktooa, P.; Locht, C.; Baulard, A.; Deprez, B., ACS Chem. Biol. 2010, 5 (11), 1007-1013. 87. Deprez-Poulain, R.; Hennuyer, N.; Bosc, D.; Liang, W. G.; Enée, E.; Marechal, X.; Charton, J.; Totobenazara, J.; Berte, G.; Jahklal, J.; Verdelet, T.; Dumont, J.; Dassonneville, S.; Woitrain, E.; Gauriot, M.; Paquet, C.; Duplan, I.; Hermant, P.; Cantrelle, F.-X.; Sevin, E.; Culot, M.; Landry, V.; Herledan, A.; Piveteau, C.; Lippens, G.; Leroux, F.; Tang, W.-J.; van Endert, P.; Staels, B.; Deprez, B., Nat. Commun. 2015, 6, 8250-8263. 315 88. Jencks, W. P., Proc. Natl. Acad. Sci. U. S. A. 1981, 78 (7), 4046-4050. 89. Farmer, P. S.; Ariëns, E. J., Trends Pharmacol. Sci. 3, 362-365. 90. Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W., Science 1996, 274 (5292), 1531-1534. 91. Bollag, G.; Tsai, J.; Zhang, J.; Zhang, C.; Ibrahim, P.; Nolop, K.; Hirth, P., Nat. Rev. Drug Discov. 2012, 11 (11), 873-886. 92. Rees, D. C.; Congreve, M.; Murray, C. W.; Carr, R., Nat. Rev. Drug Discov. 2004, 3 (8), 660-672. 93. Mondal, M.; Unver, M. Y.; Pal, A.; Bakker, M.; Berrier, S. P.; Hirsch, A. K. H., Chem. Eur. J. 2016, 22 (42), 14826-14830. 94. Murray, C. W.; Rees, D. C., Angew. Chem., Int. Ed. 2016, 55 (2), 488-492. 95. Erlanson, D. A.; Fesik, S. W.; Hubbard, R. E.; Jahnke, W.; Jhoti, H., Nat. Rev. Drug Discov. 2016, 15 (9), 605-619. 96. Congreve, M.; Carr, R.; Murray, C.; Jhoti, H., Drug Discov. Today 2003, 8 (19), 876-877. 97. Jhoti, H.; Williams, G.; Rees, D. C.; Murray, C. W., Nat. Rev. Drug Discov. 2013, 12 (8), 644-645. 98. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J., Adv. Drug Deliv. Rev. 2001, 46 (1-3), 3-26. 99. Schultes, S.; de Graaf, C.; Haaksma, E. E. J.; de Esch, I. J. P.; Leurs, R.; Krämer, O., Drug Discovery Today: Technol. 2010, 7 (3), 157-162. 100. Kumar, A.; Voet, A.; Zhang, K. Y. J., Current Med. Chem. 2012, 19 (30), 5128-5147. 101. Joseph-McCarthy, D.; Campbell, A. J.; Kern, G.; Moustakas, D., J. Chem. Inf. Model. 2014, 54 (3), 693-704. 102. Doak, B. C.; Norton, R. S.; Scanlon, M. J., Pharmacol. Ther. 2016, 167, 28-37. 103. McTigue, M. A.; Williams, D. R.; Tainer, J. A., J. Mol. Biol. 1995, 246 (1), 21-27. 104. Daunes, S.; D'Silva, C.; Kendrick, H.; Yardley, V.; Croft, S. L., J. Med. Chem. 2001, 44 (18), 2976-2983. 105. D'Silva, C.; Daunes, S.; Rock, P.; Yardley, V.; Croft, S. L., J. Med. Chem. 2000, 43 (10), 2072-2078. 106. Lim, K.; Ho, J. X.; Keeling, K.; Gilliland, G. L.; Ji, X.; Rüker, F.; Carter, D. C., Protein Sci. 1994, 3 (12), 2233-2244. 107. More, S. S.; Vince, R., J. Med. Chem. 2009, 52 (15), 4650-4656. 108. Vince, R.; Daluge, S.; Wadd, W. B., J. Med. Chem. 1971, 14, 402-404. 109. Thornalley, P. J.; Edwards, L. G.; Kang, Y.; Wyatt, C.; Davies, N.; Ladan, M. J.; Double, J., Biochem. Pharmacol. 1996, 51 (10), 1365-1372. 110. Hamilton, D. S.; Kavarana, M. J.; Sharkey, E. M.; Eiseman, J. L.; Creighton, D. J., J. Med. Chem. 1999, 42 (10), 1823-1827. 111. Kavarana, M. J.; Kovaleva, D. J.; Creighton, D. J.; Wollman, M. B.; Eiseman, J. L., J. Med. Chem. 1999, 42 (2), 221-228. 112. Murthy, N. S. R. K.; Bakeris, T.; Kavarana, M. J.; Hamilton, D. S.; Lan, Y.; Creighton, D. J., J. Med. Chem. 1994, 37 (14), 2161-2166. 113. Hoa D. Ly, S. L. C., Peter B. Sampson, and John F. Honek, Bioorg. Med. Chem. Lett. 1998, 8, 705-710. 114. Thornalley, P. J.; Edwards, L. G.; Kang, Y.; Wyatt, C.; Davies, N.; Ladan, M. J.; Double, J., Biochem. Pharmacol. 1996, 51, 1365-1372. 115. Douglas, K. T.; Al-Timari, A.; D'Silva, C.; Gohel, D. I., Biochem. J. 1982, 207 (2), 323-329. 116. Murthy, N. S. R. K.; Bakeris, T.; Kavarana, M. J.; Hamilton, D. S.; Lan, Y.; Creighton, D. J., J. Med. Chem. 1994, 37 (14), 2161-2166. 316 117. Mitsumoto, A.; Kim, K.-R.; Oshima, G.; Kunimoto, M.; Okawa, K.; Iwamatsu, A.; Nakagawa, Y., J. Biochem. 2000, 128 (4), 647-654. 118. Campaniço, A. F. d. S. Targeting bacterial glyoxalase I for the development of novel antibacterials. 2015-2016. 119. Zeitlmann, L.; Niestroj, A.; Heiser, U. Inhibitors of protein kinases. US2011224225 (A1), 2011. 120. Levy, E. J.; Anderson, M. E.; Meister, A., Proc. Natl. Acad. Sci.1993, 90 (19), 9171-9175. 121. D'Silva, C., Biochem. J. 1990, 271, 167-169. 122. O'Young, J.; Sukdeo, N.; Honek, J. F., Arch. Biochem. Biophys. 2007, 459 (1), 20-26. 123. Gasteiger E., H. C., Gattiker A., Duvaud S., Wilkins M.R., Appel R.D., Bairoch A.;, Protein Identification and Analysis Tools on the ExPASy Server;. Humana Press: p 571-607 124. Suttisansanee, U. PhD Thesis, University of Waterloo, Investigation of the Microbial Glyoxalase System. 2011. 125. Andersen, K. R.; Leksa, N. C.; Schwartz, T. U., Proteins: Struct., Funct., Bioinf. 2013, 81 (11), 1857-1861. 126. Robichon, C.; Luo, J.; Causey, T. B.; Benner, J. S.; Samuelson, J. C., Appl. Environ. Microbiol. 2011, 77 (13), 4634-4646. 127. Suttisansanee, U.; Lau, K.; Lagishetty, S.; Rao, K. N.; Swaminathan, S.; Sauder, J. M.; Burley, S. K.; Honek, J. F., J. Biol. Chem. 2011, 286 (44), 38367-38374. 128. Tolia., N. H.; Joshua-Tor., L., Nat. Methods 2006, 3 (1), 55-64. 129. Niesen, F. H.; Berglund, H.; Vedadi, M., Nat. Protocols 2007, 2 (9), 2212-2221. 130. Cameron, A. D.; Ridderström, M.; Olin, B.; Kavarana, M. J.; Creighton, D. J.; Mannervik, B., Biochemistry 1999, 38 (41), 13480-13490. 