UNIVERSIDADE DE LISBOA

FACULDADE DE CIÊNCIAS DEPARTAMENTO DE QUÍMICA E BIOQUÍMICA

Methylglyoxal effects on protein aggregation and amyloid formation

Luís Miguel Araújo da Silva Oliveira

Doutoramento em Bioquímica (Especialidade: Biofísica Molecular)

Tese orientada por: Prof. Doutor Alexandre Luís de Matos Botica Côrtes Quintas Doutor Carlos Alberto Alves Cordeiro

2009

De acordo com o disposto no artigo n°. 40 do Regulamento de Estudos Pós‐Graduados da

Universidade de Lisboa, Deliberação n° 961/2003, publicada no Diário da República – II Série n°.

153 – 5 de Julho de 2003, foram incluídos nesta dissertação os resultados dos seguintes artigos:

Luís M. A. Oliveira, Carlos Cordeiro, Ana Ponces Freire, Carla Ascenso, Alexandre Quintas. 2007.

Unveiling heme proteins conformational stability through a UV absorbance ratio method Anal

Biochem. 371: 253‐5

Alexandre Quintas, Luís M. A. Oliveira. 2008. Protein hierarchy levels of structural organization for amyloidogenesis: Biochemical and biophysical aspects underlying misfolding and disease.

Protein Misfolding in Biology and Disease T. F. Outeiro, Research Signpost: 1‐34.

Ricardo A. Gomes, Luís M. A. Oliveira, Mariana Silva, Carla Ascenso, Alexandre Quintas, Gonçalo

Costa, Ana V. Coelho, Marta Sousa Silva, António E. N. Ferreira, Ana Ponces Freire, Carlos

Cordeiro. 2008. Protein glycation in vivo: Functional and structural effects on yeast enolase. Biochem

J. 416: 317‐26

Aknowledgments/Agradecimentos

Aknowledgments/Agradecimentos

Chegada a conclusão desta etapa com mais de quatro anos, não a poderia deixar terminar sem agradecer a todos os que, directa ou indirectamente, contribuíram para a realização deste trabalho.

Em primeiro lugar quero expressar o meu mais profundo agradecimento ao Professor Doutor

Alexandre Quintas, orientador principal deste trabalho, por todo o seu apoio, amizade, compreensão e disponibilidade. Desde que começámos a trabalhar juntos que sempre manteve a nossa relação numa base de confiança, transmitindo‐me a ideia de que somos uma equipa. E uma equipa luta junta pelos seus objectivos. As suas qualidades humanas e científicas que sempre marcaram presença são uma referência. Estou certo que no futuro ainda muito mais irei aprender com ele. Por tudo, o meu mais sincero obrigado.

Ao Doutor Carlos Cordeiro, segundo orientador deste trabalho, agradeço a preocupação e ensinamentos que me fizeram crescer como investigador desde o primeiro dia que integrei o grupo de enzimologia. A sua serenidade, ponderação e conhecimento contribuíram de forma extremamente positiva para o desenrolar deste trabalho e para a minha aprendizagem como investigador.

À Professora Doutora Ana Ponces, uma palavra especial por tudo o que representa. Desde os tempos de licenciatura que é uma referência minha pelos ensinamentos, pela sua capacidade invulgar de gerar motivação, pelo apoio, amizade e postura perante a vida. Uma pessoa que admiro profundamente que me ensinou e preparou a ser o investigador que sou.

Aos amigos e colegas do Laboratório de Patologia Molecular, onde passei a maior parte do meu doutoramento, quero agradecer a simpatia, amizade, companheirismo e boa disposição que sempre pautou a nossa investigação. O meu destaque à Professora Carla Ascenso pelo apoio, ao

Carlos Família, com quem partilho o laboratório desde o início, e à Lena pela amizade. Carlos fica a frase: “Oi pessoal tudo bem? Vai um café?”. À Ana e ao Miguel pela amizade, companheirismo e espírito de grupo criado. Não esquecerei a entreajuda estabelecida convosco bem como as

Aknowledgments/Agradecimentos 

aventuras por Lausanne e Copenhaga. Espero sinceramente que não fiquem por aqui. À Joana pela alegria, amizade e bom espírito. É bom ter‐te no laboratório. Ao Gonçalo pelo espírito e amizade. À

Catarina que foi marcante pelo furacão de alegria que sempre trazia consigo quando entrava pelo laboratório. Uma amiga que ficou. E finalmente, aos colegas que, não fazendo parte do nosso grupo, estão tão perto: à Teresa e à Susana pela amizade e carinho e ao Henrique, à Inês e ao Paulo pelo bom espírito e profissionalismo que trazem sempre consigo. You make my day!!!

Agradeço também a todas as pessoas do Grupo de Enzimologia, o grupo que inicialmente me acolheu e formou. Foi muito o que aprendi junto de vós. O meu agradecimento especial ao

Professor António Ferreira pelos conhecimentos e disponibilidade que sempre mostrou. Ao Ricardo

Gomes por ser um colega espectacular. Diversão, capacidade de trabalho, entreajuda, amizade, espírito empreendedor, sempre foram qualidades que estiveram presentes. Ao Nuno Lages por ser um grande amigo com quem partilhei bons e maus momentos ao longo destes anos. Ao Hugo

Miranda pela longa amizade e por todos os desafios que superámos em conjunto ainda desde a licenciatura. À sempre sorridente Lídia pela amizade e boa disposição. E finalmente à Mariana, ao

Gonçalo e à Marta pela amizade e apoio.

Uma palavra também para o Doutor Tiago Outeiro que desde o início se revelou ser uma pessoa inspiradora que me leva sempre a tentar ser um investigador cada vez melhor. Admiro muito a sua capacidade de comunicação, trabalho e toda a ciência que produz. O meu obrigado por toda disponibilidade, serenidade e confiança que sempre esteve presente na nossa relação, bem como a ajuda fornecida para que fosse possível trabalhar no laboratório do Professor Doutor Hilal

Lashuel.

Ao Professor Doutor Hilal Lashuel o meu muito obrigado por me ter acolhido no seu laboratório. As suas qualidades científicas são do melhor que tive a possibilidade de presenciar.

Apesar do pouco tempo que estive no seu laboratório, foram inúmeros os conhecimentos que obtive e a aprendizagem que fiz. Para isso contribuiu também sem dúvida a orientação mais estreita da Doutora Katrina Paleologou com quem foi um prazer enorme trabalhar. A todos o meu muito obrigado em especial aos colegas Abid, Asad, Margot e Mirva.

Aos amigos, que estão sempre lá, para a ciência e para tudo o resto. Não há palavras para descrever a vossa importância na minha vida. Sinto‐me um privilegiado por poder chamar amigo a Aknowledgments/Agradecimentos muita gente, mas não consigo deixar de vos destacar. Obrigado João Luís, Sónia Pais, Margarida

Neves de Sousa, Ana Rita Moreira, João Lopes, Sara Gomes e a todos os Scramble por fazerem parte da minha vida.

Por último o meu agradecimento a toda a família… aos avós, aos padrinhos, aos tios e primos que, embora longe, são tão importantes e sabem estar tão perto; e duas palavras de grande carinho: para com a minha mãe por tudo o que me proporcionou desde que nasci. Este meu ínfimo sucesso é sem dúvida fruto do seu amor, trabalho e dedicação… e eu não sou fácil de aturar!!! É impossível verbalizar a importância que tens na minha vida; e para com o meu pai a quem quero muito dedicar este trabalho. A tua partida a meio deste caminho deixa‐me uma mágoa inigualável. Era capaz de escrever outra tese sobre a pessoa, o exemplo e a referência que és e como te tenho presente nos meus dias, nas minhas atitudes e decisões. Sei que ninguém estaria mais orgulhoso que tu por me ver chegar aqui e por isso sei que o que fiz também é teu.

Aknowledgments/Agradecimentos 

Aknowledgments/Agradecimentos

Table of Contents

AKNOWLEDGMENTS/AGRADECIMENTOS...... V

TABLE OF CONTENTS...... IX

SUMMARY...... XIII

RESUMO ...... XV

ABBREVIATIONS ...... XIX

CHAPTER I...... 2

1. PROTEIN STRUCTURE ...... 3 1.1 Secondary structure...... 5 1.1.1 Helical structures ...... 5 1.1.2 Beta structures ...... 6 1.1.3 Non-repetitive secondary structure...... 7 1.2 Tertiary structure...... 7 1.3 Protein dynamics and flexibility ...... 10 1.4 Natively unfolded proteins...... 11 2. THE PROTEIN FOLDING PROBLEM ...... 14 2.1 The folding code ...... 14 2.2 The folding process...... 15 3. PROTEIN MISFOLDING IN BIOLOGY AND DISEASE...... 18 3.1 Protein structure and amyloidosis...... 19 3.1.1 Intrinsically unstructured proteins in conformational diseases ...... 21 3.1.2 Monomeric proteins in conformational diseases...... 24 3.1.3 Oligomeric proteins in conformational diseases ...... 26 3.1.4 The intrinsically unstructured, monomeric and oligomeric proteins in amyloidogenesis ...... 28 3.2 Molecular mechanisms of amyloidosis ...... 29 3.3 Cellular aspects of protein misfolding and disease ...... 32 4. POST-TRANSLATIONAL PROTEIN MODIFICATIONS...... 35 4.1 Polyamine binding...... 36

Table of Contents 

4.2 Oxidation ...... 36 4.3 Nitration...... 37 4.4 Ubiquitination...... 37 4.5 Sumoylation ...... 38 4.6 Phosphorilation ...... 38 4.7 Glycation ...... 39 4.7.1 Methylglyoxal...... 42 4.7.2 Glycation and amyloidosis...... 46

CHAPTER II ...... 51

1. ABSTRACT ...... 53 2. INTRODUCTION ...... 53 3. MATERIAL AND METHODS...... 55 3.1 Methylglyoxal preparation ...... 55 3.2 Cytochrome c glycation ...... 55 3.3 Cytochrome c aggregation...... 55 3.4 Structural analysis of cytochrome c...... 56 3.5 Conformational stability measurements ...... 56 4. RESULTS AND DISCUSSION ...... 57 4.1 Glycation induces cytochrome c aggregation...... 57 4.2 Structural changes of cytochrome c upon glycation...... 59 4.3 Conformational stability of cytochrome c species ...... 63 5. CONCLUSION...... 66

CHAPTER III...... 69

1. ABSTRACT ...... 71 2. INTRODUCTION ...... 71 3. MATERIAL AND METHODS...... 73 3.1 Methylglyoxal preparation ...... 74 3.2 Yeast strains and growth conditions...... 74 3.3 Enolase purification...... 74 3.4 In vitro glycation of purified enolase by methylglyoxal...... 75 3.5 Western blot and HPLC analysis...... 75 3.6 MS analysis...... 76 3.7 Structure and stability analysis...... 77 3.8 Enolase activity assay...... 78 3.9 Protein structure...... 78 4. RESULTS AND DISCUSSION ...... 78

 x

Table of Contents

4.1 Characterization of enolase glycation by MS...... 78 4.2 Glycation effects on enolase folding structure and enzyme activity ...... 84 5. CONCLUSION...... 87 6. ACKNOWLEDGEMENTS...... 90

CHAPTER IV...... 93

1. ABSTRACT ...... 95 2. INTRODUCTION ...... 95 3. EXPERIMENTAL PROCEDURES ...... 98 3.1 Methylglyoxal preparation ...... 98 3.2 Insulin preparation ...... 98 3.3 Expression and purification of human α-synuclein ...... 99 3.4 Protein glycation ...... 100 3.5 Fibril formation ...... 100 3.6 Analysis of fibrillation kinetics ...... 100 3.7 Fluorescence measurements...... 101 3.8 Circular dichroism measurements...... 101 3.9 Conformational stability measurements ...... 101 3.10 Size-exclusion and native-PAGE experiments ...... 102 3.11 Transmission electron microscopy ...... 102 4. RESULTS AND DISCUSSION ...... 103 4.1 Methylglyoxal reduces protein fibril formation...... 103 4.2 Methylglyoxal causes protein oligomerization ...... 106 4.3 Effects of methylglyoxal on insulin structure and stability ...... 109 5. CONCLUSION...... 112 6. ACKNOWLEDGEMENTS...... 115

CHAPTER V ...... 117

APPENDIX ...... 125

REFERENCES ...... 133

Summary

Summary

Proteins are the core of life, providing functional support for all chemical processes in living cells. To perform this task, proteins must have a specific structure achieved by folding from its primary level to a three dimensional architecture. Disturbing this process inevitably leads to serious metabolic and physiological changes. Throughout evolution, molecular chaperones and complex protein quality control mechanisms evolved to keep protein structures functional in their lifespan.

When something goes astray in this process, misfolded proteins may aggregate leading to conformational diseases such as Alzheimer, Parkinson, Huntington, Andrade and Amyotrophic

Lateral Sclerosis. Several factors may contribute to protein misfolding, including point mutations, chemical stress, post‐translational modifications and impairment of protein quality control mechanisms.

Post‐translational modifications play a very important role in protein misfolding, since proteins are extensively modified to achieve three dimensional structures, to regulate its activity or to target to degradation. Contrary to controlled post‐translational modifications, like phosphorylation or glycosilation, where enzymes specifically modify proteins to produce a specific cellular effect, glycation is a non‐enzymatic process where arginine and lysine side chains are irreversibly modified by carbonyl‐containing molecules. Moreover, there are several carbonyl‐ containing molecules in vivo, like glucose and methylglyoxal that irreversibly modify proteins through the Maillard reaction. Thus, it is expected that the extensive unregulated modification of particular proteins might have a deleterious effect on protein structure and function and to be associated with cell and tissue damage observed in some pathologies and aging. The observation that AGE‐modified proteins accumulate in connexion with the clinical complications of several human pathologies links protein glycation to human diseases such as diabetes mellitus, age‐related disorders, atherosclerosis and conformational diseases. Methylglyoxal is likely to be the most important glycation agent in vivo and its synthesis occurs in all living cells, in an uncontrolled way, mainly as a non‐enzymatic by‐product of glycolysis. However, in spite of all the evidences that link protein glycation with human pathological conditions, the molecular mechanisms underlying the effects of protein glycation by methylglyoxal remains unclear.

The major goal of the investigations presented in this thesis is the understanding of the biochemical effects of methylglyoxal on structure, thermodynamic stability and aggregation of

Summary  proteins. Chapter I introduces and describes the theme. Following a brief description about protein structure and the folding problem, the importance of protein misfolding in biology and disease is presented. A review of the protein hierarchy levels of structural organization and their relation to amyloidogenesis was made. This chapter also introduces protein glycation and methylglyoxal metabolism as well as the implications of protein glycation in several human pathologies.

In chapter II, cytochrome c was chosen as a protein model to study the effects of methylglyoxal glycation. The results show a substantial decrease in conformational stability after glycation which potentiated the formation of aggregates that retained native‐like structure. This is the first evidence that methylglyoxal glycation might be associated to protein native‐like protein aggregation. In parallel a novel method for the determination of the conformational stability of heme proteins with a two state unfolding model was developed and described in the appendix.

Chapter III presents a detailed comparison between the effects of in vitro and in vivo glycation on enolase, previously demonstrated to be a major glycation target in yeast. Our results demonstrate that glycation causes enolase unfolding and structural changes, leading to loss of enzyme activity. Evidences emerge for the existence of differences between in vitro and in vivo glycation effects, particularly on the nature and molecular location of specific advanced glycation end‐products. Results regarding enolase structure, thermal stability and enzyme activity show comparable in vitro and in vivo glycation effects.

The experiments described in chapter IV use two amyloidogenic proteins as glycation targets.

Human α‐synuclein and insulin are both related to specific human diseases. In spite of having distinct structures and different fold types, the fibrillation of both proteins was decreased in the presence of methylglyoxal. Instead of amyloid fibrils, aggregation was directed to the formation of soluble aggregates. Regarding α‐synuclein, an increase in protofibrils concentration, some of them with ring‐like structures, was observed upon glycation. Similar to cytochrome c, these aggregation pathways seem to occur with only slight changes on protein structure upon glycation.

The concluding remarks in chapter V provide an integrative framework of the findings presented in this thesis. The relevance of this work and perspectives for further investigations are also highlighted in this chapter.

 xiv

Resumo

Resumo

As proteínas desempenham um papel fundamental de suporte para todos os processos químicos que ocorrem numa célula, sendo essenciais à vida. Para executar as suas funções, as proteínas possuem uma estrutura tridimensional específica que é obtida no processo de folding.

Uma perturbação deste processo leva ao aparecimento de uma estrutura diferente, levando a alterações metabólicas e fisiológicas. Ao longo da evolução, a célula desenvolveu chaperones moleculares e um complexo sistema de controlo de qualidade de proteínas com o objectivo de manter as estruturas proteicas funcionais durante o seu tempo de vida. Quando algo falha neste processo, as proteínas podem sofrer um misfold parcial e agregar levando ao aparecimento de patologias conformacionais como é o caso das doenças de Alzheimer, Parkinson, Huntington,

Andrade e Esclerose Amiotrófica Lateral. Vários factores podem contribuir para o misfold parcial de proteínas, incluindo mutações pontuais, stress químico, modificações pós‐traducionais e danos no sistema de controlo de qualidade das proteínas.

As modificações pós‐traducionais têm um papel muito importante no misfolding de proteínas, uma vez que as estas são extensivamente modificadas para adquirirem a sua estrutura tridimensional, regularem a sua actividade e serem marcadas para degradação. Geralmente as modificações pós‐traducionais são controladas por enzimas que modificam as proteínas em locais específicos de forma a produzirem um determinado efeito, como é o caso da fosforilação ou glicosilação, mas a glicação é um processo não‐enzimático. Neste caso as cadeias laterais dos resíduos de lisina e arginina são irreversivelmente modificados por moléculas que contenham grupos carbonilo. In vivo, existem várias moléculas com carbonilos capazes de modificar irreversivelmente proteínas através da reacção de Maillard, como é o caso da glucose ou do metilglioxal. Dada a influência que as modificações pós‐traducionais têm na estrutura nativa de uma proteína, é espectável que uma extensiva e não regulada modificação de determinadas proteínas possa culminar em alterações nocivas da estrutura e função dessas proteínas, que possam estar associadas a lesões em células e tecidos que são observadas em algumas patologias. Em doenças como a diabetes mellitus, aterosclerose e patologias conformacionais é possível observar a acumulação de proteínas modificadas por AGE, o que estabelece uma ligação entre estas doenças e a glicação de proteínas. O metilglioxal é o mais importante agente glicante in vivo e a sua síntese ocorre em todas as células vivas de forma não controlada, maioritariamente como um subproduto da via glicolítica. Ainda assim, e apesar de todas a observações que ligam a glicação de proteínas

Resumo 

com algumas patologias humanas, os mecanismos moleculares subjacentes aos efeitos da glicação de proteínas pelo metilglioxal mantêm‐se desconhecidos.

O principal objectivo da investigação apresentada nesta tese encontra‐se focado na compreensão dos efeitos da glicação na estrutura, estabilidade e agregação de proteínas. O capítulo

I introduz e descreve o tema. Após uma breve apresentação da estrutura de proteínas e do seu processo de folding, a importância biológica do misfolding de proteínas em contexto de patologia é discutida através de uma revisão da relação entre os níveis hierárquicos de organização estrutural de proteínas e a formação de amilóide. Este capítulo introduz ainda a glicação de proteínas, o metabolismo do metilglioxal e as relações entre glicação e diversas patologias humanas.

No capítulo II, são estudados os efeitos da glicação pelo metilglioxal no citocromo c. Os resultados mostram uma diminuição substancial da estabilidade conformacional da proteína após glicação, o que potencia a formação de agregados que mantêm uma estrutura idêntica à nativa. Este capítulo envolve pela primeira vez a glicação de proteínas pelo metilglioxal em fenómenos de agregação de proteínas que mantêm uma estrutura idêntica ao estado nativo. Paralelamente, foi desenvolvido um novo método para a determinação da estabilidade conformacional de proteínas hémicas com modelos de unfolding de dois estados, num processo que se encontra descrito em anexo.

O capítulo III apresenta uma comparação entre os efeitos da glicação in vitro e in vivo no enolase, que foi previamente demonstrado ser um alvo preferencial de glicação em levedura. Os resultados mostram que a glicação causa alterações estruturais e unfolding no enolase, o que leva à perda de actividade enzimática. As diferentes metodologias usadas para a glicação do enolase revelam a existência de diferenças ao nível da natureza e localização molecular dos produtos avançados de glicação. Os resultados obtidos para as alterações estruturais, de Tm e actividade enzimática mostram um efeito comparável entre as duas metodologias de glicação utilizadas.

As experiências descritas no capítulo IV usam duas proteínas amiloidogénicas como alvos de glicação. Quer a α–sinucleína quer a insulina são proteínas envolvidas em patologias humanas específicas. Apesar de possuírem estruturas distintas e tipos de fold diferentes, em ambos os casos a formação de fibras amilóides por partes destas proteínas foi reduzida na presença de metilglioxal.

Em vez de fibras amilóides, a agregação foi direccionada para a formação de agregados solúveis.

No caso da α–sinucleína, foi observado, após glicação, um aumento da concentração de protofibras, algumas contendo estruturas anelares. Tal como foi observado no citocromo c, estas vias de agregação parecem ocorrer apenas a partir de ligeiras alterações na estrutura.  xvi

Resumo

O capítulo V fornece um quadro integrador dos diversos resultados apresentados nesta tese, assim como realça a importância deste trabalho.

Os resultados e conclusões apresentados são relevantes não apenas porque este tipo de estudos sistemáticos aumenta o nosso conhecimento sobre os efeitos bioquímicos e estruturais da glicação em diferentes proteínas, mas também porque foi obtida informação importante que sugere que a glicação possui um papel dinâmico na estabilização de agregados oligoméricos de proteínas relevantes em contexto de patologia, com diferentes tipos de fold, o que poderá estar relacionado com um aumento da toxicidade destas proteínas. A completa compreensão dos efeitos da glicação na estrutura, estabilidade e agregação de proteínas amiloidogénicas será certamente de importância vital no desenho e criação de estratégias terapêuticas novas ou melhoradas para inibir a glicação de proteínas e combater os seus efeitos prejudiciais

Resumo 

 xviii

Abbreviations

Abbreviations

Aβ – Amyloid β‐peptide

AD – Alzheimer’s disease

AGE – Advanced glycation end‐products

Ala – Alanine

ALS – Amyotrophic lateral sclerosis

ANS – 1‐anilino‐8‐naphthalene sulfonate

Arg ‐ Arginine

CD – Circular Dichroism

CEL – Nε‐(carboxyethyl)lysine

Cys – Cysteine

DEAE – Diethylaminoethyl

DHAP – Dihydroxyacetone phosphate

DHB – 2,5‐Dihydroxibenzoic acid

DLB – Dementia with Lewy bodies

DNA – Deoxyribonucleic acid

EDTA – Ethylenediamine tetraacetic acid

ESI‐FTMS – Electrospray ionization – Fourier transformed mass spectrometry fALS – Familial amyotrophic lateral sclerosis

FAP – Familial amyloid polyneuropathies

FTICR‐MS – Fourier‐transform ion cyclotron resonance mass spectrometry

GAP – Glyceraldehyde‐3‐phosphate

GdnHCl – Guanidinium hydrochloride

Gln – Glutamine

GLO1 – Yeast glyoxalase I gene

Glu – Glutamate

Gly ‐ Glycine

HPLC – High performance liquid chromatography

HSD – Hallervorden‐Spatz disease

 xix

Abbreviations 

Hsp – Heat shock protein

Ile – Isoleucine

IR – Infrared

IUP – Intrinsically unstructured protein

LB – Lewy bodies

Leu ‐ Leucine

Lys ‐ Lysine

MAGE – Methylglyoxal‐derived advanced glycation end‐products

MALDI‐TOF‐MS – Matrix‐assisted laser‐desorption ionization–time‐of‐flight mass spectrometry)

MAPK – Mitogen‐activated protein kinase

MG–H – Hydroimidazolone

MODIC – Lysine‐arginine methylglyoxal‐derived cross‐link

MOLD – Methylglyoxal‐lysine dimmers

MSA – Multiple system atrophy

NAC – non‐Aβ component of Alzheimer’s disease

NACP – non‐Aβ component of Alzheimer’s disease precursor

NAD – Nicotinamide adenine dinucleotide (oxidized)

NADPH – Nicotinamide adenine dinucleotide phosphate (reduced)

NMR – Nuclear magnetic resonance

NRMSD – Normalized root mean square deviation

NSF – N‐ethylmaleimide‐sensitive factor

PAGE – Polyacrilamide gel electrophoresis

PD – Parkinson’s disease

PDB – Protein Data Bank

PGPH – Peptidyl glutamyl peptide hydrolase

Phe – Phenylalanine

PMSF – Phenylmethylsulfonyl fluoride

Pro – Proline

PTM – Post‐translational modification

PVDF – Polyvinylidene difluoride

RAGE – Advanced glycation end‐products receptor

RNA – Ribonucleic acid  xx

Abbreviations

ROS – Reactive oxygen species

SD – Standard deviation

SDS – Sodium dodecyl sulphate

SEC – Size exclusion chromatography

Ser – Serine

SNARE – Soluble NSF Attachment Protein Receptor

SOD – Superoxide dismutase

SOD1 – Superoxide dismutase‐1

SSA – Senile systemic amyloidosis

SUMO – Small ubiquitin‐like modifiers

TBS – Tris‐buffer saline

TEM – Transmission electron microscopy

THP – Tetrahydropyrimidine

ThT – Thioflavin T

TPI – Triose phosphate isomerase

Trp – Tryptophan

Tyr ‐ Tyrosine

TTR – Transthyretin

UCH‐L1 – Ubiquitin carboxyl‐terminal hydrolase‐L1

UPS – Ubiquitin‐proteasome system

UV – Ultra‐violet

Val – Valine

YPGlu – Yeast extract, peptone and D‐glucose growth medium

α‐CHCA – α‐Cyano‐4‐hydroxicinamic acid

ΔGº(H2O) – Protein conformational stability

 xxi

Abbreviations 

 xxii

Chapter I

Introduction

The following publication is included in this chapter:

Alexandre Quintas, Luís M. A. Oliveira. 2008. Protein hierarchy levels of structural organization for amyloidogenesis: Biochemical and biophysical aspects underlying misfolding and disease. Protein Misfolding in Biology and Disease T. F. Outeiro, Research Signpost: 1‐34.

Chapter I

1. Protein structure

Proteins are fundamental molecules in living organisms. Their structure and functions are both essential in every process within living cells. They act as highly specific catalysts for the multitude of metabolic processes, as carriers of molecules and electrons, as signal transmitters, they provide mechanical support and immune protection, generate movement and control growth and differentiation.

Proteins are the most abundant macromolecules in living cells, representing more than half of its dry weight. For that reason, in 1839, the chemist Gerardus Mulder first employed the term

“proteins” from the Greek proteios, meaning “of primary importance” on a publication (Mulder,

1839) as coined by Berzelius. At that time, the properties of those macromolecules were totally unknown since their structure and composition remained a mystery. Still, the name is very adequate, since proteins are truly essential for all living organisms.

In spite of this multitude of roles, all proteins are made of the same 22 L‐amino acids

(including selenocysteine and pyrrolysine). Although proline (Pro) is included in this list, it is in fact an iminoacid, because its side chain closes back on the amino group. The diversity comes from the fact that proteins are linear polymers of these 22 amino acids, with the length and sequence uniquely identifying each protein species. The combinatorial possibilities are amazing: for a 150‐ residue long protein (an average size protein) there are 22150 different possible sequences. Each possible sequence has the ability to fold into a three dimensional well defined structure, and the required information to do it is encoded in the sequence itself, as it has been well established by classic experiments (Anfinsen et al., 1961). Understanding how the sequence gives rise to the three dimensional structure is solving the so called “folding problem”, probably the most fundamental question in protein science, and certainly a very challenging one. Solving it would allow us to predict the structure of a protein from its sequence alone, and given the ease with which we can now engineer new gene sequences this would have enormous consequences in areas of medical sciences and biotechnology. The solution for the folding problem is still ahead of us, but it certainly depends on our ability to correctly describe the nature and magnitude of the interactions between atoms in the protein and the solvent.

 3

Introduction 

Every amino acid that composes a protein is an L‐α‐amino acid, containing both amine and carboxyl functional groups attached to the same carbon, which is called the α‐carbon. Each amino acid varies from one another by its side chain, which gives to amino acids distinctive chemical properties that can be classified in three different groups: polar, charged and hydrophobic. Amino acids are linked together in proteins by peptide bonds. The peptide bond is very close to planar

(Pauling et al., 1951) (the C‐N torsion angle is +180 degrees and only slight variations of +/‐ 2 degrees are observed, despite some exceptions found in small and cyclic peptides / proteins

(MacArthur & Thornton, 1996)) due to resonance with the C=O in the α‐carbon, which reduces the geometry around the bound to only two conformers: trans and cis. In the trans conformation, the two α‐carbon atoms are on the opposite sides of the peptide bond. In the cis conformation, these groups are on the same side of the peptide bond. Almost all peptide bonds in proteins are trans because the steric hindrance of amino acid side chains clearly penalizes the cis conformation in comparison to the trans conformation (Stewart et al., 1990; Jabs et al., 1999) (the difference in free energy of both conformers is 2.5 kcal.mol‐1 (Radzicka et al., 1988)). Only amide bonds followed by a proline residue present 5‐6.5 % of cis conformation (Stewart et al., 1990; MacArthur & Thornton,

1991; Reimer et al., 1998; Pal & Chakrabarti, 1999) (the difference in free energy of both conformers is only 0.5 kcal.mol‐1 (Maigret et al., 1970)). However, fifty per cent of all determined protein structures contain at least one cis peptide bond (Pal & Chakrabarti, 1999). A trans  cis isomerization of the concerned bonds is needed to achieve the native state of the protein and has often been found as the rate limiting step in in vitro protein folding (Brandts & Lin, 1986; Kim &

Baldwin, 1990).

Contrary to amide bonds, the phi (Φ) (between N‐Cα) and psi (Ψ) (between C‐Cα) torsion angles are relatively flexible. Nevertheless, because of steric interferences between atoms, only some possible combinations of Φ and Ψ angles are allowed. This was first analysed by

Ramachandran (Ramachandran, 1963), who used solid sphere models of amino acids to determine the range of allowed values for Φ and Ψ.

By definition, there are four levels of protein structure. The primary structure of a protein corresponds to the linear sequence of the amino acid residues in the polypeptide chain. The position of each amino acid in the sequence will determine how the chain will fold and adopt its structure. The secondary structure is the local conformation of a polypeptide chain and corresponds to the first level of organization of the polypeptide chain in space. It results from the tendency of each amino acid residue to adopt defined combinations of dihedral angles Φ and Ψ and also from  4

Chapter I the establishment of hydrogen‐bonding patterns between the carbonyls and amide protons along the chain. The tertiary structure refers to the spatial arrangement of the secondary structure units and it is the complete three‐dimensional structure of a polypeptide chain. Proteins that have association of different polypeptide chains, exhibit a fourth structural level called the quaternary structure.

While primary and quaternary structures have a clearly non‐ambiguous meaning, there is a certain amount of overlap between the definitions of secondary structure and tertiary structure, hence, a more in depth analysis is required. The term secondary structure is clearly more appropriated when discussing repetitive patterns of the local structures, like helices or sheets. Non‐ repetitive backbone conformations are often called coil or random coil. The later term is not accurate since the absence of repetitive patterns does not mean disorder, as “random” seems to imply. Atoms in coil structure have fixed well determined locations in space, not necessarily of higher mobility than regular regions. Disordered regions do occur often in experimentally determined protein structures, indicating multiple allowed conformations for that region.

1.1 Secondary structure

The repetitive patterns of local polypeptide chain conformation are one of the most notable features of protein structure. This regularity of the backbone conformation results in an approximate constancy of the main chain (Φ, Ψ) torsion angles, and the two main types of repetitive conformation match the two densely populated regions of the Ramachandran plot, approximately centered around (‐60, ‐50) and (‐120, +140) (Ramachandran, 1963). The origin of repetitive patterns is partially explained by the favourable low‐energy conformation adopted by the backbone due to hydrogen bonding and van der Waals packing interactions that play an important role in stabilizing these structures and defining their shape.

1.1.1 Helical structures

The right‐handed α‐helix, first proposed by Linus Pauling (Pauling et al., 1951), has been called the “default protein conformation” (Richardson & Richardson, 1989) due to its great stability and widespread occurrence. It is in fact the most commonly occurring type of repetitive secondary structure. There are several reasons for the high stability of the α‐helix: the characteristic torsion angles (‐60, ‐50) set the linear polypeptide chain in a minimal Gibbs energy, the –CO group of  5

Introduction 

residue n is hydrogen‐bonded to the –NH group of residue n + 4, and the tight structure of the helix allows for van der Waals interactions between neighbouring atoms, which add to helix stabilization.

The α‐helix has a distance between successive residues of 1.5 Ǻ and 3.6 residues per turn. The number of backbone atoms in a closed loop formed by two hydrogen bonded groups is 14 and therefore the α‐helix can also be called 3.614 helix (Quintas et al., 2008). The left‐handed form of α‐ helix, with angles (+60, +60) is energetically less favourable than its right‐handed counterpart, because the orientation of the side chains creates clashes with the backbone carbonyl groups, since all amino acids have the L‐configuration (except glycine) (Richardson & Richardson, 1989).

Another type of helix that can be observed in proteins is the 310 helix. As the name indicates, it contains three residues per turn and 10 backbone atoms in one hydrogen bonded loop, which corresponds to a hydrogen bond between residues n and n + 3. Bad side chain packing and poor hydrogen bond geometry (dipoles not aligned) explain why this type of helix is so rare when compared to α‐helix. The longest stretches known are 6 residues long (two turns) (Richardson &

Richardson, 1989).

1.1.2 Beta structures

Another major pattern of repetitive secondary structure is the β‐sheet structure. In contrast with α‐helix, where hydrogen bonds are formed between residues close together in the helix, strands of polypeptide chain in β conformation form hydrogen bonds with one or more neighbour strands to form a relatively flat, broad structure called β‐sheet. The low energy region of the

Ramachandran plot for β‐structures is rather shallow and large, resulting in a much wider dispersion of (Φ, Ψ) values than in the case of α‐helix. When the strands run in the same direction, the β‐sheet is said to be parallel. In antiparallel sheets the strands run in opposite directions. The parallel sheets can have only two strands, while antiparallel sheets usually have five or more strands. These and other facts point for a higher stability of antiparallel β‐sheets (Richardson &

Richardson, 1989). Mixed sheets with parallel and antiparallel strands are uncommon, probably because the optimal backbone torsion angles for the two types of conformation are different. β‐ strands usually have a pronounced right‐handed twist due to steric effects arising from the L‐amino acid configuration.

 6

Chapter I

When the polypeptide chain reverses its direction, turns are always present. Particular common are β‐turns that connect the ends of two adjacent segments of an antiparallel β‐sheet. The structure is a 180º turn involving four amino acid residues, with the carbonyl oxygen of the first residue forming a hydrogen bond with the amino‐group hydrogen of the fourth. Glycine and proline residues often occur in β‐turns because they are small and flexible to allow such torsion angles. β‐turns are often found near the surface of a protein, where the peptide groups of the central two amino acid residues in the turn can hydrogen‐bond with water.

1.1.3 Non­repetitive secondary structure

Besides helical and β‐structures, proteins have regions of aperiodic conformation usually referred as coil or random coil. Non‐repetitive secondary structure can be classified in two main types: connecting straps and compact loops. The connecting straps are extended pieces of polypeptide chain connecting repetitive elements of secondary structure, while compact loops usually have close curvatures and good side‐chain packing interactions. Side chain interactions are very important in the description of these structures, much more than periodic structures where main chain hydrogen bonding has the greatest share among the stabilizing interactions. An analysis of the set of known structures has showed that side chains have tendency to assume specific φ and

ψ angles, which are characteristic for each of the 22 amino acid side chains (Ponder & Richards,

1987) and several groups have compiled these values (Tuffery et al., 1991; Schrauber et al., 1993; De

Maeyer et al., 1997; Dunbrack & Cohen, 1997; Tuffery et al., 1997; Lovell et al., 2000). It has also been shown that the side chain hydrogen potential is roughly saturated for globular proteins, meaning that practically all potential acceptors and donors are taking part in bonds (McDonald & Thornton,

1994). These facts together, show that the “random” coil structures are far from random.

1.2 Tertiary structure

As previously defined, the term tertiary structure describes the overall topology of the folded protein chain. A folded protein is a complex object and only high resolution techniques allowed systematic studies of the protein structures that has led to the identification of general patterns of folded globular proteins that allows its classification in classes according to their topology.

The existence of repetitive secondary structure is the first simplifying principle, effectively reducing the number of variables needed to describe the outline of the protein backbone. Folded  7

Introduction 

proteins can be represented in the so called Richardson diagram (Richardson, 1985), where the backbone trace is smoothed and the secondary structure elements highlighted (Figure I.1).

Figure I.1 – Richardson diagram of the crystal structure of porcine alpha trypsin (PDBid: 1AKS), featuring the elements of secondary structure.

This type of representation is far from a complete description of the tertiary structure – side chains are missing and detailed aspects of backbone conformation have been artificially removed.

However, Richardson diagrams tell us a lot about the protein’s architecture, making the arrangement and connectivity of the secondary structure elements much more apparent than an atomic detailed representation. Figure I.1 shows how the molecule is made of two halves connected by straps of polypeptide backbone. The halves are rather compact and appear like independent structural units. Most proteins with sizes larger than 200 residues seem to be constituted by an assembly of more or less independent structural units, called domains (Gerstein, 1997), which, in most of the cases, can fold into a stable structure even when free from interactions with neighbour domains.

Some assemblies of secondary structure elements occur so often that can be seen as building blocks in their own right and are called super secondary structure elements. Examples are β‐α‐β assembly, where two parallel β‐strands are connected by segment of polypeptide chain containing an α‐helix. Another example is the Greek key, which is an assemblage of 4 anti‐parallel β‐strands with a specific connectivity.

Domains can be grouped into specific categories according to the secondary structure content. Following one of the grouping systems proposed (Levitt & Chothia, 1976), proteins can be arranged into 3 main classes: α, β and α/β. The α class contains α‐helix only proteins, the β‐class contains β‐sheet only proteins and the α/β class contains proteins in which these two types of structure occur together. The remaining elements, mostly small proteins with irregular secondary structure and high content of disulphide bridges and / or metal ions, comprise a group of its own.  8

Chapter I

Lately, a fifth group came out, called the α + β class, containing domains in which α and β structures occur, but they are well defined and separated from each other.

Within each of the classes, groups of domains displaying a similar arrangement of the structural building bocks can be identified. These arrangements are called folds. Proteins with the same fold produce in many cases nearly identical schematic diagrams, where only loops or straps show a different structure (Richardson & Richardson, 1989; Richardson et al., 1992; Thornton et al.,

1999). The total number of folds for biological proteins seems rather limited. Based in theoretical models, Govindarajan and co‐workers gave an estimate of 4000 for the total number of possible folds occurring in nature (Govindarajan et al., 1999). In spite of the limited number of folds, there are other variables that can add complexity to the system. For example, although protein folding domains are generally conserved for function across distant homologous sequences, the wrapping of backbone hydrogen bonds is not conserved. The packaging of certain conserved folds becomes progressively poorer as species diverge in some lineages (Fernandez et al., 2004a). These and other factors may elucidate the existence of chameleon sequences that can adopt several alternative three‐ dimensional configurations (Andreeva & Murzin, 2006). It is clear that the relationship between the evolution of the function of proteins with respect to the evolution of their sequence and structure is complex (Whisstock & Lesk, 2003). Very close homologues can differ in function, and there are many known examples of a given protein being reused for entirely unrelated functions, depending on the context in which they are found, in a phenomenon known as “recruitment” or

“moonlighting” (Jeffery, 2003; Sangar et al., 2007). A classic example of this is that of the duck eye lens crystallins, which are identical in sequence to liver enolase and lactate dehydrogenase

(Piatigorsky et al., 1994).

A discussion of protein tertiary structure would be incomplete without mentioning the non‐ amino acidic molecules or ions that can be an integral part of a folded protein and often necessary for the maintenance of structure and / or biological activity. The bound molecules are called prosthetic groups – one good example is the heme group in haemoglobin. Metal ions can also associate with proteins in a permanent manner and play a structural and functional role. The Ca2+ in trypsin and the Cu2+ in superoxide dismutase are examples of structural ions. Small proteins very often contain structural ions, which may add additional stability to an otherwise labile structure

(Tainer et al., 1992).

 9

Introduction 

1.3 Protein dynamics and flexibility

Although proteins have a well defined structure, they are not static molecules. Proteins are dynamic entities and possess an inherent flexibility that allows them to function through molecular interactions within the cell, among cells and even between organisms. The dynamics in a protein allow its conformation to change and respond to the presence of other molecules and / or to variations in the environment (Hvidt, 1955; Kim et al., 1993; Kim & Woodward, 1993; Mayo &

Baldwin, 1993; Bai et al., 1994). This ability to change is of primary importance in several biochemical processes such as enzyme catalysis, signal transduction, antigen recognition and protein transport.

At this point, it is important to distinguish dynamics from flexibility. The term dynamics is used for intrinsic protein molecular motions, while the term flexibility is used for the ability of a protein to adapt its structure to external stimuli. Accordingly, proteins are flexible as a consequence of their dynamics, yet their dynamics do not automatically result in flexibility (Teilum et al., 2009).

Every time a protein molecule changes its conformation, some interactions are broken and others are formed. Therefore there are several thermodynamic constraints to protein flexibility. The enthalpy and entropy contributions to the total free energy variation in a protein conformational change almost invariably counterattack each other. When a new non‐covalent interaction is formed it will decrease the enthalpy of the complex. Around the binding interface, the system will lose conformational freedom, and the entropy of the molecules will also decrease. The more interactions formed, the more ordered the complex becomes. Indeed, a linear relationship between ΔH° and ΔS° has been found for a number of protein‐ligand interactions (Gilli et al., 1994). Consequently, large values of ΔH° and ‐TΔS° for processes involving non‐covalent interactions may result in only modest values of ΔG°. This is exemplified by the interaction between HIV‐gp120 and CD4 for which ΔG° ≈ ‐12 kcal.mol‐1 (Myszka et al., 2000). This value is the result of very large and opposing

ΔH° and ‐TΔS° values of ‐63 and 51 kcal.mol‐1, respectively (Myszka et al., 2000). These evidences point to the fact that conformational changes with a significant three‐dimensional structure modification can be easily achievable in a thermodynamic point of view.

