Oder über das Jagen von Drachen

Acknowledgements IV List of abbreviations IX 1 Zusammenfassung 1-1 1 Abstract 1-2 2 Introduction 2-1 2.1 Connecting AL amyloidosis and multiple myeloma 2-1 2.1.1. Amyloidoses 2-2 2.1.2. Monoclonal gammopathies 2-3 2.1.2.1 LC can be isolated from urine 2-3 2.1.3. Myelomagenesis 2-3 2.2 LC sequence analysis 2-5 2.2.1. Light chain sequencing 2-6 2.2.2. Automated MS-based sequencing, a modern myth 2-6 2.2.3. De novo sequencing-assisted peptide mapping 2-7 2.2.4. Underlying assumption of the base peak interrogation method 2-8 2.3 Implications of immunoglobulin light chain structure 2-9 2.3.1. Light chains have a highly conserved variable and constant domains 2-10 2.3.2. Individual light chains will always be unique 2-11 2.3.3. Unique, but not as variable as one would expect 2-12 2.3.4. Structural determinants of the biophysical behavior of light chains 2-14 2.3.5. Glycosylation of light chain proteins 2-16 2.4 Amyloid 2-18 2.4.1. Protein folding and misfolding 2-19 2.4.2. Amyloid aggregation 2-20 2.5 Studies on amyloidogenic LC 2-24 2.6 The effects of (-)-epigallocatechin-(3)-gallate effect on amyloids 2-25 2.6.1. Epigallocatechin-3-gallate 2-25 2.6.2. Phenomenological observations on EGCGs interaction with amyloid 2-26 2.6.3. Redirection, remodeling & seeding competence of aggregates 2-28 2.6.4. Elucidating the mechanism of EGCGs interaction with amyloid 2-30 2.7 Scope of this thesis 2-35 3 Materials and methods 3-1 3.1 Patient material 3-1 3.1.1. Patient diagnosis 3-1 3.1.2. Patient selection for this study 3-2 3.2 Experimental procedures 3-2 3.2.1. Chemicals, material and equipment 3-2 3.2.2. Light chain protein purification 3-7 3.2.2.1 Analysis of target and most relevant matrix proteins 3-7 3.2.2.2 LC Purification from the urine of patients without albuminuria 3-8 3.2.3. Biochemical characterization of soluble light chains 3-9 3.2.3.1 Gel electrophoresis and Western blots 3-9 3.2.3.2 Gel stainings 3-9 3.2.3.3 Western blot and Immunostaining 3-11 3.2.3.4 Sample storage conditions 3-12 3.2.3.5 LC concentration and HSA content 3-12 3.2.3.6 Molecular weight determination of monomer 3-13 3.2.3.7 Dimer/monomer ratio 3-13 3.2.3.8 Proteolysis 3-13 3.2.3.9 Glycosylation analysis 3-14 3.2.4. Biophysical characterization 3-17 3.2.4.1 CD spectroscopy 3-17 3.2.4.2 Tryptophan fluorescence spectroscopy 3-17 3.2.4.3 Guanidine hydrochloride denaturations 3-17 3.2.4.4 Evaluation of unfolding data 3-18 3.2.5. Aggregation assays & aggregate characterization 3-22 3.2.5.1 Establishment of monomerization conditions 3-22 3.2.5.2 Optimization of aggregation conditions 3-22 3.2.5.3 Continuous aggregation assays with sequential shaking 3-24 3.2.5.4 Discontinuous aggregation assays with permanent shaking 3-25 3.2.5.5 Atomic force microscopy imaging 3-25 3.2.5.6 Congo red staining and fluorescence imaging 3-26 3.2.5.7 Electrophoretic-stability assays 3-26 3.2.5.8 Ultracentrifugation / solubility assay 3-26 3.2.6. MS-based protein sequencing 3-26 3.2.6.1 Development of a peptide mapping search database 3-26 3.2.6.2 LC- MS/MS measurement at the Mayo Clinic Rochester 3-27 3.2.6.3 MS/MS data analysis with the PEAKS Studio software 3-29 3.2.6.4 Prediction of amyloidogenicity 3-41 4 Results 4-1 4.1 LC purification 4-1 4.1.1. Development of a purification strategy 4-1 4.1.2. Purification of LC from urine of non- kidney-impaired patients 4-3 4.1.2.1 Sample clarification and capture purification 4-3 4.1.2.2 Pilot clarification and capture purification 