131. Baba, T.; Ara, T.; Hasegawa, M.; Takai, Y.; Okumura, Y.; Baba, M.; Datsenko, K. A.; Tomita, M.; Wanner, B. L.; Mori, H., Mol. Syst. Biol. 2006, 2 (1). 132. Andrews, J. M., The Journal of antimicrobial chemotherapy 2001, 48 Suppl 1, 5-16. 133. Balouiri, M.; Sadiki, M.; Ibnsouda, S. K., J. Pharm. Anal. 2016, 6 (2), 71-79. 134. Heatley, N. G., Biochem. J. 1944, 38 (1), 61-65. 135. Ozyamak, E.; Black, S. S.; Walker, C. A.; MacLean, M. J.; Bartlett, W.; Miller, S.; Booth, I. R., Mol. Microbiol. 2010, 78 (6), 1577-1590. 136. Cameron, A. D.; Ridderström, M.; Olin, B.; Mannervik, B., Structure 1999, 7 (9), 1067-1078. 137. Elliott, P. I. P., Chapter 1 Organometallic complexes with 1,2,3-triazole-derived ligands. In Organometallic Chemistry: Volume 39, The Royal Society of Chemistry: 2014; Vol. 39, pp 1-25. 138. Jacobsen, Ø.; Maekawa, H.; Ge, N.-H.; Görbitz, C. H.; Rongved, P.; Ottersen, O. P.; Amiry-Moghaddam, M.; Klaveness, J., The Journal of Organic Chemistry 2011, 76 (5), 1228-1238. 139. Walker, P.; Healy, J. The use of in situ click chemistry for the identification of novel GlxI inhibitors. University of Oxford, 2012. 140. Nyffeler, P. T.; Liang, C.-H.; Koeller, K. M.; Wong, C.-H., J. Am. Chem. Soc. 2002, 124 (36), 10773-10778. 141. Lau, Y. H.; Spring, D. R., Synlett 2011, 2011 (13), 1917-1919. 142. Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B., Angew. Chem., Int. Ed. 2002, 41 (14), 2596-2599. 143. Donnelly, P. S.; Zanatta, S. D.; Zammit, S. C.; White, J. M.; Williams, S. J., Chem. Commun. 2008, (21), 2459-2461. 144. Chan, T. R.; Hilgraf, R.; Sharpless, K. B.; Fokin, V. V., Org. Lett. 2004, 6 (17), 2853-2855. 317 145. Ju Myung Kwak, J. S. M., Sangei Seo, Jea I Choi, Vasu Sampath, Hee-Yoon Lee, and Hun Yeong Koh, Bull. Korean Chem. Soc. 2014, 35 (12), 3675-3678. 146. Boren, B. C.; Narayan, S.; Rasmussen, L. K.; Zhang, L.; Zhao, H.; Lin, Z.; Jia, G.; Fokin, V. V., J. Am. Chem. Soc. 2008, 130 (28), 8923-8930. 147. Boz, E.; Tüzün, N. Ş., J. Organomet. Chem. 2013, 724, 167-176. 148. Egger, J.; Weckerle, C.; Cutting, B.; Schwardt, O.; Rabbani, S.; Lemme, K.; Ernst, B., J. Am. Chem. Soc. 2013, 135 (26), 9820-9828. 149. Durrant, J. D.; McCammon, J. A., PLoS Comput. Biol. 2012, 8 (3), e1002397. 150. Trott, O.; Olson, A. J., J. Comput. Chem. 2010, 31 (2), 455-461. 151. Sousa Silva, M.; Ferreira, A. E. N.; Tomás, A. M.; Cordeiro, C.; Ponces Freire, A., FEBS J. 2005, 272 (10), 2388-2398. 152. Reiger, M.; Lassak, J.; Jung, K., FEMS Microbiol. Lett. 2015, 362 (2), 1-7. 153. Arai, M.; Nihonmatsu-Kikuchi, N.; Itokawa, M.; Rabbani, N.; Thornalley, P. J., Biochem. Soc. Trans. 2014, 42 (2), 491-494. 154. Chen, W.; Zhao, Y.; Seefeldt, T.; Guan, X., J. Pharm. Biomed. Anal. 2008, 48 (5), 1375-1380. 155. Peng, H.; Chen, W.; Cheng, Y.; Hakuna, L.; Strongin, R.; Wang, B., Sensors (Basel, Switzerland) 2012, 12 (11), 15907-15946. 156. Brückner, C.; Rettig, S. J.; Dolphin, D., Inorg. Chem. 2000, 39 (26), 6100-6106. 157. Di Antonio, M.; Biffi, G.; Mariani, A.; Raiber, E.-A.; Rodriguez, R.; Balasubramanian, S., Angew. Chem., Int. Ed. Engl. 2012, 51 (44), 11073-11078. 158. KNIME https://www.knime.org/knime-analytics-platform. 159. Beisken, S.; Meinl, T.; Wiswedel, B.; de Figueiredo, L. F.; Berthold, M.; Steinbeck, C., BMC Bioinf. 2013, 14, 257-257. 160. O’Shea, R.; Moser, H. E., J. Med. Chem 2008, 51 (10), 2871-2878. 161. Keserű, G. M.; Erlanson, D. A.; Ferenczy, G. G.; Hann, M. M.; Murray, C. W.; Pickett, S. D., J. Med. Chem. 2016, 59 (18), 8189-8206. 162. Johnson, C. N.; Erlanson, D. A.; Murray, C. W.; Rees, D. C., J. Med. Chem. 2017, 60 (1), 89-99. 163. Fielding, L., Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51 (4), 219-242. 164. Scott, D. E.; Spry, C.; Abell, C., Differential Scanning Fluorimetry as Part of a Biophysical Screening Cascade. In Fragment-based Drug Discovery Lessons and Outlook, Wiley-VCH Verlag GmbH & Co. KGaA: 2016; pp 139-172. 165. Fielding, L., Tetrahedron 2000, 56 (34), 6151-6170. 166. Gerig, J. T., Fluorine NMR. 1997. 167. Tengel, T.; Fex, T.; Emtenas, H.; Almqvist, F.; Sethson, I.; Kihlberg, J., Org. Biomol. Chem. 2004, 2 (5), 725-731. 168. Lee, Y.; Zeng, H.; Ruedisser, S.; Gossert, A. D.; Hilty, C., J. Am. Chem. Soc. 2012, 134 (42), 17448-17451. 169. Chen, H.; Viel, S.; Ziarelli, F.; Peng, L., Chem. Soc. Rev. 2013, 42 (20), 7971-7982. 170. Filler, R.; Saha, R., Future Med. Chem. 2009, 1 (5), 777-791. 171. Lea, W. A.; Simeonov, A., PLoS ONE 2012, 7 (4), e36219. 172. Kaplan, W.; Erhardt, J.; Sluis-Cremer, N.; Dirr, H.; Hüsler, P.; Klump, H., Protein Sci. 1997, 6 (2), 399-406. 173. Smith, D. B.; Johnson, K. S., Gene 1988, 67 (1), 31-40. 174. Andujar-Sanchez, M.; Smith, A. W.; Clemente-Jimenez, J. M.; Rodriguez-Vico, F.; Las Heras-Vazquez, F. J.; Jara-Perez, V.; Camara-Artigas, A., Biochemistry 2005, 44 (4), 1174-1183. 175. Lim, K.; Ho, J. X.; Keeling, K.; Gilliland, G. L.; Ji, X.; Ruker, F.; Carter, D. C., Protein Sci. 1994, 3, 2233. 318 176. Kaplan, W.; Hüsler, P.; Klump, H.; Erhardt, J.; Sluis-Cremer, N.; Dirr, H.; Protein Sci. 1997, 6, 399-406. 177. Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew, R. K.; Goodsell, D. S.; Olson, A. J., J. Comput. Chem. 2009, 30 (16), 2785-2791. 