Many other factors are important to protein dynamics and flexibility, such as interaction with water molecules, ions, the presence of post‐translational modifications, slight pH and temperature variations, etc, but a complete discussion of this subject is beyond the scope of this work. Still, to fully understand the protein function, dynamics and flexibility should be considered as a part of a

 10

Chapter I protein structure, possibly as an extra structural dimension, encoded in the amino acid sequence of a protein, just like secondary, tertiary and quaternary structural levels (Teilum et al., 2009).

1.4 Natively unfolded proteins

Natively unfolded or intrinsically unstructured proteins (IUPs) constitute a very special group of the protein kingdom. The evolutionary persistence of such proteins represents strong evidence in the favour of their importance and raises intriguing questions about the role of protein disorders in biological processes.

The term “natively unfolded” was introduced in 1994 to describe the behaviour of the tau protein (Schweers et al., 1994) and refers to an extremely flexible protein, essentially non‐compact, lacking globularity and with little or no ordered secondary structure under physiological conditions (Uversky, 2002b). In vivo NMR measurements indicate that even in the crowded environment of the cell, certain IUPs remain partially or fully unstructured (McNulty et al., 2006).

IUPs can be divided into two structurally different groups: intrinsic coils and pre molten globules

(Uversky, 2002a). Proteins from the first group have hydrodynamic dimensions typical of random coils and do not possess any ordered secondary structure. Proteins from the second group are essentially more compact, exhibiting some amount of residual secondary structure, although they are still less dense than native or molten globule proteins (Uversky, 2002a). The existence of biological active but extremely flexible proteins, questions the assumption that rigid well‐folded three‐dimensional structures are required for functioning. To overcome this problem, it has been suggested that the lack of rigid globular structure under physiological conditions might represent a considerable functional advantage for “natively unfolded” proteins, as their large plasticity allows them to interact efficiently with several different targets (Wright & Dyson, 1999; Dunker et al.,

2001).

Since the necessary information for the correct folding of a regular protein into its biologically active conformation is encoded in the amino‐acid sequence, special sequence features in IUPs are expected to be found. Some of the sequence peculiarities of these proteins include the presence of numerous uncompensated charged groups, i.e. a large net charge at neutral pH, arising from the extreme pI values in such proteins (Hemmings et al., 1984; Gast et al., 1995; Weinreb et al., 1996), and a low content of hydrophobic amino acid residues (Hemmings et al., 1984; Gast et al., 1995). The lack of hydrophobic residues prevents the formation of a hydrophobic core necessary for a stable three‐

 11

Introduction 

dimensional fold and the large number of side chain charges, present under physiological conditions, contributes to destabilize any compact state. Dunker and collaborators developed several neuronal network predictors to understand the relationship between sequence and disorder, and the results showed that disorder regions share some common sequence features over many proteins (Romero et al., 1997; Dunker et al., 1998; Garner et al., 1998; Romero et al., 1998; Li et al., 1999; Li et al., 2000; Dunker et al., 2001; Romero et al., 2001). This includes sequences enriched in

Ala, Glu, Gly, Lys, Pro, Gln, Arg and Ser and with low contents of Cys, Phe, Ile, Leu, Val, Trp and

Tyr (Dunker et al., 2001). These studies allowed the development of a variety of computer programs that predict unstructured regions from amino acid sequence. An overview of protein disorder prediction methods and a list of IUP predictors can be found in the work of Ferron and collaborators (Ferron et al., 2006).

IUPs have been described as proteins with functional plasticity. In fact, many IUPs adopt a defined three‐dimensional structure upon binding to their related partners (Wright & Dyson, 1999;

Demarest et al., 2002; Dyson & Wright, 2005). This need for interaction to fold was first reported by

Frankel (Frankel & Kim, 1991). IUPs that fold upon binding can be found among RNA binding proteins, where they play a key role in RNA recognition (Battiste et al., 1996; Mogridge et al., 1998), in vesicle trafficking and membrane fusion (Fiebig et al., 1999; Owen et al., 2004; Praefcke et al., 2004;

Schmid et al., 2006; Schmid & McMahon, 2007; Olesen et al., 2008), cell cycle regulation (Kriwacki et al., 1996) and as an integrative part of many different signalling pathways (Iakoucheva et al., 2002;

Bhattacharyya et al., 2006; Chu et al., 2007; Mathes et al., 2008; Seldeen et al., 2008). IUPs have also been described as intimately related with protein moonlighting, where its high flexibility can provide an unprecedented versatility in ligand binding that enables IUPs to have distinct functions

(Tompa et al., 2005).

Regarding the involvement of IUPs in key cellular processes and that up to 30 % of eukaryotic proteins are IUPs, or have extended disorder regions (Dunker et al., 2000; Ward et al.,

2004; Dunker et al., 2008), it was described that IUPs play central roles in mapped protein‐protein interaction networks of several different eukaryotes (Uetz et al., 2000; Ito et al., 2001; Giot et al., 2003;

Li et al., 2004; Barrios‐Rodiles et al., 2005; Rual et al., 2005; Stelzl et al., 2005; Gavin et al., 2006; Krogan et al., 2006) acting as hub proteins, which interact with an unusually large number of other partners.

These functions are achievable because of the unique properties of IUPs, like the ability to be easily post‐translational modified. Post‐translational modifications, which are essential for biological complexity and diversity by affecting protein stability, turnover, sub‐cellular localization or  12

Chapter I interaction properties and thereby have a significant impact on protein function, are often found in unstructured regions (Gsponer & Madan Babu, 2009). Together, these aspects, become particular relevant for proteins which participate in regulatory and signalling functions (Pawson & Nash,

2003; Seet et al., 2006). The inherent flexibility of unstructured segments in proteins facilitates binding of different enzymes such as kinases, phophatases, acetyltransferases, deacetylases, methylases, ubiquitin ligases and others to specific post‐translational modification sites. Thereby,

IUPs are likely to have many post‐translational modifications that are used in a combinatorial manner and mediate the specific biological responses (Gsponer & Madan Babu, 2009). Another remarkable property of IUPs is the great “capture radius” and high intermolecular association rates, because they are less compact than structured proteins, in a process known as “fly‐casting mechanism” (Shoemaker et al., 2000; Levy et al., 2004). The flexible nature of IUPs is supposed to facilitate the binding of targets to peptide motifs located in the unstructured segments. Putting all this information together, IUPs have been suggested to confer greater response variability to the cell because of the broader repertoire of structural and functional states, comparing to structured proteins. Moreover, transitions between these functional states are fast due to the unique properties of these proteins, which ultimately allow a cell to react in a much better way to intra‐ and extra‐ cellular perturbations (Gsponer & Madan Babu, 2009).

Another important topic to be discussed is how unfolded proteins are kept in the cell, successfully avoiding the protein degradation machinery at least long enough time to perform their cellular functions. Although it is intuitive that polypeptide disorder is associated with proteolytic susceptibility, a recent study of more than 3000 yeast proteins showed that protein disorder is a poor predictor of the in vivo rate of protein turnover, as protein degradation in vivo is highly regulated and influenced by many other factors (Tompa et al., 2008). Even though, the physical properties of disordered polypeptide segments allow proteins to be extensively regulated by post‐ translational modifications and that can provide the opportunity for a rapid turnover. Notably, this expected reduced lifetime may be a functional advantage, since it may constitute a component of regulation of these proteins, considering the vital roles that many IUPs play in cell cycle regulation and in transcriptional and translational processes.

 13

Introduction 

2. The protein folding problem

How a protein, after its synthesis on ribosome, will fold and adopt its native three‐ dimensional structure is currently not completely understood and a central problem of modern biophysics. Christian B. Anfinsen and Cyrus Levinthal may be considered as the historical forefathers of in vitro protein folding studies with two fundamental findings. Anfinsen first observed that all the information needed for a protein to fold may be found in its amino acid sequence (Anfinsen et al., 1961), while Levinthal pointed out that folding cannot take place via random sampling of all possible conformations (Levinthal, 1968). Thus, the traditional view of the protein folding reaction was based on the existence of a preferred route driving the denatured state to its native conformation via a sequence of consecutive intermediates, whose structural features are encoded by the primary structure. However, novel theoretical and experimental approaches have contributed to challenge this mechanism, introducing more complex and detailed models, which capture the finest details of the folding reaction (Daggett & Fersht, 2003).

2.1 The folding code

Before the mid‐1980s, the predominant view was that the protein folding code is the sum of many different small interactions (hydrogen bonds, electrostatic interactions, van der Waals interactions and hydrophobic interactions), mainly expressed through secondary structures and mainly local in the sequence (Anfinsen & Scheraga, 1975). A key idea was that primary sequence encoded secondary structures, which then encoded tertiary structures (Anfinsen & Scheraga, 1975).

However, through statistical mechanical modelling, a different view has emerged in the late 1980s claiming that there is a dominant component to the folding code – the hydrophobic interactions – and that the folding code is distributed both locally and non‐locally in the sequence (Dill, 1999).

Based on this point of view, native secondary structures are more a consequence than a cause of folding forces (Dill, 1990; Dill, 1999).

Because native proteins are only 5‐10 kcal / mol more stable than their denatured states, it is clear that no type of intermolecular force can be neglected in folding and structure prediction (Yang et al., 2007). Still it remains challenging to realize whether or not there is one dominant driving force in protein folding. Folding is not likely to be dominated by electrostatic interactions among charged side chains because most proteins have relatively few charged residues and they are concentrated in high‐dielectric regions on the protein surface. Charge mutations also typically lead to small effects on protein structure and stability. Hydrogen‐bonding interactions are important, because  14

Chapter I virtually all possible hydrogen‐bonding interactions are satisfied in native structures. Hydrogen bonds among backbone amide and carbonyl groups are key components of all secondary structures, and studies of mutations in different solvents estimate their strengths to be around 1‐4 kcal / mol or stronger (Fersht et al., 1985; Byrne et al., 1995; Deechongkit et al., 2004; Auton et al.,

2007). Similarly, tight packing in proteins implies that van der Waals interactions are important

(Chen & Stites, 2001). However, there is considerable evidence that hydrophobic interactions must play a major role in protein folding, suggesting that the folding code must be written in the side chains and not in the backbone because it is through the side chains that one protein differs from another. Protein have hydrophobic cores, implying that non‐polar amino acids are forced to be sequestered from water; model compound studies show that there is necessary 1‐2 kcal / mol for transferring a hydrophobic side chain from oil‐like media into water (Wolfenden, 2007), and proteins have many of them; proteins are readily denatured in non‐polar solvents; and sequences that are mixed up and retain only their correct hydrophobic and polar patterning, fold to their expected native states (Hecht et al., 1990; Kamtekar et al., 1993; Cordes et al., 1996; Kim et al., 1998).

Regarding secondary structures, Linus Pauling and collaborators predicted from hydrogen‐ bonding models that proteins might have α‐helices even before any protein structure was known

(Pauling et al., 1951). However, secondary structures are rarely stable on their own in solution and there are also chameleon sequences in natural proteins that assume either helical or β conformations depending on their tertiary context (Minor & Kim, 1996; Mezei, 1998). Therefore, secondary structures in proteins are substantially stabilized by the chain compactness, an indirect consequence of the hydrophobic interactions, being the only regular ways to pack a linear chain into a tight space.

2.2 The folding process

In 1968, Cyrus Levinthal realized that proteins can search and converge to native states in a microsecond scale even though they have vast possible conformational spaces (Levinthal, 1968).

Based on this observation, it was postulated that a physical mechanism of protein folding must exist. Several models have emerged to address this problem. The first – nucleation‐growth

(Wetlaufer, 1973) – proposed that tertiary structure propagates rapidly from an initial nucleus of local secondary structure. However, nucleation‐growth fell into disuse as it predicts the absence of folding intermediates, and the field of protein folding has been dominated by the study of folding

 15

Introduction 

intermediates (Kim & Baldwin, 1982; Kim & Baldwin, 1990). Two alternative models prevailed, the first of which was the framework model (Ptitsyn & Rashin, 1975; Kim & Baldwin, 1982; Kim &

Baldwin, 1990) and the related diffusion‐diffusion model (Karplus & Weaver, 1976), in which secondary structure is proposed to fold first, followed by docking of the pre‐formed secondary structural units to yield the native protein. The second was the hydrophobic collapse model

(Schellman, 1955; Kauzmann, 1959; Baldwin, 1989), in which hydrophobic collapse drives compaction of the protein so that folding can take place in a confined volume, thereby narrowing the conformational search to the native state.

However, the discovery that proteins could fold by simple two‐state kinetics, without the accumulation of folding intermediates (Jackson & Fersht, 1991) and a Φ‐value analysis of the transition state of chymotrypsin inhibitor 2, which showed that secondary and tertiary structures are formed in parallel as the protein undergoes a general collapse (Otzen et al., 1994), motivated the arising of a new unifying folding mechanism: the nucleation‐condensation (or nucleation‐collapse)

(Fersht, 1995; Itzhaki et al., 1995; Fersht, 1997) that unites features of the framework and hydrophobic collapse mechanism. Nucleation‐condensation involves the formation of long range and other native hydrophobic interactions in the transition state to stabilize the otherwise weak secondary structure. The transition state looks like a distorted form of the native structure, with the least distorted part being defined as the folding nucleus (Daggett, 2002). Some general aspects of the folding process, as postulated by the nucleation‐condensation model, are well reproduced by statistical models like funnel landscapes (discussed bellow) (Onuchic et al., 1996; Clementi et al.,

2000), still whether any or all of the classical mechanisms occur in general, and whether there is an unifying mechanism for protein folding remains unclear.

Our understanding of folding mechanisms has also been advanced by theory and simulation.

These approaches led us to recognise that folding does not involve a single microscopic pathway, but rather funnel‐shaped energy landscapes (Dill, 1985; Leopold et al., 1992; Dill & Chan, 1997), which are the mathematical functions that describe the intramolecular‐plus‐solvation free energy of a given protein as a function of the microscopic degrees of freedom. Folding processes are microscopically heterogeneous and thus are not readily probed by classical experiments because traditional experiments “measure” only average quantities. Funnels can explain several experimental observations that are otherwise paradoxical when interpreted in more classic ways.

The plasticity of folding pathways implicit in the landscape theory assumes that proteins can be rerouted through the energy landscape by mutational (Wright et al., 2003), topological (Lindberg &  16

Chapter I

Oliveberg, 2007) or solvent perturbations (Otzen & Fersht, 1998; Gianni et al., 2007) and the trajectory taken by a protein molecule is determined by thermodynamic probabilities. This explains how a protein can be kinetically trapped in a partial unfolded intermediate, which is in the base of multiple degenerative diseases.

In agreement with the statistical landscape theory, several proteins have been suggested to fold via parallel routes, such as lysozyme (Kiefhaber, 1995), some c‐type cytochromes (Gianni et al.,

2003) and the 27th immunoglobulin domain of titin (Wright et al., 2003). Also, molecular dynamic simulations have proposed the existence of parallel folding pathways for a variety of small protein domains (Lazaridis & Karplus, 1997; Borreguero et al., 2004; Caflisch, 2004; Rao et al., 2005; Juraszek

& Bolhuis, 2006; Lam et al., 2007). The funnel landscapes theory contributed to broaden our knowledge on the finest details of the folding process, but care must be taken in over‐interpreting results from molecular dynamic simulations without experimental validation (Ivarsson et al., 2008).

More recently, a new hypothesis came out claiming that the common characteristics of globular proteins suggest that the sequence alone does not shape the structure (Banavar & Maritan,

2007). In fact, in spite of having different sequences of amino acid residues, all globular proteins fold fast and reproducible driven by the same hydrophobic effect; the number of distinct folds is rather limited (Chothia & Finkelstein, 1990; Chothia, 1992) and several different sequences can adopt the same native fold (Bowie et al., 1990; Heinz et al., 1992; Matthews, 1993); also a point mutation does not lead to radical changes in the native‐state topology (Hecht et al., 1990; Sander &

Schneider, 1991; Davidson & Sauer, 1994; Dahiyat & Mayo, 1997; Wei et al., 2003). Instead, a menu of protein folds results from properties that are shared by all proteins and governed by physical laws of geometry and symmetry (Banavar & Maritan, 2007). These folds have been evolutionarily selected to provide to proteins essential properties like stability, sensitivity, diversity and dynamical accessibility. In this picture, sequences and functionality evolve to fit within the constraints of the pre determined menu of folds. These facts point to the requirement of a pre sculpted energy landscape that is shared by all proteins and has many local minima corresponding to presumed native structures (Banavar & Maritan, 2007). These local minima should not be too few because that would not lead to sufficient diversity and not too many because that would lead to an excessively rugged landscape where it would be difficult for a protein to fold rapidly and reproducibly into its native structure. The notion of a free energy landscape pre sculpted by geometry and symmetry is radically different from the conventional belief that is the sequence that  17

Introduction 

shapes a funnel‐like landscape (Bryngelson et al., 1995; Wolynes et al., 1995; Dill & Chan, 1997). The crucial role of the sequence is not to determine the native state but to choose the native conformation from a fixed, pre determined menu of folds. This hypothesis also gives a simple explanation to amyloid formation. Despite of having different sequences, virtually all proteins show an aggregational tendency in the appropriate solvent conditions showing that the specific sequence of amino acids plays again a secondary role (DuBay et al., 2004). The geometry based tube model predicts (Banavar et al., 2004) that the optimal packing of many short chains is achieved by forming cross‐linked β‐sheet structures often found in amyloid.

In summary, there has been significant progress in understanding the protein folding, either by major advances in experimental protein science with increasingly fast and accurate methods or by robust and fast computer programs. Still, the protein folding remains one of the 125 most important unsolved problems in science, as published by the Science maganize (2005).

3. Protein misfolding in biology and disease

Neurodegenerative diseases and systemic amyloidosis share a common molecular mechanism: conformational changes of soluble forms of peptides and proteins which aggregate into amyloid fibrils. These pathologies are designated conformational or misfolding diseases (Thomas et al., 1995; Kelly, 1996; Harding, 1998; Radford, 2000; Soto, 2001; Stefani & Dobson, 2003) and include other disorders in which, apparently, no amyloid fibrils are formed although the native proteins undergo pathological conformational changes (Ito & Suzuki, 2007; Jaffe & Stith, 2007). Several human diseases are associated with the formation of protein deposit with amyloid‐like characteristics, which can be neurodegenerative diseases, non‐neuropathic systemic amyloidosis or non‐neuropathic localized diseases (Chiti & Dobson, 2006). Regardless of the origin, which can be sporadic, familial or transmissible, they result, in general, from protein misfolding or unfolding

(Gerhartz et al., 1998; Quintas et al., 1999; Quintas et al., 2001; Horwich, 2002; Ma & Nussinov, 2002;

Marchal et al., 2003; Uversky & Fink, 2004). In spite of the efforts of many investigators, the molecular mechanisms of conformational diseases are still poorly understood. Besides the known common features of conformational diseases, it is thought that a vast set of unlinked observations will turn out to be the basis for proposing accurate molecular mechanisms. The major difficulties to investigate biophysical and biochemical mechanisms of misfolding and aggregation of amyloidogenic proteins are both the inherent complexity of the phenomena and the great deal of  18

Chapter I time that takes to monitor amyloid fibril formation under physiological conditions (Quintas et al.,

2001; Myers et al., 2006). Actually, biophysical and biochemical approaches to study the mechanism of amyloid formation almost always requires researchers to go beyond physiological conditions in order to ensure their feasibility. With this purpose in mind, non‐physiological conditions such as low pH values, high temperature, detergents, organic solvents or high urea concentrations are often applied (Lai et al., 1996; Ratnaswamy et al., 1999; Jiang et al., 2001; Ma & Nussinov, 2002; Souillac et al., 2002; Yamamoto et al., 2004; Wei & Shea, 2006; Librizzi et al., 2007). However, such experiments show that most proteins can assemble into amyloid‐like fibrils in vitro under extreme conditions, suggesting that the ability of a soluble protein to form amyloid‐fibrils is a generic property of the polypeptide chain once the right solution conditions are established (Chiti et al., 1999; Fandrich et al., 2001; Stefani, 2004; Trovato et al., 2006). It is surprising that even some all‐alpha‐helical proteins unrelated to the amyloid diseases, such as apo‐cytochrome c or myoglobin, can form fibrils under peculiar non‐physiological conditions (Fandrich et al., 2001; Pertinhez et al., 2001; Wain et al., 2001;

Fandrich et al., 2003). In general, these severe conditions are only populating partially unfolded states of a protein.

Inducing an unfolding stress using non‐physiological conditions to study the amyloidogenic potential of proteins is a step forward into the comprehension of conformational diseases. Yet, a larger step towards the understanding of the biochemical and biophysical aspects that turn a soluble globular protein into toxic aggregates and amyloid fibrils would be the study of the processes under the physiological and the pathological conditions of the disease. Some efforts with this purpose have been made since the late 90’s (Quintas et al., 1997; Quintas et al., 1999) and biophysical experiments carried out under physiological conditions, both in vitro and in vivo, started to be reported (Quintas et al., 2001; Outeiro & Lindquist, 2003; Myers et al., 2006; Outeiro &

Giorgini, 2006).

3.1 Protein structure and amyloidosis

In the 30s, Pauling and collaborators demonstrated that a denatured enzyme loses its activity

(Mirsky & Pauling, 1936) pointing out the protein ordered structure as the key element towards function. Two main observations relating non‐ordered structures with function, molten globule states (Dolgikh et al., 1981; Ohgushi & Wada, 1983) and intrinsically unstructured proteins (Lynch et al., 1987; Weinreb et al., 1996), have questioned the above mentioned paradigm leading to a

 19

Introduction 

scientific revolution (Wright & Dyson, 1999). The re‐assessment of the classical structure‐function paradigm has produced two major proposals namely the protein trinity (Dunker et al., 2001) and the protein quartet (Uversky, 2002a).

The protein trinity proposal suggests that a protein polypeptide chain can have three main structural and functional states: Random Coil; Molten Globule State; and Ordered State. There are three key observations underlying this protein trinity proposal: (i) protein molten globule states are characterized by persistent secondary structure and a lack of tertiary structure along with regions of non‐rigid side chain packing (Dolgikh et al., 1981); (ii) protein molten globule states can exist as stable intermediates which suggest that they can have some biological function (Bychkova et al.,

1988; Bychkova et al., 1992) and (iii) natively unfolded proteins or intrinsically disordered proteins display unique and exquisite functional roles in living organisms (Wright & Dyson, 1999).

According to Dunker and Obradovic the protein trinity proposal supports the view that the three conformational states of a protein can be the native state of the protein (Figure I.2) (Dunker &

Obradovic, 2001).

Figure I.2 ‐ Scheme showing the Protein Trinity model for structure‐function relationship. According to this model, the function arises from the three specific conformations, Random Coil; Molten Globule State; Ordered State, which are considered native states of a protein (Dunker & Obradovic, 2001).

The protein quartet model was proposed by Uversky and states that proteins can exist in any of four alternative functional conformational native states: Random Coil; Pre‐Molten Globule State;

Molten Globule State; and Ordered State (Uversky, 2002a). The main difference to the protein trinity proposal relies on the experimental observation that intrinsically unstructured proteins can be split into two groups: intrinsic coils and intrinsic pre‐molten globules. Additionally, pre‐molten globule represents a different functional state in respect to molten globule state (Figure I.3) (Uversky &

Ptitsyn, 1994).

 20

Chapter I

Figure I.3 ‐ Scheme showing the Protein Quartet model for structure‐function relationship. According to this model, the function arises from the four presented conformations, Random Coil; Pre Molten Globule; Molten Globule State; Ordered State. which are considered native states of a protein (Uversky, 2002a).

3.1.1 Intrinsically unstructured proteins in conformational diseases

The functions of IUPs are linked with the need of structural disorder (Fuxreiter et al., 2004;

Tompa, 2005). In general, the amino acid sequences of IUPs enclose a high proportion of charged residues and a low proportion of hydrophobic amino acids, originating proteins with low overall hydrophobicity and a large net charge. These seem to be crucial for the “disorder” state of natively unfolded proteins (Uversky et al., 2000). From the close to two hundred known proteins and protein domains which lack ordered three‐dimensional structure (Vucetic et al., 2005) at least twelve are associated with human amyloid diseases (Chiti & Dobson, 2006). This number represents nearly half of the proteins associated with neurodegenerative, systemic amyloidosis and non‐neuropathic localized human diseases (Chiti & Dobson, 2006).

The proportion of IUPs, monomeric and oligomeric proteins involved in human diseases associated with amyloid or intracellular inclusions with amyloid‐like characteristics are shown in table I.1. The amyloidogenic propensity of IUPs and the large number of putative IUPs in more complex genomes will be topics of discussion for the years to come.

Amongst the human amyloidogenic pathologies Alzheimer’s (AD) and Parkinson’s diseases (PD) are the most prevalent. The amyloidogenic precursor proteins in both disorders are

IUPs. The protein/peptide which aggregates in these diseases is amyloid β‐peptide (Aβ) and α‐ synuclein, respectively (table I.1).

 21

Introduction 

Table I.1 – Human diseases associated with amyloid or amyloid‐like deposits according to the native state of the precursor protein or peptide. Human diseases which precursor folding class is unknown are not mentioned. Native State Precursor Protein or Peptide Folding Class Disease

ABri IUP Familial British dementia ADan IUP Familial Danish dementia

-synuclein or variants IUP Parkinson’s disease

-synuclein IUP Dementia with Lewy bodies Nonneuropathic localized diseases Type II Amylin (IAPP) IUP diabetes Alzheimer’s disease Amyloid  peptide IUP Intrinsically Inclusion-body myositis Unstructured Proteins or Amyloid  peptide variants1 IUP Hereditary cerebral haemorrhage with amyloidosis Natively Unfolded N-terminal fragments of apolipoprotein Peptides IUP ApoAI amyloidosis AI Atrial natriuretic factor IUP Atrial amyloidosis Calcitonin IUP Medullary carcinoma of the thyroid Frafment of gelsolin variants IUP Finnish hereditary amyloidosis Huntingtin with polyQ expansion IUP Huntington’s disease

Tau IUP Frontotemporal dementia with Parkinsonism

AA amyloidosis Framents of serum amyloid A protein All- Familial Mediterranean fever

Androgen receptor with polyQ All- (residues 669- Spinal and bulbar muscular atrophy expansion 919)

All- (residues 562- Ataxins with polyQ expansion Spinocerebellar ataxias 694)

2-microglobulin All- Hemodialysis-related amyloidosis Monomeric 3D Structured Cystatin C variants + Icelandic hereditary cerebral amyloid angiopathy Proteins g-Crystallins All- Cataract

Immunoglobulin light chain fragments All- AL amyloidosis

Insulin All- Injection-localized amyloidosis

Lactoferrin + Corneal amyloidosis associated with trichiasis

Lysozyme variants + Lysozyme amyloidosis TATA box-binding protein with polyQ + Spinocerebellar ataxia 17 expansion

Superoxide Dismutase 1 All- Dimer Amyotrophic lateral sclerosis

Wild-type transthyretin All- Tetramer Senile Systemic Amyloidosis Oligomeric 3D Structured Familia amyloidotic polyneuropathy Proteins Variants of transthyretin All- Tetramer Amyloid cardiomyopathy Carpal Tunnel Syndrome

 22

Chapter I

PD is the second most prevalent neurodegenerative disease, affecting 1 ‐ 2% of the western population over the age 65. PD is a movement disorder characterized by degeneration of dopaminergic neurons in the substantia nigra in the brain. The classical symptoms are tremor, rigidity, and slowness of movements. PD degenerating neurons often present fibrils mostly composed by α‐synuclein. In spite of being predominantly sporadic, 10% of the PD cases have a genetic origin. Three missense mutations of the α‐synuclein gene were identified in patients with autossomal dominant forms of PD. The three α‐synuclein variants are A30P, E46K and A53T.

α‐synuclein is, predominantly, a pre‐synaptic protein of 140 amino acid residues with a molecular mass of 14.5 kDa. α‐synuclein has three main domains: a N‐terminal domain consisting of six repeats of an highly conserved KTKEGV motif which binds lipid membranes; a hydrophobic middle region; and a negatively charged C‐terminal domain (Ueda et al., 1993). In vitro and under physiological conditions, α‐synuclein lacks a well defined three‐dimensional structure, belonging to the class of IUPs (Biere et al., 2000; Uversky et al., 2000; Uversky, 2003). In the presence of membranes, α‐synuclein adopts a conformation with two alpha‐helices (Ulmer et al., 2005).

Although the function of α‐synuclein is still unknown, the role of the native disorder three‐ dimensional structures is generally associated with molecular recognition, protein folding inhibition, flexible linking and/or entropic springs (Dunker et al., 2001). Recent studies suggest that

α‐synuclein may act as a molecular chaperone, assisting in the folding and refolding of synaptic proteins called SNAREs (Bonini & Giasson, 2005).

In general, after ribosomal synthesis, the folding pathway of globular proteins (from the unfolded to the folded state) is overcome with a delicate balance between hydrophobic effects, intramolecular non‐covalent bonds (hydrogen bonds and salt bridges) and configurational entropy, the later being the negative counter‐balance for the folding procedure. Due to a native well organized three‐dimensional state, the general underlying molecular mechanisms of amyloidogenesis for globular proteins is associated with (i) defective folding pathways which leads to the formation and population of misfolding states or (ii) the partial unfolding of the native ordered state as a result of protein lack of protein stability. Nevertheless, IUPs do not show a native ordered state. Instead, they have a high degree of native structural disorder which is required to perform their function (Dunker & Obradovic, 2001; Dunker et al., 2002; Tompa, 2002; Radivojac et al., 2004). As stated before, IUPs show a low overall hydrophobicity and a large net charge.

Consequently the major driving effect in the folding pathway is configurational entropy which may surpass the hydrophobic effect. This balance leads to a natively disordered state of the protein. This  23

Introduction 

statement raises the question on how do IUPs form amyloid. Several authors suggest that a partial folding must happen for IUPs to undergo aggregation and amyloid fibril formation (Armen et al.,

2004; Uversky & Fink, 2004; Uversky, 2007). Experimental data from recent studies on α‐synuclein fibrillogenesis suggest the formation of a partially folded intermediate pre‐molten globule‐like as the first step of fibrillization (Uversky et al., 2001b; Uversky et al., 2001a).

The triggering cause for partial folding and formation of an IUP may be related with (i) natural propensity; (ii) covalent modification, which may lead to the development of local structure

(Wright & Dyson, 1999); or (iii) protein overexpression, which may lead to aggregation in concomitance with β‐sheet formation (Miller et al., 2004).

IUPs, also known as proteins with intrinsically disordered domains, represent a large portion of known amyloid diseases proteins. Yet, due to experimental biophysical and biochemical hindrance, the understanding of the molecular mechanism of amyloidogenesis is the object of intense debate. Nevertheless, there is a consensus around the need for partial folding of this protein class in order for amyloidogenesis to occur (Figure I.4)

Figure I.4 ‐ Molecular pathways of amyloid fibril formation from an IUP. The IUP kept in its native disordered state mainly by configurational entropy effect, suffers a partial folding. The triggering factor for this partial folding may be protein covalent modification or protein over‐expression which may leads to aggregation with a concomitant formation of β‐sheet. Eventually, the nucleus formed may progress to amyloid fibrils.

3.1.2 Monomeric proteins in conformational diseases

The molecular mechanism by which a globular protein is converted from its native ordered state into fibrils is a major challenge issue in protein biophysics and biochemistry. A group of

 24

Chapter I monomeric proteins or its variants, such as lysozyme, cystatin C, immunoglobulin light chain, prolactin, insulin, lactoferrin and γ‐crystallin, which presents a broad spectrum of fold motifs (from all‐alpha to all‐beta), have the ability to change their native three‐dimensional conformation and adopt a common β‐sheet fibrilar structure (Thomas et al., 1995; Kelly, 1996; Zerovnik, 2002; Chiti &

Dobson, 2006). Underlying the multiplicity of causes for this aberrant conduct is a common behaviour related with the lack of conformational stability of protein pathologic variants.

Human lysozyme is a monomeric protein with 130 amino acid residues which belongs to α+β fold class with two structural domains, an alpha‐domain with four alpha‐helices and one 310 helix, and a beta domain which consists mainly of an antiparallel β‐sheet (Artymiuk & Blake, 1981).

Lysozyme is a hidrolase responsible for the lysis of the peptidoglycan in the bacterial cell wall.

Interestingly, several lysozyme variants are associated with a familial form of non‐neuropathic amyloidosis, which form amyloid deposits in the liver, spleen and kidneys (Pepys et al., 1993;

Yazaki et al., 2003; Johnson et al., 2005). There are six known variants of human lysozyme, I56T,

F57I, W64R, D67H, T70N and a double mutant F57I/T70N (Pepys et al., 1993; Booth et al., 2000;

Valleix et al., 2002; Yazaki et al., 2003). Apparently, only the T70N mutation has not been associated with amyloid deposits (Yazaki et al., 2003).

Comparative conformational stability studies between wild‐type lysozyme and its amyloidogenic variants, I56T and D67H, have shown that the native states of the pathogenic variants are significantly less stable when compared to the wild‐type protein (Booth et al., 1997).

Experimental data suggest that the lesser stable native state of amyloidogenic lysozyme variants are able to undergo partial unfolding, populating an amyloidogenic intermediate state of the protein which aggregates into amyloid fibrils (Canet et al., 2002; Dumoulin et al., 2005; Johnson et al., 2005).

Experimental data regarding the molecular mechanism of fibrillation of amyloidogenic variants of human lysozyme, strongly suggest that the native states of the protein are in dynamic equilibrium with partially unfolded species. It seems that the partially unfolded intermediates of lysozyme can then undergo self‐association, leading to the formation of β‐sheet ordered aggregates and, eventually, amyloid fibrils (Canet et al., 2002; Johnson et al., 2005).

Most of amyloid fibril formation from native three‐dimensional ordered monomeric proteins seems to be related with the population of amyloidogenic intermediates due to the low conformational stability of native three‐dimensional ordered states of pathogenic variants. Figure

I.5 presents several amyloidogenic routes from native three‐dimensional ordered monomeric to different putative intermediates and, finally, to amyloid fibril formation.  25

Introduction 

Figure I.5 ‐ Possible molecular pathways for amyloid fibril formation from a monomeric three‐dimensional ordered protein. The monomeric protein in its native state, due to lesser conformational stability, can populate partial unfolded states which in turn can unfold or form directly amyloidogenic intermediates with high propensity for aggregation. Eventually, these intermediates lead to amyloid fibril formation.

For most monomeric amyloidogenic proteins the partial unfolding of the native state is the first step for amyloid fibril formation. However, some conformational diseases form partial unfolds

(misfolds) due to defective folding.

3.1.3 Oligomeric proteins in conformational diseases

In the fairly broad spectrum of pathologies associated with formation of amyloid or amyloid‐ like deposits, only few are due to oligomeric proteins. The foremost human diseases which have a multi‐subunit protein underlying the pathologies are amyotrophic lateral sclerosis (ALS), caused by superoxide dismutase‐1 (SOD1), senile systemic amyloidosis (SSA) and familial amyloid polyneuropathies (FAP) due to wild‐type transthyretin (TTR) and its variants (Table I.1).

Human SOD1 is a homodimeric protein with an eight‐stranded β‐barrel motif in each subunit and one intra‐subunit disulfide bond per monomer between Cys57 and Cys146. SOD1 needs the presence of metal ions (copper and zinc) to attain its native three‐dimensional structure and to carry out its enzymatic role. Wild‐type SOD1 has been linked to the ALS, a neurodegenerative disorder characterized by the progressive loss of motor neurons in the brain and spinal cord leading to paralysis and eventually death within 5 years of symptom onset (Watanabe et al., 2001). Most ALS cases are sporadic but 10% of the cases are familial (fALS) and are associated with variants of SOD1.

Although a full understanding of the molecular mechanisms involved in ALS are lacking, some models have been recently proposed linking partial unfolding and aggregation of SOD1 (DiDonato et al., 2003; Rakhit et al., 2004; Banci et al., 2007).

 26

Chapter I

Rakhit and co‐workers showed that both wild‐type and mutant SOD1 dissociate into monomers prior to aggregation in particular oxidative environments. In fact, oxidative damage seems to be the triggering factor for SOD destabilization and aggregation in vitro. Oxidation of both wild‐type and mutant SOD1 causes the enzyme to dissociate into monomers prior to aggregation.

According to these investigators, there is a common aggregation prone monomeric intermediate for wild type and fALS‐associated mutant SOD1 (Rakhit et al., 2004). More recently, the formation of intrasubunit disulfide bonds has been suggested to increase the tendency of SOD1 to form high molecular weight oligomers (Banci et al., 2007).

Human TTR is a homotetrameric protein with a high percentage of β‐sheet. The three‐ dimensional structure of TTR consists of two β‐sheets with four β‐strands each, forming a β‐ sandwich. Each TTR monomer has also an α‐helical segment (Blake et al., 1978). SSA and FAP are characterized by aggregation and systemic extracellular amyloid fibrils formation of wild‐type TTR or its mutant variants, respectively (Saraiva et al., 1984). While SSA affects predominantly individuals over 80 years of age, FAP is an autosomal dominant disease which, depending on the

TTR variant, can have either an early or a very early onset.

Figure I.6 ‐ Possible pathways for amyloid fibril formation of an oligomeric three‐dimensional ordered protein. The oligomeric protein in its native state can dissociate into either a native or a non‐native monomer. While the first dissociation type is reversible, the second is not. The monomeric species formed from dissociation may undergo several conformational changes due to a lack of conformational stability, from partial unfold to a total unfold. As a result of hydrophobic exposition, these species may associate to form aggregates which eventually form amyloid fibrils (Quintas et al., 2001).

In FAP the physiological model proposed in literature states that amyloid formation by TTR is triggered by irreversible tetramer dissociation to a compact non‐native monomer which, depending on its thermodynamic stability, originates less structured monomeric species (partial unfolds) with a high tendency for ordered aggregation into amyloid fibrils. Moreover, the

 27

Introduction 

amyloidogenic potential of some TTR variants is related to higher tendencies to produce partially unfolded monomeric species (Quintas et al., 2001).

Notwithstanding the lack of knowledge with respect to amyloid formation from oligomeric proteins, the few models found in the literature indicate that dissociation and thermodynamic instability of the resulting monomers are the first events for the formation of partially unfolded amyloidogenic intermediates which have high propensity to acquire β‐sheet structure upon aggregation, forming amyloid fibrils (Colon & Kelly, 1992; Lai et al., 1996; Quintas et al., 1997;

Quintas et al., 1999; Quintas et al., 2001; Rakhit et al., 2002) (Figure I.6).

3.1.4 The intrinsically unstructured, monomeric and oligomeric proteins in amyloidogenesis

In conformational disorders some proteins and peptides convert from their soluble form into insoluble amyloid fibrils. In addition, different class of folds of ordered proteins (from all‐beta to all‐alpha) and IUPs can aggregate into cross β‐sheet fibrils. These observations suggest that the propensity to form amyloid does not depend of native initial structure. However, when human diseases are grouped accordingly to their native oligomeric state, native monomeric state and native intrinsically unstructured state, different scenarios emerge. IUPs originate a larger number of amyloid diseases then native monomeric and native oligomeric proteins. In fact, native oligomeric proteins are the least represented class with only two members, SOD1, TTR and their variants. This means that, under physiological conditions, the hierarchy level of the three‐dimensional structure and the three‐dimensional ordered or unordered state of a protein are important factors for amyloidogenesis.

From the data presented in table I.1 it is clear that IUPs belong to the larger group of proteins involved in human conformational diseases. IUPs can undergo disorder to order transitions upon binding to one or more partners, being structured and unstructured during their lifetime (Dunker et al., 2001). These proteins represent an interesting aspect of molecular evolution where the balance between novel functions and the risk of aggregation is at the “Razorʹs Edge” (Merlini & Bellotti,

2003). Additionally, several questions arise regarding the nature of amyloidogenic intermediates

(Uversky & Fink, 2004): Can a completely denatured state be a more amyloidogenic intermediate?

Can a partially folded/unfolded intermediate appear in the early steps of aggregation? An IUP does not need neither to partial unfold, such as a native ordered monomeric protein does, nor to

 28

Chapter I dissociate into partially unfolded species, as a native tetrameric protein does. Besides its intrinsic propensity to aggregate, an IUP only needs the right crowded environment, the right covalent modification, or the right “chaperone” in order to form β‐sheet fibrils and lead to amyloid formation.

Finally, it is possible that oligomeric proteins show a lesser propensity for amyloidogenesis due to a higher hindrance to expose hydrophobic regions, as they need to dissociate and suffer partial unfolding. Additionally, most oligomeric proteins have higher conformational stability in relation to monomeric proteins (Backmann et al., 1998).

3.2 Molecular mechanisms of amyloidosis

Conformational diseases include a vast number of pathologies with multifactorial origins, involving genetic and/or pathophysiological/environmental causes. The understanding of the molecular mechanisms involved in protein conformational diseases needs to identify (i) the primary causes which increase the propensity of a protein for pathogenic aggregation, (ii) all possible conformational states adopted by a polypeptide chain and, (iii) the molecular routes of the amyloid fibril formation progression.

The first step is to identify the biophysical and biochemical causes which increase the propensity to turn a native soluble protein into amyloid. Etiologically, conformational disorders can be hereditary, environmental and transmissible. The increase in aggregation propensity of a protein can be triggered by several shapes: (i) the occurrence of mutations; (ii) intrinsic propensity of the protein to assume a pathological conformation at high concentrations or due to aging; (iii) proteolytic cleavage; (iv) seeding process; (v) exposure to toxins; (vi) impairment of post‐ translational modifications and (vii) impairment of the function of the cellular quality control systems (Lyubchenko et al., 2006).

In hereditary amyloidosis there is the substitution of at least one amino acid residue in the protein, which can either increase its tendency to misfold during the folding pathway or to decrease the protein conformational stability of the native fold, enhancing the tendency to unfold. Indeed, if it is true that an efficient packing of residues is often as important as the hydrophobic effect in determining the thermodynamic stability of the folded state of a protein (Zhou et al., 1992), mutations can disrupt the folding pathway or the fragile conformational stability of a protein by generating a new map of thermodynamic feasible alternative conformations, prone to form

 29

Introduction 

aggregates (Chiti et al., 1999; Quintas et al., 2001). Mutations represent one primary cause for amyloidogenesis. Amyloidogenic diseases caused by TTR, SOD1 and lysozyme variants are examples of an increased aggregation triggered by mutation. More than 80 TTR mutations have been reported, many of them amyloidogenic (Connors et al., 2000). For SOD1 more then 100 variants have been reported, several of them amyloidogenic (Antonyuk et al., 2005).