4-4 4.1.2.3 Optimization diafiltration for capture purification step 4-5 4.1.2.4 Final purified proteins 4-5 4.1.2.5 Purification of LC from patient sequenced from bone marrow-derived DNA 4-6 4.2 Biochemical characterization 4-7 4.2.1. Sample storage conditions 4-7 4.2.2. Human serum albumin content 4-8 4.2.3. Apparent molecular weight 4-9 4.2.4. Monomers & cross-linked dimers or higher order multimers 4-9 4.2.5. Susceptibility to proteolysis at 37°C 4-10 4.2.6. Glycosylation of light chain proteins 4-11 4.2.7. Summary 4-12 4.3 Biophysical characterization 4-13 4.3.1. CD spectroscopy and thermal denaturations 4-13 4.3.2. Chemical denaturations with guanidine hydrochloride 4-17 4.3.3. Summary 4-21 4.4 Aggregation behavior of Ig light chains 4-21 4.4.1. Establishment of aggregation assays 4-21 4.4.1.1 Reduction as prerequisite for LC aggregation & pH-dependence 4-22 4.4.1.2 Temperature, agitation and salt concentration 4-25 4.4.1.3 Albumin forms negligible amounts of ThT positive species 4-27 4.4.1.4 Summary 4-28 4.4.2. Analysis of LC aggregation behavior 4-28 4.4.3. The effect of epigallocatechin-3-gallate on light chain aggregation 4-33 4.4.4. Summary 4-37 4.5 Analysis of the influence of glycosylation on LC aggregation 4-37 4.5.1. Establishment of native deglycosylation 4-37 4.5.2. Secondary structure of natively deglycosylated κ-AL-1 4-38 4.5.3. Light chain stability 4-39 4.5.4. ThT aggregation assays 4-41 4.5.5. Summary 4-42 4.6 MS-based LC sequencing 4-42 4.6.1. Development of a search database for peptide mapping 4-43 4.6.2. Advancement of base peak interrogation for LC sequencing 4-45 4.6.2.1 Optimization and refinement of base peak interrogation 4-47 4.6.2.2 Method evaluation based on the sequencing of the λ-AL-1 sample 4-47 4.6.2.3 Extension of base peak interrogation method 4-49 4.6.3. The sequence of the λ-AL-1 light chain 4-50 4.6.3.1 Sequence quality, coverage and germline donor. 4-50 4.6.3.2 Mutations from germline sequence and influence on amyloidogenicity 4-52 4.6.4. On increasing the efficiency of BPC peak interrogation 4-55 4.6.4.1 Could the number of analyzed digests be reduced? 4-55 4.6.4.2 Top 5 BPC peak analysis of samples λ-AL-1, λ-AL-2 and λ-MM-3 4-56 4.6.5. Summary 4-58 5 Discussion 5-1 5.1 LC purification 5-1 5.2 Biochemical characterization 5-4 5.3 Aggregation behavior of Ig light chains 5-6 5.3.1. Relation of LC stability and aggregation 5-6 5.3.2. LC aggregation under reducing conditions 5-6 5.3.3. Analysis of LC aggregation behavior 5-7 5.3.4. The effects of EGCG on amyloid formation 5-9 5.3.4.1 EGCG affects LC aggregation in a non-sequence-specific manner 5-9 5.3.4.2 EGCG influence amyloidogenesis by a comprehensive binding mode 5-9 5.4 Analysis of the influence of the glycosylation 5-13 5.5 MS-based LC sequencing 5-13 5.5.1. Evaluation of the base peak interrogation method 5-14 5.5.2. Implications of somatic mutations in λ-AL-1 light chain 5-15 6 Bibliography 6-1 7 Academic CV 7-1 8 Appendix 8-1 8.1 LC Purification 8-1 8.1.1. Experimental procedures 8-1 8.1.1.1 Chemicals, material and equipment 8-1 8.1.1.2 LC purification from high-albumin matrices 8-1 8.1.2. Purification of first sample batch 8-4 8.1.3. Towards a purification strategy for Ig light chains from urine and serum of kidney-impaired patients 8-5 8.1.3.1 Strategy 1: Direct size exclusion chromatography. 