178. Vainio, M. J.; Johnson, M. S., J. Chem. Inf. Model. 2007, 47 (6), 2462-2474. 179. O'Boyle, N. M.; Banck, M.; James, C. A.; Morley, C.; Vandermeersch, T.; Hutchison, G. R., J. Cheminform. 2011, 3, 33. 180. James J. P. Stewart, S. MOPAC2012. HTTP://OpenMOPAC.net 181. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E., J. Comput. Chem. 2004, 25 (13), 1605-1612. 182. Vince, R.; Daluge, S., J. Med. Chem 1971, 14 (1), 35-37. 183. Chen, L.; Sun, W.; Li, J.; Liu, Z.; Ma, Z.; Zhang, W.; Du, L.; Xu, W.; Fang, H.; Li, M., Org. Biomol. Chem. 2013, 11 (2), 378-382. 184. Thornalley, P. K., Biochem. J 1991, 275 (Pt 2), 535-539. 185. Thornalley, P. J.; Ladan, M. J.; Ridgway, S. J. S.; Kang, Y., J. Med. Chem. 1996, 39 (17), 3409-3411. 186. Meyer, C.; Hoeger, B.; Chatterjee, J.; Köhn, M., Biorg. Med. Chem. 2015, 23 (12), 2848-2853. 187. Goddard-Borger, E. D.; Stick, R. V., Org. Lett. 2007, 9 (19), 3797-3800. 188. Yu, B.; Qi, P.-P.; Shi, X.-J.; Huang, R.; Guo, H.; Zheng, Y.-C.; Yu, D.-Q.; Liu, H.-M., Eur. J. Med. Chem. 2016, 117, 241-255. 189. Zhao, L.; Mao, L.; Hong, G.; Yang, X.; Liu, T., Bioorg. Med.Chem. Lett. 2015, 25 (12), 2540-2544. 190. Bouhfid, R.; Joly, N.; Essassi, E. M.; Lequart, V.; Massoui, M.; Martin, P., Synth. Commun. 2011, 41 (14), 2096-2102. 191. Walser, A.; Flynn, T.; Mason, C.; Crowley, H.; Maresca, C.; Yaremko, B.; O'Donnell, M., J. Med. Chem. 1991, 34 (3), 1209-1221. 192. Pyne, W. J.; Szabo, S. S.; Holm, R. E., J. Agric. Food Chem. 1979, 27 (3), 537-543. 193. Ouerghui, A.; Elamari, H.; Dardouri, M.; Ncib, S.; Meganem, F.; Girard, C., React. Funct. Polym. 2016, 100, 191-197. 194. Baiga, T. D., M.; Evans, R.; Kluge, A.; Lagu, B.; Miura, M.; Panigrahi, S.K.; Patane, M.; Samajdar, S.; Senaiar, R.; Takahashi,T.; WO2015US54475 20151007. 2016. 195. Ma, Z. N., F.; Woo, G. H. C.; Lo, S.-M.; Roveto, P. M.; Schaus, S. E.; Snyder, J. K., Beilstein J. Org. Chem. 2012, 8, 829–840. 196. Álvarez-Pérez, M.; Marco-Contelles, J., Arch. Org. Chem. 2011, (2), 283-296. 197. Chen, Y.-Y.; Chen, K.-L.; Tyan, Y.-C.; Liang, C.-F.; Lin, P.-C., Tetrahedron 2015, 71 (36), 6210-6218. 319 Appendix Chapter 1: Table 61. Structures of Antibiotics discussed in Chapter 1. Structure Drug class Antibiotic examples Salvarsan Sulfonamide Prontosil (shown) Penicillin R = benzyl (Penicillin G) R = phenoxymethyl (Penicillin V) R = 2,6-dimethoxybenzyl (Methicillin) R = 2-amino-2-phenyl (Ampicillin) R = 2-amino-2-(p-hydroxyphenyl) (Amoxicillin) R = 2-carboxy-2-phenyl (Carbenicillin) R = 2-(4-ethyl-2,3-dioxopiperazine-1- R carboxamido)-2-phenyl (Piperacillin) β-lactam Cephalosporin C (shown) Ceftazidime (shown) Aminoglycoside Streptomycin (shown) Kanamycin (shown) Glycopeptide Vancomycin (shown) 320 Structure Drug class Antibiotic examples Lipoglycopeptide Telavancin Peptide Chloramphenicol (shown) Cyclic peptide Cycloserine (shown) Tetracycline Chlortetracycline (shown) Macrolide Erythromycin (shown) Ketolide Telithromycin (shown) Quinolone R1 = ethyl, R2 = methyl, R3 = H, X = N (Nalidixic acid) R1 = cyclopropyl, R2 = piperazine, R3 = F, X = CH (Ciprofloxacin) R1 = ethyl, R2 = piperazine, R3 = F, X = N (Enoxacin) R1 = ethyl, R2 = piperazine, R3 = F, X = CH (Norfloxacin) Aminoacridine Proflavin (shown) Polymyxin/ Polypeptide Colistin (shown) 321 Structure Drug class Antibiotic examples Polymyxin Tyrothricin (shown) Oxazolidinone Linezolid (shown) Glycycline Tigecycline (shown) Pleuromutilin Retapamulin (shown) Macrocyclic Fidaxomicin (shown) Lipid II Teixobactin (shown) 322 Chapter 2: DNA sequencing data DNA sequencing results for pTrcgloBH6 with primers pTrcHisForward and pTrcReverse (Table 62). Table 62. DNA sequencing results for pTrccgloBH6. pTrcHisForward pTrcReverse DNA sequence CCGTGGCCTACATTACGTTGACATGGGAGAGTGCGGCAAACCCCCCTCCCTGTAAGGAGCGACATTAAAAATTACGGTGCCATGTATTAGGAATAGACTGGTTGGAGTGCAGATACTTGCCCTGAGGCACTTACTATCTATGAATTAACCGGGGAAGAACGTCCTTTAAGTTTGGCGTTCCTCACAATCGGGAAGTAGAGGACTTTATGCCGAAACCGCGAGTCAAAAAAACGACCAAAGAAGATACGTCTGATTTTCCTTAAAATATGCTCAATCCAAAATATACTTTTGATACTTTTGAAATCGGATCTGTAAACCGATTTGCACATGCTGCTTCCCTCTGCTATTGGAAACGCCCGCTAAAGCTTACAACCCTTTATGGATCTATGGGGGCGTCGGCTTAGGGAAAACACACTTAATGCATGCGATCGGCCATATGTAATATATCATAATCCTTCTGCCAAAGTGGATTATCTGTCATCTGAAAAATTTACGAACGAAATCATCAACTCTATCCGAGATAATAAAGCCGTCGATTTCCGCAATCGCTATCGAAATGTTGATGTGCTTTTGATAGATGATATTCAATTTTTACGGGGGAAAGAACAATCCCAGGAATAATTTTCCATACATTTAACACATGACACGAAGAAAGCAAACAAATCGGCATGTCGAGTGACCGGCCGCCAAACGAAATTCCGACACTTGAATACAGATTGCGCTCACGTTTTGAATGCGACTTATTACAGATATCACACCGCCTAGATCTAGAAAACGATAAATTTGCAATTTTGTACAAAAAAAGGGCCTAA TATACCCCCAAAACAGCCAGCTTTCAGTGGTGGTGGTGGTGGTGCTCGAGGAAACGGTCTTTCTTGGAGCGCAGCCACGCGAAGCGCTCCTCCGGCTGCTGCAACAGGGTTTCCTCATTGATACATTGATCAGATCGATATCTTCCGTACGCAGAAAGACGTTAATTTGGCGTTCGTTCTTCAGGATGACCGGCAGGGTAATTTGATTCTTCGCACGCAGCTCTTTGACTTTGCGATAGTAGTCATTGATGCTCAGGTCGTGCGGCAGAATGCTCAGGGCGAACTTCATGTTGGACAGCGTATACTCATGCGCGCAGCAGACCAGGGGGTCGTCCGGCAGCGCGCTCAGTTTCTTCAGAGATTGGTACTTTGGCTTGCGGGGCCTTCAAACAGACGGCCGCAACCCCCGGAAAACAGGGTGGCACCCCAAACAGGGTACGGTTTGGTAAAGTAACAATGGGGCCCAGGGTATGACCCGGGGTTGCAATAACGGAAAATTCGGGACCCAGAACGAAGGGGTTTCCCCGTTTTGACACCCGGGGGGGACCCTTTTTTGAGTTTTTTGGGACCTTCCCGACGATTTGGGGAAATTTTTCAACCATTCCTTGACCCCCCCAAATGATCCGGGGGGGAAGGGGCAAAAAAATTTTTTCCCGGTTCCCATTTTTCCCGGGGAATTGGGTTAAACCCGGTTCCGGGGCCCCCGGGGAAAAAATTGGGAAAACCCCTGTTTTTCTTCTAAAAACCCCCATTTTTTTTGTTTTTAAACCCCGGAGGTGTTTTTTTAAACCCGGGGTGGTTTCCCGGGGAAAAAATTTTTCCCCCCTAAAATTCCCCCCTTTTTTGGGCCCGGGGTTTTTTTTTTCCCCCCCCTTTTTTTTAAAAATTTTTCCCCAAACCCCTTTTTTTTTTGGGGGGACAAAAAATTTTTTTTCCAAAAAAAAGGAAGGGCCCTTTTGGGCCCCCACATTTTTTTGGGGGGTTATTTCCCCCCCCGGGCACCCCCCCACCCACCCCCCTCCGTGGTGCGTTGCTCCCCCCAAAAAAAATTTTTTGGCCCCCCCCCCCGCCGGGGAAAAAAATTTTTTTGACCACCCCCCCAAATAAATTTTTTTTTTTATATAAAAAATTAGTGTGGTGTTCCCCCCCCCCCCCCCCCCCTTTTTTTCCTAAAAAAAAAAAACCTTGGGGCCCCCCCCCTTTTTTTTTTTATAATTCTTTCCCCCCCCCCCCAGGGGGAAAACAAGGTGGTTTTGGGGGTTTTTAAGGGCCGGGGGGGGGGGGTTTTTTTCTTTTTTTTTTTGAAACCCCCGGGGGGGCA