In some conformational diseases, native proteins show an intrinsic propensity to assume a pathologic conformation which turns into amyloid fibrils at persistently high concentrations or with aging (probably due to slower protein turnover). This is the case of wild‐type TTR in SSA and of wild‐type α‐synuclein in sporadic forms of PD. SSA is a pathology associated with the deposition of wild‐type TTR fibrils in cardiac tissue which occurs predominantly in the elderly (Cornwell et al.,

1988; Westermark et al., 1990). Recent investigations have raised the hypothesis that age‐associated changes in thiol chemistry may play a role in the onset and progression of SSA (Kingsbury et al.,

2007). In PD, the natively unfolded structure of α‐synuclein can be induced to form β‐sheet structure by means of increasing concentration (Kessler et al., 2003) either by α‐synuclein overexpression or impairment of alpha‐synuclein degradation (Shtilerman et al., 2002). Taken together, these observations highlight to the need to unveil the precise triggering factor for each amyloid disease as a first step towards the understanding of the molecular mechanism of conformational diseases.

Considering all the conformational stages adopted by a polypeptide chain after synthesis until amyloid fibril formation, several authors have suggest some fairly consensual models

(Thomas et al., 1995; Kelly, 1996; Merlini et al., 2001; Quintas et al., 2001; Horwich, 2002; Merlini &

Bellotti, 2003; Uversky & Fink, 2004; Chiti & Dobson, 2006; Xu, 2007). Figure I.7 shows a general multi‐pathway molecular mechanism of amyloidogenesis concerning intrinsically unstructured, monomeric and oligomeric three‐dimensional ordered proteins.

Considering the re‐assessment of the classical structure‐function paradigm, the molecular pathways of folding and amyloid fibril formation begins with ribosomal synthesis of a polypeptide chain. According to the regular folding pathway, and depending on the amino acid sequence, a polypeptide chain can achieve one of three native states: (1) IUP (polypeptide chains with low overall hydrophobicity and large net charge), (2) 3D ordered monomeric proteins and (3) 3D ordered oligomeric proteins. The folding pathway of three‐dimensional ordered monomeric proteins can pass, or not, through intermediates (such as molten globules). The folding pathway of oligomeric proteins can be more complex and three routes are possible: the folding and subunit  30

Chapter I association occurs at the same time; the folding of the polypeptide chain takes place through a partially folded intermediate which in turn reaches the native conformation during subunit association and finally the polypeptide chain folds into a three‐dimensional ordered monomer and the native monomer associates with other native monomers.

Figure I.7 ‐ Folding pathways followed by a polypeptide chain after vectorial synthesis on ribossomes (‐) and in the context of amyloid diseases (‐‐‐). After ribosomal synthesis, a polypeptide chain can follow three different routes, depending on the information of its amino acid sequence. (1) IUPs (polypeptide chains with low overall hydrophobicity and large net charge) keep their native disorder structure in solution; (2) 3D ordered monomeric proteins are generated by the folding of polypeptide chains balanced in their hydrophobic/hydrophilic nature (with or without intermediates); (3) 3D ordered oligomeric protein form through the association of monomeric subunits which can fold prior, during or after coming together. The pathways to amyloid fibril formation can derive from the natural propensity of a protein to aggregate, mutations, proteolysis of a precursor protein, exposure to toxins and covalent modifications. Consequently (4) proteins that, instead of their normal native fold, may adopt a partially unfolded state with marginal thermodynamic stability during the folding pathway and (5) native proteins with lesser conformational stability can produce partially unfolded species which may form protofibrils by a nucleation process and, therefore, amyloid fibrils. (6) IUPs seem to adopt a partially folded conformation with high tendency for fibrillation.

Figure I.7 also shows the amyloidogenic pathways which can arise during protein folding or unfolding. During the folding, the main amyloidogenic route comes up from: (4) propensity of the unfold state to form amyloidogenic intermediates and (5) the formation of partially unfolded or misfolded intermediates with propensity to form amyloidogenic intermediates. Amyloidogenesis can take place during the unfolding, mainly due to the lack of stability of the native state protein.

Monomeric proteins can form partially unfolded intermediates with high propensity to aggregate.

Oligomeric proteins must dissociate (almost irreversibly) populating partially unfolded intermediates which in turn may aggregate into amyloid fibril. Additionally, the general

 31

Introduction 

amyloidogenic pathway for IUPs is the adoption of a partially folded conformation with high tendency for fibrillation (6).

A comprehensive mechanism of amyloidogenesis is a Herculean task that depends on biophysical and biochemical investigations to disclose the hidden molecular aspects of conformational diseases. However, and despite the scientific lack of knowledge about this subject, there is an increasing agreement pointing a general molecular motif underlying the vast majority of conformational diseases: the formation of misfolded or partially unfolded protein species associated with conformational fluctuations between these molecular species with marginal thermodynamic stability (Thomas et al., 1995; Chiti et al., 1999; Quintas et al., 1999; Zerovnik, 2002; Chiti & Dobson,

2006).

3.3 Cellular aspects of protein misfolding and disease

The survival and health of all organisms relies heavily on the ability of their proteins to fold correctly. This process is however far from perfect. Indeed a fraction of all proteins fails to fold correctly even in optimal conditions. Consequently, cells evolved mechanisms to deal with misfolded proteins, either by helping their refolding into the correct structure via molecular chaperones or by targeting them for degradation via proteolytic systems such as the proteasome, or through lysosomal or autophagic mechanisms. Whilst these quality control systems are sufficiently effective under most conditions to protect the organism from serious damage, it is clear that changes either in the environment of the cell (due to aging or oxidative damage) or in the protein sequences produced in the cells (through mutations), lead to an increased concentration of misfolded proteins, a situation that can result in different devastating disorders.

Protein misfolding underlies disease in different ways: misfolding of the protein may result in insufficient levels of functional protein; alternatively, a toxic gain of function can occur, as is the case of when toxic aggregates are formed. Protein misfolding can occur at different locations in the cell. Therefore protein quality control is a multi‐level security system to safeguard cells from aberrant proteins.

The first line of defence against protein misfolded proteins is provided by chaperones, also called heat shock proteins. Chaperones bind to unfolded stretches in proteins and keep them in a folding‐competent state while preventing aggregation. In addition they can help dissolve aggregates and target misfolded proteins for degradation. Of particular interest for the formation of

 32

Chapter I aggregates is the Hsp70 class of chaperones (Mayer & Bukau, 2005). Hsp70 bind to small hydrophobic stretches in proteins, in cooperation with a co‐chaperone of the Hsp40 family. If the load of misfolded protein increases, a stress response is initiated by these chaperones, which gears up the quality control systems (Richter‐Landsberg & Goldbaum, 2003).

The ubiquitin‐proteasome system (UPS) is the main cellular protein degradation system that tags and targets short‐lived proteins, such as transcriptional factors, as well as abnormal proteins, like incomplete, missense and misfolded proteins, for destruction (Goldberg, 2003; Varshavsky,

2005). Proteolysis occurs after a polyubiquitin chain is covalently attached to a lysine residue in the protein. Ubiquitination of proteins involves a cascade of at least three different enzymes (Gao &

Karin, 2005). The initial step in this process is the activation of ubiquitin by the enzyme E1, which forms a high‐energy thiol‐ester linkage between the final glycine of ubiquitin and a cysteine of E1.

This thiol‐linked ubiquitin is subsequently transferred to the next enzyme; the ubiquitin‐ conjugating enzyme E2; then a specific E3 ubiquitin ligase will facilitate the transfer of the activated ubiquitin to a lysine residue in the target protein. The polyubiquitinated proteins are targeted to the proteasome, a chambered protease, for destruction. The exact mechanism of the delivery of polyubiquitinated proteins to the proteasome is not fully elucidated yet, but it seems that several proteins are involved in escorting the polyubiquitinated substrates (Elsasser & Finley, 2005; Richly et al., 2005). The target protein is then unfolded inside the proteasome and degraded by three different enzymes: trypsin‐like, chymotrypsin‐like and PGPH‐like (Pickart & Cohen, 2004).

Impaired proteosomal degradation leads to an increased concentration of abnormal and misfolded proteins. Therefore, the ubiquitinated aggregates accumulate forming structures called aggresomes (Johnston et al., 1998). Aggresomes also contain molecular chaperones and proteasome subunits and are surrounded by a cage of collapsed intermediate filaments (Johnston et al., 1998;

Kopito, 2000). This growing body causes autophagy to be initiated. Autophagy is a lysosomal pathway for the turn‐over of organelles and long‐lived proteins and has been identified initially as a cellular survival response to starvation (Reggiori & Klionsky, 2005). There are two types of autophagy: selective autophagy which is triggered by intracellular components and nonselective, which is provoked by extracellular stimuli and induced by starvation (Reggiori & Klionsky, 2005).

Generally, an isolation membrane surrounds cell organelles or other cytoplasmic structures, like aggresomes. The edges of the membrane fuse resulting in a double membrane structure, known as the autophagosome. Subsequently, the outer membrane fuses with a lysosome, where the

 33

Introduction 

sequestered material is degraded into amino acids, nucleotides and free fatty acids (Levine & Yuan,

2005).

A hallmark of a large group of neurodegenerative diseases is the presence of cytosolic, nuclear or extracellular protein aggregates, which are often ubiquitinated and which may be associated with chaperones. The proteins in these aggregates are apparently tagged for degradation by the UPS, but fail to be degraded by the proteasome, becoming toxic. There is a consensus among researchers about the severity of misfolded proteins to cells, but there is disagreement about the properties of misfolded proteins that cause toxicity. The presence of inclusion bodies, composed by protein amyloid fibrils, in patients led researchers to the conclusion that the highly organized protein deposits are the primary cause of the pathological symptoms (Chiti & Dobson, 2006).

Therefore, the large fibrillar protein deposits were identified as the main cause of neuronal damage.

This hypothesis was supported by several in vitro and in vivo studies showing that fibrillar amyloid‐

β is toxic to cultured neuronal cells (Pike et al., 1991; Geula et al., 1998; Hartley et al., 1999).

However, recent studies suggest that prefibrillar aggregates, called protofibrils, rather than mature fibrils are the most potent mediators of cell damage, cytotoxicity and neurotoxicity. These conclusions are supported by the finding that the severity of cognitive impairment in protein misfolding diseases correlates with the levels of small oligomeric species and not with the large fibrillar species (Gertz et al., 1994; Lue et al., 1999; McLean et al., 1999; Sousa et al., 2001; Caughey &

Lansbury, 2003; Kayed et al., 2004; Cleary et al., 2005; Crowther et al., 2005; Danzer et al., 2007;

Lauren et al., 2009).

The paths by which misfolded proteins disturb cellular functions are not entirely understood.

The mechanisms identified that might be responsible for cellular damage are provoked by oligomeric species, which disrupts cell membranes by inserting themselves into the phospholipid bilayer, disturbing normal ion gradients (Kourie & Henry, 2002), inactivating normal folded and functional proteins, and obstructing proteasome components or chaperone proteins (Rao &

Bredesen, 2004). Oligomeric species of α‐synuclein, amyloid‐β and amylin form micelle‐like aggregates in solution (Soreghan et al., 1994) and annular pores in lipid bilayers including cell membranes (Mirzabekov et al., 1996; Lashuel et al., 2002) that initiate membrane permeabilization

(Lashuel, 2005). Subsequently, numerous pathological events are stimulated in the cells, such as

Ca2+ immobilization (Demuro et al., 2005), induction of endoplasmic reticulum stress (Lindholm et al., 2006), generation of reactive oxygen species (ROS) (Emerit et al., 2004; Zhu et al., 2007), and finally, cell death (Stefani & Dobson, 2003). Cell death may also be induced by misfolded and  34

Chapter I aggregated proteins through stimulation of apoptotic responses (Morishima et al., 2001). Apoptosis is stimulated through ROS production and increases in Ca2+ levels that are followed by the activation of caspases (Bucciantini et al., 2005). Additional, the mitogen‐activated protein kinase

(MAPK) pathway has been observed in protein misfolded diseases (Beal, 2002), as a response to various stressful conditions such as oxidative stress, cytokines and heat shock (Rattan, 2004; Zhou et al., 2006; Mitchell et al., 2007). In addition to their direct toxic effects, misfolded proteins can also promote inflammatory responses (Casserly & Topol, 2004), increasing stress responses, and neuroinflammation has been observed in patients (Manuelidis et al., 1997; Buxbaum, 2004; Aguzzi

& Heikenwalder, 2006; Finch & Morgan, 2007; Heneka & OʹBanion, 2007; Williams & Nadler, 2007).

Putting all these misfolded protein‐induced processes together, it seems clear that, when the cellular quality control systems fail to handle misfolded proteins, a very dark picture emerges leaving small chances for cell survival.

4. Post­translational protein modifications

The release of a completed polypeptide chain from a ribosome is frequently not the last chemical step in the formation of proteins. Various covalent modifications often occur, either during or after assembly of the polypeptide chain which are called post‐translational modifications

(PTMs). Knowledge of these modifications is extremely important because they may alter physical and chemical properties, folding, stability and activity and, consequently, the function of the proteins, being intimately related with disease.

Because of the invariable presence of α‐synuclein in a number of types of pathologies like

Parkinson’s disease (PD), multiple system atrophy (MSA), dementia with Lewy bodies (DLB),

Hallervorden‐Spatz disease (HSD), among many others, α‐synuclein will be discussed in this title in order to illustrate the importance of PTMs, particularly glycation. Regarding PD, above 90 % of the cases are sporadic and even though characterized by the presence of α‐synuclein aggregated into insoluble fibrillar inclusions (Spillantini et al., 1997; Spillantini et al., 1998). Thus, pathogenic modifications other than point mutations should be responsible for α‐synuclein conversion into toxic species. Analysis of Lewy bodies (LB) revealed that inclusions encompassed a wide variety of

α‐synuclein species, which greatly differ from the form found in healthy neurons. More than 300

PTMs of α‐synuclein have been reported in vivo – modifications that may alter the protein’s size, charge, structure, conformation, or binding affinities, that may interfere with protein function and  35

Introduction 

degradation (Clark et al., 2005; Beyer, 2006). Weather these contribute to aggregation or, instead, are a consequence of it, is still an open question.

4.1 Polyamine binding

The C‐term appears to be a remarkable region for PTMs. These modifications may affect its role as a negative regulator of self‐assembly and therefore greatly influence the propensity of α‐ synuclein to aggregate in vivo. For instance, the C‐term comprises acidic residues, to which positively charged molecules, such as polyamines, may bind. Polyamines are key factors for cell growth and differentiation, but at high concentrations, they produce toxic oxidative intermediates

(Auvinen et al., 1992). Polyamines were found to promote the acceleration of α‐synuclein fibrillation and aggregation (Antony et al., 2003; Goers et al., 2003; Fernandez et al., 2004b).

4.2 Oxidation

Oxidation of α‐synuclein methionine residues, which readily occurs under mild oxidative conditions, maintains α‐synuclein unfolded and less prone to oligomerize and fibrillize (Uversky et al., 2002; Yamin et al., 2003a; Glaser et al., 2005). However, extensive oxidative damage in the nigrostriatal region of PD brains has been brought to evidence. Oxidative damage of DNA, RNA, proteins and lipids were reported in the vulnerable neurons of PD and DLB (Dexter et al., 1987;

Bowling & Beal, 1995; Castellani et al., 1996; Yoritaka et al., 1996; Alam et al., 1997a; Alam et al.,

1997b; Floor & Wetzel, 1998; Castellani et al., 2002; Nunomura et al., 2002). In addition, in brain areas affected in PD, abnormally high levels of redox‐active metals such as iron or increased

Fe(III)/Fe(II) ratio have also been reported (Riederer et al., 1989). Even in the serum of live PD patients, an increase in the oxidant status together with a decrease in the anti‐oxidant capacity could be assessed (Forte et al., 2004; Hegde et al., 2004). Accordingly, experiments with antioxidant compounds revealed their anti‐α‐synuclein‐fibrillogenic and α‐synuclein‐fibril‐destabilizing effects

(Ono & Yamada, 2006). It seems clear that a tight correlation between oxidative stress and protein aggregation and neurodegeneration is observed, although it has been difficult so far to choose between free radical damage and energetic failure as a primary factor for α‐synuclein aggregation‐ mediated degeneration. Also, it is still under investigation weather oxidative stress is a cause or a consequence of the neuronal degenerative pathogenesis (Jenner, 2003).

 36

Chapter I

4.3 Nitration

Nitrated α‐synuclein has been found in LB in all of the putative sites of nitration (tyr‐39, ‐125,

‐133 and ‐136) and nitrating agents were found to promote the formation of β‐sheet structures and accelerate α‐synuclein oligomerization through covalent dityrosine crosslinking (Souza et al., 2000;

Hodara et al., 2004; Uversky et al., 2005). At low concentrations, nitrated α‐synuclein facilitates the nucleation step of α‐synuclein monomers assembly (Hodara et al., 2004), but at high concentrations, it inhibits filamentous assembly (Yamin et al., 2003b). Nitration of a single tyrosine residue, independently of its location in the sequence of α‐synuclein, is sufficient for triggering α‐synuclein dityrosine crosslinking (Norris et al., 2003). Protein nitration was suggested to be an early event in the onset and progression of the pathogenesis of PD, rather than a late‐stage phenomenon

(Reynolds et al., 2007). In addition, nitration of tyrosine 39 in the N‐term of the protein disrupts the

α‐helical conformation of α‐synuclein and its ability to bind lipid vesicles (Hodara et al., 2004).

Nitration also seems to affect the ubiquitin‐independent proteosomal degradation of the protein and could even lead to incompletely degraded C‐truncated species (Liu et al., 2003; Hodara et al.,

2004). Together, these changes lead to the amplification to the intracellular pool of α‐synuclein and may thereby promote aggregation.

4.4 Ubiquitination

α‐Synuclein ubiquitination occurs in specific residues (Nonaka et al., 2005). An autosomal dominant form of PD has been linked to a missense point mutation in the ubiquitin carboxyl‐ terminal hydrolase‐L1 (UCH‐L1), which plays a role in the mechanisms targeting proteins to proteasomal degradation by intervening in ubiquitin recycling and regulation of ubiquitin‐protein conjugation (Leroy et al., 1998; von Coelln et al., 2004). Noticeably, these familial cases did not generally develop inclusions, which induced researchers to believe that toxicity in these cases relied on the accumulation of soluble nonfibrillar α‐synuclein species. In fact, α‐synuclein is usually not ubiquitinated, but rather degraded by proteasome in an ubiquitin‐independent manner (Bennett et al., 1999; Ancolio et al., 2000; Rideout et al., 2001; Tofaris et al., 2001; David et al., 2002; Rideout &

Stefanis, 2002). Even though, ubiquitin is often found linked to α‐synuclein in LB, so it was proposed that ubiquitination occurs mainly after the assembly of α‐synuclein into oligomeric species and fibrils (Sampathu et al., 2003). Ubiquitination of α‐synuclein is believed to represent a

 37

Introduction 

disease‐specific mechanism, resulting from an unsuccessful attempt of the proteasome or of the lysosome (which requires mono‐ubiquitination) to unfold and/or degraded misfolded proteins.

4.5 Sumoylation

Lysine appears to be a common target substrate for a number of PTMs in addition to ubiquitination, including acetylation, methylation, glycation (see below) and sumoylation. SUMOs are small ubiquitin‐like modifiers, which may largely impact protein interactions and subcellular localization, and may even antagonize the proteasome pathway by competing with ubiquitin

(Dohmen, 2004; Dorval & Fraser, 2006). This PTM has been involved in a number of neurodegenerative diseases (Ueda et al., 2002; Pountney et al., 2003; Steffan et al., 2004; Pountney et al., 2005; Riley et al., 2005; Shinbo et al., 2006).

4.6 Phosphorilation

Protein phosphorylation is an essential PTM, involved in the regulation of nearly all aspects of cell life. But phosphorylation has also been spotted as a modification involved in the pathogenesis of a number of neurodegenerative diseases. Ser‐129 phosphorylation is a well established and dominant PTM in aggregated insoluble α‐synuclein (Fujiwara et al., 2002; Anderson et al., 2006). Phosphorylated α‐synuclein species constitute 90 % of the total α‐synuclein population in DLB brains (Fujiwara et al., 2002) and their amount is 30 times the amount measured in control brains (Anderson et al., 2006). Regarding the monomeric soluble α‐synuclein, only 4 % is phosphorylated. Considering the phosphorylation of Ser‐129, there has been lots of controversy regarding the toxic effects of this PTM on α‐synuclein. Expressing the α‐synuclein mutant S129D

(that mimics the phosphorylated form of α‐synuclein) in a rat model of PD, showed no toxicity, whereas the S129A mutant (that mimics the unphosphorylated form) led to degeneration

(Gorbatyuk et al., 2008). However, in drosophila and transgenic mouse models, the opposite effect was observed, with the S129D mutant displaying toxicity while the S129A did not (Chen & Feany,

2005; Freichel et al., 2007).

A limited level of phosphorylation of human α‐synuclein serine at position 87 was pointed out in 293T and PC12 cells overexpressing human α‐synuclein (Okochi et al., 2000). Modification on

Ser‐87 might have great impact on the properties of α‐synuclein, because of serine’s critical position

 38

Chapter I within the NAC domain. However, limited data about phosphorylation at this site have been generated so far.

Protein‐tyrosine kinases are involved in the regulation of synaptic function in the brain and are important players, for example, in excitability, plasticity, memory or spatial learning (Pang et al.,

1988; OʹDell et al., 1991; Gurd, 1997; Llinas et al., 1997). α‐Synuclein possesses 4 tyrosine residues

(Tyr‐39, Tyr‐125, Tyr‐133. Tyr‐136) well conserved among synuclein family members (Clayton &

George, 1998). In vitro experiments with 293T cells transfected to overexpress α‐synuclein bearing site‐specific mutation to phenylalanine, which prevents tyrosine phosphorylation, isolated Tyr‐125 as a major site of phosphorylation (Ellis et al., 2001). Phosphorylated Tyr‐125 has also been shown to be involved in regulating α‐synuclein’s affinity to metal, affect protein’s conformation, change its charge balance, and thereby influence α‐synuclein fibrillation (Liu & Franz, 2005; Liu & Franz,

2007). Phosphorylation of Tyr‐125 could also affect interaction of α‐synuclein with tau and thereby alter the dynamics of microtubular assembly (Nakamura et al., 2001).

4.7 Glycation

The glycation reaction was introduced on the scientific community for the first time in 1912 by the french chemist Louis Camille Maillard referring to a reaction between sugars and amino acids or other amino‐containing compounds (Maillard, 1912). The process involves multi‐step non‐ enzymatic reactions between carbonyl‐containing groups and amino groups leading to the formation of irreversible products named advanced glycation end‐products (AGE) (Brownlee et al.,

1984). Therefore, biomolecules with free amino groups like proteins, nucleotides and basic phospholipids can be irreversibly modified by this non‐enzymatic reaction. Regarding proteins, the side chains of arginine and lysine residues and the N‐terminal amino group are available targets for glycation. Cysteine thiol groups may also be glycated, but produce reversible and unstable adducts

(Lo et al., 1994; Westwood & Thornalley, 1997). Since the N‐terminal is usually modified in proteins, arginine and lysine side chains are the main glycation targets in proteins (Driessen et al., 1985).

The first step of the Maillard reaction involves the nucleophilic attack by the nitrogen atom of the amino group to the electrophilic carbonyl group of an aldehyde or ketone. After elimination of a water molecule, a Schiff’s base is generated. At this phase, the reaction is reversible, but the Schiff’s base may suffer an Amadori rearrangement, irreversibly forming an Amadori product: a ketoamine

(Hodge, 1955; Koenig et al., 1977; Westwood & Thornalley, 1997). Subsequently, the Amadori

 39

Introduction 

product undergoes a complex series of chemical reactions, such as intramolecular rearrangements, oxidative and non‐oxidative fragmentation and dehydration reactions to yield an irreversible adduct: the AGE (Bucala & Cerami, 1992; Vlassara et al., 1994; Westwood & Thornalley, 1997), as shown in figure I.8.

Figure I.8 – Initial steps of protein glycation by glucose, including the formation of Schiff’s base and the Amadori product. The Schiff’s base is formed by the nucleophilic attack of the amino group to the carbonyl group that subsequently undergoes the Amadori rearrangement to produce frutosamines, the Amadori product. Through a complex series of chemical reactions, irreversible protein adducts, the advanced glycation end‐products are produced. These adducts may involve protein cross‐ links by the reaction with another amino group (OʹBrien, 1997; Westwood & Thornalley, 1997).

Protein glycation depends on the concentration, specific reactivity and duration of exposure to the glycation agent (Acharya & Manning, 1980; Eble et al., 1983; McPherson et al., 1988; Farah et al., 2005), the presence of catalytic factors (as metal and buffer ions and oxygen) (Watkins et al.,

1987; Smith & Thornalley, 1992b; Smith & Thornalley, 1992a; Fu et al., 1996), the physiological pH

 40

Chapter I and temperature (Smith & Thornalley, 1992b; Smith & Thornalley, 1992a) and the protein half‐life

(Schleicher & Wieland, 1986). Also, the location of the amino acid residue in a folded protein influences the glycation reaction, either because the neighbouring amino acids influence the pKa value of the amino acid side chain undergoing glycation or restrict the binding of the glycation agent due to steric hindrance (Westwood & Thornalley, 1997; Ahmed et al., 2005).

Figure I.9 – Formation of dicarbonyl compounds in the early steps of protein glycation. These highly reactive compounds may derive from glucose autoxidation, Schiff’s base fragmentation reactions and/or Amadori product atoxidation (Hayashi et al., 1986; Wells‐Knecht et al., 1995; Thornalley et al., 1999). Methylglyoxal. Glyoxal and 3‐deoxiglucosone can also initiate protein glycation with the formation of a ketoimine that undergoes several chemical reactions to yield AGE (Hunt et al., 1993).

Being a reaction between sugars and amino groups, glycation was thoroughly studied using food sugars as glycation agents. And after the discovery of glycated haemoglobin in vivo by glucose, which increases as a function of the mean glycaemia of diabetic patients (being now used as a glycaemia marker), glycation by glucose became the main target for experimental investigation

(Bunn et al., 1975). However glucose is the least reactive of all sugars, being speculated that this was the main reason why glucose was selected as the major metabolic fuel during evolution (Bunn &

Higgins, 1981), and its intracellular concentration is negligible. Glucose autoxidation, Schiff’s base  41

Introduction 

fragmentation and Amadori product’s autoxidation yield highly reactive dicarbonyl compounds, such as 3‐deoxyglucosone, glyoxal and methylglyoxal (Figure I.9) (Hayashi et al., 1986; Kato et al.,

1987; Wells‐Knecht et al., 1995; Thornalley et al., 1999). These compounds also react with protein amino groups and are much more reactive (Figure I.9) (Hunt et al., 1993; Lo et al., 1994; Westwood &

Thornalley, 1997). Accordingly, more recent investigations showed that in physiological conditions, reactive dicarbonyl compounds are key intermediates of protein modification by the Maillard reaction (Thornalley, 1994; Thornalley, 1996; Chang & Wu, 2006). These observations focused the attention on methylglyoxal, since this α‐oxoaldehyde is present in all cells and is considered to be the most important glycation agent in vivo (Thornalley, 1994; Thornalley, 1996; Chang & Wu, 2006).

4.7.1 Methylglyoxal

Methylglyoxal is an unavoidable product of cellular metabolism either in normal or pathological conditions. The biochemical research on methylglyoxal started with the discovery of an enzymatic system which converts α‐oxoaldehydes into α‐hidroxyacids – the glyoxalase system

(Dakin & Dudley, 1913a; Dakin & Dudley, 1913b; Neuberg, 1913). First, methylglyoxal was considered to be a glycolitic intermediate (Neuberg & Kobel, 1928), but subsequent works led to a gradual dismissal of the role of methylglyoxal as a metabolic intermediate (Lohmann, 1932; Racker,

1951). Later, Szent‐Gyorgyi introduced the hypothesis that glyoxalase I and methylglyoxal regulate cell division and might be involved with carcinogenesis (Szent‐Gyorgyi, 1965), giving a sudden impulse to the research in this area.

Methylglyoxal is synthesized through different enzymatic and non‐enzymatic metabolic pathways, such as reactions involved in L‐threonine metabolism (Ray & Ray, 1987; Lyles &

Chalmers, 1992), catabolism of ketone bodies (Casazza et al., 1984; Koop & Casazza, 1985;

Aleksandrovskii, 1992); enzymatically by methylglyoxal synthase (an ezyme only found in some bacteria) (Cooper & Anderson, 1970), through Maillard reaction (Thornalley et al., 1999), lipoperoxidation (Esterbauer et al., 1982) and as a by‐product of glycolysis (Richard, 1993).

 42

Chapter I

Figure I.10 – Methylglyoxal formation from the triose phosphates dihydroxiacetone phosphate (DHAP) and glyceraldehyde 3‐phosphate (GAP). Triose phosphates are unstable molecules and the β‐elimination reaction of the phosphate group from the common enediolate phosphate intermediate irreversible yields methylglyoxal. The stabilization of this intermediate by triose phosphate isomerase (TPI) is essential to avoid methylglyoxal formation. However, enediolate phosphate intermediate can leak from the enzyme active site forming methylglyoxal in a paracatalytic reaction. Adapted from (Richard, 1993).

The most important pathway for methylglyoxal formation in eukaryotic cells is the glycolitic by‐pass, where the β‐elimination of the phosphate group from triose phosphate intermediates glyceraldehyde‐3‐phosphate (GAP) and dihydroxyacetone phosphate (DHAP) produces methylglyoxal (Richard, 1993) (figure I.10). At physiological pH, phosphorylated trioses are much more reactive towards the loss of α‐carbonyl protons than the corresponding triose, producing an enediolate phosphate, which has a low energy barrier for the phosphate group removal (Richard,

1993) and thus forming methylglyoxal (Richard, 1993). The stabilization of the enzyme‐bound enediolate phosphate by triose phosphate isomerase (TPI, D‐glyceraldehyde‐3‐phosphate aldose‐ ketone‐isomerase, EC. 5.3.1.1) is an absolute requirement to avoid the substrate degradation into methylglyoxal, being the protonation of enediol 106 faster than the removal of the phosphate group

(Richard, 1993). Still, estimation of the methylglyoxal non‐enzymatic formation rate is given as 0.1 mM per day (Richard, 1993), even though the rate of formation of methylglyoxal depends on the  43

Introduction 

organism, tissue, cell, metabolism and physiological conditions. Its formation is related to the glycolitic flux, confirming that triose phosphate degradation is the main pathway for the production of methylglyoxal (Altenberg & Greulich, 2004).

Figure I.11 – The glyoxalase system. This enzymatic pathway comprises two enzymes (glyoxalase I and glyoxalase II) responsible for the GSH‐dependent catabolism of methylglyoxal, producing D‐lactate (Racker, 1951; Thornalley, 1990). Glyoxalase I catalyses the formation of S‐D‐lactoylglutathione from the hemithioacetal produced by the non‐enzymatic reaction between methylglyoxal and GST (Vander Jagt et al., 1975; Thornalley, 1990; Thornalley, 1993). Then GSH is regenerated by glyoxalase II, which catalyses the formation of D‐lactate from S‐D‐lactoylglutathione (Thornalley, 1990; Vander Jagt, 1993).

Methylglyoxal, a highly reactive and toxic compound, irreversibly damages proteins and nucleic acids through the Maillard reaction (Lo et al., 1994; Westwood & Thornalley, 1997; Oya et al.,

1999). High doses of methylglyoxal cause cell death while, with sub lethal concentrations, a cell growth delay is observed (Okado et al., 1996; Kalapos, 1999; Ponces Freire et al., 2003; Maeta et al.,

2005). Therefore, protective enzymatic mechanisms evolved to prevent the damage of biomolecules by this unavoidable product of cell metabolism. The glyoxalase system is, by far, the most investigated catabolic route for this α–oxoaldehyde. However, there are other pathways to catabolise methylglyoxal, namely α–oxoaldehyde dehydrogenase (2‐oxoaldehyde:NAD(P)+ 2‐ oxidoreductase, EC. 1.2.1.23) (Monder, 1967), aldehyde dehydrogenase (aldehyde:NAD+ oxidoreductase, EC. 1.2.1.3) (Izaguirre et al., 1998), aldose reductase (alditol:NAD(P)+ 1‐ oxidoreductase, EC. 1.1.1.21) (Vander Jagt et al., 1992), methylglyoxal reductase (D‐  44

Chapter I lactaldehyde:NAD+ oxidoreductase, EC. 1.1.1.78) (Ray & Ray, 1984) and pyruvate dehydrogenase

(pyruvate:dihydrolipoyllysine‐residue‐acetyltransferase‐lipoyllysine 2 oxidoreductase, EC. 1.2.4.1)

(Baggetto & Lehninger, 1987). Although the real significance of each of these pathways in methylglyoxal catabolism in vivo is controversial, it is consensual that the glyoxalase system and the aldose reductase enzyme are the most relevant methylglyoxal catabolic pathways.

The glyoxalase system was fully characterized for the first time as a system composed by two different enzymes by Racker (Racker, 1951). The glyoxalase system comprises glyoxalase I (S‐D‐ lactoylglutathione methylglyoxal‐lyase, EC. 4.4.1.5) and glyoxalase II (S‐2‐hydroxyacylglutathione hydrolase, EC. 3.1.2.6), and converts methylglyoxal to D‐lactate using reduced glutathione as a specific cofactor (Racker, 1951; Thornalley, 1990) as shown in figure I.11.

Aldose reductase is a member of the aldo‐keto reductase superfamily. It was first described by Hers, who observed the NADPH‐dependent reduction of glucose and other aldehydes to polyols by extracts of seminal vesicles and placenta (Hers, 1956; Ginsburg & Hers, 1960). This enzyme exhibits broad substrate specificity for a variety of aldehydes and shows a high affinity for methylglyoxal (Km of 8 μM and kcat/Km of 1.8 x 107 M‐1.min‐1), suggesting that this enzyme may be relevant in methylglyoxal detoxification (Vander Jagt et al., 1992). Interestingly, aldose reductase contains a putative glutathione binding site near the active site, which may have a major role in methylglyoxal catabolism. In the presence of high glutathione concentrations, the efficiency of reduction of methylglyoxal by aldose reductase increases and the site of reduction switches from the aldehyde to the ketone carbonyl, as illustrated in figure I.12.

Figure I.12 – Methylglyoxal catabolism by aldose reductase. Depending on the presence of GSH, aldose reductase catalyses the NADPH‐dependent reduction of methylglyoxal to acetol (reduction of aldehyde group) or to lactaldehyde (reduction of the ketone group). These compounds are then converted to propanediol by another NADPH‐dependent reaction catalysed by aldose reductase. Adapted by (Vander Jagt & Hunsaker, 2003).

 45

Introduction 

Methylglyoxal has two functional groups: a highly reactive aldehyde group and an electron acceptor ketone group, being an excellent electrophile molecule. In biological systems, methylglyoxal can modify amine groups of proteins, nucleic acids and basic phospholipids through the Maillard reaction. However, the main targets for methylglyoxal glycation are the amino acid side chains of arginine and lysine. Irreversible reactions may occur with lysine and arginine residues leading to the formation of irreversible adducts on proteins known as MAGE

(methylglyoxal‐derived advanced glycation end‐products) (Gomes et al., 2005b). The reaction between methylglyoxal and lysine residues leads to the formation of Nε‐(carboxyethyl)lysine (CEL), a MAGE found in human lens proteins with increased concentrations over ageing (Ahmed et al.,

1997). Methylglyoxal is also responsible for protein cross‐links, one of the major consequences of protein glycation (Sell & Monnier, 1989; Fu et al., 1994). With lysine residues, methylglyoxal forms methylglyoxal‐lysine dimers (MOLD) (Nagaraj et al., 1996), which can be observed in lens protein as well as in diabetic patients (Nagaraj et al., 1996; Frye et al., 1998). Besides the lysine cross‐links, a lysine‐arginine methylglyoxal‐derived cross‐link, named MODIC, was also described (Lederer &

Klaiber, 1999). In fact, methylglyoxal preferred targets are arginine residues (Lo et al., 1994; Oya et al., 1999). Different MAGE can be found from reaction between methylglyoxal and arginine residues, namely 5‐hydroimidazolones [Nδ‐(5‐methyl‐imidazolone‐2‐yl)‐ornithine] (Henle et al.,

1994), which is believed to be the major product of methylglyoxal derived glycation; THP

[tetrahydropyrimidine, Nδ‐(4‐carboxy‐4,6‐dimethyl‐5,6‐dihydroxy‐1,4,5,6‐ tetrahydropyrimidin‐2‐ yl)‐ornithine (Oya et al., 1999); and argpyrimidine [Nδ‐(5‐hydroxy‐4,6‐dimethylpyrimidine‐2‐yl)‐L‐ ornithine] (Shipanova et al., 1997; Oya et al., 1999). The later one is also a fluorescence dye used to monitor protein glycation.

4.7.2 Glycation and amyloidosis

Amyloidosis is a generic term used to designate a group of clinical and biochemical diseases characterized by protein deposition into insoluble amyloid fibrils (Sipe, 1992; Ghiso et al., 1994).

There are over 40 human diseases associated with the formation of extracellular amyloid deposits or intracellular inclusions with amyloid‐like characteristics (Chiti & Dobson, 2006). Several types of amyloidosis and their amyloid protein precursors were identified (Ghiso et al., 1994), and although there is no obvious sequence homology between the different amyloidogenic proteins, all amyloid deposits share particular biochemical features such as high insolubility and proteolysis resistance, a

 46

Chapter I

β‐pleated sheet structure and similar colouring properties like apple‐green birefringence under polarized light after Congo red staining and yellow‐green fluorescence with thioflavin T (Sipe,

1992; Ghiso et al., 1994). These similarities suggest that common mechanisms are involved in this type of disorders. Numerous point mutations are associated with the amyloidogenic behaviour of several proteins like a‐synuclein (Pankratz & Foroud, 2004; Valente et al., 2004), transthyretin

(Saraiva, 2001) and Aβ peptide (Haass et al., 1994). In familiar amyloid polyneuropathy (FAP), more than 80 transthyretin point mutations were associated with amyloid fibril formation (Saraiva, 2001).

However, wild type transthyretin can also form amyloid deposits (Westermark et al., 1990), hinting for the complexity of amyloid fibril formation pathways where several factors beyond genetic determinants may play an important role. The abnormal proteolytic processing and/or post‐ translational modifications including glycation, among many others, are probably involved in disease. Neuropathological amyloid deposits and glycated protein share several structural properties such as high insolubility, protease resistance, characteristic cross‐link structures, fluorescence and the presence of brown coloured compounds, and for this reason it was proposed that protein glycation might account for amyloid formation in vivo (Colaco & Harrington, 1994;

Harrington & Colaco, 1994).

In agreement with this hypothesis, AGE‐modified proteins were detected in amyloid deposits from several amyloidosis such as Alzheimer’s (Smith et al., 1994; Yan et al., 1994), Parkinson’s diseases (Castellani et al., 1996; Munch et al., 2000) and FAP (Gomes et al., 2005a). In Alzheimer’s disease, both extracellular Aβ‐peptide amyloid plates and intracellular neurofibrillary tangles of tau protein are highly modified with AGE (Smith et al., 1994; Vitek et al., 1994; Yan et al., 1994). It was therefore suggested that protein modifications through the Maillard reaction could stabilize the amyloid deposits, accounting for their high insolubility and protease resistance (Smith et al., 1994).

In addition, AGE‐modified tau leads to an increase in the production and secretion of Aβ, followed by reactive oxygen species formation (Yan et al., 1995). Methylglyoxal glycation of Aβ promotes the formation of β‐sheet, oligomers and protofibrils (Chen et al., 2006). Interestingly, the expression levels of glyoxalase I in the brains of AD patients are lower in later stages of the disease, as confirmed both at the level of mRNA and protein (Kuhla et al., 2007). Since this is the main methylglyoxal catabolic pathway, a higher carbonyl stress is expected that together with an increase in AGE content, oxidative stress, sustained inflammation, plaques and tangles formation can ultimately lead to cell death.

 47

Introduction 

In PD patients, glycation was first reported in the substantia nigra and locus ceruleus displaying higher immunoreactivity at the periphery of LB (Castellani et al., 1996). In agreement with this report, co‐localization of AGE with α‐synuclein in very early LB was observed, where α‐synuclein is also present at the periphery of LB (Munch et al., 2000). These results suggest that glycation may be involved in the chemical cross‐linking, increasing the proteolytic resistance of the protein deposits. Comparing glycation levels in cerebral cortex, amygdale and substancia nigra with old control individuals, the number and extension of glycated proteins were significantly higher in PD patients (Dalfo et al., 2005).

In FAP patients, argpyrimidine, a specific methylglyoxal‐derived advanced glycation end‐ product, was found in amyloid fibrils from FAP patients and it was not detected in control subjects

(Gomes et al., 2005a).

Even though glycation is involved in amyloidosis, it is still controversial whether glycation of susceptible proteins could be an initial event in amyloid fibril formation or merely a result of amyloid fibril accumulation due to the longevity of the protein components which, as a result of their insolubility and protease resistance, persist in the body for long periods of time. Nevertheless, several lines of evidence suggest that glycation directly promotes or accelerates abnormal protein deposition in β‐fibrils characteristic of these pathologies. Munch and co‐workers found AGE in very early Lewy bodies suggesting its involvement in protein cross‐link and the formation of insoluble, non‐degradable aggregates (Munch et al., 2000). The accumulation of AGE‐modified proteins also leads to inflammation and propagation of tissue damage by several mechanisms like oxidative stress and increase and release of pro‐inflammatory cytokines mediated be AGE:RAGE interaction. Regardless of the exact chronology of AGE accumulation, it is becoming evident that

AGE in these deposits are not static by‐products of disease, but rather dynamic participants in neuronal dysfunction, inducing several cellular responses that may lead to cell dysfunction and death.

 48

Chapter I

 49

Chapter II

Cytochrome c glycation: Aggregation without unfolding

Luís M. A. Oliveira, Carlos Família, António E. N. Ferreira, Carlos Cordeiro, Ana Ponces Freire and Alexandre Quintas. Structure and stability comprehensive study of native‐like aggregation of cytochrome c upon glycation. (manuscript in preparation).