8-6 8.1.3.2 Strategy 2: Depletion of albumin and IgG 8-7 8.1.3.3 Strategy 3: LC-enrichment non-antibody affinity chromatography 8-8 8.1.3.4 Strategy 4: Enrichment of LC via with Immune affinity chromatography 8-10 8.1.3.5 Summary 8-13 8.1.3.6 Discussion: LC purification from high albumin matrices 8-13 8.1.3.7 Proposed purification strategy for urine of kidney impaired patients 8-14 8.3 Biochemical characterization 8-16 8.3.1. Data: Molecular weight determination 8-16 8.4 Biophysical characterization 8-18 8.4.1. Data: Thermal denaturations (CD) 8-18 8.4.2. GdnHCl denaturations 8-28 8.4.2.1 Optimizations for GdnHCl-induced LC denaturations 8-28 8.4.2.2 Reversibility of GdnHCl denaturations 8-30 8.4.2.3 Data: Trp fluorescence 8-32 8.4.2.4 Data: CD spectroscopy 8-43 8.5 Aggregation behavior of Ig light chains 8-46 8.5.1. Experimental procedures 8-46 8.5.1.1 Aggregation assays & aggregate characterization 8-46 8.5.2. Data: LC aggregation kinetics 8-48 8.5.3. EGCG affects LC aggregation non-sequence-specific 8-49 8.5.3.1 Establishment of conditions to evaluate EGCGs influence on LC aggregation 8-49 8.5.3.3 Data: Aggregation in presence of EGCG 8-54 8.6 MS-based sequencing 8-56 8.6.1. Optimization and refinement of base peak interrogation 8-56 8.6.2. Peak tables and alignment tables 8-60 8.6.2.1 λ-AL-1 sample 8-60 8.6.2.2 λ-AL-2 & λ-MM-3 sample 8-86AL amyloidosis and multiple myeloma (MM) is caused by plasma cells that form a clonal population that produces monoclonal light chains (LC). A clonal population exceeding 10% of all bone marrow plasma cells is considered as MM, a form of bone marrow cancer. Only about 46% of AL patients progress to MM. In contrast to MM AL is defined by amyloid deposits formed by the LC. Amyloid fibrils contain a common cross-β-sheet motif which binds to amyloidophilic dyes like Thioflavin T (ThT). Epigallocatechin-3-gallate (EGCG) was discovered as a possible therapeutic agent in amyloidosis. It reduced cardiac amyloid deposits in AL patients suggesting its potential as secondary treatment option. The aim of my thesis was to characterize structures and stabilities of authentic amyloidogenic light chains, to analyze their aggregation behavior in vitro, and to investigate EGCG as a possible treatment against amyloidogenic deposits in patients suffering from AL. I isolated nine LC proteins (2x κ-AL, 2x λ-AL, 2x κ-AL, 3x λ-AL) from patient’s urine. SDS-PAGE resolved LC into monomers and disulfide-linked dimers. Thermal and chemical stabilities of LC proteins correlated weakly to MM or AL, suggesting that kinetic factors might determine amyloid formation. All light chains formed ThT-positive high molecular weight aggregates. Aggregation kinetics displayed two distinct phases. The first aggregation phase was initiated by reduction of inter- molecular disulfide bonds. It corresponded to conversion into oligomers, while the second phase corresponded to cross-β-sheets formation. At low LC concentrations, all aggregation kinetics showed only weak concentration dependence, suggesting that a conformational change was rate-limiting. At higher concentrations AL-LC, but not MM-LC displayed, concentration dependence typical for amyloid formation kinetics. EGCG inhibited the second aggregation phase and induced SDS-stable aggregates in all nine LC. Glycosylation of one LC did not affect its aggregation. In addition to the analysis of LC aggregation I determined the sequence of one LC (λ-AL-1) via mass spectrometry. I advanced the previously described base peak interrogation method, which infers LC peptide sequences after combined peptide mapping and de novo sequencing from LC-MS/MS