DNA sequencing results for pET3a-gloA with primer T7 (Table 63). Table 63. DNA sequencing results for pET3a-gloA. DNA sequence CCCGGGCTTTTCATCATGAATAATTTTGTTTACTTTAAGAAGGAGATATACATATGCGTCTGTTGCATACTATGTTGCGTGTTGGCGATCTGCAACGTAGCATCGACTTTTATCCAAGGTGCTGGGCATGAAGCTGCTGCGCACGAGCGAAAATCCGGAATACAAGTATAGCCTGGCATTTGTGGGCTACGGTCCGGAAACCGAGGAGGCAGTTATCGAGCTGACCTATACTGGGGTGTTGACAAATACGAGCTGGGTACGGCGTACGGTCATATTGCTTTGAGCGTGGACAACGCGGCGGAAGCGTGTGAGAAAATCCGTCAGAACGGCGGTAACGTGACTCGTGAGGCAGGCCCGGTGAAAGGCGGTACCACCGTCATCGCGTTCGTCGAAGATCGGATGGCTACAAGATTGAACTGATGAGGAGAAAGACGCTGGTCGTGGTCTGGGTAACCTCGAGCACCACCACCACCACTGAACTAGTTTTGGATCCGGCTGCTAACAAAGCCCGAAAGGAAGCTGAGTTGGCTGCTGCCACCGCTGAGCAATAACTAGCATAACCCCTTGGGGCCTCTAAACGGGTCTTGAGGGGTTTTTTGCTGAAAGGAGGAACTATATCCGGATATCCACAGGACGGGTGTGGTCGCCATGATCGCGTAGTCGATAGTGGCTCCAAGTAGCGAAGCGAGCAGGACTGGGCGGCGGCCAAAGCGGTCGGACAGTGCTCGAGAACGGGTGCGCATAGAAATTGCATCAACGCATATAGCGCTAGCAGCACGCCATAGTGACTGGCGATGCTGTCGTAATGGACGATATCCCGCAAGAGGCCCGGTAGTACCGGCATAACCAAGCCTATGCCTACAGCATCCAGTTGACGATGCCGAGGATGACGATGAGCGTATTGTTTATATTTCATACTTGGTGGCCTGATTGTGTTTATTAATTCAACTGTGATTATATCACCGTAATTAAAGTTCATCGTTGTTTAATTCGTCTAAACTTGTCTATTCTTTGAAGAACTAAAGCTCCCTCTTAATTCCGCCTATTTTAAATATGATTTCTGTATTGCGATATAACGGTCTCCTTTCCTCTCCTTCTC 323 Chapter 3: Copper addition bar graphs AlkF3 was an alkyne hit for GlxII. This triazole (56), which was derived from AlkF3, showed the ‘best’ binding and inhibition data with GlxII. This alkyne was still observed to be a hit under the standard kTGS conditions for GlxII and not GlxI (Fig. 154). 56 did show binding affinity with GlxI, with a thermal shift of 3.48 °C, but no inhibition. Upon the addition of copper the % triazole formation increased under all conditions. The effect under protein conditions was the greatest and in all cases the % triazole in the presence of protein was at least double to the controls resulting in the compound being a hit for GlxII. With 0.2, 0.5, 1 and 2 mM copper addition the alkyne was also a hit for GlxI. With 0.5 mM being the optimal copper concentration for GlxII and 2 mM for GlxI. Fig. 154. Bar graphs of % triazole formation between AlkF3 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. AlkC1 was an alkyne hit for GlxI. This triazole (74), derived from AlkC1, had the highest binding affinity for GlxI. This alkyne was still observed to be a hit under the standard kTGS conditions for GlxI but also for GlxII (Fig. 155). 74 was shown not to have binding affinity with GlxII, due to a thermal shift of -2.40 °C, but did show inhibition under both enzyme kinetics assay. Upon the addition of copper the % triazole formation increased under control conditions but not protein conditions. With 0.5 mM being the optimal copper concentration for GlxII and 0 mM for GlxI. GlxIGlxIIBSABuffer0.000.050.100.150.20AlkF3 AlkF3 AlkF3 AlkF3 AlkF3 AlkF3GlxI 0.01 0.02 0.02 0.04 0.07 0.04GlxII 0.03 0.03 0.16 0.06 0.04 0.11BSA 0.01 0.00 0.01 0.02 0.01 0.01Buffer 0.00 0.00 0.00 0.00 0.01 0.03% Triazole formation324 Fig. 155. Bar graphs of % triazole formation between AlkC1 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. AlkB4 was not an alkyne hit for either GlxI or GlxII. This alkyne was observed to be a hit under the standard kTGS conditions in this experiment, confirming the effect a change in chemical environment. Upon the addition of copper the % triazole formation increased under all conditions (Fig. 156). The effect under protein conditions was the greatest. As the copper concentration increased from 0 to 0.2 mM the % triazole formation increased in all condition with the protein at least double to the controls resulting in the compound being a hit. As the copper concentration increased to 0.5 and 1 mM, the alkyne was not considered a hit and 2 and 4 mM copper concentration resulted in the alkyne being a hit for GlxII but not GlxI. Fig. 156. Bar graphs of % triazole formation between AlkB4 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Alk5 was not an alkyne hit for either GlxI or GlxII. This alkyne was observed to be a hit under the standard kTGS conditions for GlxI, in this experiment. Upon addition of copper the % triazole formation increased 3 times under GlxI and buffer conditions but had little GlxIGlxIIBSABu…0.001.002.00AlkC1 AlkC1 AlkC1 AlkC1 AlkC1 AlkC1GlxI 0.03 0.04 0.03 0.04 0.04 0.04GlxII 1.61 1.27 1.43 0.60 0.96 1.30BSA 0.02 0.03 0.01 0.03 0.05 0.04Buffer 0.01 0.02 0.01 0.02 0.04 0.07% Triazole formationGlxIGlxIIBSABuffer0.000.100.200.30AlkB4 AlkB4 AlkB4 AlkB4 AlkB4 AlkB4GlxI 0.03 0.12 0.10 0.10 0.09 0.08GlxII 0.05 0.14 0.26 0.13 0.17 0.29BSA 0.00 0.05 0.10 0.08 0.08 0.07Buffer 0.00 0.05 0.05 0.05 0.06 