1. Abstract

After ribosomal synthesis a polypeptide chain, usually achieves a native three‐dimensional structure. However, under some hereditary or pathophysiological conditions like point mutations and post‐translational modifications, proteins may aggregate originating conformational diseases.

The hallmark of conformational diseases is protein misfolding/unfolding which leads to the formation of high molecular mass species and later to β‐sheet amyloid fibrils. Although, some of these aggregates retain native‐like structure which suggests protein pathologic association mechanisms other than cross β‐strand of amyloid fibrils often called native‐like aggregation.

Native‐like aggregation mechanism is still an undisclosed subject. Here it is revealed that cytochrome c non‐covalently associates into high molecular mass native‐like species under glycation conditions. Moreover it is shown that methylglyoxal glycation of cytochrome c also results in the formation of a partial unfolding species. The data collected from analytical and spectroscopic techniques allows the proposal of a comprehensive thermodynamic model for native‐ like aggregation of methylglyoxal glycated cytochrome c.

2. Introduction

In general, after ribosomal synthesis, the folding pathway is overcome with a delicate balance between hydrophobic effects, intramolecular non‐covalent bonds (hydrogen bonds and salt bridges) and configurational entropy, the later being the negative counter‐balance for the folding procedure. Depending on its polypeptide chain sequence, a protein achieves one of three native states: an intrinsic unstructured (typically resulting from polypeptide chains with low overall hydrophobicity and large net charge that contains little persistent structure (Uversky et al., 2000)), a

3D ordered monomeric globular structure or, by association with other polypeptide chains, an oligomeric structure. However, under inherited or pathophysiological conditions, including (i) the occurrence of mutations, (ii) intrinsic propensity of the protein to assume a pathological conformation at high concentrations or due to aging, (iii) proteolytic cleavages, (iv) seeding process,

(v) toxins exposure, (vi) impairment of post‐translational modifications, such as phosphorylation and advanced glycation and (vii) impairment function of proteasome (Lyubchenko et al., 2006), a polypeptide chain can adopt and interconvert between a multitude of conformational states

Aggregation without unfolding 

comprising misfolds/partial unfolds, aggregates and ‐sheet amyloid fibrils originating conformational or misfolding diseases (Thomas et al., 1995; Kelly, 1996; Stefani & Dobson, 2003).

Most of human conformational diseases are associated with the formation of amyloid protein deposits as a result of protein misfolding and the formation of ‐structures (Saraiva et al., 1984;

Booth et al., 1997; Quintas et al., 1999; Quintas et al., 2001). These can be neurodegenerative diseases

(e.g. Alzheimer’s and Parkinson’s diseases), non‐neuropathic systemic amyloidosis (e.g. familial amyloidotic polyneuropathies) or non‐neuropathic localized diseases (e.g. hereditary cerebral haemorrhage with amyloidosis). Nevertheless, misfolding diseases also include other disorders in which, apparently, no ‐sheet amyloid fibrils are formed even though the native proteins undergo pathological conformational changes with clinical significance (Ho et al., 2006; Marcon et al., 2006;

Plakoutsi et al., 2006; Rousseau et al., 2006). These protein aggregates retain native‐like structural properties indicative of the presence of mechanisms other than misfold/partial unfolds and ‐strand assembly.

As stated before, impaired post‐translational modifications, such as phosphorylation, advanced glycation, deamidation, racemization, among others, are triggering factors for misfolding/unfolding and protein aggregation. An important irreversible post‐translational modification found in living cells is glycation, whereby protein amino groups form advanced glycation end‐products (AGE) by non‐enzymatic reaction with carbonyl compounds (Monnier &

Cerami, 1981). Glycation was found to induce structural modifications on different proteins (Seidler

& Seibel, 2000; Argirova & Breipohl, 2002; Seidler & Kowalewski, 2003; Kumar et al., 2004) and the formation of amyloid cross‐β structure in albumin (Bouma et al., 2003). Methylglyoxal, a non‐ enzymatic by‐product of dihydroxyacetone phosphate and D‐glyceraldehyde‐3‐phosphate during glycolysis in eukaryotic cells, is the main intracellular glycation agent (Thornalley, 1994; Thornalley,

1996; Chang & Wu, 2006).

Cytochrome c is an all‐alpha hemic protein, strongly conserved, located in the mitochondrial intermembrane area and participates in electron transfer chain. When released from mitochondria, cytochrome c is an important triggering factor of caspase activation pathway in apoptosis (Liu et al.,

1996; Kluck et al., 1997; Li et al., 1997; Yang et al., 1997; Slee et al., 1999).

In this study, we show that cytochrome c is converted into native‐like non‐covalent aggregates and partially unfolded species upon glycation suggesting that, apart from the known effect on the formation of a cross‐β structure, glycation also plays a critical role in protein native‐ like aggregation mechanisms.  54

Chapter II

3. Material and methods

3.1 Methylglyoxal preparation

High purity methylglyoxal was prepared by acid hydrolysis of methylglyoxal 1,1‐ dimethylacetal (Sigma) as reported by Kellum (Kellum et al., 1978) followed by fractional distillation under reduced pressure in nitrogen atmosphere. The concentration of methylglyoxal in stock solutions was determined by end point enzymatic assay as described by Racker (Racker,

1951). Purity was verified by HPLC (Cordeiro & Ponces Freire, 1996) and NMR analysis on a Bruker

Avance 400.

3.2 Cytochrome c glycation

Horse heart cytochrome c (10 mg/ml; Sigma) was incubated with methylglyoxal (10 mM) in

50 mM potassium phosphate buffer, pH 7.0 supplemented with 150 mM of NaF, at 37 ºC in sterile conditions. Control samples were treated in the same way but without methylglyoxal addition.

Aliquots were collected in sterile conditions at defined times from 0 to 400 hours and immediately analysed.

3.3 Cytochrome c aggregation

Aggregation of cytochrome c upon glycation by methylglyoxal was monitored by size exclusion chromatography (SEC) and SDS‐PAGE. Samples were analysed by SEC at incubation times of 0 and 7 days with a LKB Bromma 2150 isocratic pump with an UV detector JASCO 2075.

The mobile phase was 5 mM phosphate buffer pH 7.0 with 150 mM NaF. Separation was achieved on a molecular exclusion analytical column (Amersham‐Pharmacia Superdex™ 75 10/300 GL) at a flow rate of 0.4 ml/min. Eluting peaks were monitored at 280 nm.

Samples with 7 days incubation and the eluting peaks collected from gel filtration were separated by SDS‐PAGE on a Bio‐Rad Mini‐Protean 3 system, using a 15 % separation gel and a 4 % stacking gel. Proteins were stained with Comassie Briliant Blue (Wilson, 1979).

 55

Aggregation without unfolding 

3.4 Structural analysis of cytochrome c

After 7 days incubation with methylglyoxal, glycated and control samples of cytochrome c were analysed by circular dichroism (CD), fluorescence spectroscopy and HPLC. Secondary structure analysis was performed by far‐UV (185‐260 nm) CD in a Jasco J810 Spectropolarimeter equipped with a temperature control unit Julabo F25. Far UV CD spectra were recorded with 0.1 cm

(linear) path length quartz cuvette at 25 ºC. For each spectrum, three scans were averaged and protein concentration was determined by absorbance at 410 nm (ε410=1.06x105 M‐1 cm‐1) in a UV‐

Visible spectrophotometer Jasco V‐530. For protein secondary structure estimation, CD spectra were deconvoluted using the CDSSTR (Johnson, 1999) deconvolution algorithm on Dichroweb

(Lobley et al., 2002; Whitmore & Wallace, 2004). Tertiary structure changes were probed by tryptophan red shift fluorescence analysis. Fluorescence spectra were measured in a SLM Aminco

MC400 instrument equipped with a Julabo F12 temperature control unit. Fluorescence emission spectra were recorded between 300 and 400 nm with excitation at 280 nm in 1 cm path length quartz cuvette at 25 ºC. Alterations on protein surface area were analysed by reverse phase HPLC.

HPLC analysis was performed with protein solutions at a final concentration of 1 mg/ml on a

Beckman Coulter System Gold, equipped with a UV diode‐array detector (Beckman Coulter System

Gold 168) and a fluorescence detector (Jasco FP 2020 Plus). The mobile phase consisted of 0.08 %

(V/V) TFA in type I water (solvent A) and 0.08 % (V/V) TFA in acetonitrile (solvent B), and the elution gradient program was 5 % to 40 % solvent B in 20 min. Separation was achieved on a reversed phase analytical column (LiChrospher 100 Merck RP‐18, 5 μm) at a flow rate of 1 ml/min.

Eluting species were monitored by absorbance at 400 nm.

3.5 Conformational stability measurements

Conformational stability of glycated and non‐glycated cytochrome c species, after 7 days incubation at 37 ºC, was measured by CD in guanidinium hydrochloride (GdnHCl)‐induced protein unfolding experiments at 25 ºC after 2 days incubation. CD denaturation curves were constructed using the ellipticity at 222 nm. Monomeric and dimeric species were analysed according to a two‐ state unfolding model M ↔ U and D2 ↔ 2U respectively using the linear extrapolation method

(Pace, 1986) in a non‐linear least squares fitting procedure and yielded values for ΔGº(H2O), the conformational stability, and m, the dependence of ΔGº on denaturant concentration. Cm, the denaturant concentration at the midpoint of the unfolding transition was calculated as

 56

Chapter II

Cm  Gº H 2O / m . Denaturation curves for monomeric species were analysed considering the equation developed by Santoro & Bolen (Bolen & Santoro, 1988; Santoro & Bolen, 1988).

For dimeric species, denaturation curves were analysed considering equation II.1 U y  yº m GdnHCl   yº m GdnHCl U (equation II.1) D2 D2 U U KU where y represents the experimental variable being used to follow the transition; yD2 and yU are the values of y for D2 and U respectively and [U] is the concentration of U. [U] is given by equation II.2

2 K KU 16KU CT U   U  (equation II.2) 4 4 where CT is total protein concentration used and KU is the equilibrium unfolding constant.

Parameters were obtained by fitting the derived equations to the experimental data by non‐linear regression using the Solver add‐on for Microsoft Exel. Standard errors and the variance‐covariance matrix were computed under the assumption of normal distribution of residuals.

To determine the presence of intermediates in the unfolding pathway of the glycated dimeric form of cytochrome c, gel filtrations in the presence of GdnHCl were performed. Samples (100 μl) were incubated in mobile phase supplemented with GdnHCl at various concentrations (1.5 ‐ 2.5 M) for 2 days. Gel filtration analysis was carried out as described in Cytochrome c aggregation section.

GdnHCl was present in gel filtration mobile phase at sample’s concentration.

4. Results and discussion

4.1 Glycation induces cytochrome c aggregation

A molecular exclusion chromatography was performed for native and methylglyoxal‐ glycated cytochrome c samples at different incubation times (figure II.1). Native cytochrome c appears as a single molecular species, with an apparent molecular mass of 12.9 kDa (elution volume

– 12.96 ml) (figures II.1 A). Gel filtration elution profiles of glycated protein reveal an increasing complexity of post‐glycation cytochrome c forms with the emergence of four new main species.

These species show apparent molecular masses consistent with the dimeric, trimeric and tetrameric forms of cytochrome c (figure II.1 B, fractions D, T3 and T4, respectively). Moreover, another molecular species appears to be present at a higher elution volume (figure II.1 B, fraction U).

 57

Aggregation without unfolding 

Figure II.1 ‐ Gel filtration chromatograms of native and glycated cytochrome c. A – Native cytochrome c; B – Cytochrome c after incubation with methylglyoxal. All samples have the same protein concentration (0.85 mM) and were eluted with a flow rate of 0.4 ml/min. The labels in chromatograms have the following meaning: T4 – Glycated tetramer; T3 – Glycated trimer; D – Glycated dimer; M – Glycated monomer; U – High elution volume species.

Protein cross‐link formation due to glycation does occur (Chellan & Nagaraj, 1999; Verzijl et al., 2002), but when using methylglyoxal, only the lysine‐lysine dimer MOLD or the lysine‐arginine methylglyoxal‐derived cross‐link MODIC may be formed. These are minor advanced glycation end‐products when compared to other AGE. The relative amounts of cytochrome c in aggregated forms suggest that major non‐covalent interactions are likely to be involved. The nature of the interactions in cytochrome c aggregates was first evaluated by SDS‐PAGE (figure II.2).

Figure II.2 ‐ SDS‐PAGE of the glycated and non‐glycated gel filtration fractions. A – Without thermal denaturation. B – With thermal denaturation at 100 ºC for 5 minutes. The labels in gels have the following meaning: LMW, low molecular weight markers (Bio‐Rad); U, high elution volume species; M, glycated monomer; D, glycated dimer; T3+T4 corresponds to fractions containing glycated trimer an tetramer; CytC, native cytochrome c; WP, glycated cytochrome c before gel filtration.

SDS‐PAGE results (figure II.2) are in agreement with the gel filtration data, confirming the presence of the monomeric (M), dimeric (D) and trimeric plus tetrameric (T) glycated cytochrome c

 58

Chapter II forms as the main constituents of the collected chromatographic fractions. Interestingly, fraction D migrated primarily as a dimer but also as a monomer, and fraction T3+T4 migrated primarily as a trimer and tetramer, but also as a dimer and monomer (figure II.2 A). Moreover, boiling the samples for 5 min prior to loading in the gel augmented the dissociation effect (figure II.2 B), showing that cytochrome c aggregates are susceptible to dissociation under thermal denaturation conditions. According to these results, we can assume that, at some extent, non‐covalent aggregates are present in the aggregated forms of glycated cytochrome c. These non‐covalent interactions are strong since the presence of denaturing reagents, as SDS, was not enough to promote aggregates dissociation. Previously, in the works of McCutchen and Goldsteins a similar behaviour was observed when studying native TTR (McCutchen et al., 1995; Goldsteins et al., 1997). In fact, the tetrameric TTR does not completely dissociate in similar denaturating conditions, appearing as a dimer in a SDS‐PAGE gel, even though these dimers are not covalently bound.

The higher retention time form of cytochrome c (figure II.2 – lane U) migrates in SDS‐PAGE gel with an apparent molecular mass corresponding to a monomeric form of cytochrome c in contrast to the apparent molecular mass determined by gel filtration. This result discards the idea of protein fragments and suggests that this form is in fact a monomer of the glycated cytochrome c.

4.2 Structural changes of cytochrome c upon glycation

To evaluate the secondary structure of the different species of post‐glycation cytochrome c, far‐UV CD was used. Figure II.3 shows the CD spectra of the native cytochrome c together with the monomeric, dimeric and trimeric forms of glycated protein, at 25 ºC and physiological ionic strength and pH.

A direct analysis of the CD spectra shows that the content in secondary structure of both native and glycated forms of cytochrome c is apparently very similar. Spectral minima at 208 and

222 nm and a positive band bellow 200 nm, which dominates all spectra, are characteristic of α‐ helical structures (Sreerama & Woody, 2004). Nevertheless, CD spectra of aggregates show some differences when compared to monomers. Aggregates spectra show no loss in secondary structure, but a slightly readjustment of secondary structural elements distribution in relation to monomeric forms. CD spectra deconvolution of native and glycated species using the CDSSTR algorithm

(Johnson, 1999), allowed determination of secondary structure contents presented in table II.1.

 59

Aggregation without unfolding 

Figure II.3 ‐ Far‐UV CD spectra of native and different glycated forms of cytochrome c in 5 mM phosphate buffer pH 7 and 150 mM NaF at 25 ºC after 7 days incubation with methylglyoxal. Native cytochrome c is represented in black, glycated monomer in blue, glycated dimer in red and the glycated trimer is represented in green.

Table II.1 – Distribution of the structural element fractions obtained by deconvolution of CD spectra using CDSSTR algorithm available on Dichroweb. The NRMSD parameter represents the normalized root mean square deviance. Structural element α‐Helix β‐Sheet β‐Turns Unordered structure NRMSD

Native Cytochrome c 0.32 0.08 0.18 0.41 0.024

Glycated Monomer 0.32 0.08 0.19 0.40 0.021

Glycated Dimer 0.25 0.19 0.15 0.40 0.056

Glycated Trimer 0.24 0.19 0.15 0.41 0.057

Deconvolution data confirmed the mainly helical content of native cytochrome c and glycated monomeric forms of cytochrome c. In contrast, upon deconvolution, the data concerning the dimeric and trimeric forms of glycated cytochrome c revealed an increase in β‐sheet content with a corresponding loss of α–helix and β–turns. These results demonstrate that glycated cytochrome c aggregation induces β‐sheet formation, which is usually found in the protein aggregation processes. However, the predominant secondary structure of the post‐glycated cytochrome c high molecular mass aggregates remains α‐helix, suggesting the presence of native‐like aggregates. It is known that the formation of β‐strand conformation during the association of two or more polypeptide chains is a common pathway to defeat the decrease of entropy inherent to the aggregation process, maximizing the enthalpic contribution due to hydrogen bonds formation, an underlying mechanism of amyloidogenesis. Accordingly, the slight increase of β‐sheet during

 60

Chapter II aggregation of post‐glycated cytochrome c can be driven by inter‐monomer hydrogen bond formation.

CD spectra deconvolution of native cytochrome c also shows the presence of a small fraction of ‐sheet content. Although not detected either by X‐ray diffraction or by NMR, Calvert and colleagues have observed a similar occurrence by FTIR (Calvert et al., 1997).

The CD spectrum of the monomeric higher retention time form of cytochrome c (designated as fraction U) shows a deep loss of secondary regular structural elements (data not shown), consistent with the appearance of a partial unfolded form of glycated cytochrome c. This new populated state of monomeric protein species shows a profound effect of methylglyoxal on cytochrome c secondary conformation.

Secondary structure modifications usually produce changes in tertiary structure. To evaluate alterations on the three‐dimensional structure resultant of structural alterations at secondary level, tryptophan intrinsic fluorescence and reverse phase HPLC were used. Monitoring fluorescence changes around tryptophan residues environment can elucidate possible tertiary structure alterations. On the other hand, reversed phase HPLC analysis allows a comparison of the hydrophobic degree of the protein.

Figure II.4 ‐ Fluorescence emission spectra of native cytochrome c (dash), glycated monomer (solid) and glycated dimer (dash dot dash) in 5 mM phosphate buffer pH 7 and 150 mM NaF at 25 ºC after 7 days incubation with methylglyoxal.

Figure II.4 shows an overlay of the fluorescence spectra obtained with native cytochrome c and with the glycated monomeric and dimeric species fluorescence spectra. Fluorescence spectra show that no deviance in native cytochrome c fluorescence emission λmax occurs either upon glycation or with aggregation. These results indicate that tryptophan environment is not affected by glycation neither by aggregation.  61

Aggregation without unfolding 

The hydrophobic properties of the multiple cytochrome c species found in gel filtration assay were characterized by reverse phase HPLC. Figure II.5 shows an overlay of native and glycated cytochrome c chromatogram [A] and an overlay of all the glycated forms of cytochrome c [B].

Figure II.5 – Reverse phase HPLC chromatograms of the different cytochrome c species. A – Native cytochrome c (solid) and glycated cytochrome c without separation in gel filtration chromatography (dash) chromatogram. Peaks are eluted at 15’22’ and 16’15’’ respectively. B – Chromatographic profiles of all glycated cytochrome c forms. All species have the same elution time (16’15’’).

Reverse phase HPLC analysis reveals different chromatographic profiles between native and glycated cytochrome c (figure II.5 A). Native and glycated cytochrome c were eluted at 15’22’’ and

16’15’’ respectively, which is in line with an increased hydrophobicity acquired by the protein upon glycation. This observation can be explained by the loss of superficial charges that occurs at the initial steps of glycation with the Schiff base formation. However, the sample containing the mixture of the different glycated cytochrome c forms reveals only a single peak, differing from gel

 62

Chapter II filtration chromatogram. Figure II.5 B shows that all the glycated cytochrome c forms, separated by gel filtration chromatography, elute in a single band by reverse phase HPLC. This indicates that the different glycated species have the same level of interaction with the reverse phase column and most likely an identical distribution of hydrophobic and hydrophilic areas.

4.3 Conformational stability of cytochrome c species

To understand thermodynamically how glycation can induce aggregation, we have characterized the conformational stability of the glycated monomeric and dimeric forms of cytochrome c and compared with the native protein. Figure II.6 shows the circular dichroism equilibrium denaturation curves at pH 7 and 25 ºC of native and glycated cytochrome c using guanidinium hydrochloride (GdnHCl) as denaturant.

Monomeric cytochrome c unfolds by a two‐state model (M ↔ U) (Hammack et al., 1998), but there are no evidences of glycation‐induced aggregates. To determine the unfolding pathway of glycated dimer, gel filtration experiments in the presence of increasing amounts of GdnHCl were performed.

Gel filtration experiments (figure II.7) support the hypothesis that during the dimeric denaturation process, cytochrome c progressively unfolds into a monomeric unfolded protein, as compared to a similar experiment with monomeric cytochrome c. The total absence of an unfolded dimeric form shows that aggregates are in fact non‐covalent and that cross‐links are not observable in this denaturant concentration range. In the absence of GdnHCl, the glycated dimer appears as single species with a retention time of 28 min (data not shown). This is consistent with an apparent molecular mass of 25 kDa, as expected for the dimeric form of glycated cytochrome c. In the 1.5‐2.5

M GdnHCl concentration range, the dimeric form is in the transition state between native and unfolded forms (figure II.7) and elutes as two distinct peaks, corresponding to the native and unfolded form. Retention times decrease progressively with the increase of denaturant concentration. Hence, within the worked concentration range, the denaturant promotes a partial expansion of the hydrodynamic radius of the protein in both dimeric and monomeric forms.

 63

Aggregation without unfolding 

Figure II.6 ‐ Guanidinium hydrochloride equilibrium denaturation curves of native cytochrome c (A), glycated monomer (B) and glycated dimer (C) at pH 7 and 25 ºC monitored by ellipticity at 222 nm. For the monomeric forms of cytochrome c, the equation used represents a two‐sate unfolding model (M↔U), while for the dimeric form the equation used represents a two‐sate unfolding model (D2↔2U). The curves are non‐linear least squares fits to the equations representing the entire denaturation curve and using a linear extrapolation method to the experimental circular dichroism data. The insets are the residues plots.

To establish the thermodynamic parameters for native and glycated cytochrome c (monomer and dimer), monomeric species were analysed according to a two‐state unfolding model M ↔ U.

Dimeric species was analysed considering a two state model D2 ↔ 2U, as judged by the absence of unfolding intermediates observed in gel filtration experiments. GdnHCl‐induced denaturation was found to be reversible for all cytochrome c forms studied, as judged by circular dichroism

 64

Chapter II experiments. Both fits were made using the linear extrapolation method (Pace, 1986) in a non‐linear least squares fitting procedure and yielded values for ΔGº(H2O), the conformational stability, and m, the dependence of ΔGº on denaturant concentration. Table II.2 shows the values obtained from the curves in figure II.6 for ΔGº(H2O), m and Cm, the denaturant concentration at the midpoint of the unfolding transition. The magnitude of ΔGº(H2O) for the native cytochrome c fits well within typical values obtained for denaturation of similar size proteins (Bartalesi et al., 2002) and the value of m explains the cooperative transition between the two well defined states represented in native cytochrome c curve. When glycation occurs, a significant decrease on cytochrome c conformational stability from 8.07 ± 0.63 to 3.11 ± 0.56 kCal.mol‐1 is observed. Additionally, Cm also suffers a decrease from 2.53 ± 0.39 to 2.21 ± 0.84 M, confirming the reduction of conformational stability upon glycation. The m value reflects the extent of newly exposed surface area upon unfolding (Myers et al., 1995). Though, the observed decrease upon glycation in the m value from 3.19 ± 0.25 to 1.41 ±

0.22 kcal∙mol‐1.M‐1 may indicate a less compact structure for the glycated monomer compared with the native one. The observed loss of conformational stability and the appearance of a less compact structure, validated by the results, justify the aggregation potential of the glycated monomer.

Figure II.7 ‐ Unfolding of dimeric glycated cytochrome c monitored by gel filtration chromatography. Glycated dimer was previous collected from gel filtration in non‐denaturant conditions and incubated with 1.5 (solid), 2 (dot) and 2.5 M (dash) of GdnHCl at 25 ºC for 2 days. Then aliquots were analysed by gel filtration and eluted with the same incubation buffer. The inset represents an overlay of the glycated monomer (solid) and dimer (dash) at 1 M of GdnHCl. Graphic shows that both species unfold to the same denatured monomeric form.

 65

Aggregation without unfolding 

Table II.2 – Thermodynamic parameters from GdnHCl unfolding studies of native and glycated cytochrome c. Parameters were obtained by a direct fit of the model equations to experimental data in figure II.8. ΔGº(H2O) is the protein conformational stability; m is the dependence of ΔGº on denaturant concentration; Cm is the denaturant concentration at the midpoint of the unfolding transition.

ΔGº(H2O) m Cm

‐1 ‐1) (kcal∙mol‐1) (kcal∙mol .M (M)

Native 8.07 ± 0.63 3.19 ± 0.25 2.53 ± 0.39

Glycated Monomer 3.11 ± 0.56 1.41 ± 0.22 2.21 ± 0.84

Glycated Dimer 3.56 ± 1.09 2.13 ± 0.41 1.67 ± 0.49

When compared, glycated monomer and dimer show only a slight increase of conformational stability with aggregation. This distribution for the conformational stability values of the studied cytochrome c species is in agreement with a thermodynamic model for aggregation based on the search for the most stable form. The results on conformational stability of native and glycated forms of cytochrome c explain the aggregation process observed in gel filtration chromatography (figure

II.1). The decrease in conformational stability of cytochrome c upon glycation is balanced with an increase of stability gain upon aggregation. The dimeric form of post‐glycated cytochrome c has a slightly higher conformational stability than the monomeric form.

5. Conclusion

In recent years, the molecular mechanisms of protein aggregation have been the focus of wide attention due to their implications in widespread human and animal pathologies. So far, the major research in the conformational diseases area has been focussed in the implications of protein self‐ association occurring as a result of misfolding/partial unfolding and leads to the formation of aggregates and cross β‐sheet amyloid fibrils. Recently, some papers have reported pathologic association mechanisms of proteins other than cross β‐strand of amyloid fibrils in which aggregates retain native‐like structure, such is the case of human pancreatitis associated protein and of acylphosphatase from Sulfolobus solfataricus (Ho et al., 2006; Plakoutsi et al., 2006).

Here we show that, once glycated with methylglyoxal, cytochrome c begins an aggregation process wherein the formed high molecular mass aggregates retain native‐like structure. Post‐ glycated cytochrome c aggregates have a slight increase of ‐sheet content as judge by CD.

 66

Chapter II

However, the main secondary structure remains α‐helix. Additionally, tryptophan intrinsic fluorescence did not show any differences between native and glycated (monomer and dimer) cytochrome c species. Conformational stability studies show a decrease in ΔGº(H2O) from 8.07 ±

0.63 to 3.11 ± 0.56 kCal.mol‐1 of post‐glycated monomeric form when compared with native monomeric form of cytochrome c. Reduced protein stability causes unfolding and aggregation, still aggregation takes place through native‐like association of post‐glycated cytochrome c. We propose a novel model for native‐like aggregation of methylglyoxal glycated cytochrome c where the aggregation pathway occurs directly through monomer addition in a process that is thermodynamically favoured and where the post‐glycated cytochrome c unfolded species is an off pathway, a clear difference from the amyloidogenic aggregation pathways (figure II.8).

Figure II.8 ‐ Molecular model for cytochrome c aggregation upon methylglyoxal glycation. The native protein is primarily irreversibly glycated by methylglyoxal giving a glycated monomer. This monomer, which has a reduced stability, can follow two different pathways. It can form a rapid equilibrium with a partial unfolded monomeric form, or it can irreversibly go through an aggregation pathway by the sequential addition of glycated monomers.

As stated before, glycation is a relevant post‐translational modification for amyloidogenesis

(Bouma et al., 2003). The present model discloses, for the first time, a native‐like aggregation of a protein induced by glycation which can be implied in complications on apoptotic signalization mechanisms and on electron transfer reaction in mitochondrial electron transfer chain.

 67

Aggregation without unfolding 

 68

Chapter III

Enolase glycation: The in vitro in vivo debate

Ricardo A. Gomes, Luís M. A. Oliveira, Mariana Silva, Carla Ascenso, Alexandre Quintas, Gonçalo Costa, Ana V. Coelho, Marta Sousa Silva, António E. N. Ferreira, Ana Ponces Freire and Carlos Cordeiro. Protein glycation in vivo: functional and structural effects on yeast enolase (2008) Biochem J. 416(3): 317‐26.

Chapter III

1. Abstract

Protein glycation is involved in structure and stability changes that impair protein functionality, which is associated with several human diseases, such as diabetes and amyloidotic neuropathies (Alzheimer’s disease, Parkinson’s disease and Andrade’s syndrome). To understand the relationship of protein glycation with protein dysfunction, unfolding and β‐fibre formation, numerous studies have been carried out in vitro. All of these previous experiments were conducted in non‐physiological or pseudo‐physiological conditions that bear little to no resemblance to what may happen in a living cell. In vivo, glycation occurs in a crowded and organized environment, where proteins are exposed to a steady‐state of glycation agents, namely methylglyoxal, whereas in vitro, a bolus of a suitable glycation agent is added to diluted protein samples. In the present study, yeast was shown to be an ideal model to investigate glycation in vivo since it shows different glycation phenotypes and presents specific protein glycation targets. A comparison between in vivo glycated enolase and purified enolase glycated in vitro revealed marked differences. All effects regarding structure and stability changes were enhanced when the protein was glycated in vitro.

The same applies to enzyme activity loss, dimer dissociation and unfolding. However, the major difference lies in the nature and location of specific advanced glycation end‐products. In vivo, glycation appears to be a specific process, where the same residues are consistently modified in the same way, whereas in vitro several residues are modified with different advanced glycation endproducts.

2. Introduction

Protein glycation is a non‐enzymatic post‐translational modification where arginine and lysine side‐chain amino groups are irreversibly modified by carbonyl compounds, forming AGE

(Westwood & Thornalley, 1997; Grillo & Colombatto, 2008). Increased protein glycation is associated with several human pathologies such as diabetes mellitus and related clinical complications (retinopathy, nephropathy and diabetic vascular diseases) (Brownlee, 1995), uraemia

(Miyata et al., 2000), atherosclerosis (Kume et al., 1995), age‐related disorders (Brownlee, 1995) and neurodegenerative diseases of the amyloid type (Miyata et al., 1994; Vitek et al., 1994; Yan et al.,

1994; Shibata et al., 2002). Glycated proteins are present in β‐amyloid and τ deposits of Alzheimer’s

 71

The in vitro in vivo debate 

disease (Vitek et al., 1994; Yan et al., 1994), in Lewy inclusion bodies of α‐synuclein in Parkinson’s disease (Castellani et al., 1996), in transthyretin amyloid deposits in familial amyloidotic polyneuropathy (Gomes et al., 2005a) and in amyloidotic lateral sclerosis (Shibata et al., 2002). In all of these amyloid pathologies, β‐sheet fibril structure and the presence of AGE are common features, suggesting a possible role for glycation in amyloid formation and pathogenesis. In diabetes mellitus and related clinical complications, D‐glucose was considered to be the main glycation agent leading to irreversible changes in extracellular proteins (Brownlee, 1995). However, several lines of evidence suggests that the increased glycation levels observed in several pathological conditions are not directly related to a higher D‐glucose concentration, but to the increased concentrations of highly reactive low‐molecular‐mass carbonyl compounds, leading to what is described as carbonyl stress (Baynes & Thorpe, 1999; Miyata et al., 2000). In fact, in several diseases where glycation products accumulate, glycaemia is normal (Gomes et al., 2005a). Moreover, glycation is often found in intracellular proteins, as in Alzheimer’s (Yan et al., 1994) and Parkinson’s (Castellani et al., 1996) diseases, where the D‐glucose concentration is negligible. The dicarbonyl compound methylglyoxal is the most significant glycation agent in vivo, considering its high reactivity and continuous formation, mainly by the irreversible β‐elimination of the phosphate group of dihydroxyacetone phosphate and D‐glyceraldehyde‐3‐phosphate (Richard, 1984; Richard, 1993; Thornalley, 1996).

Albeit non‐enzymatic, this is a physiological process that happens alongside glycolysis, hence methylglyoxal formation and protein glycation occur in all living cells. By reacting with the guanine group of arginine, methylglyoxal forms argpyrimidine, MG‐H1 (hydroimidazolones) and THP

(tetrahydropyrimidine) (Westwood & Thornalley, 1997; Grillo & Colombatto, 2008). From its reactions with the ε‐amino group of lysine, it forms CEL [Nε‐(carboxyethyl)lysine] and a MOLD

(methylglyoxallysine dimer) (Westwood & Thornalley, 1997; Grillo & Colombatto, 2008). These specific markers of protein glycation by methylglyoxal are globally known as MAGE. These MAGE have been identified in vivo associated with diabetes (Kilhovd et al., 2003) and in neurodegenerative disorders of the amyloid‐type such as FAP and ALS (Shibata et al., 2002; Gomes et al., 2005a).

Previous evidence suggests that methylglyoxal and not other glycation agents specifically alter the function of several proteins (Portero‐Otin et al., 2002; Speer et al., 2003; Pedchenko et al., 2005). One example is the specific modification of Arg188 of Hsp27 with the formation of argpyrimidine, which is essential to the anti‐apoptotic activity of this protein (Sakamoto et al., 2002). Moreover, glycation by methylglyoxal of small Hsps appears to be essential for its activation, suggesting a

 72

Chapter III mechanism that allows cells to detect and react to carbonyl and unfolding stress (Sakamoto et al.,

2002; Nagaraj et al., 2003; Oya‐Ito et al., 2006; Gomes et al., 2008a).

Research has been conducted in the last decades whereby the effects of glycation on protein structure and function were investigated in conditions that cannot be assigned to what happens in vivo (Luthra & Balasubramanian, 1993; Bouma et al., 2003; Seidler & Kowalewski, 2003; Bakhti et al.,

2007). Use of concentrations of glycation agents up to the molar range is not uncommon. These studies are in sharp contrast with what happens inside living cells, where the methylglyoxal concentration is in the nano to micromolar range, whereas proteins are at a high concentration in a crowded and organized milieu (Srere, 1987). To investigate the effects of glycation on protein structure and function in vivo, cellular models must be sought where glycation conditions may be controlled and yeast offers an ideal cell model to investigate glycation in vivo, now that glycation phenotypes and protein glycation targets have been uncovered (Gomes et al., 2005b; Gomes et al.,

2006; Gomes et al., 2008a). Among these, enolase2p (2‐phospho‐D‐glycerate hydrolase, EC 4.2.1.11), the main glycation target, shows a glycation‐dependent enzyme activity loss (Gomes et al., 2006).

In the present study, enolase2p was purified from Saccharomyces cerevisiae cells in glycation conditions (glycated in vivo) and in non‐glycation conditions. The native enzyme was modified by methylglyoxal in vitro, in conditions often found in the literature. A detailed investigation of the effects of methylglyoxal‐mediated glycation in vivo and in vitro on the structure, thermal stability and enzyme activity of yeast enolase was performed. MAGE location and identification was made by a bottom‐up approach, using MALDI–TOF‐MS (matrix‐assisted laser‐desorption ionization– time‐of‐flight MS) and high mass accuracy FTICR‐MS (Fourier‐transform ion cyclotron resonance

MS).

3. Material and methods

Peptone, yeast extract and agar were from Difco, whereas D‐glucose (microbiology grade),

KCl and MgSO4 were obtained from Merck. Ammonium sulfate, NaH2PO4, Na2HPO4, NaCl, NaF,

DTT (dithiothreitol), iodoacetamide and TFA (trifluoroacetic acid) were from Sigma. Tris, 20%

(w/v) SDS and glycine were from Bio‐Rad. EDTA was from BDH chemicals, whereas phosphoenolpyruvate, methylglyoxal 1,1‐dimethyl acetal and DHB (2,5 dihydroxybenzoic acid) were from Fluka. Modified trypsin was from Promega. GELoader tips were from Eppendorf.

POROS 10 R2 reversed‐phase chromatography medium was from PerSeptive Biosystems. α‐CHCA  73

The in vitro in vivo debate 

(α‐cyano‐4‐hydroxicinamic acid), 3,5 dimethoxy‐4‐hydroxycinnamic acid (sinapinic acid), PepMix1

(MS peptide standards) and ProMix3 (MS protein standards) were obtained from LaserBiolabs.

Amicon filters were purchased from Millipore. BSA protein digest was from Bruker Daltonics.

Acetonitrile and methanol were HPLC‐gradient grade and were obtained from Riedel de Haen; ultrapure water (type I) was produced in a Millipore Milli‐Q system.

3.1 Methylglyoxal preparation

Methylglyoxal was prepared by acid hydrolysis of methylglyoxal 1,1‐dimethylacetal as reported by Kellum and co‐workers (Kellum et al., 1978) and purified by fractional distillation under reduced pressure in a nitrogen atmosphere (McLellan et al., 1992). Once prepared, methylglyoxal solutions were standardized by enzyme assay with glyoxalase I and II, as described previously (Racker, 1951). Purity was verified by HPLC and NMR analysis on a Bruker Advance

400.

3.2 Yeast strains and growth conditions

S. cerevisiae strains, Euroscarf collection (Frankfurt, Germany), were: BY4741 (genotype

BY4741 MATa; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0) and ΔGLO1 (isogenic to BY4741 with

YML004c::KanMX4). Strains were kept in YPGlu agar slopes [0.5% (w/v) yeast extract, 1% (w/v) peptone, 2% (w/v) agar and 2% (w/v) D‐glucose] at 4ºC and cultured in liquid YPGlu medium. To induce protein glycation, ΔGLO1 strain was cultured for 9 days to reach the stationary phase of growth (Gomes et al., 2005b). The reference BY4741 strain was collected at the end of the exponential phase of growth (18 h).

3.3 Enolase purification

Native enolase was purified from BY4741 yeast cells at the end of the exponential phase of growth (18 h), whereas glycated enolase was purified from ΔGLO1 culture at the stationary phase

(9 days), a condition where glycation was previously observed (Gomes et al., 2005b). Purification was achieved by anion‐exchange chromatography and size‐exclusion chromatography after ammonium sulphate protein precipitation from crude extracts, based on a previously described

 74

Chapter III method (Kustrzeba‐Wojcicka & Golczak, 2000). Cells were disrupted by sonication (5 cycles of 1 min at 100W with 1 min cooling on ice). The extract was centrifuged at 40000 g at 4ºC for 30 min to eliminate cell debris and adjusted to 50% saturation of ammonium sulfate. Saturation was subsequently adjusted to 67% by adding solid ammonium sulfate. After centrifugation at 40000 g at

4ºC for 30 min, the supernatant was made 100% saturated and centrifuged again. The pellet, containing enolase, was resuspended in 20 mM Tris/HCl (pH 8.2), containing 5 mM MgSO4 and 1 mM EDTA. Sample was dialysed overnight at 4ºC against the same buffer to remove ammonium sulfate and loaded on to an ion‐exchange chromatography DEAE‐Sephadex A‐50 column equilibrated with 20 mM Tris/HCl (pH 8.2), containing 5 mM MgSO4 and 1 mM EDTA. Proteins were eluted with a linear NaCl gradient (0–0.5 M) at a flow rate of 1 ml ∙ min−1 and the eluate was monitored at 280 nm. Protein‐containing fractions were collected and probed by dot blot analysis using an antiyeast enolase antibody (a gift from Dr H. M. Park, Department of Microbiology,

Chungnam National University, Korea). Fractions containing enolase were collected, concentrated by ultrafiltration using Amicon filters and applied to a gel‐filtration CM‐Sephadex C‐50 column, equilibrated with 50 mM NaH2PO4/Na2HPO4 buffer (pH 7.4) containing 150 mM NaCl, 5 mM

MgSO4 and 1 mM EDTA. Proteins were eluted with the same buffer at a flow rate of 1 ml ∙ min−1.

Again, fractions containing enolase, probed by dot blot, were collected and combined. In the purification of glycated enolase, the protein fractions were also probed by dot blot with an anti‐

MAGE antibody (a gift from Dr. Ram Nagaraj, CaseWestern University, Cleveland, OH, U.S.A.).

Enolase purity was evaluated by SDS/PAGE.

3.4 In vitro glycation of purified enolase by methylglyoxal

Purified native enolase (5 μM) was incubated with 10 mM methylglyoxal in 100 mM potassium phosphate buffer (pH 7.4) at 30 ºC for 5 days in sterile conditions. Enolase concentration was determined spectrophotometrically (ε280 = 0.89 ml ∙ mg−1 ∙ cm−1) (Huang & Dong, 2003) in a

UV–visible Jasco V‐530 spectrophotometer.

3.5 Western blot and HPLC analysis

Proteins (30 μg of protein per lane) were separated by SDS/PAGE in a Mini‐protean 3 system

(Bio‐Rad), using a 12% polyacrylamide separation gel and a 6% polyacrylamide stacking gel.

Proteins were transferred on to PVDF membranes (Hybond‐P; Amersham Pharmacia Biotech),  75

The in vitro in vivo debate 

using the Mini Trans‐Blot system (Bio‐Rad). Transfer was performed with 39 mM glycine, 48 mM

Tris, 0.0375% SDS and 20% (v/v) methanol. Prestained standard proteins (Bio‐Rad) were also loaded on to the gel. Total proteins were stained with Ponceau S solution [0.5% Ponceau S in 1%

(v/v) glacial acetic acid] to confirm protein transfer. For the dot blot assay, purified proteins were applied directly on to PVDF membranes previously activated with methanol and equilibrated with transfer buffer. The membranes were blocked overnight at 4 ºC in 1% (v/v) blocking solution

(Roche) in TBS [50 mM Tris/HCl with 150 mM NaCl (pH 7.5)]. The anti‐MAGE antibody was used diluted 1:5000 in 0.5% blocking solution in TBS for 3 h, whereas the anti‐enolase antibody was used diluted 1:10000 in the same conditions. Washes, secondary antibody and detection procedures were performed using the BM Chemiluminescence Western Blotting Kit (Roche) following the manufacturer’s instructions.

Detection of glycation‐induced fluorescence was monitored by reversed‐phase HPLC on a

Beckman–Coulter System Gold equipped with a Beckman–Coulter high‐pressure binary gradient pump 126, a Beckman–Coulter 168‐diode‐array detector (1 nm resolution, 200–600 nm) and a fluorescence detector FP‐2020 Plus (Jasco). The mobile phase consisted of 0.08% TFA in type I water

(solvent A) and 0.08% TFA in acetonitrile (solvent B), and the elution gradient program was: 10–

80% solvent B in 30 min and 80–100% solvent B in 10 min. Separation was achieved on a reversed‐ phase analytical column (LiChrospher 100 Merck RP‐18, 5 μm) at a flow rate of 1 ml ∙ min−1. Eluting species were monitored by the fluorescence signal at λexc/λem of 320/385 nm, characteristic of argpyrimidine.