data: To confidently determine the sequence of the λ-AL-1 LC an IMGT-derived germline LC sequence database, alignment to the IMGT numbering system, and a scoring system for sequence quality were introduced. I identified 10 mutations of the λ-AL-1 LC compared to its germline parent. Strikingly, most of these mutations lay at the protein- protein interface inside the LC-dimers. This supports my interpretation of the kinetic data that conformational change rather than self-assembly determines LC amyloid formation.Sowohl AL Amyloidose als auch Multiples Myelom (MM) werden durch klonale Plasmazellen verursacht, die monoklonale Leichtketten produzieren. Diese klonalen Plasmazellen gelten als MM, sobald sie einen Anteil von über 10% des Knochenmarks ausmachen. Dies ist jedoch nur bei 46% der AL-Patienten der Fall. Deshalb wird für die Diagnose AL das Vorhandensein von amyloiden Proteinablagerungen herangezogen. Diese Ablagerungen besitzen ein cross-β-Faltblatt Motiv, an welches spezifische Farbstoffe wie Thioflavin T (ThT) binden können. Das Polyphenol Epigallocatechin-3-gallat (EGCG), das an das selbe cross-β-Faltblatt Motiv bindet, ist ein möglicher Wirkstoff zur Behandlung von AL. EGCG war in der Lage Amyloidablagerungen im Herzen einzelner Patienten aufzulösen. Meine Arbeit analysiert die Amyloidbildung von Leichtketten aus dem Urin von AL und MM Patienten und untersucht zudem den mechanistischen Effekt von EGCG auf die Leichtkettenaggregation. Dazu wurden sowohl die Struktur und Stabilität von Leichtketten als auch ihr Aggregationsmechanismus untersucht. Zunächst habe ich ein Protokoll zur Isolierung von neun Leichtkettenproteine (2x κ-AL, 2x λ-AL, 2x κ-AL, 3x λ-AL) aus dem Urin von AL- und MM-Patienten etabliert. Die aufgereinigten Leichtketten lagen zu 50% als Monomere und zu 50% als Disulfid-verbrückte Dimere vor. Ihre Stabilität korrelierte in der thermischen und chemischen Denaturierung nur schwach zur Diagnose MM oder AL, was nahelegt, dass die unterschiedliche Aggregation von AL und MM Leichtketten in vivo auf kinetische Faktoren zurückzuführen ist. In vitro formten sowohl AL- als auch MM- Leichtketten ThT-bindende Aggregate in einer zweiphasigen Aggregationskinetik. Voraussetzung für die erste Phase war die Reduzierung intermolekularer Disulfidbrücken. Hierbei wandelten sich monomere Leichtketten in aggregationsfähige Oligomere um. In der zweiten Phase bildete das Protein cross-β-Faltblätter. Dabei hing die erste Phase nicht von der Leichkettenkonzentration ab. Dies deutet darauf hin, dass hier eine Konformationsänderung geschwindigkeitsbestimmend ist. Im Gegensatz zu MM- Leichtketten, war die zweite Aggregationsphase von AL-Leichketten abhängig von der Leichtkettenkonzentration. EGCG inhibierte die zweite Aggregationsphase aller neun Leichtketten und induzierte SDS-stabile Aggregate. Die Glykosylierung einer Leichtkette hatte keinen Einfluss auf deren Aggregationsverhalten. Weiterhin habe ich ein massenspektrometrisches Verfahren zur Sequenzierung von Leichtketten weiterentwickelt. Mittels Basispeak-Abfrage und de novo Sequenzierung assistiertem Peptidmapping in einer IMGT-abgeleiteten LC-Sequenzdatenbank konnte ich die Sequenz der λ-AL-1 Leichtkette und ihre Sequenzgüte zuverlässig bestimmen. Ich konnte 10 Mutationen im Vergleich zum Genvorläufer bestimmen, von denen viele im Protein-Protein-Interface innerhalb von LC Dimeren lagen. Dies stützt meine Interpretation, das eine Konformationsänderung der Bildung von LC-Amyloid bestimm