Expression, inhibition and in silico modelling of human neuronal nitric oxide synthase

Nitric oxide (NO) is a free radical, gaseous molecule that is involved in a vast range of biological signalling processes including long term memory formation, neurogenesis, vasodilation and inflammation. The molecule is produced by a family of enzymes called nitric oxide synthases (NOS). There are three major isoforms of these NO producing enzymes, endothelial NOS (eNOS), inducible NOS (iNOS) and neuronal NOS (nNOS). eNOS is constitutively expressed in the circulatory system, while iNOS is inducibly expressed in response to cellular invasion by a pathogen. nNOS, which was the isoform targeted in this work, is constitutively expressed in the brain and central nervous system, where the NO it produces regulates many neurogenic pathways. Over activation of nNOS leads to elevated levels of NO in the brain. This has been implicated in an array of serious physiological conditions including schizophrenia, Alzheimer’s and Parkinson’s disease. Modulation of nNOS over activation therefore represents a potential therapeutic avenue for treatment of a variety of debilitating and life threatening conditions. The field of nNOS inhibition currently focuses almost exclusively on the arginine binding site of the rat nNOS isoform of the enzyme. The arginine binding site in rat nNOS does differ from that of human nNOS, and as such is not an ideal system with which to study human nNOS inhibition. Therefore, this work incorporated computational studies to model and examine the binding sites of human nNOS. These included construction and verification of a human nNOS homology model. This offered the best option for a detailed examination of the protein as no crystal structure of human nNOS was initially available. The model was used to screen for novel and selective human nNOS inhibitors. This structure based approach was complimented by the use of a ligand based design approach. To do this, a pharmacophore was constructed and used for high throughput screening for the identification of novel inhibitory compounds. The pharmacokinetic properties of the hit compounds were considered, as was their potential selectivity for human nNOS over human eNOS and human iNOS. Over 20 million compounds from the Maybridge and Zinc databases were screened with the various computational methods, and 72 were chosen for biological evaluation. This high throughput in silico screening approach has not been widely utilised in the search for human nNOS inhibitors, with only one such study reported. This publication did not attempt to biologically evaluate the potential nNOS inhibitors. A functional assay was required to test the compounds identified from the in silico screens. This necessitated the recombinant expression of human nNOS, iNOS and eNOS proteins. The literature routinely reports rat nNOS, murine iNOS and bovine eNOS to test inhibitory compounds, so the establishment of an all human high throughput assay system represented a major step forward in the field. Bacterial cells were transformed with the DNA sequences for all three human NOS isoforms. The proteins were recombinantly expressed and purified from the resulting E. coli lysate. The procedure for human nNOS expression was optimised, and the co-expression of human nNOS with human calmodulin was achieved. Calmodulin is a vital co-factor for the production of NO from NOS and its co-expression with human nNOS has not previously been reported. The same approach was applied to eNOS and iNOS. The expressed human NOS proteins were then used to establish a functional, high throughput NOS assay using the Griess reaction. Once established, the assay was used to screen the computationally derived hits for human nNOS inhibitory activity. Some hits could be purchased, but many had to be synthesised. The synthesised compounds were loosely classified into three classes, i) piperazine containing compounds, ii) urea containing compounds and iii) modified arginines and thioethers. The high throughput expression/assay system for human nNOS worked well, and although no novel inhibitory compounds were identified, it allowed for the testing of 72 in silico derived compounds. The assay system also allowed for determination of the IC50 of the non-selective NOS inhibitor, L-NNA. This work represents a valuable contribution to the field of nNOS inhibition as it is among the first projects to attempt large scale screening for human nNOS inhibitory compounds. The projects also succeed in the expression of all three human NOS proteins with human calmodulin, and a high throughput expression/assay system was established for human nNOS inhibition. Novel compounds were successfully synthesised, characterised and evaluated. This work provides a useful roadmap for the future development of human nNOS inhibitory compounds, as well as providing an established computational screening methodology and a high throughput functional human nNOS assays/expression system