0.17% Triazole formation325 to no effect on BSA and GlxII (Fig. 157). This alkyne was considered a hit, where % triazole formation was at least double under protein conditions compared to control conditions, at 0, 0.2, 0.5 and 1 mM copper concentration but not 2 and 4 mM. Fig. 157. Bar graphs of % triazole formation between Alk5 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. AlkD2-2 was a hit for GlxII. Under standard kTGS conditions this alkyne was still a hit for GlxII. The triazole (67), derived from AlkD2-2, destabilised GlxII with a negative thermal shift but did show inhibition of GlxII activity. Upon the addition of copper the % triazole formation decreased under protein conditions and remained the same for control conditions for 0, 0.2, 0.5 and 1 mM and increased at 2 and 4 mM, resulting in the alkyne only being considered a hit at 0 and 0.5 mM. With 0.5 mM being the optimal copper concentration for GlxII and 0 mM for GlxI (Fig. 158). Fig. 158. Bar graphs of % triazole formation between AlkD2-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. GlxIGlxIIBSABu…0.000.200.40Alk5 Alk5 Alk5 Alk5 Alk5 Alk5GlxI 0.06 0.20 0.18 0.21 0.17 0.17GlxII 0.00 0.00 0.00 0.00 0.00 0.00BSA 0.02 0.03 0.09 0.07 0.09 0.11Buffer 0.02 0.07 0.07 0.06 0.06 0.19% Triazole formationGlxIGlxIIBSABuffer0.000.050.100.150.200.25AlkD2-2 AlkD2-2 AlkD2-2 AlkD2-2 AlkD2-2 AlkD2-2GlxI 0.10 0.07 0.08 0.08 0.14 0.08GlxII 0.22 0.07 0.14 0.11 0.16 0.14BSA 0.06 0.05 0.03 0.07 0.22 0.17Buffer 0.05 0.03 0.04 0.05 0.07 0.19% Triazole formation326 AlkC1-2 was a hit for GlxII. The triazole (66), derived from AlkC1-2, showed binding affinity and inhibition of GlxII activity. Under standard kTGS conditions this alkyne was not a hit for GlxII or GlxI, which could be explained by the change in chemical environment in this experiment. Upon addition of copper, no effect was observed at 0.2 mM. At 0.5 and 1 mM the % triazole formation increased under protein conditions resulting in the alkyne being a bit for GlxI and GlxII. At 2 and 4 mM the control conditions also showed an increase in % triazole formation resulting in this compound not being a hit. With 1 mM being the optimal copper concentration for GlxII and 0.5 - 2 mM for GlxI (Fig. 159). Fig. 159. Bar graphs of % triazole formation between AlkC1-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. AlkB1-2 was not a hit for GlxI or GlxII during the initial kTGS studies. Under standard kTGS conditions this alkyne was a hit for GlxII and GlxI, which could be explained by the change in chemical environment in this experiment. Upon addition of copper this alkyne remained a hit for GlxI and II as there was no change in % triazole formation at 0.2, 0.5 and 1 mM. The % triazole formation increased in the controls at 2 and 4 mM resulting in the compound not being considered a hit. 0 mM copper concentration was optimal for GlxI and GlxII for this alkyne to be considered a hit (Fig. 160). GlxIGlxIIBSABuffer0.000.501.001.502.00AlkC1-2 AlkC1-2 AlkC1-2 AlkC1-2 AlkC1-2 AlkC1-2GlxI 0.14 0.13 0.41 0.59 0.74 0.39GlxII 0.15 0.12 0.38 1.62 0.38 0.32BSA 0.12 0.12 0.12 0.10 0.24 0.25Buffer 0.11 0.08 0.09 0.07 0.07 0.36% Triazole formation327 Fig. 160. Bar graphs of % triazole formation between AlkB1-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. AlkA5-2 was not a hit for GlxI or GlxII during the initial kTGS studies. Under standard kTGS conditions this alkyne was a hit for GlxII and GlxI, due to a change in the chemical environment in this experiment. Upon addition of copper the % triazole formation remained similar at 0.2, 0.5 and 1 mM copper concentration. At 2 and 4 mM the % triazole formation increased under all condition resulting in the compound no longer being considered a hit. With 0.5 mM being the optimal copper concentration for GlxII and GlxI (Fig. 161). Fig. 161. Bar graphs of % 1,4-triazole formation between AlkA5-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. The LC-MS/MS showed the potential 1,5-triazole which under GlxI and control conditions appeared to increase its % triazole formation and decrease under GlxII conditions upon increased copper concentration (Fig. 162). GlxIGlxIIBSABuffer0.000.100.200.300.40AlkB1-2 AlkB1-2 AlkB1-2 AlkB1-2 AlkB1-2 AlkB1-2GlxI 0.21 0.12 0.12 0.12 0.18 0.21GlxII 0.25 0.19 0.22 0.17 0.23 0.35BSA 0.08 0.06 0.07 0.08 0.16 0.20Buffer 0.09 0.06 0.07 0.08 0.06 0.18%Triazole formationGlxIGlxIIBSABuffer0.000.020.040.060.080.100.120.14AlkA5-2 AlkA5-2 AlkA5-2 AlkA5-2 AlkA5-2 AlkA5-2GlxI 0.06 0.04 0.05 0.04 0.09 0.07GlxII 0.06 0.03 0.07 0.06 0.08 0.13BSA 0.03 0.02 0.02 0.02 0.10 0.11Buffer 0.03 0.02 0.02 0.02 0.03 0.13% Triazole formation328 Fig. 162. Bar graphs of % 1,5-triazole formation between AlkA5-2 and GAz with increasing amounts of copper: 0, 0.2, 0.5, 1, 2 and 4 mM. Chapter 4: Pre-screening of fragment libraries using DSF Table 64. Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Alkyne library 1. (ΔTm). NR = no result, N/A = not tested. Fragment concentration 1 mM 1 mM + 0.01% TWEEN20 5 mM 5 mM + 0.01% TWEEN20 ID Structure Tm ΔTm Tm ΔTm Tm ΔTm Tm ΔTm GlxII 65.08 0 69.42 0 66.36 0 68.83 0 GSH 67.08 2.01 70.15 1.36 52.6 -14.9 72.19 3.37 AlkA1 65.82 0.66 68.92 -0.5 69.35 2.99 NR NR AlkB1 64.36 -0.72 67.99 -1.43 61.06 -5.31 NR NR AlkC1 65.4 0.32 70.45 1.02 40.93 -25.4 NR NR AlkD1 65.34 0.26 69.59 0.17 48.35 -18 NR NR AlkE1 65.39 0.31 69.24 -0.19 85.79 19.43 NR NR AlkF1 65.51 0.43 69.7 0.27 66.86 0.5 NR NR AlkG1 65.36 0.28 69.05 -0.37 69.48 3.11 NR NR AlkH1 63.78 -1.3 69.25 -0.17 66.29 -0.07 NR NR AlkA2 66.09 1.01 70 0.58 67.97 1.61 NR NR AlkB2 65.54 0.46 69.67 0.24 66.63 0.27 NR NR AlkC2 64.98 -0.1 70.15 0.73 62.18 -4.18 NR NR GlxIGlxIIBSABuffer0.000.20AlkA5-2 1.5AlkA5-2 1.5AlkA5-2 1.5AlkA5-2 1.5AlkA5-2 1.5AlkA5-2 1.5GlxI 0.00 0.01 0.01 0.00 0.02 0.03GlxII 0.17 0.05 0.08 0.09 0.05 0.01BSA 0.00 0.00 0.00 0.01 0.03 0.07Buffer 0.00 0.00 0.00 0.01 0.01 0.07% Triazole formation329 Fragment concentration 1 mM 1 mM + 0.01% TWEEN20 5 mM 5 mM + 0.01% TWEEN20 ID Structure Tm ΔTm Tm ΔTm Tm ΔTm Tm ΔTm AlkD2 65.98 0.9 69.84 0.42 NR NR NR NR AlkF2 66.35 1.27 69.41 -0.01 41.91 -24.5 NR NR AlkG2 66.39 1.31 68.38 -1.05 63.1 -3.27 NR NR AlkH2 65.62 0.54 70.9 1.48 69.11 2.74 NR NR AlkA3 64.91 -0.17 69.3 -0.12 63.9 -2.47 NR NR AlkB3 65.69 0.61 69.47 0.05 64.85 -1.52 NR NR AlkC3 64.25 -0.83 68.52 -0.91 NR NR NR NR AlkD3 65.22 0.14 70.3 0.88 66.82 0.46 NR NR AlkE3 66.27 1.19 70.62 1.2 62.87 -3.5 59.53 -9.3 AlkF3 66.44 1.36 72 2.58 35.31 -31.1 NR NR AlkG3 66.34 1.26 69.97 0.55 41.98 -24.4 NR NR AlkH3 65.58 0.5 69.83 0.4 67.29 0.93 NR NR AlkA4 66.6 1.52 69.4 -0.02 67.88 1.52 NR NR AlkB4 66.48 1.4 69.78 0.36 68.45 2.08 NR NR AlkC4 65.39 0.31 70.23 0.8 67.84 1.47 NR NR AlkD4 64.17 -0.91 70.22 0.8 56.04 -10.3 NR NR AlkE4 65.92 0.84 69.3 -0.13 66.78 0.42 NR NR AlkF4 40.89 -24.2 70.35 0.93 41.23 -25.1 NR NR AlkG4 63.73 -1.35 73.41 3.99 40.24 -26.1 NR NR AlkH4 65.56 0.48 70.18 0.76 59.73 -6.63 NR NR AlkA5 64.32 -0.76 68.5 -0.92 65.45 -0.91 NR NR AlkB5 67.09 2.01 69.43 0.01 67.09 0.73 NR NR AlkC5 65.13 0.05 68.95 -0.48 63.5 -2.86 66.58 -2.3 AlkD5 68.76 3.68 69.06 -0.36 67.65 1.29 NR NR AlkE5 67.07 1.99 69.29 -0.13 66.28 -0.08 68.04 -0.8 AlkF5 68.07 2.99 69.32 -0.1 66.2 -0.17 NR NR Alk-1 65.96 -0.42 70. -0.87 NR NR NR NR Alk-2 66.03 -0.35 65.79 -1.72 NR NR NR NR Alk-3 65.18 -1.21 45.68 -21.8 NR NR NR NR 330 Fragment concentration 1 mM 1 mM + 0.01% TWEEN20 5 mM 5 mM + 0.01% TWEEN20 ID Structure Tm ΔTm Tm ΔTm Tm ΔTm Tm ΔTm Alk-4 66.3 -0.08 56.5 -11 NR NR NR NR Alk-5 66.35 -0.03 69.07 0.27 61.88 -5.63 NR NR Alk-6 66.17 -0.22 70.14 1.34 66.12 -1.39 NR NR Alk-7 66.2 -0.19 69.19 0.39 66.74 -0.77 NR NR Alk-8 66.06 -0.32 69.16 0.37 66.34 -1.17 NR NR 1% DMSO 66.34 -0.04 69.01 -0.41 N/A N/A N/A N/A 5% DMSO N/A N/A N/A N/A 66.86 0.5 62.3 -6.52 CM-A-12 66.78 1.7 N/A N/A N/A N/A N/A N/A Fig. 163. Bar graphs of thermal melt temperature (Tm, °C), thermal shift (ΔTm) with respect to GlxII with Alkyne library 1 at 1mM, with AlkF4 data removed. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Fig. 164. Bar graphs of thermal melt temperature (Tm, °C) and thermal shift (ΔTm) with respect to GlxII with Alkyne library 1 at 5mM. Significance of changes evaluated by a Student’s t-test (where *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). 331 Fig. 165. Bar graphs of thermal melt temperature (Tm, °C) and thermal shift (ΔTm) with respect to GlxII with Alkyne library 1 at 1mM + Tween20. Significance of changes evaluated by a Student’s t-test (where * p ≤ 0.05, ● n = 1, ●● = NR). Table 65: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Azide library 1. ID Structure Tm ΔTm ID Structure Tm ΔTm GlxII 65.27 0.00 AZH03 64.99 -0.28 DMSO 65.34 -0.05 AZA04 68.59 3.32 GSH 63.02 -2.25 AZB04 64.71 -0.56 CM-A-12 65.10 -0.30 AZC04 65.48 0.08 AZA01 66.14 0.87 AZD04 64.80 -0.59 AZB01 66.59 1.32 AZ-1 65.92 0.52 AZC01 66.00 0.73 AZ-2 66.91 1.51 AZD01 65.15 -0.12 AZ-3 66.27 0.87 AZE01 65.52 0.25 AZ-4 64.69 -0.71 AZF01 66.32 1.05 AZ-5 65.98 0.58 AZG01 68.04 2.77 AZ-6 64.79 -0.61 332 ID Structure Tm ΔTm ID Structure Tm ΔTm AZH01 64.92 -0.35 AZ-7 64.85 -0.55 AZA02 67.98 2.71 AZ-8 65.80 0.41 AZB02 65.50 0.23 AZ-9 67.09 1.70 AZC02 66.61 1.34 AZ-10 64.86 -0.54 AZD02 64.78 -0.49 AZ-11 66.64 1.24 AZE02 67.20 1.93 AZ-12 67.25 1.86 AZF02 64.56 -0.71 AZ-13 67.74 2.34 AZG02 72.23 6.96 AZ-14 66.38 0.99 AZH02 64.77 -0.50 AZ-15 65.53 0.13 AZA03 69.72 4.45 AZ-16 67.02 1.62 AZB03 63.33 -1.94 AZ-17 65.46 0.06 AZC03 71.68 6.41 AZ-18 65.59 0.20 AZD03 65.71 0.44 AZ-19 68.13 2.74 AZE03 65.89 0.62 AZ-20 66.92 1.53 AZF03 66.44 1.17 AZ-21 65.27 -0.13 AZG03 64.64 -0.63 AZ-22 65.95 0.55 333 Fig. 166. Bar graphs of thermal shift (ΔTm, °C) of Azide library 1 in the presence of GlxII. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 66: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Alkyne library 2. ID Structure Tm ΔTm ID Structure Tm ΔTm GlxII 68.92 0 GlxII 68.82 0 DMSO 67.97 0 DMSO 67.72 0 GSH 67.26 -0.72 GAz 67.56 -0.41 GAz 67.19 -0.53 AlkA1-2 68.64 0.67 AlkA4-2 68.15 0.43 AlkA2-2 68.87 0.9 AlkA5-2 68.41 0.69 AlkA3-2 68.15 0.17 AlkA6-2 68.39 0.67 AlkB1-2 68.16 0.19 AlkA7-2 67.17 -0.55 AlkB2-2 67.97 0 AlkB4-2 68.38 0.66 AlkB3-2 68.81 0.84 AlkB5-2 67.7 -0.02 AlkC1-2 68.27 0.29 AlkB6-2 68.48 0.76 AlkC2-2 69.24 1.27 AlkB7-2 67.69 -0.03 AlkC3-2 68.35 0.38 AlkC4-2 68.07 0.35 334 ID Structure Tm ΔTm ID Structure Tm ΔTm AlkD1-2 66.25 -1.73 AlkC5-2 67.56 -0.16 AlkD2-2 68.61 0.64 AlkC6-2 68.56 0.84 AlkD3-2 61.29 -6.68 AlkC7-2 67.2 -0.52 AlkE1-2 68.26 0.29 AlkD4-2 66.61 -1.11 AlkE2-2 68.1 0.13 AlkD5-2 67.88 0.16 AlkE3-2 68.31 0.34 AlkD6-2 69.15 1.43 AlkF1-2 68.21 0.24 AlkD7-2 67.63 -0.09 AlkF2-2 68.75 0.78 AlkE4-2 68.05 0.33 AlkF3-2 67.39 -0.59 AlkE5-2 67.49 -0.23 AlkG1-2 68.12 0.15 AlkE6-2 67.44 -0.28 AlkG2-2 66.04 -1.93 AlkF4-2 68.16 0.44 AlkG3-2 68.15 0.18 AlkF5-2 68.26 0.54 AlkH1-2 68.18 0.21 AlkF6-2 67.93 0.21 AlkH2-2 68.42 1.16 AlkG4-2 67.48 -0.24 AlkH3-2 68.37 0.4 AlkG5-2 67.66 -0.06 AlkG6-2 67.03 -0.69 AlkH4-2 68.01 0.29 AlkH5-2 67.93 0.21 AlkH6-2 68.18 0.46 335 Fig. 167. Bar graphs of thermal shift (ΔTm, °C) of Alkyne library 2 in the presence of GlxII. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 67: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxII for Azide library 2. ID Structure Tm ΔTm ID Structure Tm ΔTm GlxII 68 0 GlxII 67.45 0 DMSO 66.55 0 DMSO 65.63 0 GSH 65.28 -1.27 GAz 66.8 0.26 GAz 65.69 0.06 AzA1-2 65.93 -0.62 AzA4-2 65.26 -0.37 AzA2-2 67.87 1.33 AzA5-2 65.9 0.28 AzA3-2 67.33 0.78 AzA6-2 66.06 0.43 AzB1-2 67.33 0.78 AzA7-2 66.01 0.39 AzB2-2 66.61 0.06 AzB4-2 66.13 0.5 336 ID Structure Tm ΔTm ID Structure Tm ΔTm AzB3-2 66.84 0.3 AzB5-2 65.91 0.28 AzC1-2 67.23 0.68 AzB6-2 66.67 1.05 AzC2-2 67 0.45 AzB7-2 64.97 -0.66 AzC3-2 66.91 0.36 AzC4-2 66.24 0.62 AzD1-2 65.91 -0.63 AzC5-2 65.64 0.02 AzD2-2 66.98 0.43 AzC6-2 64.85 -0.78 AzD3-2 67 0.46 AzC7-2 64.51 -1.12 AzE1-2 66.91 0.36 AzD4-2 65.81 0.18 AzE2-2 67.7 1.16 AzD5-2 64.88 -0.74 AzE3-2 67.08 0.54 AzD6-2 66 0.37 AzF1-2 67.07 0.53 AzD7-2 65.78 0.15 AzF2-2 67.02 0.48 AzE4-2 66.43 0.81 AzF3-2 67.04 0.5 AzE5-2 66.31 0.69 AzG1-2 66.32 -0.23 AzE6-2 65.61 -0.02 AzG2-2 66 -0.55 AzF5-2 66.9 1.28 AzG3-2 67.44 0.89 AzF6-2 66.32 0.69 AzH1-2 67.31 0.77 AzG4-2 66.56 0.93 337 ID Structure Tm ΔTm ID Structure Tm ΔTm AzH2-2 66.87 1.59 AzG5-2 65.82 0.19 AzH3-2 66.8 0.26 AzG6-2 65.44 -0.19 AzH4-2 66.4 0.78 AzH5-2 66.96 1.33 AzH6-2 68.01 2.38 Fig. 168. Bar graphs of thermal shift (ΔTm, °C) of Azide library 2 in the presence of GlxII. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 68: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Alkyne library 1. ID Structure Tm ΔTm ID Structure Tm ΔTm GlxI 62.58 0.00 GlxI 63.20 0.00 1% DMSO 62.48 0.00 1% DMSO 62.34 0.00 GSH 63.40 0.93 GSH 63.26 0.92 338 ID Structure Tm ΔTm ID Structure Tm ΔTm GAz 63.43 0.96 GAz 64.95 2.61 AlkA1 62.49 0.02 AlkA4 62.40 0.06 AlkA2 63.61 1.13 AlkA5 62.36 0.02 AlkA3 62.31 -0.17 AlkB4 62.62 0.28 AlkB1 62.27 -0.20 AlkB5 62.71 0.37 AlkB2 63.02 0.55 AlkC4 62.71 0.37 AlkB3 62.72 0.24 AlkC5 62.59 0.25 AlkC1 63.63 1.15 AlkD4 63.89 1.55 AlkC2 63.01 0.54 AlkD5 62.91 0.57 AlkC3 62.66 0.18 AlkE4 62.92 0.58 AlkD1 62.15 -0.33 AlkE5 62.60 0.26 AlkD2 62.52 0.04 AlkF4 60.70 -1.64 AlkD3 62.65 0.18 AlkF5 63.04 0.70 AlkE1 62.89 0.41 AlkG4 62.44 0.10 339 ID Structure Tm ΔTm ID Structure Tm ΔTm AlkE2 63.64 1.16 AlkH4 62.99 0.65 AlkE3 60.17 -2.31 Alk-1 62.42 0.08 AlkF1 62.96 0.49 Alk-2 62.59 0.25 AlkF2 63.44 0.97 Alk-3 56.09 -6.25 AlkF3 62.44 -0.04 Alk-4 60.34 -2.00 AlkG1 63.93 1.45 Alk-5 62.82 0.48 AlkG2 59.74 -2.74 Alk-6 62.77 0.43 AlkG3 62.21 -0.27 Alk-7 62.49 0.15 AlkH1 62.06 -0.42 Alk-8 62.33 -0.01 AlkH2 63.06 0.58 AlkH3 63.11 0.63 340 Fig. 169. Bar graphs of thermal shift (ΔTm, °C) of Alkyne library 1 in the presence of GlxI. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 69: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Alkyne library 2. ID Structure Tm ΔTm ID Structure Tm ΔTm GlxI 62.05 0 GlxI 62.15 0 DMSO 62.03 0 DMSO 61.72 0 GSH 62.66 0.63 GAz 64.23 2.2 GAz 63.22 1.5 AlkA1-2 62.12 0.09 AlkA4-2 62.32 0.6 AlkA2-2 62.99 0.96 AlkA5-2 61.7 -0.02 AlkA3-2 61.83 -0.2 AlkA6-2 61.45 -0.27 AlkB1-2 62.01 -0.02 AlkA7-2 60.58 -1.14 AlkB2-2 63.74 1.71 AlkB4-2 61.76 0.04 AlkB3-2 61.96 -0.07 AlkB5-2 61.77 0.05 341 ID Structure Tm ΔTm ID Structure Tm ΔTm AlkC1-2 62.15 0.12 AlkB6-2 61.87 0.15 AlkC2-2 61.36 -0.67 AlkB7-2 61.86 0.14 AlkC3-2 62.93 0.9 AlkC4-2 61.77 0.05 AlkD1-2 62.1 0.07 AlkC5-2 61.58 -0.14 AlkD2-2 61.96 -0.07 AlkC6-2 62.41 0.69 AlkD3-2 50.63 -11.4 AlkC7-2 61.44 -0.28 AlkE1-2 62.45 0.42 AlkD4-2 61.97 0.25 AlkE2-2 62.74 0.71 AlkD5-2 60.97 -0.75 AlkE3-2 62.12 0.09 AlkD6-2 62.15 0.43 AlkF1-2 62.59 0.56 AlkD7-2 62.22 0.5 AlkF2-2 59.71 -2.32 AlkE4-2 62.58 0.86 AlkF3-2 62.49 0.46 AlkE5-2 62.56 0.84 AlkG1-2 62.49 0.46 AlkE6-2 62.62 0.9 AlkG2-2 61.62 -0.41 AlkF4-2 62.24 0.52 AlkG3-2 62.94 0.91 AlkF5-2 62.81 1.09 AlkH1-2 62.21 0.18 AlkF6-2 63.11 1.39 AlkH2-2 62.48 -0.18 AlkG4-2 61.15 -0.57 AlkH3-2 62.46 0.43 AlkG5-2 62.5 0.78 342 ID Structure Tm ΔTm ID Structure Tm ΔTm AlkG6-2 62.31 0.59 AlkH4-2 63.14 1.42 AlkH5-2 62.27 0.55 AlkH6-2 62.73 1.01 Fig. 170. Bar graphs of thermal shift (ΔTm, °C) of Alkyne library 2 in the presence of GlxI. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 70: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Azide library 1. ID Structure Tm ΔTm ID Structure Tm ΔTm GlxI 60.57 0 GlxI 61.19 0 DMSO 60.8 0 DMSO 61.4 0 GSH 61.72 0.92 GSH 61.95 0.54 GAz 62.39 1.59 GAz 63.83 2.43 AzA1 60.82 0.02 AzA4 60.79 -0.61 343 ID Structure Tm ΔTm ID Structure Tm ΔTm AzA2 60.8 0 Az1 60.82 -0.58 AzA3 60.83 0.03 Az2 60.64 -0.77 AzB1 65.67 4.87 Az3 60.87 -0.53 AzB2 60.38 -0.42 AzB4 61.07 -0.33 AzB3 61.19 0.39 Az4 61.01 -0.39 AzC1 61.4 0.6 Az5 61.07 -0.34 AzC2 61.01 0.21 Az6 60.86 -0.54 AzC3 61.76 0.96 AzC4 61.11 -0.29 AzD1 61.33 0.53 Az7 61.22 -0.18 AzD2 61.37 0.57 Az8 61.32 -0.08 AzD3 61.73 0.93 Az9 60.6 -0.8 AzE1 60.97 0.17 AzD4 61.22 -0.19 AzE2 61.71 0.91 Az10 61.4 -0.01 AzE3 61.05 0.25 Az11 59.52 -1.88 344 ID Structure Tm ΔTm ID Structure Tm ΔTm AzF1 60.8 0 Az12 61.44 0.03 AzF2 61.26 0.46 Az13 61.28 -0.12 AzF3 60.85 0.05 Az14 61.17 -0.24 AzG1 61.32 0.52 Az15 60.97 -0.44 AzG2 61.76 0.96 Az16 60.85 -0.55 AzG3 60.91 0.11 Az17 61.16 -0.24 AzH1 60.91 0.11 Az18 60.82 -0.58 AzH2 60.79 -0.93 Az19 61.03 -0.37 AzH3 60.67 -0.13 Az20 60.93 -0.51 Az21 60.82 -0.58 Az22 61.06 -0.34 345 Fig. 171. Bar graphs of thermal shift (ΔTm, °C) of Azide library 1 in the presence of GlxI. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). Table 71: Thermal melt temperature (Tm, °C) and thermal shift with respect to GlxI for Azide library 2. ID Structure Tm ΔTm ID Structure Tm ΔTm GlxI 62.08 0 GlxI 62.12 0 DMSO 62.61 0 DMSO 62.32 0 GSH 62.77 0.16 GAz 64.88 2.27 GAz 64.16 1.84 AzA1-2 62.61 0 AzA4-2 62.01 -0.32 AzA2-2 63.13 0.52 AzA5-2 62.07 -0.26 AzA3-2 62.77 0.16 AzA6-2 62.1 -0.22 AzB1-2 62.6 -0.01 AzA7-2 61.82 -0.5 AzB2-2 62.87 0.26 AzB4-2 62.07 -0.25 346 ID Structure Tm ΔTm ID Structure Tm ΔTm AzB3-2 62.65 0.04 AzB5-2 62.07 -0.25 AzC1-2 62.88 0.27 AzB6-2 61.69 -0.63 AzC2-2 62.87 0.26 AzB7-2 61.9 -0.42 AzC3-2 63.46 0.85 AzC4-2 61.94 -0.38 AzD1-2 62.4 -0.21 AzC5-2 62.26 -0.06 AzD2-2 62.71 0.1 AzC6-2 62.44 0.12 AzD3-2 62.41 -0.2 AzC7-2 61.72 -0.6 AzE1-2 62.45 -0.16 AzD4-2 62.39 0.06 AzE2-2 62.64 0.03 AzD5-2 61.19 -1.13 AzE3-2 62.32 -0.29 AzD6-2 63.72 1.39 AzF1-2 62.46 -0.15 AzD7-2 62.04 -0.28 AzF2-2 62.42 -0.19 AzE4-2 62.07 -0.25 AzF3-2 62.08 -0.53 AzE5-2 62.51 0.18 AzG1-2 61.64 -0.97 AzE6-2 62.44 0.12 AzG2-2 62.34 -0.27 AzF5-2 62.52 0.2 AzG3-2 62.36 -0.25 AzF6-2 61.88 -0.45 AzH1-2 62.6 -0.01 AzG4-2 62.18 -0.14 347 ID Structure Tm ΔTm ID Structure Tm ΔTm AzH2-2 63.35 0.58 AzG5-2 62.24 -0.08 AzH3-2 62.04 -0.57 AzG6-2 62.04 -0.29 AzH4-2 62.15 -0.18 AzH5-2 62.17 -0.15 AzH6-2 62.29 -0.04 Fig. 172. Bar graphs of thermal shift (ΔTm, °C) of Azide library 2 in the presence of GlxI. Significance of changes evaluated by a Student’s t-test (where **** p ≤ 0.0001, *** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05). 348 Chapter 5: DNA sequencing for GST The DNA sequence for pGEX-2T with primers pGEX forward and pGEX reverse (Table 72). Table 72. DNA sequence of pGEX. pGEX forward sequence pGEX reverse sequence TGGGGTTGGTGGCGACCATCCTCCAAAATCGGATCTGGTTCCGCGTGGATCCCCGGGAATTCATCGTGACTGACTGACGATCTGCCTCGCGCGTTTCGGTGATGACGGTGAAAACCTCTGACACATGCAGCTCCCGGAGACGGTCACAGCTTGTCTGTAAGCGGATGCCGGGAGCAGACAAGCCCGTCAGGGCGCGTCAGCGGGTGTTGGCGGGTGTCGGGGCGCAGCCATGACCCAGTCACGTAGCGATGCGGAGTGTATAATTCTTGAAGACGAAAGGGCCTCGTGATACGCCTATTTATAGGTTAATGTCATGATAATAATGGTTTCTTAGACGTCAGGTGGCACTTTCGGGGAAATGTGCGCGGAACCCCTATTTGTTTATTTTTCTAAATACATTCAAATATGTATCCGCTCATGAGACAATAACCCTGATAAATGCTTCAATAATATTGAAAAAGGAAGAGTATGAGTATTCAACATTTCCGTGTCGCCCTTATTCCCTTTTTTGCGGCATTTTGCCTTCCTGTTTTTGCTCACCCAGAAACGCTGGTGAAAGTAAAAGATGCTGAAGATCAGTTGGGTGCACGAGTGGGTTACATCGAACTGGATCTCAACAGCGGTAAGATCCTTGAGAGTTTTCGCCCCGAAGAACGTTTTCCAATGATGAGCACTTTTAAAGTTCTGCTATGTGGCGCGGTATTATCCCGTGTTGACGCCGGGCAAGAGCAACTCGGTCGCCGCATACACTATTCTCAGAATGACTGGTTGAGTACTCACCAGTCACAGAAAAGCATCTTACGGATGGCATGACAGTAAGAGAATTATGCAGTGCTGCCATAACCATGAGTGATAACACTGCGGCCAACTTACTTCTGACAACGATCGGAGGACCGAAGGAGCTAACCGCTTTTTTGCACAACATGGGGGATCATGTAACTCGCCTTGATCGTTGGGAACCGTAGCTGAATGAAGCCATACCAAACGACGAGCGTGACACCACGATGTCTGCAGTAATGGCAACAACGTTGCGCAAACTATTAACTGGCTGAACTACTTACTCTAGCTTCCCGGCCACAATTAATAGACTGGATGGGAGGTGTATAAAGTTGTAGGACCACTCTTGCGTTCGGCCCTTCCGGTTGGTTGGTTTATATGTGGTAACAT CCCCCTCCTCCCTCGTACCGCGCGAGGCAGATCGTCAGTCAGTCACGATGAATTCCCGGGGATCCACGCGGAACCAGATCCGATTTTGGAGGATGGTCGCCACCACCAAACGTGGCTTGCCAGCCCTGCAAAGGCCATGCTATATACTTGCTGGATTTCAAGTACTTATCAATTTGTGGGATAGCTTCAATCGTTTTTTAAAACAAACTAATTTTGGGAACGCATCCAGGCACATTGGGTCATGTATAAAACAACATCAAGAGCGTCATACAACATGAAGTCAGGATGGGTTACATGATCACCATTTAAATATGTTTTATGACATAAACGATCTTCGAACATTTTCAGCATTTCAGGTAGCTTGCAAGAAAATCAACTTTGAGAGTTTCAAAGTCTTTACTATATGCAATTCTCGAAACACCGTATCTAATATCCAAAACCGCTCCTTCAAGCATTGAAATCTCTGCACGCTCTTTTGGACAACCACCCAACATGTTGTGCTTGTCAGCTATATAACGTATGATGGCCATAGACTGTGTTAATTTAACATCACCATCAATATAATAAGGAAGATTGGGAAACTCCAAACCCAACAAACTTTTTGTTTCGCCATTTATCACCTTCATCGCGCTCATACAAATGCTCTCATATTTTTCTTCAAGATATTCCAAAAGAAGTCGAGTGGGTTGCACAAGGCCCTTAATTTTCCAATAACCTAGTATAGGGGACATGAATACTGTTTCCTGTGTGAAAT TGTTATCCGCTCACAATTCCACACATTATACGAGCCGATGATTAATTGTCAACAGCTCATTTCAGAATATTTGCCAGAACCGTTATGATGTCGGCGCAAAAAACATTATCCAGAACGGGAGTGCGCCTTGAGCGACACGAATTATGCAGTGATTTACGACCTGCACAGCCATACCACAGCTTCCGATGGCTGCCTGACGCCAGAAGCATTGGTGCACCGTGCAGTCGATAAGCTAATTCCAACGCCATCAAAAATAATTCGCGTCTGGCCTTCCTGTAGCCAGCTTTCATCAACATTAAATGTGAGCGAGTAACAACCCGTCGGATTCTCCGTGGGAACAAACGGCGGATTGACCGTAATGGGATAAGTCACGTTGGTGTAGATGGGCGCATCGTAACCGTGCATCTGCCAGTTTGAGGGGACGACGACAGTATCGGCCTCAGGAAGAACGCACTCCAGCCAGTTTTCCGGAACCCTTTTGGGGCCGGAAACCAGGAAAAGGGCCTTTCCCCTTCAGGTTGCCACTTGTGGGAAAGGCGATCGGGGGGGCCCTTTCCTATTAACCCATTGGCAAAAGGGGTAGTCA

Inhibitors and probes targeting endo-glycosidases

Abstract

Extracted data

Inhibitors and probes targeting endo-glycosidases

Abstract

Extracted data

Related items

Related items