3.6 MS analysis

MALDI–TOF mass spectra were acquired in a Voyager‐DE STR MALDI–TOF (Applied

Biosystems). FTICR‐MS mass spectra were obtained in a Bruker Apex Ultra with a 7 Tesla magnet

(Bruker Daltonics). For intact protein mass measurement, sinapinic acid (20 mg ∙ ml−1) prepared in

70% (v/v) acetonitrile with 0.1% TFA was used as the matrix, and MALDI–TOF spectra were obtained in positive linear mode. To identify the purified proteins and assign the glycated amino acid residues, a peptide mass fingerprint was performed. Protein bands were excised and subjected to reduction, alkylation and digestion with sequencing‐grade modified trypsin in gel, according to

Pandey and co‐workers (Pandey et al., 2000). The peptide mixture was purified, concentrated by R2 pore microcolumns (Gobom et al., 1999) and eluted directly to the MALDI target plate with 0.8 μl of

 76

Chapter III recrystallized matrix α‐CHCA (10 mg ∙ ml−1) prepared in 70% (v/v) acetonitrile with 0.1% TFA.

Monoisotopic peptide masses were used to search for homologies and protein identification with

Peptide Mass Fingerprint of Mascot (http://www.matrixscience.com). The identification of glycated amino acid residues was performed as described previously (Gomes et al., 2006). Briefly, a glycated enolase peptide should have a miscleavage associated with the defined mass increment of a specific

MAGE. Moreover, an arginine modification should have a miscleavage in an arginine residue and the same holds true for lysine modifications (Gomes et al., 2006). The analysis of glycated peptides was also performed by FTICR‐MS. In this case, besides in‐gel digestion, proteins were hydrolysed in‐solution, essentially as described (Olsen & Mann, 2004). The resulting peptide mixture was also purified using PerfectPure C‐18 tips (Eppendorf) and diluted in 50% (v/v) methanol with 1% (v/v) formic acid for ESI (electrospray ionization)–FTICR‐MS analysis. The peptide mixture was analysed by MALDI–FTICR‐MS, using DHBmatrix [10 mg ∙ ml−1 prepared in 70% (w/v) acetonitrile with 0.1%

TFA] in an Anchorchip MALDI target (Bruker Daltonics).

3.7 Structure and stability analysis

Structural analysis was performed by CD spectroscopy and gel filtration chromatography.

Prior to CD analysis, individual protein species were separated by gel filtration on an analytical column (Amersham‐Pharmacia SuperdexTM 75 10/300 GL) with 10 mM phosphate buffer (pH 7) containing 100 mM NaF as the mobile phase at a flow rate of 0.4 ml ∙ min−1 (LKB Bromma 2150 isocratic pump with a UV detector JASCO 2075). Eluting peaks were monitored at 280 nm and individual protein fractions were collected for further analysis. Secondary structure analysis was performed by far‐UV (185–240 nm) CD in a Jasco J810 spectropolarimeter at 25 ºC (Julabo F25 temperature control unit) with a 0.1 cm path length. CD spectra were deconvoluted using the

CDSSTR algorithm (Johnson, 1999) on Dichroweb (Lobley et al., 2002; Whitmore & Wallace, 2004)

(http://www.cryst.bbk.ac.uk/cdweb/html/home.html). Molar ellipticity was calculated on the basis of a mean residue mass of 107.13 Da. All spectra were solvent baseline corrected. Conformational stability measurements were performed by thermal‐induced protein unfolding. CD denaturation curves were constructed by raising the temperature from 20 to 85 ºC and measuring the ellipticity at

222 nm. The Tm value (temperature at which 50% of denaturation occurs) of native and glycated enolase was calculated as previously described (Pace et al., 1990).

 77

The in vitro in vivo debate 

3.8 Enolase activity assay

Enolase activity was determined at 30 ºC in a 1.5 ml reaction volume, on a Beckman DU‐7400 diode array spectrophotometer, with temperature control and magnetic stirring, essential to maintain isotropic conditions during the assay. Enolase activity was followed by measuring phosphoenolpyruvate consumption at 240 nm and its concentration was calculated using ε=1.42 mM−1 ∙ cm−1 (the present study). The reaction mixture, containing 50 mM Tris/HCl (pH 7.4), 100 mM

KCl, 1 mM MgSO4, 0.01 mM EDTA and a known amount of protein, was pre‐incubated for 10 min and the reaction was started by the addition of phosphoenolpyruvate.

3.9 Protein structure

The enolase dimer structure was represented by PDB entry 1ebh. Molecular graphic images were produced using the UCSF Chimera package from the Resource for Biocomputing,

Visualization, and Informatics at the University of California, San Francisco, CA, U.S.A. [supported by NIH (National Institutes of Health) P41 RR‐01081] (Pettersen et al., 2004). Relative solvent surface accessibility was calculated according to Gerstein (Gerstein, 1992).

4. Results and discussion

4.1 Characterization of enolase glycation by MS

Native enolase was purified from the BY4741 reference strain at the end of the exponential phase of growth and glycated in vitro by incubation with 10 mM methylglyoxal in potassium phosphate buffer, the most common glycation condition (Ahmed et al., 1997; Shipanova et al., 1997;

Kang, 2003). Purified native enolase also served as a control for enzyme activity, secondary structure composition and thermal stability. In vivo‐glycated enolase was purified from the ΔGLO1 strain, lacking the glyoxalase I gene and enzyme activity, hence with a higher intracellular methylglyoxal concentration than the reference strain. At the stationary phase of growth, glycation was observed (Gomes et al., 2005b). Protein purity and identity were verified by SDS/PAGE and western blot analysis with anti‐yeast enolase antibody respectively. Protein identity was further confirmed by peptide mass fingerprint after in‐gel trypsin digestion and MALDI–TOF analysis of the resulting peptide mixture (results not shown).

 78

Chapter III

AB Eno2 (126‐138) 25 Eno2 1578.9 1497.8 (88‐102) 1.60

Eno2 1557.8 Eno2 2471.4 (241‐254) (415‐436) Intensity Intensity Eno2 (32‐53)

Eno2 (32‐53) 2328.3

2344.3

0 0 1500 1580 m/z 2150 m/z 2450 E:\Ricardo TRABALHO\PHD\MS\Nova temporada\17_11_2006\Ricardotxt\3.1.txt (15:25 03/12/07) Native enolase C:\PHD\MS\Nova temporada\17_11_2006\Ricatemporada\17_11_2006\Ricardotxt\2.1.txtrdotxt\2.1.txt (11:45 07/14/07) Native enolase 8.00 1498.0 Eno2 (126‐138) 0.50

Eno2 (415‐436) 2471.2

80 Da Eno2 + Eno2 2441.2

(241‐254) Intensity SERLAKLNQLLR (88‐102) Eno2 Intensity (403‐414) (32‐53)

1557.9 1578.9 1520.3 2328.2

2211.2

0 0 2150 2450 1500 m/z 1580 m/z

C:\PHD\MS\Nova temporada\17_11_2006\gly vivo.massml (00:37 07/14/07) C:\PHD\MS\Nova temporada\17_11_2006\Ricardotxt\4.1.txt (11:40 07/14/07) Spectrum Notes: Usar este para a teseEnolase glycated in vivo Enolase glycated in vivo

Eno2 Eno2 1557.8 (126‐138) (241‐254) 4.00 Eno2 18 (415‐436) 54 Da Eno2 + SVYDSRGNPTVEVEL (32‐53) 2328.2 1497.9 Eno2 TTEK (9‐27) (88‐102) 2471.3 54 Da + 2178.2 54 Da SGETEDTFIADLVVGLRTG 1578.9 + GVFRSIVPSGASTGV QIK (375‐396) HEALEMR (28‐49) 2441.3 Intensity 80 Da 2196.2 Intensity + SERLAKLNQLLR 2354.3 2403.4 (403‐414)

1520.5

0 0 2150 1500 m/z 1580 m/z 2450 C:\PHD\MS\Nova temporada\17_11_2006\Ricardotxt\5.1.txt (11:37 07/14/07) Enolase glycated in vitro Enolase glycated in vitro

Figure III.1. Chemical detection and molecular location of MAGE in enolase. Glycated peptides show miscleavages associated with specific mass increments characteristics of a given MAGE (Gomes et al., 2006). Figure shows sections of MALDI‐TOF mass spectrum where the appearance of new peptides with MAGE in enolase glycated in vivo (A) and in vitro (B) are observed, in comparison with the native protein. The complete analysis of the mass spectrometry data is presented on table III.2 and III.3.

 79

The in vitro in vivo debate 

As previously shown, peptide mass fingerprint data contain hidden information regarding

MAGE nature and location (Gomes et al., 2006). Since only lysine and arginine residues are modified, tryptic digestion of glycated proteins will produce peptides with at least one miscleavage associated with a defined mass increase corresponding to a specific MAGE (Figure III.1). Proteins were also trypsin hydrolysed in solution to increase the sequence coverage and therefore improve

MAGE identification. The resulting peptide mixtures were analysed by MALDI–TOF, ESI–FTICR‐

MS and MALDI–FTICR‐MS. With the combination of these MS techniques, sequence coverage of approx. 70% was obtained in all cases.

When enolase is glycated in vitro, ten out of 14 arginine residues are modified by methylglyoxal in the form of MG‐H, whereas only one lysine residue, either Lys336 or Lys337, is modified to CEL, as observed by MALDI–TOF and MALDI–FTICR (Table III.1). These results are consistent with previous studies that point to MG‐Hs as the most abundant modifications (Ahmed et al., 2003; Ahmed et al., 2005). The analysis of the in‐solution digest by ESI–FTICR‐MS also revealed that some arginine residues even form different MAGE on different enolase molecules, such as Arg14 which may be modified as MG‐H (observed mass of 2178.056 Da, corresponding to enolase peptide 9‐27 with m/z of 2124.045 plus 54.011 Da of a MG‐H modification) or as THP

(observed mass of 2268.088 corresponding to the same enolase peptide 9‐27 plus 144.042 Da of a

THP modification) (Table III.2). The same MAGE replacement was observed for Arg119 and Arg184

(Table III.2).

This molecular heterogeneity can be seen in a linear mode MALDI–TOF mass spectrum of in vitro‐glycated enolase, showing a large mass increase and peak broadening compared with the molecular mass of native enolase (Figure III.2). The formation of different MAGE on the same amino acid residues hints that glycation is not specific in vitro. When enolase is glycated in vivo, only five arginine residues are found to be modified: Arg402 or Arg405, Arg119 and Arg8 as MG‐

Hs and Arg405 and Arg414 as argpyrimidine (Table III.1). Only one lysine residue was found as

CEL. These modifications appear to be specific since no other MAGE were found at these positions.

In this case, the protein molecular mass increase is negligible, as observed by linear mode MALDI–

TOF, indicating a lesser glycation extent, and no peak broadening or asymmetry, consistent with a homogeneous distribution of enolase molecular species (Figure III.2).

 80

Chapter III

Table III.1. Chemical identification and molecular localization of MAGE in enolase by MALDI‐TOF and MALDI‐FTICR mass spectrometry. Glycated residues are underlined. MG‐H, hydroimidazolone; Argp, argpyrimidine; CEL, N‐ (carboxyethyl)lysine.

Observed Theoretical Mass Glycated Glycation mass peptide Peptide sequence increase MAGE residue (Da) mass (Da) (Da) VYARSVYDSR 1269.67 1215.61 54.01 MG‐H R8 (5‐14) TGAPARSERLAK R402 or 1310.63 1256.71 54.01 MG‐H (397‐408) R405 SERLAKLNQLLR 1520.14 1440.86 80.03 Argp R405 (403‐414) LNQLLRIEEELGDK 1724.02 1669.92 54.01 MG‐H R414 (409‐422) IATAIEKKAADALLL K336 or 1741.92 1669.03 72.02 CEL K (330‐345) K337 LGANAILGVSMAAA 2010.19 1956.07 54.01 MG‐H R119 In vitro RAAAAEK (105‐125) TFAEAMRIGSEVYHN 2020.12 1965.99 54.01 MG‐H R184 LK (178‐194) SVYDSRGNPTVEVEL 2178.18 2124.05 54.01 MG‐H R14 TTEK (9‐27) GVFRSIVPSGASTGVH 2354.32 2300.18 54.01 MG‐H R31 EALEMR (28‐49) SGETEDTFIADLVVGL 2403.42 2349.23 54.01 MG‐H R391 RTGQIK (375‐396) TAGIQIVADDLTVTN 2635.57 2581.42 PARIATAIEK 54.01 MG‐H R329 (312‐336) TGAPARSERLAK R402 or 1310.63 1256.71 54.01 MG‐H (397‐408) R405 SERLAKLNQLLR 1520.23 1440.86 80.03 Argp R405 (403‐414) IATAIEKKAADALLL K336 or 1741.90 1669.03 72.02 CEL K (330‐345) K337

SKLGANAILGVSMAA In vivo 2252.20 2171.20 ARAAAAEK 54.01 MG‐H R119 (103‐125) LNQLLRIEEELGDK 1750.00‡ 1669.96 80.03 Argp R414 (409‐422) AVSKVYARSVYDSR 1654.83‡ 1600.84 54.01 MG‐H R8 (1‐14) ‡ Only observed by MALDI‐FTICR‐MS

 81

The in vitro in vivo debate 

Table III.2. Identification and localization of MAGE in enolase glycated in vitro by methylglyoxal using ESI‐FTICR‐MS. Glycated residues are underlined. Noticeably, the same glycated amino acid residue appears with different chemical modifications. MG‐H, hydroimidazolone; THP, tetrahydropyrimidine. Theoretical Mass Observe Glycated Charge [M+H]+ peptide Peptide sequence Increas MAGE d m/z residue mass (Da) e (Da) SKLGANAILGV 742.388 +3 2225.207 2171.196 SMAAARAAAA 54.011 MG‐H R119 EK (103‐125) LGANAILGVSM 1005.508 +2 2010.080 1956.070 AAARAAAAEK 54.011 MG‐H R119 (105‐125) SKLGANAILGV 772.375 +3 2315.239 2171.196 SMAAARAAAA 144.042 THP R119 EK (103‐125) SVYDSRGNPTV 1089.478 +2 2178.056 2124.046 54.011 MG‐H R14 EVELTTEK (9‐27) SVYDSRGNPTV 726.67 +3 2178.056 2124.046 54.011 MG‐H R14 EVELTTEK (9‐27) SVYDSRGNPTV 879.771 +3 2637.315 2583.305 EVELTTEKGVFR 54.011 MG‐H R14 (9‐31) SVYDSRGNPTV 756.68 +3 2268.088 2124.046 144.042 THP R14 EVELTTEK (9‐27) TFAEAMRIGSE 1010.459 +2 2019.996 1965.985 VYHNLK 54.011 MG‐H R184 (178‐194) TFAEAMRIGSE 1055.471 +2 2110.027 1965.985 VYHNLK 144.042 THP R184 (178‐194)

This molecular heterogeneity can be seen in a linear mode MALDI–TOF mass spectrum of in vitro‐ glycated enolase, showing a large mass increase and peak broadening compared with the molecular mass of native enolase (Figure III.2). The formation of different MAGE on the same amino acid residues hints that glycation is not specific in vitro. When enolase is glycated in vivo, only five arginine residues are found to be modified: Arg402 or Arg405, Arg119 and Arg8 as MG‐Hs and

Arg405 and Arg414 as argpyrimidine (Table III.1). Only one lysine residue was found as CEL.

These modifications appear to be specific since no other MAGE were found at these positions. In this case, the protein molecular mass increase is negligible, as observed by linear mode MALDI–

TOF, indicating a lesser glycation extent, and no peak broadening or asymmetry, consistent with a homogeneous distribution of enolase molecular species (Figure III.2).

 82

Chapter III

AB C Native enolase Enolase glycated in vitro Enolase glycated in vivo

Figure III.2. Analysis of intact protein mass by linear MALDI‐TOF of native (A), in vitro glycated (B) and in vivo glycated enolase (C). For all mass spectra, a peak with m/z similar to theoretical yeast enolase 2 molecular mass (46 782 Da) was obtained. With glycation in vitro, a broadening of the peak is observed indicating higher sample heterogeneity.

To gain insights into the susceptibility of arginine residues towards methylglyoxal‐derived glycation, its partial solvent exposure was calculated according to Gerstein (Gerstein, 1992) (Figure

III.3). For arginine modifications, no obvious relationship exists between the partial solvent exposure of amino groups and the susceptibility towards glycation. It is quite interesting to notice that arginine residues with reduced surface exposure of both side chain amino groups, such as arginine residues 119, 391, 405 and 414, are glycated. Meanwhile, the two arginine residues that show the highest surface exposure (Arg200 and Arg288) were not found to be glycated. By contrast, solvent exposure appears to be a determinant factor for glycation of lysine residues since Lys336 and Lys337 show the highest solvent exposure of all lysine residues in enolase (results not shown).

35

30

25

20

15

Surface exposureSurface 10

5

0 R8 R14 R31 R49 R119 R184 R329 R391 R402 R405 R414 R200 R288 R374 Glycated arginine residues Non-glycated arginine residues Figure III.3. Surface exposure of arginine side chain in yeast enolase, calculated according to Gerstein (Gerstein, 1992). Glycated and non‐glycated arginine residues are shown.

 83

The in vitro in vivo debate 

4.2 Glycation effects on enolase folding structure and enzyme activity

The effects of glycation on enolase structure were evaluated by far‐UV CD spectroscopy after size‐exclusion chromatography separation of each molecular form of enolase in solution (native, in vivo‐ and in vitro‐glycated). Gel‐filtration chromatograms clearly show two major peaks, one eluting at 20–25 min and the other at 35–40 min (Figure III.4 A). The first peak corresponds to the enzymatically active enolase dimer (90 kDa) with native secondary structure elements (Figures III.4

B and C). The second peak immunoreacts with an anti‐yeast enolase antibody, but shows no CD signal and no enzyme activity, indicating the presence of unfolded, inactive enolase (Figures III.4 B

– D). Lack of absorption at 222 nm and 208 nm in the CD spectrum indicates a complete loss of regular secondary structural elements, consistent with the absence of enzyme activity. Similar results were obtained for folded and unfolded protein fractions of glycated enolase (results not shown). A comparison of the size‐exclusion chromatograms show that, when glycation occurs, a higher fraction of unfolded inactive enolase is observed relative to the dimeric, folded enzyme

(Figure III.4 A). For native enolase, the unfolded to folded area ratio is approx. one, whereas a 2‐ fold increase in this ratio is observed for in vivo‐glycated enolase. When enolase is glycated in vitro, an even higher amount of unfolded enolase is observed, a 9‐fold increase relative to the native enzyme.

If glycation promotes protein dissociation and unfolding, then both protein fractions should be glycated. To confirm this hypothesis, these fractions were separated by size‐exclusion chromatography, analysed by reversed‐phase HPLC with fluorescence detection at λexc 320 and

λem 385 nm (characteristic of argpyrimidine) and by western blot analysis using an anti‐MAGE antibody. Indeed, both protein fractions derived from enolase glycated in vivo and in vitro contain argpyrimidine and other MAGE (Figure III.4 E). These results clearly show that enolase glycation causes protein unfolding and since glycation is an irreversible process, unfolded protein may not be refolded back to the active enzyme form. The effect is far more severe when the protein is glycated in vitro.

Once the different enolase molecular species were separated by size‐exclusion chromatography, samples were analysed by far‐UV CD spectroscopy. Striking differences were observed, particularly within the regions of 195 nm, 208 nm and 222 nm (Figure III.5 A). In native enolase, the α‐helical content was 40%, β‐sheet was 20%, 21% of turns and 19% of unordered structure (Table III.3). These values are in agreement with the values of 37.6% for α‐helix, 21% of β‐ sheet, 26% of turns and 15.4% unordered structures obtained by Fourier transformed IR  84

Chapter III spectroscopy (Huang & Dong, 2003). X‐ray crystallography analysis of yeast enolase, estimated an

α‐helical content of 37.3% and 17.2% of β‐sheet (Stec & Lebioda, 1990). In vivo‐glycated enolase shows little to no secondary structure loss. However, a redistribution of secondary structure elements is evident, with an increase in unordered structure from 19% to 25%, a reduction in the α‐ helical content from 40%to 35%, whereas β‐sheet content remains unchanged (Table III.3). When enolase is glycated in vitro, a distinct scenario emerges. There is a much higher loss of α‐helix content, from 40% to 17%, and a large increase in β‐sheet, from 20% to 32%, compared with in vivo‐ glycated enolase (Table III.3). Unordered structure elements also increase, from 19% to 32%.

Native In vivo glycated In vitro glycated A 0.025 I 0.02 II II

m 0.015 II I 0.01 Abs 280n

0.005 I

0 10 15 20 25 30 35 40 45 50 10 15 20 25 30 35 40 45 50 10 15 20 25 30 35 40 45 50 Time (min) Time (min) Time (min)

BC30000 0.85 PeakFraction 1 I Structure analysisFractionPeak 1 I Enzyme activity Peak 2 25000 FractionPeak 2 II 0.83 Fraction II 0.81 20000 0.79 15000 0.77 10000 0.75 5000

[PEP] mM[PEP] 0.73 0 0.71

Mean residue ellipticity -5000 0.69

-10000 0.67

-15000 0.65 185 195 205 215 225 235 0 5 10 15 20 Wavelength (nm) Time (min) DE0.0016 0.0016 0.0014 0.0014 In vivo In vitro

Native glycated glycated 0.0012 0.0012 Fraction I

0.0010 0.0010 Fraction I Fraction II 0.0008 0.0008 Volts Volts 0.0006 0.0006

Fraction II 0.0004 0.0004

0.0002 0.0002

0.0000 0.0000

3 4 5 6 7 8 9 10 11 12 13 Minutes Figure III.4. Characterization of the different enolase species. A) Size exclusion chromatography of native, glycated in vivo and glycated in vitro enolase, showing the presence of two major protein fractions (marked as fraction I and II). B) Far‐UV CD spectra of fraction I (black) and fraction II (light gray) collected from gel filtration of native enolase. Contrary to fraction I, no CD signal was observed in fraction II, showing a complete loss of secondary structure elements. C) Enolase activity assays of fraction I (black) and II (light gray). Consistent with the lack of secondary structure, no enzyme activity was detected in fraction II. D) Dot‐blot analysis of both fractions with anti‐yeast enolase antibody, with positive results. E) Glycation analysis by HPLC and dot‐blot of fraction I (black) and fraction II (light gray) from enolase glycated in vivo. Both fractions show a fluorescent peak around 6 min at wavelengths characteristic of argpyrimidine (ex.320/em.385) indicating that both fractions are glycated. The positive signal obtained by dot‐blot analysis with anti‐MAGE antibody also indicates that both fractions are glycated.

 85

The in vitro in vivo debate 

Table III.3. Distribution of secondary structure elements for native, in vivo and in vitro glycated enolase, obtained by CD spectra deconvolution with the CDSSTR algorithm (Dichroweb; http://www.cryst.bbk.ac.uk/cdweb/html/home.html). The NRMSD parameter is the normalized root mean square deviation. Unordered Structural elements ‐Helix ‐Sheet ‐Turns NRMSD structure Native 0.40 0.20 0.21 0.19 0.011 Glycated in vivo 0.35 0.20 0.20 0.25 0.017 Glycated in vitro 0.17 0.32 0.19 0.32 0.026

To evaluate the glycation effects on the structural stability of enolase, thermal denaturation of the native, in vivo‐ and in vitro‐glycated enolase was monitored by CD spectroscopy. Glycation shifts the thermal denaturation curve of enolase to higher temperatures, indicating an increased resistance to thermal unfolding (Figure III.5 B). In fact, the Tm for native enolase is 53.6 ºC, whereas in vivo‐glycated enolase shows a Tm of 58.6 ºC and the in vitro‐glycated enolase an even higher Tm of 61.4 ºC. Enolase thermal denaturation is an irreversible process, as confirmed by protein aggregation and lack of secondary structure analysed by CD when the temperature was returned to

25 ºC (results not shown), hence the determination of thermodynamic parameters could not be performed.

A B 30000 1

25000 0.9 NativeNon-glycated Native Glycated in vivo 20000 Glycated in vivo 0.8 Glycated in vivo Glycated in vitro Glycated in vitro Glycated in in vitro vitro 15000 0.7

0.6 10000 0.5 5000 0.4 0 0.3 Mean residue ellipticity residue Mean

-5000 value) (normalized mDEG 0.2 -10000 0.1

-15000 0 185 195 205 215 225 235 20 40 60 80 Wavelength (nm) Temp. (ºC)

Figure III.5. Enolase structure and stability. A) Far‐UV CD spectra of native, in vivo glycated and in vitro glycated enolase between 185 and 240 nm. B) Thermal denaturation of native, in vivo and in vitro glycated enolase. Upon glycation, a shift towards higher melting temperatures is observed.

Secondary structure changes are associated with protein function modifications. Therefore glycation‐induced conformational changes are likely to have pronounced effects on enolase activity.

Indeed, we observed a marked decrease in enolase activity upon glycation (Figure III.6). In vivo‐

 86

Chapter III glycated enolase shows a 65% activity loss compared with native enolase. When enolase is glycated in vitro, an even more severe activity loss is observed (84%). As glycation causes enolase denaturation with a consequent enzyme inactivation, it could be argued that the loss of enzyme‐ specific activity may be solely explained by the higher amount of unfolded inactive protein in the sample. This implies that, if the activity of glycated folded enolase remained the same, the specific activity would decrease. To investigate this hypothesis, folded and unfolded fractions from in vivo and in vitro‐glycated enolase were separated by size‐exclusion chromatography and enzyme activity was determined for each individual fraction (fraction I being folded active and glycated enolase, whereas fraction II is unfolded glycated enolase). In both cases, no enolase activity was detected in fraction II, consistent with the lack of secondary structure. In fraction I, enolase activity was detected, albeit the specific activity was again much lower than that of the native enzyme

(Figure III.6).

300 )

-1 250 .mg -1 200 M.min  150

100

50 Enolase activity (

0 Native123456Glycated Glycated Glycated Glycated in vivo in vitro in vivo in vitro (Fraction I) (Fraction I) Figure III.6. Glycation effects on enolase activity. Enolase activity was determined by the consumption of phosphoenolpyruvate, as described. Glycation induces a considerable activity loss. Fraction I contains folded active enolase. Data shown are averages from three independent activity assays ± SD.

5. Conclusion

Arginine residues have a probability of approx. 20% of being located in ligand‐ and substrate‐ binding sites of proteins (Gallet et al., 2000). Hence, methylglyoxal‐derived arginine glycation is expected to have significant effects on protein structure and function, being related to several human pathologies (Vitek et al., 1994; Brownlee, 1995; Castellani et al., 1996; Miyata et al., 2000;

Gomes et al., 2005a). Therefore this post‐translational modification has been the subject of intensive  87

The in vitro in vivo debate 

research, where in vitro‐glycation of clinically relevant and model proteins was investigated (Luthra

& Balasubramanian, 1993; Bouma et al., 2003; Ahmed et al., 2005). The major drawback of this approach lies on the differences between glycation conditions in vitro and in vivo. In vitro, non‐ physiological concentrations of glycation agents are used, from millimolar to even molar concentrations (Luthra & Balasubramanian, 1993; Bouma et al., 2003; Seidler & Kowalewski, 2003;

Bakhti et al., 2007). Additionally, protein interactions are not taken into account. Also, protein turnover and the action of molecular chaperones, some of which are activated upon glycation by methylglyoxal (Nagaraj et al., 2003; Gomes et al., 2006; Oya‐Ito et al., 2006), is absent. These differences highlight the importance of investigating weather or not protein glycation mechanisms and their biochemical effects in vitro are reliable comparing to what happens in a living cell. For protein glycation in vivo, our previous studies validated yeast as an eukaryotic cell model to investigate protein glycation in vivo (Gomes et al., 2005b; Gomes et al., 2006). Enolase, the major glycation target, shows glycation‐dependent activity loss (Gomes et al., 2006), providing an important model for studying glycation effects in vivo.

MS analysis has shown that protein glycation is different in vivo and in vitro, indicating that enolase glycation in vivo is site‐specific whereby only a few amino acid residues are consistently modified with the same MAGE. By contrast, glycation in vitro is a heterogeneous process, resulting in the formation of a complex population of enolase molecules with different glycation profiles. In vitro, different MAGE may be present at the same arginine residue, in different protein molecules.

Glycation specificity is not related to the partial solvent exposure of arginine residues (Figure III.3).

We previously suggested that the arginine‐rich deep crevice in enolase protein structure, accessible to the solvent and located at the dimer interface, may be a favourable hotspot for the occurrence of glycation (Gomes et al., 2006). Several glycated arginine residues identified after in vitro‐glycation and almost all glycated arginines in vivo are located in this cleft (Figures III.7 A and B).

Interestingly, glycation was not detected in the two most exposed arginine residues (Arg200 and

Arg288), outside the cleft (Figure III.7 C).

In vivo‐glycated enolase shows an increase of unordered structure. α‐Helical content decreases and Tm increases with glycation, suggesting that hydrophobic effects have a more pronounced effect. When enolase was glycated in vitro, besides an enhanced increase in unordered structure and a decrease in α‐helix, a marked gain of β‐sheet was also observed and the Tm increase is more pronounced.

 88

Chapter III

Glycation also leads to enolase unfolding. This is in agreement with our previous model of glycation‐induced enolase inactivation in which the glycation of the critical arginine residue 414 disrupted an ionic pair formed with glutamate residue 20, essential for dimer stability (Gomes et al.,

2006). A modification of Arg414 was observed in the present study, both as a consequence of in vivo‐ and in vitro glycation. When enolase is glycated in vitro, unfolding is much more pronounced, which can be due to the higher glycation extent. The observed changes in protein structure and stability are related to the glycation‐dependent activity loss, 65% inactivation in vivo and 85% activity loss in vitro.

AB C

Figure III.7. Surface landscape of dimeric yeast enolase, showing the glycated (red) and non‐glycated (yellow) arginine residues. For greater clarity, the surface of one of the subunits is shown in light gray. A) Enolase glycated in vitro showing glycated arginine residue in a cleft at the dimer interface. B) In vivo glycated enolase showing four out of five glycated arginine residues in the cleft between the two monomers. C) Enolase structure showing the highest solvent exposed arginine residues R200 and R288 that were not glycated. Interestingly, these arginine residues are not located at the dimer interface.

The results of the present study reveal important differences between glycation in vivo and in vitro in the conditions used, which may be related to diverse glycation specificities. This observation highlights the importance of investigating protein glycation in vivo in a model system as yeast, already validated in the research of amyloidotic neurodegenerative diseases (Outeiro & Lindquist,

2003).

 89

The in vitro in vivo debate 

6. Acknowledgements

We thank Dr H. M. Park for the gift of the anti‐yeast‐enolase polyclonal antibody and Dr

Ram Nagaraj for the gift of the anti‐MAGE antibody. We wish to acknowledge Dr Ana Varela

Coelho for providing data from the Laboratório de espectrometria de massa at the Instituto de

Tecnologia Química e Biológica, Universidade Nova de Lisboa, Oeiras, Portugal. Work was supported by grants SFRH/BD/13884/2003 (R.A.G.), SFRH/BD/23604/2005 (L.M.A.O.) and

SFRH/BPD/28345/2006 (M.S.S.) from the Fundação para a Ciência e a Tecnologia, Ministério da

Ciência e Tecnologia, Portugal.

 90

Chapter III

 91

Chapter IV

α­synuclein and insulin glycation: Glycation prevents amyloid formation

1. Abstract

Neurodegenerative disorders, like Alzheimer’s, Parkinson’s, amyotrophic lateral sclerosis, and prion diseases, are debilitating and yet incurable disorders that demand intensive research.

These diseases are characterized by a slow and progressive loss of neuronal cells, and by the deposition of misfolded and aggregated proteins in the brain, either intra‐ or extracellularly. Protein structural changes and aggregation are the molecular characteristics defining such diseases and in all cases, advanced glycation end‐products are present in the fibrils found in these amyloidotic pathologies. It is now believed that protein glycation plays a major role in these disorders although the underlying mechanisms are still unknown. Importantly, in all these pathological conditions glycaemia is normal, leading to the introduction of a new type of cellular stress, carbonyl stress, which is caused by a generalized increase in the concentration of reactive carbonyl AGE‐precursors.

Among these, methylglyoxal is the most important glycation agent in vivo, mainly derived from glycolysis in an unavoidable process which has been extensively studied in yeast. In this study, methylglyoxal was investigated for its effects on the structural and fibril‐forming properties of two amyloid proteins: α‐synuclein and insulin. Methylglyoxal was found to inhibit protein fibrillation and to induce oligomerization in both proteins. Methylglyoxal induced an α  β structural change in insulin and a reduced conformational stability was observed, which led to the formation of insulin aggregates. Regarding glycated α‐synuclein, protein species were separated into three fractions: monomers, dimers and high molecular mass oligomers. High molecular mass oligomers were morphologically characterized by transmission electron microscopy and revealed to have mainly a ring structure.

2. Introduction

Diseases like Alzheimer’s, Parkinson’s or prion’s share a common feature, the formation of insoluble protein deposits in the form of β‐fibrils (Glenner, 1980b; Glenner, 1980a; Hardy & Allsop,

1991; Blake & Serpell, 1996; Goedert & Spillantini, 1998; Olanow & Tatton, 1999; Selkoe, 1999). In each case, a different protein accumulates, in different tissues and organs, leading to the specific clinical complications of each disease. The underlying mechanisms beneath the structural transition from a soluble and functional protein to a highly structured and toxic protein aggregate are still unknown. Neuropathological amyloid deposits and glycated proteins share several properties such

Glycation prevents amyloid formation 

as high insolubility and protease resistance, and for this reason it was proposed that protein glycation might account for amyloid formation in vivo (Colaco & Harrington, 1994; Harrington &

Colaco, 1994). In agreement with this hypothesis, AGE‐modified proteins were detected in amyloid deposits from several amyloidosis. In AD, there is increase AGE content in Aβ plaques (Vitek et al.,

1994) and glycation promotes tau fibrillation (Ledesma et al., 1994). In ALS, glycation was first detected in sporadic and inherited ALS patients both in spinal cord and brain samples, and further studies revealed that AGE levels were higher in the presence of Sod1 mutation, while control human and mouse cases did not display AGE immunoreactivity (Shibata et al., 2002). AGE were also reported and quantified in FAP (Gomes et al., 2005a) and the same stands for PD, where glycation was first reported in the substatia nigra and locus ceruleus displaying higher immunoreactivity at the periphery of LBs of PD patients (Castellani et al., 1996).

Insulin is a small protein hormone that is crucial for the control of glucose metabolism and in type I diabetes treatment, where autoimmune response causes a progressive permanent destruction of the insulin‐producing cells in the pancreas. Insulin regulates blood glucose levels by stimulating glucose transport across the cell membrane and by down regulation of enzymes involved in gluconeogenesis (Taylor, 1991; Shepherd & Kahn, 1999; Zierath et al., 2000). It is composed of two polypeptide chains, the A‐chain (21 residues) and the B‐chain (30‐residues) linked together by two disulfide bonds (Blundell et al., 1972; Baker et al., 1988). In solution, insulin exists as an equilibrium mixture of monomers, dimers, tetramers and hexamers, and possibly higher association states, depending on concentration, pH, metal ions, ionic strength and solvent composition (Brange, 1987).

In the secretory vesicles of the pancreas, the predominant form of insulin is a zinc‐coordinated hexamer, formed by the association of three dimers, and stabilized by two to four zinc ions, but when released to the blood stream, insulin is present in its biologically active form, i. e. the monomeric form (Nystrom & Quon, 1999; Ottensmeyer et al., 2000). Insulin forms amyloid‐like fibrils in vitro, which are promoted by elevated temperatures, low pH, and increased ionic strength

(Brange et al., 1997; Ahmad et al., 2005); and it is the key protein in a clinical condition observed in insulin‐dependent diabetic patients, called insulin injection amyloidosis (Westermark et al., 2005).

In this rare condition, full‐length insulin molecules are found in fibrillar form at the site of frequent insulin injections (Storkel et al., 1983; Dische et al., 1988; Brange et al., 1997). Recently, serum samples from patients with Parkinson’s disease have been found to display an autoimmune response to insulin oligomers and fibrils (Wilhelm et al., 2007), possibly indicating the presence of insulin aggregates in this disease as well. Upon fibrillation, the molecule of insulin undergoes  96

Chapter IV structural changes from a predominantly α‐helical state to a β‐sheet rich conformation. The fibrillar

β‐sheets have been described as either parallel (Burke & Rougvie, 1972; Bouchard et al., 2000;

Nettleton et al., 2000) or antiparallel (Yu et al., 1974; Turnell & Finch, 1992; Sawaya et al., 2007). The

α‐ to β‐ transition appears only to occur upon fibril assembly, as proposed by Jimenez and co‐ workers (Jimenez et al., 2002) and recently, Vestergaard and co‐workers proposed that insulin fibrils are formed primarily by α‐helical oligomers (Vestergaard et al., 2007).

α‐Synuclein was first cloned from the Torpedo californium (Maroteaux et al., 1988) and later identified in humans as the non‐Aβ component of AD (NAC) precursor (NACP) (Ueda et al., 1993).

Attention focused on α‐synuclein when Polymeropoulos and co‐workers reported that in a family with familial PD, a mutation, A53T, was associated with disease (Polymeropoulos et al., 1997). A number of groups quickly reported that α‐synuclein is a major component of LB (Spillantini et al.,

1997; Wakabayashi et al., 1997; Baba et al., 1998; Irizarry et al., 1998; Takeda et al., 1998). Two additional missense mutations, A30P (Kruger et al., 1998) and E46K (Zarranz et al., 2004) are also associated with familial PD. α‐Synuclein is a member of a highly conserved family of proteins consisting of α‐, β‐ and γ‐synuclein. It is a 14.5 kDa protein with 140 amino acid residues (Lavedan,

1998), and the α‐synuclein gene contains seven exons, five of which are protein coding (Clayton &

George, 1998; George, 2002; Norris et al., 2004). α‐Synuclein may be subdivided in three different domains: (i) the N‐terminal (residues 1‐65) is composed almost entirely of six repeats of a degenerate 11‐amino acid consensus motif with slight variations of the sequence KTKEGV. The 11‐ mer repeats comprise an apolipoprotein‐like‐class‐A2 helix, which mediates binding to vesicles of phospholipids. In neutral solution, α‐synuclein is unfolded, but binding to lipid shifts the protein secondary structure from ≈3 % to > 70 % α‐helix (Davidson et al., 1998; George, 2002; Kahle et al.,

2002; Lee et al., 2002; Bussell & Eliezer, 2003); (ii) the central region (residues 66‐95) contains two of the 11‐mer repeats and the hydrophobic NAC portion (non‐Aβ component of AD amyloid). The

NAC region of α‐synuclein appears to be crucial to fibrillation, because α‐synuclein lacking this domain and β‐synuclein, which lacks 11 amino acid residues in the NAC region, have reduced susceptibility to form fibrils (Ueda et al., 1993; el‐Agnaf & Irvine, 2002); (iii) the C‐terminal domain

(residues 96‐140) is less conserved than the N‐terminus and contains a preponderance of acidic residues, being negatively charged. The functions of α‐synuclein are still poorly understood, nevertheless is thought to be associated with synaptic function and plasticity modulation, dopamine regulation, vesicular trafficking and is found to assist vesicle fusion and protein folding

(Bonini & Giasson, 2005; Chandra et al., 2005; Klein & Lohmann‐Hedrich, 2007).  97

Glycation prevents amyloid formation 

Protein glycation is a post‐translational modification whereby amino groups in lysine and arginine side chains react irreversibly with carbonyl molecules forming advanced glycation end‐ products. AGE formation in proteins exerts a profound effect on protein structure, stability and function. Methylglyoxal is the most significant glycation agent in vivo, being one of the most reactive dicarbonyl molecules in living cells. This compound is an unavoidable by‐product of glycolysis, arising from the non‐enzymatic β‐elimination reaction of the phosphate group of dihydroxyacetone phosphate and D‐glyceraldehyde 3‐phosphate (Richard, 1993). Methylglyoxal irreversibly reacts with amino groups in lipids, nucleic acids and proteins, forming methylglyoxal‐ derived advanced glycation end‐products. In Aβ, glycation by methylglyoxal promotes the formation of β‐sheets, oligomers and protofibrils and also increases the size of the aggregates (Chen et al., 2006). Argpyrimidine is a specific methylglyoxal modification occurring in arginine residues, and was recently associated with neurodegenerative diseases (Gomes et al., 2005a). However, little is known about the detailed effects of methylglyoxal glycation on the fibrillation of insulin and α‐ synuclein. Here, we describe the effects of methylglyoxal on the fibrillation properties of both proteins in vitro, and its effects on protein structure and stability. Apparently, methylglyoxal glycation reduces fibrillation in these two amyloid forming proteins.

3. Experimental Procedures

3.1 Methylglyoxal preparation

Methylglyoxal was prepared by acid hydrolysis of methylglyoxal 1,1‐dimethylacetal as reported by Kellum and co‐workers (Kellum et al., 1978) and purified by fractional distillation under reduced pressure in a nitrogen atmosphere (McLellan et al., 1992). Once prepared, methylglyoxal solutions were standardized by enzyme assay with glyoxalase I and II, as described previously (Racker, 1951). Purity was verified by HPLC (Cordeiro & Ponces Freire, 1996) and NMR analysis on a Bruker Avance 400.

3.2 Insulin preparation

Insulin exists in solution as an equilibrium mixture of monomers, dimers, tetramers and hexamers, and possibly higher associated states, depending on concentration, pH, metal ions, ionic strength and solvent composition (Brange, 1987). To prepare a solution containing only insulin in  98

Chapter IV the monomeric form, human insulin (Sigma) was dissolved in type I water to a final concentration of 6 mg/ml and acidified with H3PO4 to a pH of 5. Insulin at pH 5 was then incubated for 15 min at room temperature and subsequently neutralized to pH 7 with NaOH 0.1 M. Protein final concentration was determined by absorbance at 275 nm (ε275 = 4560 M‐1 cm‐1) in a UV‐Visible spectrophotometer Jasco V‐530. Insulin was proven to be in the monomeric form after pH neutralization as evaluated from size exclusion chromatography experiments. Also circular dichroism experiments showed that no structural changes or unfolding occurred with pH variations.

3.3 Expression and purification of human α­synuclein

The expression and purification procedure of human α‐synuclein was a modified version of a previously described method (Kessler et al., 2003). Briefly, cells of E. coli strain BL‐21 (GE

Healthcare, NJ, USA) were transformed with the appropriate expression vector, and expression was induced by the addition of isopropyl D‐thiogalactopyranoside at a final concentration of 1 mM.

Cells were harvested, resuspended in 50 mM Tris (pH 8.5), 50 mM KCl, 5 MgAc, 0.1 % NaN3 and

300 μM PMSF, and lysed by three passages through a French cell press. The extract was centrifuged at 18000 g at 4 ºC for 30 min to eliminate cell debris. The supernatant was saved and boiled for 20 min. the boiled extract was centrifuged at 45000 g at 4 ºC for 45 min and supernatant was filtered with a 0.2 μm filter to remove possible cell contamination. The α‐synuclein containing extract was loaded on to an ion‐exchange chromatography Q SepharoseTM (GE Healthcare, NJ, USA) fast flow column equilibrated with 20 mM Tris/HCl (pH 8.0). Proteins were eluted with a linear NaCl gradient (0.12 ‐ 0.5 M) at a flow rate of 1.5 ml.min‐1 and the eluate was monitored at 280 nm.

Protein‐containing fractions were collected and probed by western blot analysis using Syn‐1 anti‐α‐ synuclein antibody (BD Transduction Laboratories, CA, USA). Fractions containing α‐synuclein were collected, concentrated by centrifugation using Amicon filters (Millipore) and applied to a gel filtration Superdex 75 column (GE Healthcare, NJ, USA), equilibrated with 50 mM sodium phosphate buffer containing 150 mM NaCl. Proteins were eluted with the same buffer at a flow rate of 1 ml.min‐1. Again, fractions containing α‐synuclein, probed by western blot, were collected and combined for dialysis against water and then lyophilized for future analysis. α‐synuclein concentration was determined spectrophotometrically (ε275 = 5974 M‐1 cm‐1) in a UV‐Visible Jasco V‐

530 spectrometer.

 99

Glycation prevents amyloid formation 

3.4 Protein glycation

Human α‐synuclein (1 mg.ml‐1) insulin (3 mg.ml‐1) were incubated with methylglyoxal (1 to 5 mM) in 50 mM potassium phosphate buffer, pH 7.4 supplemented with 150 mM of NaF, at 37 ºC in sterile conditions. Control samples were treated in the same way but without methylglyoxal addition. To evaluate the effects of methylglyoxal on insulin stability and secondary structure changes, samples were incubated without stirring, while for fibrillation kinetic studies, samples were incubated with vigorous agitation. Aliquots were collected in sterile conditions at defined times from 0 to 72 hours and analysed at once.

3.5 Fibril formation

Solutions of 0.6 ml of monomeric α‐synuclein and insulin at pH 7.4 in 50 mM of sodium phosphate buffer supplemented with 150 mM of sodium fluoride were incubated and stirred at 37

ºC in the presence of 0, 1 and 5 mM of methylglyoxal. Fibril formation was monitored with the thioflavin T (ThT) binding assay (Naiki et al., 1989; Naiki et al., 1990). Aliquots of 10 μl were removed from the incubated samples and added to 1.0 ml of 25 μM ThT in 50 mM Tris buffer (pH

8.0).

3.6 Analysis of fibrillation kinetics

The kinetics of α‐synuclein and insulin fibril formation could be described as sigmoidal curves defined by an initial lag phase, in which a negligible change in ThT fluorescence intensity was observed; a subsequent exponential growth phase, in which ThT fluorescence increased; and a final equilibrium phase, in which ThT fluorescence reached a plateau, indicating the end of fibril formation. ThT fluorescence measurements were plotted as a function of time and fitted to a curve described by equation IV.1.

y f  m f x Y  yi  mi x  xx (equation IV.1)  0 1 e  where Y is the fluorescence intensity and x0 is the time to 50 % of maximal fluorescence. The initial base line during the lag phase is described by yi + mix. The final base line after the growth phase had ended is described by yf + mfx. The apparent first‐order rate constant (kapp) for the growth of fibrils is calculated as 1/τ, and the lag time is calculated as xo‐2τ. This expression is unrelated to the

 100

Chapter IV underlying molecular events, but provides a convenient method for comparison of the kinetics fibrillation.

3.7 Fluorescence measurements

Fluorescence measurements were performed in quartz cuvettes with 1 cm excitation light path using a Perkin Elmer LS50B spectrofluorimeter. ThT fluorescence was recorded immediately after addition of the aliquots to the ThT mixture from 470 to 530 nm with excitation at 450 nm, an increment of 0.5 nm, an integration time of 1 s and 5 nm slits for both excitation and emission. For each sample, the signal was obtained as the ThT intensity at 482 nm from which was subtracted a blank measurement recorded prior to addition of insulin to the ThT solution. Tyrosine intrinsic fluorescence emission spectra were recorded from 280 to 340 nm with excitation at 275 nm, an increment of 0.5 nm, an integration time of 1 s and slits of 7 nm for excitation and 5 nm for emission. ANS emission spectra were recorded from 460 to 600 nm with excitation at 350 nm, an increment of 0.5 nm, an integration time of 1 s and slits of 7 nm for excitation and 5 nm for emission.

3.8 Circular dichroism measurements

Secondary structure analysis was performed by far‐UV (185‐260 nm) CD in a Jasco J810 spectropolarimeter equipped with a temperature control unit Julabo F25 using an α‐synuclein concentration of 1 mg.ml‐1 and an insulin concentration of 3 mg.ml‐1. Far UV CD spectra were recorded with 0.01 cm (linear) path length quartz cuvette at 25 ºC. For each spectrum, three scans were averaged and protein concentration was determined by absorbance at 275 nm using the above mentioned insulin extinction coefficient in a UV‐Visible spectrophotometer Jasco V‐530. For protein secondary structure estimation, CD spectra were deconvoluted using the CDSSTR (Johnson, 1999) deconvolution algorithm on Dichroweb (Lobley et al., 2002; Whitmore & Wallace, 2004). CD spectra of the appropriate buffers were recorded and subtracted from the protein spectra.

3.9 Conformational stability measurements

CD denaturation curves were constructed using the ellipticity at 222 nm, monitored at 25 ºC after 24 h incubation with GdnHCl. The denaturation of glycated and non‐glycated insulin could be  101

Glycation prevents amyloid formation 

described as sigmoidal curves and were analysed according to a two‐state unfolding model M ↔

U using the linear extrapolation method (Pace, 1986) in a non‐linear least squares fitting procedure and yielded values for ΔGº(H2O), the conformational stability, and m, the dependence of ΔGº on denaturant concentration. Cm, the denaturant concentration at the midpoint of the unfolding transition was calculated as Cm  Gº H 2O/ m . Denaturation curves for monomeric species were analysed considering the equation developed by Santoro & Bolen (Bolen & Santoro, 1988;

Santoro & Bolen, 1988).

3.10 Size­exclusion and native­PAGE experiments

Aggregation of human α‐synuclein and insulin upon glycation by methylglyoxal was monitored by size exclusion chromatography and Native‐PAGE. Samples were analysed by SEC at defined incubation times with HPLC Jasco PU‐2080 Plus isocratic pump with an UV detector

JASCO 2075. The mobile phase was 50 mM sodium phosphate buffer pH 7.0 with 150 mM NaF.

Separation was achieved on a molecular exclusion analytical column (Amersham‐Pharmacia

Superdex™ 75 10/300 GL) at a flow rate of 0.4 ml/min. Eluting peaks were monitored at 275 nm.

Insulin samples were also separated by Native‐PAGE on a Bio‐Rad Mini‐Protean 3 system, using a 12 % separation gel and a 4 % stacking gel. Proteins were stained with Comassie Briliant

Blue (Wilson, 1979).

3.11 Transmission electron microscopy

5 μl of sample was applied to carbon‐coated Formvar 200 mesh grids (Electron Microscopy

Sciences) and incubated at room temperature for 60 s. The grids were then washed sequentially by depositing 10‐μl droplets of double distilled sterile water (2 times) followed by a 10‐μl droplet of fresh 2 % (w/v) uranyl acetate, which remained on the grid for 30 s. After each step, the excess solution was blotted with Whatman filter paper, and the grids were vacuum‐dried from the edges.

The samples were analysed using a Phillips CM‐10 TEM microscope operated at 100 kV acceleration voltage.

 102

Chapter IV

4. Results and discussion

4.1 Methylglyoxal reduces protein fibril formation

Methylglyoxal was incubated with α‐synuclein and insulin, and the aggregation of both proteins was induced. To investigate the effect of methylglyoxal on protein fibril formation, the fibrillation was monitored in the presence of different concentrations of methylglyoxal by thioflavin

T fluorescence at 37 ºC and pH 7.4. Figure IV.1 represents time‐dependent changes in the ThT fluorescence during the process of α‐synuclein and insulin fibril formation at 37 ºC as a function of methylglyoxal concentration. The kinetics of the ThT fluorescence intensity at 482 nm, exhibit characteristic sigmoidal curves, which have an initial lag phase, a subsequent growth phase, and a final equilibrium phase. Such curves are consistent with a nucleation‐dependent polymerization model, in which the lag corresponds to the nucleation phase, and the exponential part to fibril growth (elongation) (Buchet et al., 1991; Jarrett & Lansbury, 1992; Fahmy et al., 1993; Jarrett &

Lansbury, 1993; Schmid, 1995; Lomakin et al., 1997; Wood et al., 1999). Figure IV.1 shows the effect of increasing the concentration of methylglyoxal on the fibrillation properties of both insulin and α‐ synuclein.

700 1000 A B 600 900 800 500 700

400 600 500 300 400 ThT fluorescence

200 ThT Fluorescence 300

200 100 100

0 0 0 5 10 15 20 25 0 10203040506070 Time (h) Time (h) 10.000 1.600 16 0.4 8.000 1.200 12 0.3 6.000 0.800 8 0.2 4.000 kapp (h-1) kapp kapp (h-1) Lag Time(h) Lag Time (h) (h)Lag Time 2.000 0.400 4 0.1

0.000 0.000 0 0 0246 0246 0246 0246 [MGO] (mM) [MGO] (mM) [MGO] (mM) [MGO] (mM)

Figure IV.1. Kinetics of fibrillation of insulin (A) and α–synuclein (B) in the presence of methylglyoxal at different concentrations – 0 mM (blue), 1 mM (green) and 5 mM (red) – monitored by the enhancement of ThT fluorescence intensity. Protein concentration was 3 mg / ml for insulin and 1 mg / ml for α–synuclein, and measurements were performed at 37 ºC and pH 7.4 in 50 mM sodium phosphate buffer containing 150 mM NaF. ThT fluorescence was excited at 450 nm, and the emission wavelength was 482 nm. The curves are non‐linear least squares fits to equation IV.1. The insets represent the methylglyoxal dependence of kinetic parameters of insulin and α–synuclein: lag time and apparent rate constant of elongation. Lag time is taken as x0‐2τ and the kapp is given by 1/τ.

 103

Glycation prevents amyloid formation 

Previous reports showed that AGE formation accelerates the fibrillation of several proteins including β‐amyloid peptide, tau, and albumin, implying a close association with the development of proteinaceous inclusions in amyloid‐related diseases (Ledesma et al., 1994; Loske et al., 2000). In contrast, our data shows that AGE modifications of insulin and α‐synuclein by methylglyoxal resulted in a substantial inhibition of protein fibril formation. Hashimoto and collaborators also observed that sugar‐induced modification of β2‐microglobulin exhibited inhibitory effects on the extension of fibrils of the protein, suggesting an inhibitory role of AGE products against amyloid formation (Hashimoto et al., 1999) and a recent work reported similar observations for α‐synuclein

(Lee et al., 2009). We assume that this difference might be due to inherent properties of the native structure distinct to each protein, or differential structural changes induced by AGE modifications.

Regarding α‐synuclein the lag time seems to increase from 5.4 h to 14.8 h in the presence of 5 mM methylglyoxal, suggesting a longer nucleation phase, while the apparent rate constant of fibril formation apparently does not depend on methylglyoxal concentration. The same stands for insulin, where no methylglyoxal dependence of the kinetic parameters was observed. Still, further investigation is necessary, varying more methylglyoxal concentrations, to understand how nucleation and fibril elongation are individually affected by methylglyoxal. This fibril formation reduction may happen either because the monomeric form of glycated protein is more stable than the non‐glycated monomer, because there is stabilization of intermediate species on fibrillation pathway or because pathways other than fibril formation may occur upon glycation.

Protein fibril formation was also monitored by circular dichroism (figure IV.2). As represented in figure IV.2, α‐synuclein has an initial far‐UV CD spectrum typical of an unfolded polypeptide chain, reflecting the lack of ordered secondary structure at physiological conditions, while insulin presents a mainly α‐helical secondary structure with spectral local minima at 222 and

208 nm and a positive band bellow 200 nm, which are characteristic of α‐helical conformations.

CD spectra collected at several time points showed that fibrillation is accompanied by a transition from a “random” structure, in the case of α‐synuclein, and from an α‐helical structure, in the case of insulin, to a β‐sheet conformation (figure IV.2 A). This conversion was most extensive when methylglyoxal was absent. The inset for α‐synuclein shows a sigmoidal transition from random to β‐sheet which is consistent with the relative rates of fibril formation observed by ThT fluorescence. Regarding insulin, protein precipitation due to fibril formation did not allow the CD spectra measurements of the longer incubation times.

 104

Chapter IV

60 A 0204060 02468 60 0 0

-10 -5 40 40 -20 -10 -15

222nm (mdeg) -30 222nm (mdeg) 222nm   -20 20 -40 20 -25 time (h) time (h)

0 (mdeg)

 0 (mdeg) 185 205 225 245  185 205 225 245 -20 -20 -40

-40 -60 w avelenght (nm) wavelength (nm) B

60 60 0246 0246 0 0 -5 -10 40 40 -10 -20 -15 -30

222nm (mdeg) -20 222nm (mdeg)  20  -40 20 -25 [MGO] (mM) [MGO] (mM) (mdeg) (mdeg) 

0  0 185 205 225 245 185 205 225 245

-20 -20

-40 -40 w avelength (nm) wavelength (nm)

Figure IV.2. β‐Sheet formation of α–synuclein (left panel) and insulin (right panel) followed by far‐UV circular dichroism. (A) CD spectroscopy detects the random coil‐to‐β‐sheet (in α–synuclein) and the α‐helical‐to‐β‐sheet (in insulin) transition at increasing incubation times: 0, 7, 18, 24, 39 and 43 h (in order of the increase in negative θ222nm values for α–synuclein CD spectra) and 0, 3, 5 and 7 h (in order of the decrease in negative θ222nm values for insulin CD spectra). The insets show the dependence of θ222nm on time. (B) Far‐UV CD circular dichroism spectra of 1 mg / ml α–synuclein and 3 mg / ml insulin in the presence of 0 (blue), 1 (green), 2 (black) and 5 mM (red) of methylglyoxal after 24 and 7 h incubation for α–synuclein and insulin respectively. The insets show the dependence of θ222nm on methylglyoxal concentration. All measurements were made in 50 mM of sodium phosphate buffer pH 7.4 containing 150 mM NaF with a cell path length of 0.1 mm at 37 ºC.

Figure IV.2 B shows the data on the effect of adding increasing concentrations of methylglyoxal on the spectral properties of α‐synuclein and insulin after a defined incubation time.

When methylglyoxal is absent, the shape and intensity of the far‐UV CD spectra change drastically after a defined incubation period, whereas when methylglyoxal is added to the protein, these spectral changes are much smaller. In fact, figure IV.2 B shows that methylglyoxal stabilizes protein native structure which is in agreement with the inhibition of fibril formation observed in ThT kinetic measurements (figure IV.1). The insets shows that this stabilization is methylglyoxal depend as it increases with methylglyoxal increasing concentration.

 105

Glycation prevents amyloid formation 

4.2 Methylglyoxal causes protein oligomerization

To investigate the early steps of protein aggregation, samples were collected and analysed by size exclusion chromatography and Native‐PAGE (figure IV.3). Non‐glycated Glycated

[MGO]‐0 mM 1 mM 5 mM kDa 170 130 100 70 55 40 35 25

15 10 0h 2h 5h 10h 18h 24h 0h 2h 5h 10h 18h 24h 0h 2h 5h 10h 18h 24h Figure IV.3. Early steps of insulin aggregation monitored by size exclusion chromatography and Native‐PAGE. SEC experiments were performed at different incubation times of non‐glycated and glycated human insulin with 5 mM of methylglyoxal in 50 mM sodium phosphate buffer pH 7.4 containing 150 mM NaF. Aliquots were also collected and analysed through PAGE gels in absence of SDS and β‐mercaptoethanol.  106

Chapter IV

Non‐glycated insulin appears as a single molecular species with an apparent molecular mass of 5.9 kDa (elution volume – 14.10 ml), which corresponds to the monomeric form of human insulin. The same was observed for glycated insulin at time 0, as it can be observed either by gel filtration or gel electrophoresis. During incubation time, the non‐glycated insulin monomer is changed into amyloid fibrils as it can be observed by the decrease in SEC peak intensity. Gel electrophoresis also shows the reduction on monomeric species, but a new band appears on the beginning of the lane, corresponding to high molecular mass species (fibrils). An interesting phenomenon is that this transition from monomer to fibrils appears to be direct, and intermediate species seem to be absent or in undetectable concentrations. When methylglyoxal is added to the medium, SEC peak intensity also becomes reduced, but other species are present on the chromatogram corresponding to insulin soluble aggregates. These aggregates are also observed in gel electrophoresis and show apparent molecular masses consistent with dimeric, tetrameric and hexameric forms of insulin. Moreover, high molecular mass species are only detected in the later incubation times. These results show that methylglyoxal reduces insulin fibril formation and promotes the population of oligomeric states.

Regarding α‐synuclein, the same effect on protein methylglyoxal‐induced oligomerization was observed by size exclusion chromatography (figure IV.4). Since the methylglyoxal reaction with α‐synuclein generated a heterogeneous mixture of modified protein species, including monomers, dimers and oligomers, a homogeneous sample containing α‐synuclein oligomers was prepared by separation through SEC.

Samples containing α‐synuclein fibrils formed with and without the presence of methylglyoxal, as well as the modified α‐synuclein oligomers isolated from SEC, were subsequently subjected to morphological characterization through transmission electron microscopy (TEM) (figure IV.5).

 107

Glycation prevents amyloid formation 

Figure IV.4. Separation of methylglyoxal‐modified α‐synuclein species. The heterogeneous mixture of α‐synuclein was subjected to a size exclusion chromatography using a Superose 6 column previously equilibrated in 50 mM sodium phosphate buffer containing 150 mM NaF. α‐synuclein were separated in three fractions corresponding to α‐synuclein oligomers (1), dimers (2) and monomers (3).

TEM images of glycated and non‐glycated α‐synuclein show clear differences (figure IV.5 A and B). Consistent with the substantial inhibition of protein fibrillation observed by adding an amount of methylglyoxal to soluble monomeric α‐synuclein, images obtained from transmission electron microscopy showed no fibril‐like structure when methylglyoxal was added to α‐synuclein.

In contrast with the mature fibrils observed for non‐modified α‐synuclein, glycated α‐synuclein formed amorphous aggregates that did not present the typical cross‐β amyloid structure. Also, a larger population of small spherical aggregates was observed. This finding is of particular interest since soluble small aggregates were identified as the main toxic species in neurodegeneration. The possibility that a molecular species other than the amyloid fibril could be pathogenic arose when oligomeric species rich in β‐sheet structure (protofibrils) were found to be discrete intermediates in the fibrillation of β‐amyloid and α‐synuclein in vitro (Lansbury, 1999; Goldberg & Lansbury, 2000).

Toxic protofibrils have been implicated in other neurodegenerative diseases as well as in systemic amyloidosis such as type II diabetes (in which the amyloid protein is IAPP) (Janson et al., 1999) and familial amyloidotic polyneuropathy (in which is transthyretin) (Sousa et al., 2001).

 108

Chapter IV

AB

C

Figure IV.5. Transmission electron microscopy of α‐synuclein aggregates: (A) α‐synuclein fibrils formed with no methylglyoxal; (B) α‐synuclein aggregates formed in the presence of methylglyoxal 5 mM; (C) α‐synuclein oligomers produced after incubation with methylglyoxal 5 mM and subsequently isolated through size exclusion chromatography. Scale bar = 200 nm

The heterogeneous population of glycated α‐synuclein species was fractioned by size exclusion chromatography. Analysis of the fraction containing α‐synuclein oligomeric species

(protofibrils) by electron microscopy revealed the presence of annular species with pore‐like morphology. These annular species, considered as a subpopulation of amyloid protofibrils, have been assigned to explain the pore activity of α‐synuclein protofibrils in vesicle‐permeabilization models (Volles & Lansbury, 2002).

4.3 Effects of methylglyoxal on insulin structure and stability

To understand what methylglyoxal‐derived structural changes might be associated to fibril inhibition and stabilization of oligomeric species, insulin was incubated at 37 ºC and pH 7.4 in 50

 109

Glycation prevents amyloid formation 

mM sodium phosphate buffer containing 150 mM NaF at a final concentration of 3 mg / ml without agitation, a condition that does not promote aggregation. In these conditions, insulin is glycated but remains almost entirely in monomeric form. Figure IV.6 represents the far‐UV CD spectra of non‐ glycated and glycated insulin with two different methylglyoxal concentrations (1 and 5 mM) measured at pH 7.4 and 37 ºC and at different incubation times.

Figure IV.6. Effects of methylglyoxal on insulin structure. Insulin (3 mg/ml) was incubated with different methylglyoxal concentrations, namely 1 and 5 mM at 37 ºC without stirring for 48h and compared with appropriate control. Insulin secondary structure was monitored by far‐UV CD. Circular dichroism spectra were recorded as a function of time at different methylglyoxal concentrations (A – 0 mM; B – 1 mM and C – 5 mM). Spectra were collected at time zero (blue) and after 24h (red) and 48h (green) incubation. The insets represent tyrosine intrinsic fluorescence as function of time at different methylglyoxal concentrations (same legend as CD spectra). Deconvolution of the CD spectra are presented in table IV.I.

Table IV.1 – Distribution of the structural element fractions for native and glycated insulin along time obtained by deconvolution of CD spectra using CDSSTR algorithm available on Dichroweb (Dichroweb; http://www.cryst.bbk.ac.uk/cdweb/html/home.html) (Lobley et al., 2002; Whitmore & Wallace, 2004). The NRMSD parameter represents the normalized root mean square deviance. [MGO] (mM) Time (h) α‐Helix β‐Sheet β‐Turns Unordered structure NRMSD 0 31 23 22 24 0.028 0 24 33 23 21 23 0.033 48 32 22 22 24 0.029 0 31 24 21 24 0.027 1 24 28 26 22 26 0.032 48 24 27 22 27 0.036 0 32 22 22 24 0.022 5 24 23 27 21 27 0.029 48 23 28 21 27 0.035

At time zero, every form of insulin shows far‐UV CD spectra typical of a substantial α‐helical structure, consistent with the crystal structure of the insulin hexamers (Blundell et al., 1971). This includes characteristic minimums in the vicinity of 208 and 222 nm and a positive band below 200 nm. In contrast with the results obtained in the “effects of methylglyoxal on insulin fibrillation” section the CD spectra of non‐glycated insulin remains unchanged during the incubation period, while glycated insulin seems to undergo slight spectral changes. Spectra deconvolution shows a

 110

Chapter IV redistribution of secondary structure elements in glycated insulin either with 1 mM or 5 mM methylglyoxal; with a respective increase in β‐sheet content from 24 % to 27 % and from 22 % to 28

%, an increase from 24 % to 27 % in unordered structure and a reduction from 31 % to 24 % and 32

% to 23 % in the relative α‐helical content.

Insulin structural effects were also evaluated at tertiary structure by monitoring tyrosine intrinsic fluorescence. Once again, no changes were observed in non‐glycated insulin, while in glycated insulin tyrosine quenching was observed. Tyrosine quenching was considered previously to occur upon contact with water (Noronha et al., 2004), which suggests that non‐glycated insulin is a more compact and closed structure.

To evaluate the effects of methylglyoxal glycation on insulin conformational stability, we determined the conformational stability of the glycated insulin and compared with the native protein. Figure IV.7 shows the circular dichroism equilibrium denaturation at pH 7.4 and 25 ºC of native and glycated insulin using GdnHCl as denaturant.

[GdnHCl] (mM) [GdnHCl] (mM)

012345678 012345678 0 0

‐2 AB‐2

‐4 ‐4 ‐6 ‐6 ‐8 ‐8 ‐10 ‐10 ‐12 1 2 Ellipticity (mdeg) Ellipticity 1.5 (mdeg) Ellipticity ‐12 0.5 ‐14 1 ‐14 0 ‐16 0.5 0 ‐0.5 ‐0.5 ‐18 ‐16 ‐1 ‐1 ‐20 ‐18

Figure IV.7. Guanidinium hydrochloride equilibrium denaturation curves of native (A) and glycated (B) insulin at pH 7.4 and 25 ºC, monitored by ellipticity at 222 nm. The curves are non‐linear least squares fits to a two‐state unfolding model equation (Santoro & Bolen, 1988) representing the entire denaturation curve and using a linear extrapolation method to the experimental circular dichroism data (Pace, 1986). The insets are the residues plot.

GdnHCl‐induced denaturation was found to be reversible, as judged by CD experiments after dialysis of GdnHCl‐denatured insulin. Fits were made using the linear extrapolation method

(Pace, 1986) in a non‐linear least squares fitting procedure and yielded values for ΔGº(H2O), the conformational stability, and m, the dependence of ΔGº on denaturant concentration. Table IV.2 shows the values obtained from the curves in figure IV.7 for ΔGº(H2O), m, and Cm, the denaturant concentration at the midpoint of the unfolding transition.  111

Glycation prevents amyloid formation 

Glycation resulted in a weaker [GdnHCl] dependence of unfolding (smaller m‐value), which may indicate some deviation from the two‐state model used to fit the data, though this deviation is not detectable by the fitting procedure. According to parameters on table IV.2, glycated insulin has a smaller conformational stability. In fact, the ΔGº(H2O) for native insulin is 3.34 ± 0.33, whereas glycated insulin shows a ΔGº(H2O) of 2.66 ± 0.27; and this decrease in conformational stability is also supported by the smaller Cm value of glycated insulin. The observed loss of conformational stability and the appearance of a less compacted structure, as indicated by the decrease in m value, suggests a higher susceptibility to different unfolding and aggregation pathways, what may explain the existence of aggregation pathways other than amyloid as a consequence of the slight increase in

β‐sheet content upon methylglyoxal glycation.

Table IV.2 – Thermodynamic parameters from GdnHCl unfolding studies of native and glycated insulin. Parameters were obtained by a direct fit of the model equations to experimental data in figure II.8. ΔGº(H2O) is the protein conformational stability; m is the dependence of ΔGº on denaturant concentration; Cm is the denaturant concentration at the midpoint of the unfolding transition.

ΔGº(H2O) m Cm

‐1 ‐1) (kcal∙mol‐1) (kcal∙mol .M (M)

Insulin 3.34 ± 0.33 0.63 ± 0.10 5.31 ± 0.98

Glycated Insulin 2.66 ± 0.27 0.52 ± 0.09 5.10 ± 0.98

5. Conclusion

In spite of efforts in understanding the molecular mechanisms underlying the aggregation of amyloidogenic proteins, these are still far from being understood. Structural and stability alterations provoked by point mutations are considered to be a major cause in aggregation, but the occurrence of sporadic cases of neurodegenerative diseases suggests that it might be a multi‐ factorial phenomenon. Post‐translational modifications have a profound impact on protein’s structure and may also contribute to protein misfolding and aggregation; and glycation might be one important factor as it has been reported in several neurodegenerative diseases.

Other matter of intense investigation is whether the amyloid fibril deposits or the prefibrillar aggregates, called protofibrils, are the most potent mediators of cell damage, cytotoxicity and neurotoxicity. The finding that the severity of cognitive impairment in protein misfolding diseases correlates with the levels of small oligomeric species and not with the large fibrillar species has led  112

Chapter IV researchers to the conclusion that the soluble small aggregates are the primary cause of the pathological symptoms (Caughey & Lansbury, 2003; Crowther et al., 2005; Danzer et al., 2007;

Lauren et al., 2009).

Here we investigated the effects of methylglyoxal, the most important glycation agent in vivo, on structural and fibril forming properties of two amyloidogenic proteins: human α‐synuclein and insulin. Interestingly, the modification of both proteins had a significant inhibitory effect on the protein fibrillation. These results are in contrast with previous reports where it was observed that

AGE modifications accelerated the fibrillation of several proteins including β‐amyloid peptide, tau and albumin (Ledesma et al., 1994; Loske et al., 2000), showing that the inherent properties of the native structure and fold, distinct to each protein, are key starting points to understand what causes the structural transition from a soluble and functional protein to a highly structured and toxic protein deposit. Regarding α‐synuclein, these results also suggests that the presence of AGE in LB might be a secondary event subsequent to the long‐term deposition of aggregates, rather than the cause of fibrillation in the early steps of inclusion body formation, as suggested before (Hashimoto et al., 1999; Lee et al., 2009).

Furthermore, glycated α‐synuclein and insulin undergo an aggregation pathway where oligomeric species are apparently stabilized, instead of mature fibrils. Moreover, aggregation of glycated α‐synuclein resulted in an increase of oligomeric species, some of them, with an annular structure. These ring structures have been considered to disrupt cell membranes by inserting themselves into the phospholipid bilayer, disturbing normal ion gradients (Kourie & Henry, 2002), initiating cell permeabilization (Lashuel, 2005) and causing cell death.

Insulin glycation caused slight changes on protein’s structure with an increase in β‐sheet content from 22 to 28 % with 5 mM methylglyoxal and the formation of a less compact structure as evaluated by tyrosine quenching. These minor changes resulted in a small decrease of insulin conformational stability from 3.34 ± 0.33 to 2.66 ± 0.27 kcal.mol‐1. These results suggests that minor modifications in the balance between hydrophobic effects, intramolecular non‐covalent bonds

(hydrogen bonds and salt bridges) and configurational entropy, that regulates the shape of the protein’s folding funnel, may have deep changes on protein’s biophysical behaviour.

No direct evidence has been reported about the contribution of glycation and methylglyoxal to amyloid formation and cell damage in a context of disease. However, in diabetes a high level of glycation agents are present due to hyperglycaemia, and neurons are a major site of oxygen consumption and use glucose almost exclusively for energy generation. Moreover, oxidative stress,  113

Glycation prevents amyloid formation 

known as a major causative factor for the pathogenesis of conformational diseases, was reported to enhance the generation of glycation agents (Shinpo et al., 2000). The decrease of cellular glutathione

(Sian et al., 1994) might also cause a reduction in the activity of glyoxalase system, which is the principal catabolic pathway for dicarbonyl detoxification in cells. Therefore, methylglyoxal could be accumulated in various cells and tissues, triggering protein glycation in short‐lived monomeric proteins, and have an important role on the early events of protein aggregation.

Although the exact molecular mechanism for methylglyoxal‐induced inhibition of amyloid fibril formation remains unclear, the anti‐fibrillogenic activity of methylglyoxal is here described for two amyloidogenic proteins with different fold types in a way that suggests that glycation unlocks pathways for protein aggregation others than amyloid formation as consequence of less stable and less compact monomeric structures that are not accessible to be a constituent part of a highly rigid and organized structure as an amyloid fibril.

 114

Chapter IV

6. Acknowledgements

We thank Dr. Tiago F. Outeiro for the gift of the E.coli expression vector containing α‐ synuclein cDNA. We wish to thank Dr. Hilal Lashuel and Dr. Katerina Paleologou for providing transmission electron microscopy images from the Laboratory of Molecular Neurobiology and

Functional Neuroproteomics, EPFL, Lausanne, Switzerland. Work was supported by the grant

SFRH/BD/23604/2005 (L.M.A.O.) from the Fundação para a Ciência e a Tecnologia, Ministério da

Ciência e Tecnologia, Portugal.

 115

Glycation prevents amyloid formation 

 116

Chapter V

Concluding Remarks

Chapter V

In spite of significant progress in understanding the pathology of major neurodegenerative disorders such as Alzheimer’s (AD), Parkinson’s (PD), Huntington’s (HD) diseases, and

Amyotrophic Lateral Sclerosis (ALS), the molecular mechanism(s) leading to neuronal degeneration remain elusive. In each disease, the pathological processes lead to dysfunction and degeneration in distinct subsets of neurons in different brain areas (Morrison et al., 1998; Cowan & Raymond, 2006).

Proteins specific to each disease are responsible for degeneration such as amyloid precursor protein

(APP) and tau in basal forebrain cholinergic neurons in AD; α‐synuclein in dopaminergic neurons in the substantia nigra in PD; huntingtin (Htt) in striatal and cortical neurons in HD; and Cu‐Zn superoxide dismutase (SOD1) in motor neurons in ALS (Mattson & Pedersen, 1998; Chinta &

Andersen, 2005; McKinney & Jacksonville, 2005; Boillee et al., 2006; Heese & Akatsu, 2006; Lee &

Trojanowski, 2006; Gonzalez de Aguilar et al., 2007; Ramaswamy et al., 2007). Dominant mutations in APP, tau, α‐synuclein, Htt, and SOD1 cause these proteins to acquire novel structural conformations and misfold, increasing their propensity to aggregate (Kopito & Ron, 2000;

Muchowski, 2002). Indeed, the presence of insoluble aggregates formed by mutant polypeptides is a hallmark of many neurodegenerative disorders (Thompson & Barrow, 2002; Ross & Poirier, 2004).

However, amyloid diseases can also be found as a result of protein aggregation in the absence of genetic mutations. In many tauopathies, aggregation proceeds in the absence of point mutations in tau protein (Hernandez & Avila, 2007). Mutant α‐synuclein protein does aggregate, but so does its wild‐type counterpart (Spillantini & Goedert, 2000; McLean et al., 2001). In ALS, mutated SOD1 protein does not cause well‐defined and detectable inclusions (Wang et al., 2002; Valentine & Hart,

2003; Wood et al., 2003). Thus, point mutations are not the only cause for protein aggregation as originally thought. Other aspects related to protein modification and environmental stress may be on the basis of disease, inducing a transition from a monomeric soluble and functional polypeptide chain to highly organized and toxic aggregates. The release of a complete polypeptide chain from a ribosome is frequently not the last chemical step in the formation of proteins. Various post‐ translational modifications often occur, either during or after assembly of the polypeptide chain.

Knowledge of these modifications is extremely important because they may alter physical and chemical properties, folding, stability and activity, and consequently, the function of the proteins.

Protein glycation is a post‐translational modification that has gained an increased attention in the context of several human pathologies, such as age‐related disorders and neurodegenerative diseases of amyloid type, from the earlier observations that glycated haemoglobin increases as a function of glycaemia in diabetic patients. Glycated proteins are characterized by modified  119

Concluding Remarks 

biochemical features often associated with neuropathological amyloid lesions like high insolubility, protease resistance and, in some cases, fluorescence and brown colour (Colaco & Harrington, 1994;

Harrington & Colaco, 1994). These modified proteins undergo changes in their structure, mainly caused by the removal of positive surface charges as a consequence of the nucleophilic addition of the glycation agent to the amino groups of lysine and arginine side chains. As a consequence, function loss and cellular toxicity occurs. Furthermore, protein glycation has been reported in several neurodegenerative diseases being involved in amyloidosis. A major unanswered question is whether glycation of susceptible proteins is one of the triggering events in the disease process or just an outcome from glycation reactions with low turnover aggregated protein species. Several studies suggest that glycation may be an early event promoting or accelerating the abnormal protein deposition into β–fibrils, followed by increased protease resistance and insolubility. In PD,

Munch and co‐workers found AGE in very early Lewy bodies (Munch et al., 2000), and in prions disease AGE occurs in astrocytes prior to the formation of the abnormal disease‐associated conformation of the prion protein (PrPres) (Choi et al., 2004). Argpyrimidine, a specific marker of protein glycation by methylglyoxal was also found in transthyretin amyloid deposits in FAP patients (Gomes et al., 2005a). Accumulation of AGE‐modified proteins also leads to inflammation and propagation of tissue damage by several mechanisms like oxidative stress and increase and release of pro‐inflammatory cytokines mediated by AGE:RAGE interaction. These factors suggests that glycation my act as a dynamic participant in protein misfolding and aggregation and consequently in cell dysfunction and death.

The present work started with the aim of assessing the effects of glycation on protein’s structure, stability and aggregation pattern. First the effects on structure and thermodynamic stability on cytochrome c upon glycation by methylglyoxal were investigated. Cytochrome c glycation resulted in a significant loss of protein conformational stability from 8.07 ± 0.63 to 3.11 ±

0.56 kcal.mol‐1. This loss of conformational stability is also associated to a less compact structure, as judged by the dependence of ΔGº on denaturant concentration determined in the chaotropic unfolding experiments. Although not related to aggregation and amyloid formation phenomena at physiological conditions, monomeric cytochrome c undergoes an aggregation pathway at pH 7 and

37 ºC upon methylglyoxal glycation, where dimers, trimers and tetramers were detected. Compared to the glycated cytochrome c monomer, there is an increase in protein conformational stability upon

‐1 aggregation since glycated cytochrome c dimmer has a ΔGº(H2O) of 3.56 ± 1.09 kcal.mol .

Cytochrome c unfolding was also observed by the appearance in size exclusion chromatography of  120

Chapter V a monomeric form of cytochrome c with no CD signal. HPLC analysis of the different cytochrome c species revealed that there is an increase of protein hydrophobicity, which is explained by the loss of charges on the protein’s surface due to glycation reaction, and no changes in protein tertiary structure were detected. Regarding cytochrome c structure at secondary level, there is a slight increase of protein’s β‐sheet content. However, the main secondary structure remains α‐helix. This data allowed us to propose a molecular model where methylglyoxal‐glycated cytochrome c starts an aggregation process wherein the formed high molecular mass aggregates retain native‐like structure. This type of protein aggregation is different from the usual amyloid formation and was recently related to some proteins, such is the case of human pancreatitis associated protein and of acylphosphatase from Sulfolobus solfataricus (Ho et al., 2006; Plakoutsi et al., 2006).

Next, protein glycation in vitro was compared with protein glycation in vivo, taken yeast enolase as target (Gomes et al., 2008b). In contrast to the common belief that glycation is a random process glycation in yeast was found to occur at specific protein targets (Gomes et al., 2005b; Gomes et al., 2006). Three glycolytic enzymes were shown to display MAGE: enolase 2, phosphoglycerate mutase, and aldolase (Gomes et al., 2005b; Gomes et al., 2006). Among these, enolase is the major glycation target, showing glycation‐dependent activity loss (Gomes et al., 2006) and providing an important model for studying glycation effects in vivo. In vivo is site‐specific whereby only a few amino acid residues are consistently modified with the same MAGE. By contrast, glycation in vitro is a heterogeneous process, resulting in the formation of a complex population of enolase molecules with different glycation profiles. A different extension in glycation reaction and a different glycation pattern could lead to differences in the glycation effects on protein’s stability, structure and function. In vivo‐glycation effects on enolase resulted in protein dissociation and unfolding, decrease in α‐helical content and increase in β‐sheet content, increase in Tm, and activity loss. When enolase was glycated in vitro, the same effects were observed, yet to a larger extent. These observations underline the importance of investigating protein glycation in vivo. Still, protein glycation in vitro and in vivo appears to share comparable qualitative results regarding protein structure, stability and function.

As mentioned above, glycation may play an important role in the misfolding and aggregation of proteins, being intimately related to neurodegeneration. Thus, we investigated the effects of methylglyoxal glycation in two different amyloidogenic proteins: α‐synuclein and insulin. In contrast with previous studies where glycation was found to promote fibrillation and the development of protein inclusion bodies (Ledesma et al., 1994; Loske et al., 2000), data obtained for  121

Concluding Remarks 

α‐synuclein and insulin shows a reduction of protein fibril formation upon glycation. Instead of forming fibrils, both α‐synuclein and insulin undergoes aggregation pathways where oligomeric species are stabilized. Additionally, aggregation of glycated α‐synuclein resulted in an increase of oligomeric species some of them with an annular structure. This finding is of relevant since soluble small aggregates were identified as the main toxic species in neurodegeneration. Amongst different protofibrilar species, these ring‐like structures have been indentified as a probable major toxic species, disrupting cell membranes, disturbing normal ion gradients (Kourie & Henry, 2002), initiating cell permeabilization (Lashuel, 2005) and causing cell death.

The effects of methylglyoxal glycation on insulin’s structure and stability were also investigated and only small changes were observed when compared to the native protein. These results show a similar behaviour to the one observed for cytochrome c, where a native‐like aggregation pathway was described upon glycation. Still, additional experiments are necessary to fully characterize the structural effects of glycation on insulin, namely probing the nature of insulin aggregates.

Glycation appears to promote aggregation pathways other than amyloid formation or stabilize protofibrils in amyloid formation pathway. The disparity between the results obtained for

α–synuclein and insulin, and the published in literature for β–amyloid peptide, tau and albumin brings back to debate the importance of the inherent properties of the native structure and fold, distinct to each protein, as a key starting point to understand protein aggregation and amyloidogenesis. An interesting study would be a detailed and systematic investigation relating the effects of glycation by methylglyoxal with amyloidogenic proteins of different fold types. This knowledge would be very important to describe the molecular mechanisms on the basis of protein aggregation and the consequent development of improved therapeutic strategies to minimize protein toxicity in conformational diseases.

 122

Chapter V

 123

Concluding Remarks 

 124

Appendix

A new UV absorbance ration method for measure protein conformational stability

Luís M. A. Oliveira, Carlos Cordeiro, Ana Ponces Freire, Carla Ascenso and Alexandre Quintas. Unveiling heme proteins conformational stability through a UV absorbance ratio method. (2007) Anal Biochem. 371(2): 253‐5.

Appendix

Monitoring protein unfolding by fluorescence or circular dichroism spectroscopy is one of the most important approaches to determine the thermodynamic stability of a protein. Because specialized equipment is required, these are not widespread methodologies. Single‐wavelength UV spectroscopy can also be used to monitor protein unfolding but has severe shortcomings, including a poor signal/noise ratio and low response sensibility. Here we propose an absorbance ratio method to monitor protein unfolding that overcomes these issues, producing high quality data for hemic proteins with a two‐state unfolding mechanism.

In proteins, a CD signal will be observed when the protein is an asymmetrical environment by virtue of the adopted three‐dimensional structure. The absorption in the far‐UV (≤240 nm) is due mainly to the peptide bond, where there is a weak but broad n  п* transition centred around 220 nm and a more intense п  п* transition around 190 nm. The different types of regular secondary structure found in native proteins give rise to characteristic CD spectra in the far UV in contrast to the unfolded state, which has no CD signal. In an unfolding curve measured by far‐UV CD, the dependent variable correlates with the loss of secondary structure from the native to the unfolded state (Pace, 1986; Fasman, 1996; A. Rodger, 1997; Kelly & Price, 2000; N. Berova, 2000).

On the other hand, fluorescence spectroscopy responds to changes in the environment of the tryptophan and tyrosine residues, hence, to changes in tertiary structure. It usually is not possible to predict how the intensity of fluorescence will be altered with conformational changes. Some proteins exhibit large increases in their intensities and lifetimes, whereas others exhibit decreases, during polypeptide chain unfolding. Proteins with heme prosthetic groups typically exhibit an increase in intensity on unfolding due to the loss of the heme group, which quenches the tryptophan emission in the native state by Forster energy transfer. The average energy of the emission of the tryptophan residues usually suffers red shifts on unfolding due to solvent exposure; thus solvent relaxation is increased in the unfolded state. The magnitude of the energy shift

 127

Novel UV absorbance ration method 

depends on the extent to which the protein is buried in the native state. In the unfolded state most tryptophan residues in proteins have spectra with maximum at approximately 355 nm. In addition, fluorescence spectroscopy is a very sensitive technique but should not be used to monitor thermal unfolding because the pre‐ and post‐transition baselines are steep and highly temperature sensitive

(Becktel & Schellman, 1987; Schellman, 1987).

The equilibrium unfolding curves measured by UV spectroscopy share the same theoretical principles with both spectroscopic techniques mentioned above. In all of these methods the dependent variable (y) is measured as a function of the independent variable, which is often temperature (for thermal denaturation curves) or denaturant concentration (urea or guanidine hydrochloride). The denaturation curves usually show a sigmoidal cooperative transition from the native to the denatured state. The UV spectroscopy, as fluorescence, responds to changes on the tryptophan environment, correlating with tertiary structure modifications. However, these changes produce small variations in the UV signal, making this spectroscopic method a poor technique to monitor unfolding transitions. In hemic proteins, there are two defined UV spectral regions that are sensitive to conformational changes: the region around 280 nm, for which the tryptophan and tyrosine residues are responsible, and the hemic region around 410 nm. To determine the conformational stability of hemic proteins by UV spectroscopy, we propose a method based on the ratio between the absorbance at 400 nm and 287 nm. This method increases the sensibility of the technique because it enhances the transition slope which is directly related to the limit detection. It also increases signal/noise ratio because systematic errors are cancelled.

The proposed method is illustrated by measuring the conformational stability of cytochrome c (horse heart, Sigma) and myoglobin (horse muscle, Sigma). In both cases, protein solutions were prepared at 0.1 mg / ml in 50 mM sodium phosphate buffer at pH 7 containing 150 mM of NaF. Protein unfolding was monitored by UV spectroscopy in a Jasco V‐530

 128

Appendix spectrophotometer with 1 cm path length quartz cuvette and by CD in a Jasco J810 spectropolarimeter with 0.1 cm (linear) path length quartz cuvette. Guanidinium hydrochloride

(GdnHCl)‐induced protein unfolding experiments were performed at 25 ºC after 1 day incubation with several protein solutions and increasing guanidinium chloride concentration in the range of 0 to 6 M for cytochrome c and 0 to 3 M for myoglobin. Equilibrium unfolding curves were analysed according to a two‐state unfolding model using the linear extrapolation method (Greene & Pace,

1974; Santoro & Bolen, 1992). A non‐linear least squares fitting procedure yielded values for

ΔGº(H2O), the conformational stability, and m, the dependence of ΔGº on denaturant concentration.

Cm, the denaturant concentration at the midpoint of the unfolding transition was calculated as

Cm  Gº H 2O / m . Denaturation curves were analysed considering the equation developed by

Santoro and Bolen (Bolen & Santoro, 1988; Santoro & Bolen, 1988). Parameters in table VI.1 were obtained by fitting the derived equations to experimental data by non‐linear regression using the

Solver add‐on for Microsoft Excel. Standard errors and the variance‐covariance matrix were computed under the assumption of normal distribution of residuals.

The curves shown in figure VI.1 present sigmoidal transitions for both spectroscopic methods used. The ratio between the absorbance at 400 nm and 287 nm has effectively increased the transition sensibility in a well‐defined unfolding curve. The denaturation curves for cytochrome c and myoglobin obtained by the ratio between the absorbance at 400 nm and 287 nm are quite similar to those obtained by CD, in contrast to the curves obtained by single‐wavelength UV spectroscopy. Figure VI.1 A show an improvement of the transition region in the cytochrome c unfolding curve. This shows an increase of the method sensibility when compared to the single wavelength method. In the myoglobin unfolding curve (figure VI.1 B), the proposed method produces changes mainly in the pre and post‐transition baselines, illustrating the importance of cancelling systematic errors. Thermodynamic parameters determined by the proposed method are

 129

Novel UV absorbance ration method 

also identical to those determined by CD when compared with the parameters obtained by UV spectroscopy in a single‐wavelength variation, thereby validating the obtained results with the proposed method.

Table VI.1 – Conformational stability (ΔGº(H2O), dependence of ΔGº on denaturant concentration (m) and denaturant concentration at the mid point of the unfolding transition (Cm) for cytochrome c and myoglobin.

ΔGº(H2O) m Cm Protein Method (kcal∙mol‐1) (kcal∙mol‐1.M‐1) (M)

Absorbance Ratio Method 7.34 ± 1,79 2.89 ± 0,68 2.54 ± 0,87

Circular Dichroism 7.33 ± 0,96 2.65 ± 0,34 2.76 ± 0.51 Cytochrome c 287 nm 5.64 ± 1.18 2.49 ± 0.39 2.27 ± 0.59 UV 400 nm 5.26 ± 1.18 1.71 ± 0.39 3.08 ± 0.98

Absorbance Ratio Method 10.41 ± 0.58 5.72 ± 0,32 1.82 ± 0.14

Circular Dichroism 9.57 ± 1.37 5.33 ± 0,76 1.80 ± 0.36 Myoglobin 287 nm 26.93 ± 12.96 14.56 ± 7.20 1.85 ± 1.28 UV 400 nm 10.62 ± 0.81 5.94 ± 0.45 1.79 ± 0.19

Note. Values were determined from data obtained by the proposed absorbance ratio method and by circular dichroism spectroscopy and UV spectroscopy at single wavelength monitoring. The equation used to fit the experimental data represents a two‐sate unfolding model (Bolen & Santoro, 1988; Santoro & Bolen, 1988).

By adopting this methodology, it is possible to improve the reliability of the results obtained in the equilibrium unfolding curves measured by UV spectroscopy, thereby making the UV equilibrium unfolding curves a suitable technique that is accessible virtually to all laboratories on the measurement of the conformational stability of a hemic protein with a two‐state unfolding mechanism.

 130

Appendix

Figure VI.1 ‐ GdnHCl equilibrium denaturation curve of cytochrome c (A) and myoglobin (B). Each protein was dissolved at 0.1 mg / ml in 50 mM sodium phosphate buffer pH 7, containing 150 mM NaF. All data was collected at 25 ºC by monitoring the ellipticity at 222 nm ( ) and the ratio between the absorbance at 400 nm and the absorbance at 287 nm (Δ). The insets show the denaturation curves measured by classic UV spectroscopy at single wavelength. A1 and A2 are the curves for cytochrome c at 287 nm and 400 nm, respectively, whereas B1 and B2 are the curves measured for myoglobin at the corresponding wavelengths.

This work was supported by grant SFRH/BD/23604/2005 (L.M.O.) from the Fundação para

Ciência e a Tecnologia – Ministério da Ciência, Tecnologia e Ensino Superior, Portugal .

 131

Novel UV absorbance ration method 

 132

References

References

 133

References

References

(2005). ʺSo much more to know.ʺ Science 309(5731): 78‐102.

A. Rodger, B. N. (1997). Circular Dichroism and Linear Dichroism. Oxford, Oxford Univ. Press.

Acharya, A. S. and J. M. Manning (1980). ʺReactivity of the amino groups of carbonmonoxyhemoglobin S with

glyceraldehyde.ʺ J Biol Chem 255(4): 1406‐1412.

Aguzzi, A. and M. Heikenwalder (2006). ʺPathogenesis of prion diseases: current status and future outlook.ʺ Nat Rev

Microbiol 4(10): 765‐775.

Ahmad, A., V. N. Uversky, D. Hong and A. L. Fink (2005). ʺEarly events in the fibrillation of monomeric insulin.ʺ J Biol

Chem 280(52): 42669‐42675.

Ahmed, M. U., E. Brinkmann Frye, T. P. Degenhardt, S. R. Thorpe and J. W. Baynes (1997). ʺN‐epsilon‐

(carboxyethyl)lysine, a product of the chemical modification of proteins by methylglyoxal, increases with age in

human lens proteins.ʺ Biochem J 324 ( Pt 2): 565‐570.

Ahmed, N., D. Dobler, M. Dean and P. J. Thornalley (2005). ʺPeptide mapping identifies hotspot site of modification in

human serum albumin by methylglyoxal involved in ligand binding and esterase activity.ʺ J Biol Chem 280(7):

5724‐5732.

Ahmed, N., P. J. Thornalley, J. Dawczynski, S. Franke, J. Strobel, G. Stein and G. M. Haik (2003). ʺMethylglyoxal‐derived

hydroimidazolone advanced glycation end‐products of human lens proteins.ʺ Invest Ophthalmol Vis Sci 44(12):

5287‐5292.

Alam, Z. I., S. E. Daniel, A. J. Lees, D. C. Marsden, P. Jenner and B. Halliwell (1997a). ʺA generalised increase in protein

carbonyls in the brain in Parkinsonʹs but not incidental Lewy body disease.ʺ J Neurochem 69(3): 1326‐1329.

Alam, Z. I., A. Jenner, S. E. Daniel, A. J. Lees, N. Cairns, C. D. Marsden, P. Jenner and B. Halliwell (1997b). ʺOxidative

DNA damage in the parkinsonian brain: an apparent selective increase in 8‐hydroxyguanine levels in substantia

nigra.ʺ J Neurochem 69(3): 1196‐1203.

Aleksandrovskii, Y. A. (1992). ʺAntithrombin III, C1 inhibitor, methylglyoxal, and polymorphonuclear leukocytes in the

development of vascular complications in diabetes mellitus.ʺ Thromb Res 67(2): 179‐189.

Altenberg, B. and K. O. Greulich (2004). ʺGenes of glycolysis are ubiquitously overexpressed in 24 cancer classes.ʺ

Genomics 84(6): 1014‐1020.

Ancolio, K., C. Alves da Costa, K. Ueda and F. Checler (2000). ʺAlpha‐synuclein and the Parkinsonʹs disease‐related mutant

Ala53Thr‐alpha‐synuclein do not undergo proteasomal degradation in HEK293 and neuronal cells.ʺ Neurosci Lett

285(2): 79‐82.

 135

References 

Anderson, J. P., D. E. Walker, J. M. Goldstein, R. de Laat, K. Banducci, R. J. Caccavello, R. Barbour, J. Huang, K. Kling, M.

Lee, L. Diep, P. S. Keim, X. Shen, T. Chataway, M. G. Schlossmacher, P. Seubert, D. Schenk, S. Sinha, W. P. Gai

and T. J. Chilcote (2006). ʺPhosphorylation of Ser‐129 is the dominant pathological modification of alpha‐

synuclein in familial and sporadic Lewy body disease.ʺ J Biol Chem 281(40): 29739‐29752.

Andreeva, A. and A. G. Murzin (2006). ʺEvolution of protein fold in the presence of functional constraints.ʺ Curr Opin Struct

Biol 16(3): 399‐408.

Anfinsen, C. B., E. Haber, M. Sela and F. H. White, Jr. (1961). ʺThe kinetics of formation of native ribonuclease during

oxidation of the reduced polypeptide chain.ʺ Proc Natl Acad Sci U S A 47: 1309‐1314.

Anfinsen, C. B. and H. A. Scheraga (1975). ʺExperimental and theoretical aspects of protein folding.ʺ Adv Protein Chem 29:

205‐300.

Antony, T., W. Hoyer, D. Cherny, G. Heim, T. M. Jovin and V. Subramaniam (2003). ʺCellular polyamines promote the

aggregation of alpha‐synuclein.ʺ J Biol Chem 278(5): 3235‐3240.

Antonyuk, S., J. S. Elam, M. A. Hough, R. W. Strange, P. A. Doucette, J. A. Rodriguez, L. J. Hayward, J. S. Valentine, P. J.

Hart and S. S. Hasnain (2005). ʺStructural consequences of the familial amyotrophic lateral sclerosis SOD1 mutant

His46Arg.ʺ Protein Sci 14(5): 1201‐1213.

Argirova, M. and W. Breipohl (2002). ʺComparison between modifications of lens proteins resulted from glycation with

methylglyoxal, glyoxal, ascorbic acid, and fructose.ʺ J Biochem Mol Toxicol 16(3): 140‐145.

Armen, R. S., D. O. Alonso and V. Daggett (2004). ʺAnatomy of an amyloidogenic intermediate: conversion of beta‐sheet to

alpha‐sheet structure in transthyretin at acidic pH.ʺ Structure 12(10): 1847‐1863.

Artymiuk, P. J. and C. C. Blake (1981). ʺRefinement of human lysozyme at 1.5 A resolution analysis of non‐bonded and

hydrogen‐bond interactions.ʺ J Mol Biol 152(4): 737‐762.

Auton, M., L. M. Holthauzen and D. W. Bolen (2007). ʺAnatomy of energetic changes accompanying urea‐induced protein

denaturation.ʺ Proc Natl Acad Sci U S A 104(39): 15317‐15322.

Auvinen, M., A. Paasinen, L. C. Andersson and E. Holtta (1992). ʺOrnithine decarboxylase activity is critical for cell

transformation.ʺ Nature 360(6402): 355‐358.

Baba, M., S. Nakajo, P. H. Tu, T. Tomita, K. Nakaya, V. M. Lee, J. Q. Trojanowski and T. Iwatsubo (1998). ʺAggregation of

alpha‐synuclein in Lewy bodies of sporadic Parkinsonʹs disease and dementia with Lewy bodies.ʺ Am J Pathol

152(4): 879‐884.

Backmann, J., G. Schafer, L. Wyns and H. Bonisch (1998). ʺThermodynamics and kinetics of unfolding of the thermostable

trimeric adenylate kinase from the archaeon Sulfolobus acidocaldarius.ʺ J Mol Biol 284(3): 817‐833.

Baggetto, L. G. and A. L. Lehninger (1987). ʺIsolated tumoral pyruvate dehydrogenase can synthesize acetoin which inhibits

pyruvate oxidation as well as other aldehydes.ʺ Biochem Biophys Res Commun 145(1): 153‐159.

Bai, Y., J. S. Milne, L. Mayne and S. W. Englander (1994). ʺProtein stability parameters measured by hydrogen exchange.ʺ

Proteins 20(1): 4‐14.

 136

References

Baker, E. N., T. L. Blundell, J. F. Cutfield, S. M. Cutfield, E. J. Dodson, G. G. Dodson, D. M. Hodgkin, R. E. Hubbard, N.

W. Isaacs, C. D. Reynolds and et al. (1988). ʺThe structure of 2Zn pig insulin crystals at 1.5 A resolution.ʺ Philos

Trans R Soc Lond B Biol Sci 319(1195): 369‐456.

Bakhti, M., M. Habibi‐Rezaei, A. A. Moosavi‐Movahedi and M. R. Khazaei (2007). ʺConsequential Alterations in

Haemoglobin Structure upon Glycation with Fructose: Prevention by Acetylsalicylic Acid.ʺ J Biochem (Tokyo)

141(6): 827‐833.

Baldwin, R. L. (1989). ʺHow does protein folding get started?ʺ Trends Biochem Sci 14(7): 291‐294.

Banavar, J. R., T. X. Hoang, A. Maritan, F. Seno and A. Trovato (2004). ʺUnified perspective on proteins: a physics

approach.ʺ Phys Rev E Stat Nonlin Soft Matter Phys 70(4 Pt 1): 041905.

Banavar, J. R. and A. Maritan (2007). ʺPhysics of proteins.ʺ Annu Rev Biophys Biomol Struct 36: 261‐280.

Banci, L., I. Bertini, A. Durazo, S. Girotto, E. B. Gralla, M. Martinelli, J. S. Valentine, M. Vieru and J. P. Whitelegge (2007).

ʺMetal‐free superoxide dismutase forms soluble oligomers under physiological conditions: a possible general

mechanism for familial ALS.ʺ Proc Natl Acad Sci U S A 104(27): 11263‐11267.

Barrios‐Rodiles, M., K. R. Brown, B. Ozdamar, R. Bose, Z. Liu, R. S. Donovan, F. Shinjo, Y. Liu, J. Dembowy, I. W. Taylor,

V. Luga, N. Przulj, M. Robinson, H. Suzuki, Y. Hayashizaki, I. Jurisica and J. L. Wrana (2005). ʺHigh‐throughput

mapping of a dynamic signaling network in mammalian cells.ʺ Science 307(5715): 1621‐1625.

Bartalesi, I., I. Bertini, K. Ghosh, A. Rosato and P. Turano (2002). ʺThe unfolding of oxidized c‐type cytochromes: the

instructive case of Bacillus pasteurii.ʺ J Mol Biol 321(4): 693‐701.

Battiste, J. L., H. Mao, N. S. Rao, R. Tan, D. R. Muhandiram, L. E. Kay, A. D. Frankel and J. R. Williamson (1996). ʺAlpha

helix‐RNA major groove recognition in an HIV‐1 rev peptide‐RRE RNA complex.ʺ Science 273(5281): 1547‐1551.

Baynes, J. W. and S. R. Thorpe (1999). ʺRole of oxidative stress in diabetic complications: a new perspective on an old

paradigm.ʺ Diabetes 48(1): 1‐9.

Beal, M. F. (2002). ʺOxidatively modified proteins in aging and disease.ʺ Free Radic Biol Med 32(9): 797‐803.

Becktel, W. J. and J. A. Schellman (1987). ʺProtein stability curves.ʺ Biopolymers 26(11): 1859‐1877.

Bennett, M. C., J. F. Bishop, Y. Leng, P. B. Chock, T. N. Chase and M. M. Mouradian (1999). ʺDegradation of alpha‐

synuclein by proteasome.ʺ J Biol Chem 274(48): 33855‐33858.

Beyer, K. (2006). ʺAlpha‐synuclein structure, posttranslational modification and alternative splicing as aggregation

enhancers.ʺ Acta Neuropathol 112(3): 237‐251.

Bhattacharyya, R. P., A. Remenyi, B. J. Yeh and W. A. Lim (2006). ʺDomains, motifs, and scaffolds: the role of modular

interactions in the evolution and wiring of cell signaling circuits.ʺ Annu Rev Biochem 75: 655‐680.

Biere, A. L., S. J. Wood, J. Wypych, S. Steavenson, Y. Jiang, D. Anafi, F. W. Jacobsen, M. A. Jarosinski, G. M. Wu, J. C.

Louis, F. Martin, L. O. Narhi and M. Citron (2000). ʺParkinsonʹs disease‐associated alpha‐synuclein is more

fibrillogenic than beta‐ and gamma‐synuclein and cannot cross‐seed its homologs.ʺ J Biol Chem 275(44): 34574‐

34579.

 137

References 

Blake, C. and L. Serpell (1996). ʺSynchrotron X‐ray studies suggest that the core of the transthyretin amyloid fibril is a

continuous beta‐sheet helix.ʺ Structure 4(8): 989‐998.

Blake, C. C., M. J. Geisow, S. J. Oatley, B. Rerat and C. Rerat (1978). ʺStructure of prealbumin: secondary, tertiary and

quaternary interactions determined by Fourier refinement at 1.8 A.ʺ J Mol Biol 121(3): 339‐356.

Blundell, T. L., J. F. Cutfield, S. M. Cutfield, E. J. Dodson, G. G. Dodson, D. C. Hodgkin and D. A. Mercola (1972). ʺThree‐

dimensional atomic structure of insulin and its relationship to activity.ʺ Diabetes 21(2 Suppl): 492‐505.

Blundell, T. L., J. F. Cutfield, S. M. Cutfield, E. J. Dodson, G. G. Dodson, D. C. Hodgkin, D. A. Mercola and M. Vijayan

(1971). ʺAtomic positions in rhombohedral 2‐zinc insulin crystals.ʺ Nature 231(5304): 506‐511.

Boillee, S., C. Vande Velde and D. W. Cleveland (2006). ʺALS: a disease of motor neurons and their nonneuronal

neighbors.ʺ Neuron 52(1): 39‐59.

Bolen, D. W. and M. M. Santoro (1988). ʺUnfolding free energy changes determined by the linear extrapolation method. 2.

Incorporation of delta G degrees N‐U values in a thermodynamic cycle.ʺ Biochemistry 27(21): 8069‐8074.

Bonini, N. M. and B. I. Giasson (2005). ʺSnaring the function of alpha‐synuclein.ʺ Cell 123(3): 359‐361.

Booth, D. R., M. B. Pepys and P. N. Hawkins (2000). ʺA novel variant of human lysozyme (T70N) is common in the normal

population.ʺ Hum Mutat 16(2): 180.

Booth, D. R., M. Sunde, V. Bellotti, C. V. Robinson, W. L. Hutchinson, P. E. Fraser, P. N. Hawkins, C. M. Dobson, S. E.

Radford, C. C. Blake and M. B. Pepys (1997). ʺInstability, unfolding and aggregation of human lysozyme variants

underlying amyloid fibrillogenesis.ʺ Nature 385(6619): 787‐793.

Borreguero, J. M., F. Ding, S. V. Buldyrev, H. E. Stanley and N. V. Dokholyan (2004). ʺMultiple folding pathways of the

SH3 domain.ʺ Biophys J 87(1): 521‐533.

Bouchard, M., J. Zurdo, E. J. Nettleton, C. M. Dobson and C. V. Robinson (2000). ʺFormation of insulin amyloid fibrils

followed by FTIR simultaneously with CD and electron microscopy.ʺ Protein Sci 9(10): 1960‐1967.

Bouma, B., L. M. Kroon‐Batenburg, Y. P. Wu, B. Brunjes, G. Posthuma, O. Kranenburg, P. G. de Groot, E. E. Voest and M.

F. Gebbink (2003). ʺGlycation induces formation of amyloid cross‐beta structure in albumin.ʺ J Biol Chem 278(43):

41810‐41819.

Bowie, J. U., J. F. Reidhaar‐Olson, W. A. Lim and R. T. Sauer (1990). ʺDeciphering the message in protein sequences:

tolerance to amino acid substitutions.ʺ Science 247(4948): 1306‐1310.

Bowling, A. C. and M. F. Beal (1995). ʺBioenergetic and oxidative stress in neurodegenerative diseases.ʺ Life Sci 56(14): 1151‐

1171.

Brandts, J. F. and L. N. Lin (1986). ʺProline isomerization studied with proteolytic enzymes.ʺ Methods Enzymol 131: 107‐126.

Brange, J. (1987). Galenics of insulin. The physicochemical and pharmaceutical aspects of insulin and insulin preparations.

Berlin, Springer‐Verlag.

Brange, J., L. Andersen, E. D. Laursen, G. Meyn and E. Rasmussen (1997). ʺToward understanding insulin fibrillation.ʺ J

Pharm Sci 86(5): 517‐525.  138

References

Brownlee, M. (1995). ʺAdvanced protein glycosylation in diabetes and aging.ʺ Annu Rev Med 46: 223‐234.

Brownlee, M., H. Vlassara and A. Cerami (1984). ʺNonenzymatic glycosylation and the pathogenesis of diabetic

complications.ʺ Ann Intern Med 101(4): 527‐537.

Bryngelson, J. D., J. N. Onuchic, N. D. Socci and P. G. Wolynes (1995). ʺFunnels, pathways, and the energy landscape of

protein folding: a synthesis.ʺ Proteins 21(3): 167‐195.

Bucala, R. and A. Cerami (1992). ʺAdvanced glycosylation: chemistry, biology, and implications for diabetes and aging.ʺ

Adv Pharmacol 23: 1‐34.

Bucciantini, M., S. Rigacci, A. Berti, L. Pieri, C. Cecchi, D. Nosi, L. Formigli, F. Chiti and M. Stefani (2005). ʺPatterns of cell

death triggered in two different cell lines by HypF‐N prefibrillar aggregates.ʺ FASEB J 19(3): 437‐439.

Buchet, R., S. Varga, N. W. Seidler, E. Molnar and A. Martonosi (1991). ʺPolarized infrared attenuated total reflectance

spectroscopy of the Ca(2+)‐ATPase of sarcoplasmic reticulum.ʺ Biochim Biophys Acta 1068(2): 201‐216.

Bunn, H. F., D. N. Haney, K. H. Gabbay and P. M. Gallop (1975). ʺFurther identification of the nature and linkage of the

carbohydrate in hemoglobin A1c.ʺ Biochem Biophys Res Commun 67(1): 103‐109.

Bunn, H. F. and P. J. Higgins (1981). ʺReaction of monosaccharides with proteins: possible evolutionary significance.ʺ

Science 213(4504): 222‐224.

Burke, M. J. and M. A. Rougvie (1972). ʺCross‐ protein structures. I. Insulin fibrils.ʺ Biochemistry 11(13): 2435‐2439.

Bussell, R., Jr. and D. Eliezer (2003). ʺA structural and functional role for 11‐mer repeats in alpha‐synuclein and other

exchangeable lipid binding proteins.ʺ J Mol Biol 329(4): 763‐778.

Buxbaum, J. N. (2004). ʺThe systemic amyloidoses.ʺ Curr Opin Rheumatol 16(1): 67‐75.

Bychkova, V. E., R. Berni, G. L. Rossi, V. P. Kutyshenko and O. B. Ptitsyn (1992). ʺRetinol‐binding protein is in the molten

globule state at low pH.ʺ Biochemistry 31(33): 7566‐7571.

Bychkova, V. E., R. H. Pain and O. B. Ptitsyn (1988). ʺThe ʹmolten globuleʹ state is involved in the translocation of proteins

across membranes?ʺ FEBS Lett 238(2): 231‐234.

Byrne, M. P., R. L. Manuel, L. G. Lowe and W. E. Stites (1995). ʺEnergetic contribution of side chain hydrogen bonding to

the stability of staphylococcal nuclease.ʺ Biochemistry 34(42): 13949‐13960.

Caflisch, A. (2004). ʺProtein folding: simple models for a complex process.ʺ Structure 12(10): 1750‐1752.

Calvert, J. F., J. L. Hill and A. Dong (1997). ʺRedox‐dependent conformational changes are common structural features of

cytochrome c from various species.ʺ Arch Biochem Biophys 346(2): 287‐293.

Canet, D., A. M. Last, P. Tito, M. Sunde, A. Spencer, D. B. Archer, C. Redfield, C. V. Robinson and C. M. Dobson (2002).

ʺLocal cooperativity in the unfolding of an amyloidogenic variant of human lysozyme.ʺ Nat Struct Biol 9(4): 308‐

315.

Casazza, J. P., M. E. Felver and R. L. Veech (1984). ʺThe metabolism of acetone in rat.ʺ J Biol Chem 259(1): 231‐236.

 139

References 

Casserly, I. and E. Topol (2004). ʺConvergence of atherosclerosis and Alzheimerʹs disease: inflammation, cholesterol, and

misfolded proteins.ʺ Lancet 363(9415): 1139‐1146.

Castellani, R., M. A. Smith, P. L. Richey and G. Perry (1996). ʺGlycoxidation and oxidative stress in Parkinson disease and

diffuse Lewy body disease.ʺ Brain Res 737(1‐2): 195‐200.

Castellani, R. J., G. Perry, S. L. Siedlak, A. Nunomura, S. Shimohama, J. Zhang, T. Montine, L. M. Sayre and M. A. Smith

(2002). ʺHydroxynonenal adducts indicate a role for lipid peroxidation in neocortical and brainstem Lewy bodies

in humans.ʺ Neurosci Lett 319(1): 25‐28.

Caughey, B. and P. T. Lansbury (2003). ʺProtofibrils, pores, fibrils, and neurodegeneration: separating the responsible

protein aggregates from the innocent bystanders.ʺ Annu Rev Neurosci 26: 267‐298.

Chandra, S., G. Gallardo, R. Fernandez‐Chacon, O. M. Schluter and T. C. Sudhof (2005). ʺAlpha‐synuclein cooperates with

CSPalpha in preventing neurodegeneration.ʺ Cell 123(3): 383‐396.

Chang, T. and L. Wu (2006). ʺMethylglyoxal, oxidative stress, and hypertension.ʺ Can J Physiol Pharmacol 84(12): 1229‐1238.

Chellan, P. and R. H. Nagaraj (1999). ʺProtein crosslinking by the Maillard reaction: dicarbonyl‐derived imidazolium

crosslinks in aging and diabetes.ʺ Arch Biochem Biophys 368(1): 98‐104.

Chen, J. and W. E. Stites (2001). ʺPacking is a key selection factor in the evolution of protein hydrophobic cores.ʺ

Biochemistry 40(50): 15280‐15289.

Chen, K., J. Maley and P. H. Yu (2006). ʺPotential inplications of endogenous aldehydes in beta‐amyloid misfolding,

oligomerization and fibrillogenesis.ʺ J Neurochem 99(5): 1413‐1424.

Chen, L. and M. B. Feany (2005). ʺAlpha‐synuclein phosphorylation controls neurotoxicity and inclusion formation in a

Drosophila model of Parkinson disease.ʺ Nat Neurosci 8(5): 657‐663.

Chinta, S. J. and J. K. Andersen (2005). ʺDopaminergic neurons.ʺ Int J Biochem Cell Biol 37(5): 942‐946.

Chiti, F. and C. M. Dobson (2006). ʺProtein misfolding, functional amyloid, and human disease.ʺ Annu Rev Biochem 75:

333‐366.

Chiti, F., P. Webster, N. Taddei, A. Clark, M. Stefani, G. Ramponi and C. M. Dobson (1999). ʺDesigning conditions for in

vitro formation of amyloid protofilaments and fibrils.ʺ Proc Natl Acad Sci U S A 96(7): 3590‐3594.

Choi, Y. G., J. I. Kim, Y. C. Jeon, S. J. Park, E. K. Choi, R. Rubenstein, R. J. Kascsak, R. I. Carp and Y. S. Kim (2004).

ʺNonenzymatic glycation at the N terminus of pathogenic prion protein in transmissible spongiform

encephalopathies.ʺ J Biol Chem 279(29): 30402‐30409.

Chothia, C. (1992). ʺProteins. One thousand families for the molecular biologist.ʺ Nature 357(6379): 543‐544.

Chothia, C. and A. V. Finkelstein (1990). ʺThe classification and origins of protein folding patterns.ʺ Annu Rev Biochem 59:

1007‐1039.

Chu, I., J. Sun, A. Arnaout, H. Kahn, W. Hanna, S. Narod, P. Sun, C. K. Tan, L. Hengst and J. Slingerland (2007). ʺp27

phosphorylation by Src regulates inhibition of cyclin E‐Cdk2.ʺ Cell 128(2): 281‐294.

 140

References

Clark, R. S., H. Bayir and L. W. Jenkins (2005). ʺPosttranslational protein modifications.ʺ Crit Care Med 33(12 Suppl): S407‐

409.

Clayton, D. F. and J. M. George (1998). ʺThe synucleins: a family of proteins involved in synaptic function, plasticity,

neurodegeneration and disease.ʺ Trends Neurosci 21(6): 249‐254.

Cleary, J. P., D. M. Walsh, J. J. Hofmeister, G. M. Shankar, M. A. Kuskowski, D. J. Selkoe and K. H. Ashe (2005). ʺNatural

oligomers of the amyloid‐beta protein specifically disrupt cognitive function.ʺ Nat Neurosci 8(1): 79‐84.

Clementi, C., H. Nymeyer and J. N. Onuchic (2000). ʺTopological and energetic factors: what determines the structural

details of the transition state ensemble and ʺen‐routeʺ intermediates for protein folding? An investigation for small

globular proteins.ʺ J Mol Biol 298(5): 937‐953.

Colaco, C. A. and C. R. Harrington (1994). ʺGlycation: a pathological modification in neuropathies?: a hypothesis.ʺ

Neuroreport 5(8): 859‐861.

Colon, W. and J. W. Kelly (1992). ʺPartial denaturation of transthyretin is sufficient for amyloid fibril formation in vitro.ʺ

Biochemistry 31(36): 8654‐8660.

Connors, L. H., A. M. Richardson, R. Theberge and C. E. Costello (2000). ʺTabulation of transthyretin (TTR) variants as of

1/1/2000.ʺ Amyloid 7(1): 54‐69.

Cooper, R. A. and A. Anderson (1970). ʺThe formation and catabolism of methylglyoxal during glycolysis in Escherichia

coli.ʺ FEBS Lett 11(4): 273‐276.

Cordeiro, C. and A. Ponces Freire (1996). ʺMethylglyoxal assay in cells as 2‐methylquinoxaline using 1,2‐diaminobenzene as

derivatizing reagent.ʺ Anal Biochem 234(2): 221‐224.

Cordes, M. H., A. R. Davidson and R. T. Sauer (1996). ʺSequence space, folding and protein design.ʺ Curr Opin Struct Biol

6(1): 3‐10.

Cornwell, G. G., 3rd, K. Sletten, B. Johansson and P. Westermark (1988). ʺEvidence that the amyloid fibril protein in senile

systemic amyloidosis is derived from normal prealbumin.ʺ Biochem Biophys Res Commun 154(2): 648‐653.

Cowan, C. M. and L. A. Raymond (2006). ʺSelective neuronal degeneration in Huntingtonʹs disease.ʺ Curr Top Dev Biol 75:

25‐71.

Crowther, D. C., K. J. Kinghorn, E. Miranda, R. Page, J. A. Curry, F. A. Duthie, D. C. Gubb and D. A. Lomas (2005).

ʺIntraneuronal Abeta, non‐amyloid aggregates and neurodegeneration in a Drosophila model of Alzheimerʹs

disease.ʺ Neuroscience 132(1): 123‐135.

Daggett, V. (2002). ʺMolecular dynamics simulations of the protein unfolding/folding reaction.ʺ Acc Chem Res 35(6): 422‐

429.

Daggett, V. and A. R. Fersht (2003). ʺIs there a unifying mechanism for protein folding?ʺ Trends Biochem Sci 28(1): 18‐25.

Dahiyat, B. I. and S. L. Mayo (1997). ʺDe novo protein design: fully automated sequence selection.ʺ Science 278(5335): 82‐87.

Dakin, H. D. and H. W. Dudley (1913a). ʺAn enzyme concerned with the formation of hydroxy acids from ketonic

aldehydes.ʺ J Biol Chem 14: 155‐157.  141

References 

Dakin, H. D. and H. W. Dudley (1913b). ʺOn glyoxalase.ʺ J Biol Chem 14: 423‐431.

Dalfo, E., M. Portero‐Otin, V. Ayala, A. Martinez, R. Pamplona and I. Ferrer (2005). ʺEvidence of oxidative stress in the

neocortex in incidental Lewy body disease.ʺ J Neuropathol Exp Neurol 64(9): 816‐830.

Danzer, K. M., D. Haasen, A. R. Karow, S. Moussaud, M. Habeck, A. Giese, H. Kretzschmar, B. Hengerer and M. Kostka

(2007). ʺDifferent species of alpha‐synuclein oligomers induce calcium influx and seeding.ʺ J Neurosci 27(34): 9220‐

9232.

David, D. C., R. Layfield, L. Serpell, Y. Narain, M. Goedert and M. G. Spillantini (2002). ʺProteasomal degradation of tau

protein.ʺ J Neurochem 83(1): 176‐185.

Davidson, A. R. and R. T. Sauer (1994). ʺFolded proteins occur frequently in libraries of random amino acid sequences.ʺ

Proc Natl Acad Sci U S A 91(6): 2146‐2150.

Davidson, W. S., A. Jonas, D. F. Clayton and J. M. George (1998). ʺStabilization of alpha‐synuclein secondary structure

upon binding to synthetic membranes.ʺ J Biol Chem 273(16): 9443‐9449.

De Maeyer, M., J. Desmet and I. Lasters (1997). ʺAll in one: a highly detailed rotamer library improves both accuracy and

speed in the modelling of sidechains by dead‐end elimination.ʺ Fold Des 2(1): 53‐66.

Deechongkit, S., P. E. Dawson and J. W. Kelly (2004). ʺToward assessing the position‐dependent contributions of backbone

hydrogen bonding to beta‐sheet folding thermodynamics employing amide‐to‐ester perturbations.ʺ J Am Chem

Soc 126(51): 16762‐16771.

Demarest, S. J., M. Martinez‐Yamout, J. Chung, H. Chen, W. Xu, H. J. Dyson, R. M. Evans and P. E. Wright (2002). ʺMutual

synergistic folding in recruitment of CBP/p300 by p160 nuclear receptor coactivators.ʺ Nature 415(6871): 549‐553.

Demuro, A., E. Mina, R. Kayed, S. C. Milton, I. Parker and C. G. Glabe (2005). ʺCalcium dysregulation and membrane

disruption as a ubiquitous neurotoxic mechanism of soluble amyloid oligomers.ʺ J Biol Chem 280(17): 17294‐17300.

Dexter, D. T., F. R. Wells, F. Agid, Y. Agid, A. J. Lees, P. Jenner and C. D. Marsden (1987). ʺIncreased nigral iron content in

postmortem parkinsonian brain.ʺ Lancet 2(8569): 1219‐1220.

DiDonato, M., L. Craig, M. E. Huff, M. M. Thayer, R. M. Cardoso, C. J. Kassmann, T. P. Lo, C. K. Bruns, E. T. Powers, J. W.

Kelly, E. D. Getzoff and J. A. Tainer (2003). ʺALS mutants of human superoxide dismutase form fibrous

aggregates via framework destabilization.ʺ J Mol Biol 332(3): 601‐615.

Dill, K. A. (1985). ʺTheory for the folding and stability of globular proteins.ʺ Biochemistry 24(6): 1501‐1509.

Dill, K. A. (1990). ʺDominant forces in protein folding.ʺ Biochemistry 29(31): 7133‐7155.

Dill, K. A. (1999). ʺPolymer principles and protein folding.ʺ Protein Sci 8(6): 1166‐1180.

Dill, K. A. and H. S. Chan (1997). ʺFrom Levinthal to pathways to funnels.ʺ Nat Struct Biol 4(1): 10‐19.

Dische, F. E., C. Wernstedt, G. T. Westermark, P. Westermark, M. B. Pepys, J. A. Rennie, S. G. Gilbey and P. J. Watkins

(1988). ʺInsulin as an amyloid‐fibril protein at sites of repeated insulin injections in a diabetic patient.ʺ

Diabetologia 31(3): 158‐161.

 142

References

Dohmen, R. J. (2004). ʺSUMO protein modification.ʺ Biochim Biophys Acta 1695(1‐3): 113‐131.

Dolgikh, D. A., R. I. Gilmanshin, E. V. Brazhnikov, V. E. Bychkova, G. V. Semisotnov, S. Venyaminov and O. B. Ptitsyn

(1981). ʺAlpha‐Lactalbumin: compact state with fluctuating tertiary structure?ʺ FEBS Lett 136(2): 311‐315.

Dorval, V. and P. E. Fraser (2006). ʺSmall ubiquitin‐like modifier (SUMO) modification of natively unfolded proteins tau and

alpha‐synuclein.ʺ J Biol Chem 281(15): 9919‐9924.

Driessen, H. P., W. W. de Jong, G. I. Tesser and H. Bloemendal (1985). ʺThe mechanism of N‐terminal acetylation of

proteins.ʺ CRC Crit Rev Biochem 18(4): 281‐325.

DuBay, K. F., A. P. Pawar, F. Chiti, J. Zurdo, C. M. Dobson and M. Vendruscolo (2004). ʺPrediction of the absolute

aggregation rates of amyloidogenic polypeptide chains.ʺ J Mol Biol 341(5): 1317‐1326.

Dumoulin, M., D. Canet, A. M. Last, E. Pardon, D. B. Archer, S. Muyldermans, L. Wyns, A. Matagne, C. V. Robinson, C.

Redfield and C. M. Dobson (2005). ʺReduced global cooperativity is a common feature underlying the

amyloidogenicity of pathogenic lysozyme mutations.ʺ J Mol Biol 346(3): 773‐788.

Dunbrack, R. L., Jr. and F. E. Cohen (1997). ʺBayesian statistical analysis of protein side‐chain rotamer preferences.ʺ Protein

Sci 6(8): 1661‐1681.

Dunker, A. K., C. J. Brown, J. D. Lawson, L. M. Iakoucheva and Z. Obradovic (2002). ʺIntrinsic disorder and protein

function.ʺ Biochemistry 41(21): 6573‐6582.

Dunker, A. K., E. Garner, S. Guilliot, P. Romero, K. Albrecht, J. Hart, Z. Obradovic, C. Kissinger and J. E. Villafranca

(1998). ʺProtein disorder and the evolution of molecular recognition: theory, predictions and observations.ʺ Pac

Symp Biocomput: 473‐484.

Dunker, A. K., J. D. Lawson, C. J. Brown, R. M. Williams, P. Romero, J. S. Oh, C. J. Oldfield, A. M. Campen, C. M. Ratliff,

K. W. Hipps, J. Ausio, M. S. Nissen, R. Reeves, C. Kang, C. R. Kissinger, R. W. Bailey, M. D. Griswold, W. Chiu,

E. C. Garner and Z. Obradovic (2001). ʺIntrinsically disordered protein.ʺ J Mol Graph Model 19(1): 26‐59.

Dunker, A. K. and Z. Obradovic (2001). ʺThe protein trinity‐‐linking function and disorder.ʺ Nat Biotechnol 19(9): 805‐806.

Dunker, A. K., Z. Obradovic, P. Romero, E. C. Garner and C. J. Brown (2000). ʺIntrinsic protein disorder in complete

genomes.ʺ Genome Inform Ser Workshop Genome Inform 11: 161‐171.

Dunker, A. K., C. J. Oldfield, J. Meng, P. Romero, J. Y. Yang, J. W. Chen, V. Vacic, Z. Obradovic and V. N. Uversky (2008).

ʺThe unfoldomics decade: an update on intrinsically disordered proteins.ʺ BMC Genomics 9 Suppl 2: S1.

Dyson, H. J. and P. E. Wright (2005). ʺIntrinsically unstructured proteins and their functions.ʺ Nat Rev Mol Cell Biol 6(3):

197‐208.

Eble, A. S., S. R. Thorpe and J. W. Baynes (1983). ʺNonenzymatic glucosylation and glucose‐dependent cross‐linking of

protein.ʺ J Biol Chem 258(15): 9406‐9412. el‐Agnaf, O. M. and G. B. Irvine (2002). ʺAggregation and neurotoxicity of alpha‐synuclein and related peptides.ʺ Biochem

Soc Trans 30(4): 559‐565.

 143

References 

Ellis, C. E., P. L. Schwartzberg, T. L. Grider, D. W. Fink and R. L. Nussbaum (2001). ʺalpha‐synuclein is phosphorylated by

members of the Src family of protein‐tyrosine kinases.ʺ J Biol Chem 276(6): 3879‐3884.

Elsasser, S. and D. Finley (2005). ʺDelivery of ubiquitinated substrates to protein‐unfolding machines.ʺ Nat Cell Biol 7(8):

742‐749.

Emerit, J., M. Edeas and F. Bricaire (2004). ʺNeurodegenerative diseases and oxidative stress.ʺ Biomed Pharmacother 58(1):

39‐46.

Esterbauer, H., K. H. Cheeseman, M. U. Dianzani, G. Poli and T. F. Slater (1982). ʺSeparation and characterization of the

aldehydic products of lipid peroxidation stimulated by ADP‐Fe2+ in rat liver microsomes.ʺ Biochem J 208(1): 129‐

140.

Fahmy, K., F. Jager, M. Beck, T. A. Zvyaga, T. P. Sakmar and F. Siebert (1993). ʺProtonation states of membrane‐embedded

carboxylic acid groups in rhodopsin and metarhodopsin II: a Fourier‐transform infrared spectroscopy study of

site‐directed mutants.ʺ Proc Natl Acad Sci U S A 90(21): 10206‐10210.

Fandrich, M., M. A. Fletcher and C. M. Dobson (2001). ʺAmyloid fibrils from muscle myoglobin.ʺ Nature 410(6825): 165‐166.

Fandrich, M., V. Forge, K. Buder, M. Kittler, C. M. Dobson and S. Diekmann (2003). ʺMyoglobin forms amyloid fibrils by

association of unfolded polypeptide segments.ʺ Proc Natl Acad Sci U S A 100(26): 15463‐15468.

Farah, M. A., S. Bose, J. H. Lee, H. C. Jung and Y. Kim (2005). ʺAnalysis of glycated insulin by MALDI‐TOF mass

spectrometry.ʺ Biochim Biophys Acta 1725(3): 269‐282.

Fasman, G. D. (1996). Circular Dichroism and The Conformational Analysis of Biomolecules. New York, Plenum Press.

Fernandez, A., R. Scott and R. S. Berry (2004a). ʺThe nonconserved wrapping of conserved protein folds reveals a trend

toward increasing connectivity in proteomic networks.ʺ Proc Natl Acad Sci U S A 101(9): 2823‐2827.

Fernandez, C. O., W. Hoyer, M. Zweckstetter, E. A. Jares‐Erijman, V. Subramaniam, C. Griesinger and T. M. Jovin

(2004b). ʺNMR of alpha‐synuclein‐polyamine complexes elucidates the mechanism and kinetics of induced

aggregation.ʺ EMBO J 23(10): 2039‐2046.

Ferron, F., S. Longhi, B. Canard and D. Karlin (2006). ʺA practical overview of protein disorder prediction methods.ʺ

Proteins 65(1): 1‐14.

Fersht, A. R. (1995). ʺOptimization of rates of protein folding: the nucleation‐condensation mechanism and its implications.ʺ

Proc Natl Acad Sci U S A 92(24): 10869‐10873.

Fersht, A. R. (1997). ʺNucleation mechanisms in protein folding.ʺ Curr Opin Struct Biol 7(1): 3‐9.

Fersht, A. R., J. P. Shi, J. Knill‐Jones, D. M. Lowe, A. J. Wilkinson, D. M. Blow, P. Brick, P. Carter, M. M. Waye and G.

Winter (1985). ʺHydrogen bonding and biological specificity analysed by protein engineering.ʺ Nature 314(6008):

235‐238.

Fiebig, K. M., L. M. Rice, E. Pollock and A. T. Brunger (1999). ʺFolding intermediates of SNARE complex assembly.ʺ Nat

Struct Biol 6(2): 117‐123.

 144

References

Finch, C. E. and T. E. Morgan (2007). ʺSystemic inflammation, infection, ApoE alleles, and Alzheimer disease: a position

paper.ʺ Curr Alzheimer Res 4(2): 185‐189.

Floor, E. and M. G. Wetzel (1998). ʺIncreased protein oxidation in human substantia nigra pars compacta in comparison

with basal ganglia and prefrontal cortex measured with an improved dinitrophenylhydrazine assay.ʺ J

Neurochem 70(1): 268‐275.

Forte, G., B. Bocca, O. Senofonte, F. Petrucci, L. Brusa, P. Stanzione, S. Zannino, N. Violante, A. Alimonti and G.

Sancesario (2004). ʺTrace and major elements in whole blood, serum, cerebrospinal fluid and urine of patients

with Parkinsonʹs disease.ʺ J Neural Transm 111(8): 1031‐1040.

Frankel, A. D. and P. S. Kim (1991). ʺModular structure of transcription factors: implications for gene regulation.ʺ Cell 65(5):

717‐719.

Freichel, C., M. Neumann, T. Ballard, V. Muller, M. Woolley, L. Ozmen, E. Borroni, H. A. Kretzschmar, C. Haass, W.

Spooren and P. J. Kahle (2007). ʺAge‐dependent cognitive decline and amygdala pathology in alpha‐synuclein

transgenic mice.ʺ Neurobiol Aging 28(9): 1421‐1435.

Frye, E. B., T. P. Degenhardt, S. R. Thorpe and J. W. Baynes (1998). ʺRole of the Maillard reaction in aging of tissue proteins.

Advanced glycation end product‐dependent increase in imidazolium cross‐links in human lens proteins.ʺ J Biol

Chem 273(30): 18714‐18719.

Fu, M. X., J. R. Requena, A. J. Jenkins, T. J. Lyons, J. W. Baynes and S. R. Thorpe (1996). ʺThe advanced glycation end

product, Nepsilon‐(carboxymethyl)lysine, is a product of both lipid peroxidation and glycoxidation reactions.ʺ J

Biol Chem 271(17): 9982‐9986.

Fu, M. X., K. J. Wells‐Knecht, J. A. Blackledge, T. J. Lyons, S. R. Thorpe and J. W. Baynes (1994). ʺGlycation, glycoxidation,

and cross‐linking of collagen by glucose. Kinetics, mechanisms, and inhibition of late stages of the Maillard

reaction.ʺ Diabetes 43(5): 676‐683.

Fujiwara, H., M. Hasegawa, N. Dohmae, A. Kawashima, E. Masliah, M. S. Goldberg, J. Shen, K. Takio and T. Iwatsubo

(2002). ʺalpha‐Synuclein is phosphorylated in synucleinopathy lesions.ʺ Nat Cell Biol 4(2): 160‐164.

Fuxreiter, M., I. Simon, P. Friedrich and P. Tompa (2004). ʺPreformed structural elements feature in partner recognition by

intrinsically unstructured proteins.ʺ J Mol Biol 338(5): 1015‐1026.

Gallet, X., B. Charloteaux, A. Thomas and R. Brasseur (2000). ʺA fast method to predict protein interaction sites from

sequences.ʺ J Mol Biol 302(4): 917‐926.

Gao, M. and M. Karin (2005). ʺRegulating the regulators: control of protein ubiquitination and ubiquitin‐like modifications

by extracellular stimuli.ʺ Mol Cell 19(5): 581‐593.

Garner, E., P. Cannon, P. Romero, Z. Obradovic and A. K. Dunker (1998). ʺPredicting Disordered Regions from Amino

Acid Sequence: Common Themes Despite Differing Structural Characterization.ʺ Genome Inform Ser Workshop

Genome Inform 9: 201‐213.

 145

References 

Gast, K., H. Damaschun, K. Eckert, K. Schulze‐Forster, H. R. Maurer, M. Muller‐Frohne, D. Zirwer, J. Czarnecki and G.

Damaschun (1995). ʺProthymosin alpha: a biologically active protein with random coil conformation.ʺ

Biochemistry 34(40): 13211‐13218.

Gavin, A. C., P. Aloy, P. Grandi, R. Krause, M. Boesche, M. Marzioch, C. Rau, L. J. Jensen, S. Bastuck, B. Dumpelfeld, A.

Edelmann, M. A. Heurtier, V. Hoffman, C. Hoefert, K. Klein, M. Hudak, A. M. Michon, M. Schelder, M.

Schirle, M. Remor, T. Rudi, S. Hooper, A. Bauer, T. Bouwmeester, G. Casari, G. Drewes, G. Neubauer, J. M.

Rick, B. Kuster, P. Bork, R. B. Russell and G. Superti‐Furga (2006). ʺProteome survey reveals modularity of the

yeast cell machinery.ʺ Nature 440(7084): 631‐636.

George, J. M. (2002). ʺThe synucleins.ʺ Genome Biol 3(1): REVIEWS3002.

Gerhartz, B., I. Ekiel and M. Abrahamson (1998). ʺTwo stable unfolding intermediates of the disease‐causing L68Q variant

of human cystatin C.ʺ Biochemistry 37(49): 17309‐17317.

Gerstein, M. (1992). ʺA resolution‐sensitive procedure for comparing protein surfaces and its application to the comparison

of antigen‐combining sites.ʺ Acta Cryst A 48: 271.

Gerstein, M. (1997). ʺA structural census of genomes: comparing bacterial, eukaryotic, and archaeal genomes in terms of

protein structure.ʺ J Mol Biol 274(4): 562‐576.

Gertz, H. J., A. Siegers and J. Kuchinke (1994). ʺStability of cell size and nucleolar size in Lewy body containing neurons of

substantia nigra in Parkinsonʹs disease.ʺ Brain Res 637(1‐2): 339‐341.

Geula, C., C. K. Wu, D. Saroff, A. Lorenzo, M. Yuan and B. A. Yankner (1998). ʺAging renders the brain vulnerable to

amyloid beta‐protein neurotoxicity.ʺ Nat Med 4(7): 827‐831.

Ghiso, J., T. Wisniewski and B. Frangione (1994). ʺUnifying features of systemic and cerebral amyloidosis.ʺ Mol Neurobiol

8(1): 49‐64.

Gianni, S., Y. Ivarsson, P. Jemth, M. Brunori and C. Travaglini‐Allocatelli (2007). ʺIdentification and characterization of

protein folding intermediates.ʺ Biophys Chem 128(2‐3): 105‐113.

Gianni, S., C. Travaglini‐Allocatelli, F. Cutruzzola, M. Brunori, M. C. Shastry and H. Roder (2003). ʺParallel pathways in

cytochrome c(551) folding.ʺ J Mol Biol 330(5): 1145‐1152.

Gilli, P., V. Ferretti, G. Gilli and P. A. Borea (1994). ʺEnthalpy–entropy compensation in drug‐receptor binding.ʺ J Phys

Chem 98: 1515‐1518.

Ginsburg, V. and H. G. Hers (1960). ʺOn the conversion of fructose to glucose by guinea pig intestine.ʺ Biochim Biophys

Acta 38: 427‐434.

Giot, L., J. S. Bader, C. Brouwer, A. Chaudhuri, B. Kuang, Y. Li, Y. L. Hao, C. E. Ooi, B. Godwin, E. Vitols, G.

Vijayadamodar, P. Pochart, H. Machineni, M. Welsh, Y. Kong, B. Zerhusen, R. Malcolm, Z. Varrone, A. Collis,

M. Minto, S. Burgess, L. McDaniel, E. Stimpson, F. Spriggs, J. Williams, K. Neurath, N. Ioime, M. Agee, E.

Voss, K. Furtak, R. Renzulli, N. Aanensen, S. Carrolla, E. Bickelhaupt, Y. Lazovatsky, A. DaSilva, J. Zhong, C.

A. Stanyon, R. L. Finley, Jr., K. P. White, M. Braverman, T. Jarvie, S. Gold, M. Leach, J. Knight, R. A. Shimkets,

 146

References

M. P. McKenna, J. Chant and J. M. Rothberg (2003). ʺA protein interaction map of Drosophila melanogaster.ʺ

Science 302(5651): 1727‐1736.

Glaser, C. B., G. Yamin, V. N. Uversky and A. L. Fink (2005). ʺMethionine oxidation, alpha‐synuclein and Parkinsonʹs

disease.ʺ Biochim Biophys Acta 1703(2): 157‐169.

Glenner, G. G. (1980a). ʺAmyloid deposits and amyloidosis. The beta‐fibrilloses (first of two parts).ʺ N Engl J Med 302(23):

1283‐1292.

Glenner, G. G. (1980b). ʺAmyloid deposits and amyloidosis: the beta‐fibrilloses (second of two parts).ʺ N Engl J Med

302(24): 1333‐1343.

Gobom, J., E. Nordhoff, E. Mirgorodskaya, R. Ekman and P. Roepstorff (1999). ʺSample purification and preparation

technique based on nano‐scale reversed‐phase columns for the sensitive analysis of complex peptide mixtures by

matrix‐assisted laser desorption/ionization mass spectrometry.ʺ J Mass Spectrom 34(2): 105‐116.

Goedert, M. and M. G. Spillantini (1998). ʺLewy body diseases and multiple system atrophy as alpha‐synucleinopathies.ʺ

Mol Psychiatry 3(6): 462‐465.

Goers, J., V. N. Uversky and A. L. Fink (2003). ʺPolycation‐induced oligomerization and accelerated fibrillation of human

alpha‐synuclein in vitro.ʺ Protein Sci 12(4): 702‐707.

Goldberg, A. L. (2003). ʺProtein degradation and protection against misfolded or damaged proteins.ʺ Nature 426(6968): 895‐

899.

Goldberg, M. S. and P. T. Lansbury, Jr. (2000). ʺIs there a cause‐and‐effect relationship between alpha‐synuclein

fibrillization and Parkinsonʹs disease?ʺ Nat Cell Biol 2(7): E115‐119.

Goldsteins, G., K. Andersson, A. Olofsson, I. Dacklin, A. Edvinsson, V. Baranov, O. Sandgren, C. Thylen, S.

Hammarstrom and E. Lundgren (1997). ʺCharacterization of two highly amyloidogenic mutants of transthyretin.ʺ

Biochemistry 36(18): 5346‐5352.

Gomes, R., M. Sousa Silva, A. Quintas, C. Cordeiro, A. Freire, P. Pereira, A. Martins, E. Monteiro, E. Barroso and A.

Ponces Freire (2005a). ʺArgpyrimidine, a methylglyoxal‐derived advanced glycation end‐product in familial

amyloidotic polyneuropathy.ʺ Biochem J 385(Pt 2): 339‐345.

Gomes, R. A., H. V. Miranda, M. S. Silva, G. Graca, A. V. Coelho, A. E. Ferreira, C. Cordeiro and A. P. Freire (2006). ʺYeast

protein glycation in vivo by methylglyoxal. Molecular modification of glycolytic enzymes and heat shock

proteins.ʺ FEBS J 273(23): 5273‐5287.

Gomes, R. A., H. V. Miranda, M. Sousa Silva, G. Graca, A. V. Coelho, A. E. do Nascimento Ferreira, C. Cordeiro and A. P.

Freire (2008a). ʺProtein glycation and methylglyoxal metabolism in yeast: finding peptide needles in protein

haystacks.ʺ FEMS Yeast Res 8(1): 174‐181.

Gomes, R. A., L. M. Oliveira, M. Silva, C. Ascenso, A. Quintas, G. Costa, A. V. Coelho, M. Sousa Silva, A. E. Ferreira, A.

Ponces Freire and C. Cordeiro (2008b). ʺProtein glycation in vivo: functional and structural effects on yeast

enolase.ʺ Biochem J 416(3): 317‐326.

 147

References 

Gomes, R. A., M. Sousa Silva, H. Vicente Miranda, A. E. Ferreira, C. A. Cordeiro and A. P. Freire (2005b). ʺProtein

glycation in Saccharomyces cerevisiae. Argpyrimidine formation and methylglyoxal catabolism.ʺ FEBS J 272(17):

4521‐4531.

Gonzalez de Aguilar, J. L., A. Echaniz‐Laguna, A. Fergani, F. Rene, V. Meininger, J. P. Loeffler and L. Dupuis (2007).

ʺAmyotrophic lateral sclerosis: all roads lead to Rome.ʺ J Neurochem 101(5): 1153‐1160.

Gorbatyuk, O. S., S. Li, L. F. Sullivan, W. Chen, G. Kondrikova, F. P. Manfredsson, R. J. Mandel and N. Muzyczka (2008).

ʺThe phosphorylation state of Ser‐129 in human alpha‐synuclein determines neurodegeneration in a rat model of

Parkinson disease.ʺ Proc Natl Acad Sci U S A 105(2): 763‐768.

Govindarajan, S., R. Recabarren and R. A. Goldstein (1999). ʺEstimating the total number of protein folds.ʺ Proteins 35(4):

408‐414.

Greene, R. F., Jr. and C. N. Pace (1974). ʺUrea and guanidine hydrochloride denaturation of ribonuclease, lysozyme, alpha‐

chymotrypsin, and beta‐lactoglobulin.ʺ J Biol Chem 249(17): 5388‐5393.

Grillo, M. A. and S. Colombatto (2008). ʺAdvanced glycation end‐products (AGEs): involvement in aging and in

neurodegenerative diseases.ʺ Amino Acids 35(1): 29‐36.

Gsponer, J. and M. Madan Babu (2009). ʺThe rules of disorder or why disorder rules.ʺ Prog Biophys Mol Biol.

Gurd, J. W. (1997). ʺProtein tyrosine phosphorylation: implications for synaptic function.ʺ Neurochem Int 31(5): 635‐649.

Haass, C., A. Y. Hung, D. J. Selkoe and D. B. Teplow (1994). ʺMutations associated with a locus for familial Alzheimerʹs

disease result in alternative processing of amyloid beta‐protein precursor.ʺ J Biol Chem 269(26): 17741‐17748.

Hammack, B., K. Attfield, D. Clayton, E. Dec, A. Dong, C. Sarisky and B. E. Bowler (1998). ʺThe magnitude of changes in

guanidine‐HCl unfolding m‐values in the protein, iso‐1‐cytochrome c, depends upon the substructure containing

the mutation.ʺ Protein Sci 7(8): 1789‐1795.

Harding, J. J. (1998). ʺCataract, Alzheimerʹs disease, and other conformational diseases.ʺ Curr Opin Ophthalmol 9(1): 10‐13.

Hardy, J. and D. Allsop (1991). ʺAmyloid deposition as the central event in the aetiology of Alzheimerʹs disease.ʺ Trends

Pharmacol Sci 12(10): 383‐388.

Harrington, C. R. and C. A. Colaco (1994). ʺAlzheimerʹs disease. A glycation connection.ʺ Nature 370(6487): 247‐248.

Hartley, D. M., D. M. Walsh, C. P. Ye, T. Diehl, S. Vasquez, P. M. Vassilev, D. B. Teplow and D. J. Selkoe (1999).

ʺProtofibrillar intermediates of amyloid beta‐protein induce acute electrophysiological changes and progressive

neurotoxicity in cortical neurons.ʺ J Neurosci 19(20): 8876‐8884.

Hashimoto, N., H. Naiki and F. Gejyo (1999). ʺModification of beta 2‐microglobulin with D‐glucose or 3‐deoxyglucosone

inhibits A beta 2M amyloid fibril extension in vitro.ʺ Amyloid 6(4): 256‐264.

Hayashi, T., S. Mase and M. Namiki (1986). ʺFormation of 3‐Carbon Sugar Fragment at an Early Stage of the Browning

Reaction of Sugar with Amines or Amino‐Acids.ʺ Agricultural and Biological Chemistry 50: 1959‐1964.

Hecht, M. H., J. S. Richardson, D. C. Richardson and R. C. Ogden (1990). ʺDe novo design, expression, and characterization

of Felix: a four‐helix bundle protein of native‐like sequence.ʺ Science 249(4971): 884‐891.  148

References

Heese, K. and H. Akatsu (2006). ʺAlzheimerʹs disease‐‐an interactive perspective.ʺ Curr Alzheimer Res 3(2): 109‐121.

Hegde, M. L., P. Shanmugavelu, B. Vengamma, T. S. Rao, R. B. Menon, R. V. Rao and K. S. Rao (2004). ʺSerum trace

element levels and the complexity of inter‐element relations in patients with Parkinsonʹs disease.ʺ J Trace Elem

Med Biol 18(2): 163‐171.

Heinz, D. W., W. A. Baase and B. W. Matthews (1992). ʺFolding and function of a T4 lysozyme containing 10 consecutive

alanines illustrate the redundancy of information in an amino acid sequence.ʺ Proc Natl Acad Sci U S A 89(9):

3751‐3755.

Hemmings, H. C., Jr., A. C. Nairn, D. W. Aswad and P. Greengard (1984). ʺDARPP‐32, a dopamine‐ and adenosine 3ʹ:5ʹ‐

monophosphate‐regulated phosphoprotein enriched in dopamine‐innervated brain regions. II. Purification and

characterization of the phosphoprotein from bovine caudate nucleus.ʺ J Neurosci 4(1): 99‐110.

Heneka, M. T. and M. K. OʹBanion (2007). ʺInflammatory processes in Alzheimerʹs disease.ʺ J Neuroimmunol 184(1‐2): 69‐

91.

Henle, T., A. W. Walter, R. Haessner and H. Klostermeyer (1994). ʺDetection and Identification of a Protein‐Bound

Imidazolone Resulting from the Reaction of Arginine Residues and Methylglyoxal.ʺ Zeitschrift Fur Lebensmittel‐

Untersuchung Und‐Forschung 199: 55‐58.

Hernandez, F. and J. Avila (2007). ʺTauopathies.ʺ Cell Mol Life Sci 64(17): 2219‐2233.

Hers, H. G. (1956). ʺ[The mechanism of the transformation of glucose in fructose in the seminal vesicles.].ʺ Biochim Biophys

Acta 22(1): 202‐203.

Ho, M. R., Y. C. Lou, W. C. Lin, P. C. Lyu, W. N. Huang and C. Chen (2006). ʺHuman pancreatitis‐associated protein forms

fibrillar aggregates with a native‐like conformation.ʺ J Biol Chem 281(44): 33566‐33576.

Hodara, R., E. H. Norris, B. I. Giasson, A. J. Mishizen‐Eberz, D. R. Lynch, V. M. Lee and H. Ischiropoulos (2004).

ʺFunctional consequences of alpha‐synuclein tyrosine nitration: diminished binding to lipid vesicles and increased

fibril formation.ʺ J Biol Chem 279(46): 47746‐47753.

Hodge, J. E. (1955). ʺThe Amadori rearrangement.ʺ Adv Carbohydr Chem 10: 169‐205.

Horwich, A. (2002). ʺProtein aggregation in disease: a role for folding intermediates forming specific multimeric

interactions.ʺ J Clin Invest 110(9): 1221‐1232.

Huang, P. and A. Dong (2003). ʺThermal, chemical and chemothermal denaturation of yeast enolase.ʺ Spectroscopy 17: 453‐

467.

Hunt, J. V., M. A. Bottoms and M. J. Mitchinson (1993). ʺOxidative alterations in the experimental glycation model of

diabetes mellitus are due to protein‐glucose adduct oxidation. Some fundamental differences in proposed

mechanisms of glucose oxidation and oxidant production.ʺ Biochem J 291 ( Pt 2): 529‐535.

Hvidt, A. (1955). ʺDeuterium exchange between ribonuclease and water.ʺ Biochim Biophys Acta 18(2): 306‐308.

Iakoucheva, L. M., C. J. Brown, J. D. Lawson, Z. Obradovic and A. K. Dunker (2002). ʺIntrinsic disorder in cell‐signaling

and cancer‐associated proteins.ʺ J Mol Biol 323(3): 573‐584.

 149

References 

Irizarry, M. C., W. Growdon, T. Gomez‐Isla, K. Newell, J. M. George, D. F. Clayton and B. T. Hyman (1998). ʺNigral and

cortical Lewy bodies and dystrophic nigral neurites in Parkinsonʹs disease and cortical Lewy body disease contain

alpha‐synuclein immunoreactivity.ʺ J Neuropathol Exp Neurol 57(4): 334‐337.

Ito, D. and N. Suzuki (2007). ʺMolecular pathogenesis of seipin/BSCL2‐related motor neuron diseases.ʺ Ann Neurol 61(3):

237‐250.

Ito, T., T. Chiba, R. Ozawa, M. Yoshida, M. Hattori and Y. Sakaki (2001). ʺA comprehensive two‐hybrid analysis to explore

the yeast protein interactome.ʺ Proc Natl Acad Sci U S A 98(8): 4569‐4574.

Itzhaki, L. S., D. E. Otzen and A. R. Fersht (1995). ʺThe structure of the transition state for folding of chymotrypsin inhibitor

2 analysed by protein engineering methods: evidence for a nucleation‐condensation mechanism for protein

folding.ʺ J Mol Biol 254(2): 260‐288.

Ivarsson, Y., C. Travaglini‐Allocatelli, M. Brunori and S. Gianni (2008). ʺMechanisms of protein folding.ʺ Eur Biophys J

37(6): 721‐728.

Izaguirre, G., A. Kikonyogo and R. Pietruszko (1998). ʺMethylglyoxal as substrate and inhibitor of human aldehyde

dehydrogenase: comparison of kinetic properties among the three isozymes.ʺ Comp Biochem Physiol B Biochem

Mol Biol 119(4): 747‐754.

Jabs, A., M. S. Weiss and R. Hilgenfeld (1999). ʺNon‐proline cis peptide bonds in proteins.ʺ J Mol Biol 286(1): 291‐304.

Jackson, S. E. and A. R. Fersht (1991). ʺFolding of chymotrypsin inhibitor 2. 1. Evidence for a two‐state transition.ʺ

Biochemistry 30(43): 10428‐10435.

Jaffe, E. K. and L. Stith (2007). ʺALAD porphyria is a conformational disease.ʺ Am J Hum Genet 80(2): 329‐337.

Janson, J., R. H. Ashley, D. Harrison, S. McIntyre and P. C. Butler (1999). ʺThe mechanism of islet amyloid polypeptide

toxicity is membrane disruption by intermediate‐sized toxic amyloid particles.ʺ Diabetes 48(3): 491‐498.

Jarrett, J. T. and P. T. Lansbury, Jr. (1992). ʺAmyloid fibril formation requires a chemically discriminating nucleation event:

studies of an amyloidogenic sequence from the bacterial protein OsmB.ʺ Biochemistry 31(49): 12345‐12352.

Jarrett, J. T. and P. T. Lansbury, Jr. (1993). ʺSeeding ʺone‐dimensional crystallizationʺ of amyloid: a pathogenic mechanism

in Alzheimerʹs disease and scrapie?ʺ Cell 73(6): 1055‐1058.

Jeffery, C. J. (2003). ʺMoonlighting proteins: old proteins learning new tricks.ʺ Trends Genet 19(8): 415‐417.

Jenner, P. (2003). ʺOxidative stress in Parkinsonʹs disease.ʺ Ann Neurol 53 Suppl 3: S26‐36; discussion S36‐28.

Jiang, X., J. N. Buxbaum and J. W. Kelly (2001). ʺThe V122I cardiomyopathy variant of transthyretin increases the velocity of

rate‐limiting tetramer dissociation, resulting in accelerated amyloidosis.ʺ Proc Natl Acad Sci U S A 98(26): 14943‐

14948.

Jimenez, J. L., E. J. Nettleton, M. Bouchard, C. V. Robinson, C. M. Dobson and H. R. Saibil (2002). ʺThe protofilament

structure of insulin amyloid fibrils.ʺ Proc Natl Acad Sci U S A 99(14): 9196‐9201.

 150

References

Johnson, R. J., J. Christodoulou, M. Dumoulin, G. L. Caddy, M. J. Alcocer, G. J. Murtagh, J. R. Kumita, G. Larsson, C. V.

Robinson, D. B. Archer, B. Luisi and C. M. Dobson (2005). ʺRationalising lysozyme amyloidosis: insights from

the structure and solution dynamics of T70N lysozyme.ʺ J Mol Biol 352(4): 823‐836.

Johnson, W. C. (1999). ʺAnalyzing protein circular dichroism spectra for accurate secondary structures.ʺ Proteins 35(3): 307‐

312.

Johnston, J. A., C. L. Ward and R. R. Kopito (1998). ʺAggresomes: a cellular response to misfolded proteins.ʺ J Cell Biol

143(7): 1883‐1898.

Juraszek, J. and P. G. Bolhuis (2006). ʺSampling the multiple folding mechanisms of Trp‐cage in explicit solvent.ʺ Proc Natl

Acad Sci U S A 103(43): 15859‐15864.

Kahle, P. J., C. Haass, H. A. Kretzschmar and M. Neumann (2002). ʺStructure/function of alpha‐synuclein in health and

disease: rational development of animal models for Parkinsonʹs and related diseases.ʺ J Neurochem 82(3): 449‐457.

Kalapos, M. P. (1999). ʺMethylglyoxal in living organisms: chemistry, biochemistry, toxicology and biological implications.ʺ

Toxicol Lett 110(3): 145‐175.

Kamtekar, S., J. M. Schiffer, H. Xiong, J. M. Babik and M. H. Hecht (1993). ʺProtein design by binary patterning of polar

and nonpolar amino acids.ʺ Science 262(5140): 1680‐1685.

Kang, J. H. (2003). ʺModification and inactivation of human Cu,Zn‐superoxide dismutase by methylglyoxal.ʺ Mol Cells 15(2):

194‐199.

Karplus, M. and D. L. Weaver (1976). ʺProtein‐folding dynamics.ʺ Nature 260(5550): 404‐406.

Kato, H., D. B. Shin and F. Hayase (1987). ʺ3‐Deoxyglucosone Cross‐Links Proteins under Physiological Conditions.ʺ

Agricultural and Biological Chemistry 51: 2009‐2011.

Kauzmann, W. (1959). ʺSome factors in the interpretation of protein denaturation.ʺ Adv Protein Chem 14: 1‐63.

Kayed, R., Y. Sokolov, B. Edmonds, T. M. McIntire, S. C. Milton, J. E. Hall and C. G. Glabe (2004). ʺPermeabilization of

lipid bilayers is a common conformation‐dependent activity of soluble amyloid oligomers in protein misfolding

diseases.ʺ J Biol Chem 279(45): 46363‐46366.

Kellum, M. W., B. Oray and S. J. Norton (1978). ʺA convenient quantitative synthesis of methylglyoxal for glyoxalase I

assays.ʺ Anal Biochem 85(2): 586‐590.

Kelly, J. W. (1996). ʺAlternative conformations of amyloidogenic proteins govern their behavior.ʺ Curr Opin Struct Biol 6(1):

11‐17.

Kelly, S. M. and N. C. Price (2000). ʺThe use of circular dichroism in the investigation of protein structure and function.ʺ

Curr Protein Pept Sci 1(4): 349‐384.

Kessler, J. C., J. C. Rochet and P. T. Lansbury, Jr. (2003). ʺThe N‐terminal repeat domain of alpha‐synuclein inhibits beta‐

sheet and amyloid fibril formation.ʺ Biochemistry 42(3): 672‐678.

Kiefhaber, T. (1995). ʺKinetic traps in lysozyme folding.ʺ Proc Natl Acad Sci U S A 92(20): 9029‐9033.

 151

References 

Kilhovd, B. K., I. Giardino, P. A. Torjesen, K. I. Birkeland, T. J. Berg, P. J. Thornalley, M. Brownlee and K. F. Hanssen

(2003). ʺIncreased serum levels of the specific AGE‐compound methylglyoxal‐derived hydroimidazolone in

patients with type 2 diabetes.ʺ Metabolism 52(2): 163‐167.

Kim, D. E., H. Gu and D. Baker (1998). ʺThe sequences of small proteins are not extensively optimized for rapid folding by

natural selection.ʺ Proc Natl Acad Sci U S A 95(9): 4982‐4986.

Kim, K. S., J. A. Fuchs and C. K. Woodward (1993). ʺHydrogen exchange identifies native‐state motional domains important

in protein folding.ʺ Biochemistry 32(37): 9600‐9608.

Kim, K. S. and C. Woodward (1993). ʺProtein internal flexibility and global stability: effect of urea on hydrogen exchange

rates of bovine pancreatic trypsin inhibitor.ʺ Biochemistry 32(37): 9609‐9613.

Kim, P. S. and R. L. Baldwin (1982). ʺSpecific intermediates in the folding reactions of small proteins and the mechanism of

protein folding.ʺ Annu Rev Biochem 51: 459‐489.

Kim, P. S. and R. L. Baldwin (1990). ʺIntermediates in the folding reactions of small proteins.ʺ Annu Rev Biochem 59: 631‐

660.

Kingsbury, J. S., R. Theberge, J. A. Karbassi, A. Lim, C. E. Costello and L. H. Connors (2007). ʺDetailed structural analysis

of amyloidogenic wild‐type transthyretin using a novel purification strategy and mass spectrometry.ʺ Anal Chem

79(5): 1990‐1998.

Klein, C. and K. Lohmann‐Hedrich (2007). ʺImpact of recent genetic findings in Parkinsonʹs disease.ʺ Curr Opin Neurol

20(4): 453‐464.

Kluck, R. M., E. Bossy‐Wetzel, D. R. Green and D. D. Newmeyer (1997). ʺThe release of cytochrome c from mitochondria: a

primary site for Bcl‐2 regulation of apoptosis.ʺ Science 275(5303): 1132‐1136.

Koenig, R. J., S. H. Blobstein and A. Cerami (1977). ʺStructure of carbohydrate of hemoglobin AIc.ʺ J Biol Chem 252(9):

2992‐2997.

Koop, D. R. and J. P. Casazza (1985). ʺIdentification of ethanol‐inducible P‐450 isozyme 3a as the acetone and acetol

monooxygenase of rabbit microsomes.ʺ J Biol Chem 260(25): 13607‐13612.

Kopito, R. R. (2000). ʺAggresomes, inclusion bodies and protein aggregation.ʺ Trends Cell Biol 10(12): 524‐530.

Kopito, R. R. and D. Ron (2000). ʺConformational disease.ʺ Nat Cell Biol 2(11): E207‐209.

Kourie, J. I. and C. L. Henry (2002). ʺIon channel formation and membrane‐linked pathologies of misfolded hydrophobic

proteins: the role of dangerous unchaperoned molecules.ʺ Clin Exp Pharmacol Physiol 29(9): 741‐753.

Kriwacki, R. W., L. Hengst, L. Tennant, S. I. Reed and P. E. Wright (1996). ʺStructural studies of p21Waf1/Cip1/Sdi1 in the

free and Cdk2‐bound state: conformational disorder mediates binding diversity.ʺ Proc Natl Acad Sci U S A 93(21):

11504‐11509.

Krogan, N. J., G. Cagney, H. Yu, G. Zhong, X. Guo, A. Ignatchenko, J. Li, S. Pu, N. Datta, A. P. Tikuisis, T. Punna, J. M.

Peregrin‐Alvarez, M. Shales, X. Zhang, M. Davey, M. D. Robinson, A. Paccanaro, J. E. Bray, A. Sheung, B.

Beattie, D. P. Richards, V. Canadien, A. Lalev, F. Mena, P. Wong, A. Starostine, M. M. Canete, J. Vlasblom, S.

 152

References

Wu, C. Orsi, S. R. Collins, S. Chandran, R. Haw, J. J. Rilstone, K. Gandi, N. J. Thompson, G. Musso, P. St Onge,

S. Ghanny, M. H. Lam, G. Butland, A. M. Altaf‐Ul, S. Kanaya, A. Shilatifard, E. OʹShea, J. S. Weissman, C. J.

Ingles, T. R. Hughes, J. Parkinson, M. Gerstein, S. J. Wodak, A. Emili and J. F. Greenblatt (2006). ʺGlobal

landscape of protein complexes in the yeast Saccharomyces cerevisiae.ʺ Nature 440(7084): 637‐643.

Kruger, R., W. Kuhn, T. Muller, D. Woitalla, M. Graeber, S. Kosel, H. Przuntek, J. T. Epplen, L. Schols and O. Riess

(1998). ʺAla30Pro mutation in the gene encoding alpha‐synuclein in Parkinsonʹs disease.ʺ Nat Genet 18(2): 106‐108.

Kuhla, B., K. Boeck, A. Schmidt, V. Ogunlade, T. Arendt, G. Munch and H. J. Luth (2007). ʺAge‐ and stage‐dependent

glyoxalase I expression and its activity in normal and Alzheimerʹs disease brains.ʺ Neurobiol Aging 28(1): 29‐41.

Kumar, M. S., P. Y. Reddy, P. A. Kumar, I. Surolia and G. B. Reddy (2004). ʺEffect of dicarbonyl‐induced browning on

alpha‐crystallin chaperone‐like activity: physiological significance and caveats of in vitro aggregation assays.ʺ

Biochem J 379(Pt 2): 273‐282.

Kume, S., M. Takeya, T. Mori, N. Araki, H. Suzuki, S. Horiuchi, T. Kodama, Y. Miyauchi and K. Takahashi (1995).

ʺImmunohistochemical and ultrastructural detection of advanced glycation end products in atherosclerotic lesions

of human aorta with a novel specific monoclonal antibody.ʺ Am J Pathol 147(3): 654‐667.

Kustrzeba‐Wojcicka, I. and M. Golczak (2000). ʺEnolase from Candida albicans‐purification and characterization.ʺ Comp

Biochem Physiol B Biochem Mol Biol 126(1): 109‐120.

Lai, Z., W. Colon and J. W. Kelly (1996). ʺThe acid‐mediated denaturation pathway of transthyretin yields a conformational

intermediate that can self‐assemble into amyloid.ʺ Biochemistry 35(20): 6470‐6482.

Lam, A. R., J. M. Borreguero, F. Ding, N. V. Dokholyan, S. V. Buldyrev, H. E. Stanley and E. Shakhnovich (2007). ʺParallel

folding pathways in the SH3 domain protein.ʺ J Mol Biol 373(5): 1348‐1360.

Lansbury, P. T., Jr. (1999). ʺEvolution of amyloid: what normal protein folding may tell us about fibrillogenesis and disease.ʺ

Proc Natl Acad Sci U S A 96(7): 3342‐3344.

Lashuel, H. A. (2005). ʺMembrane permeabilization: a common mechanism in protein‐misfolding diseases.ʺ Sci Aging

Knowledge Environ 2005(38): pe28.

Lashuel, H. A., D. Hartley, B. M. Petre, T. Walz and P. T. Lansbury, Jr. (2002). ʺNeurodegenerative disease: amyloid pores

from pathogenic mutations.ʺ Nature 418(6895): 291.

Lauren, J., D. A. Gimbel, H. B. Nygaard, J. W. Gilbert and S. M. Strittmatter (2009). ʺCellular prion protein mediates

impairment of synaptic plasticity by amyloid‐beta oligomers.ʺ Nature 457(7233): 1128‐1132.

Lavedan, C. (1998). ʺThe synuclein family.ʺ Genome Res 8(9): 871‐880.

Lazaridis, T. and M. Karplus (1997). ʺʺNew viewʺ of protein folding reconciled with the old through multiple unfolding

simulations.ʺ Science 278(5345): 1928‐1931.

Lederer, M. O. and R. G. Klaiber (1999). ʺCross‐linking of proteins by Maillard processes: characterization and detection of

lysine‐arginine cross‐links derived from glyoxal and methylglyoxal.ʺ Bioorg Med Chem 7(11): 2499‐2507.

 153

References 

Ledesma, M. D., P. Bonay, C. Colaco and J. Avila (1994). ʺAnalysis of microtubule‐associated protein tau glycation in paired

helical filaments.ʺ J Biol Chem 269(34): 21614‐21619.

Lee, D., C. W. Park, S. R. Paik and K. Y. Choi (2009). ʺThe modification of alpha‐synuclein by dicarbonyl compounds

inhibits its fibril‐forming process.ʺ Biochim Biophys Acta 1794(3): 421‐430.

Lee, H. J., C. Choi and S. J. Lee (2002). ʺMembrane‐bound alpha‐synuclein has a high aggregation propensity and the ability

to seed the aggregation of the cytosolic form.ʺ J Biol Chem 277(1): 671‐678.

Lee, V. M. and J. Q. Trojanowski (2006). ʺMechanisms of Parkinsonʹs disease linked to pathological alpha‐synuclein: new

targets for drug discovery.ʺ Neuron 52(1): 33‐38.

Leopold, P. E., M. Montal and J. N. Onuchic (1992). ʺProtein folding funnels: a kinetic approach to the sequence‐structure

relationship.ʺ Proc Natl Acad Sci U S A 89(18): 8721‐8725.

Leroy, E., R. Boyer, G. Auburger, B. Leube, G. Ulm, E. Mezey, G. Harta, M. J. Brownstein, S. Jonnalagada, T. Chernova, A.

Dehejia, C. Lavedan, T. Gasser, P. J. Steinbach, K. D. Wilkinson and M. H. Polymeropoulos (1998). ʺThe

ubiquitin pathway in Parkinsonʹs disease.ʺ Nature 395(6701): 451‐452.

Levine, B. and J. Yuan (2005). ʺAutophagy in cell death: an innocent convict?ʺ J Clin Invest 115(10): 2679‐2688.

Levinthal, C. (1968). ʺAre there pathways for protein folding?ʺ J Chem Phys 65: 44‐45.

Levitt, M. and C. Chothia (1976). ʺStructural patterns in globular proteins.ʺ Nature 261(5561): 552‐558.

Levy, Y., P. G. Wolynes and J. N. Onuchic (2004). ʺProtein topology determines binding mechanism.ʺ Proc Natl Acad Sci U

S A 101(2): 511‐516.

Li, P., D. Nijhawan, I. Budihardjo, S. M. Srinivasula, M. Ahmad, E. S. Alnemri and X. Wang (1997). ʺCytochrome c and

dATP‐dependent formation of Apaf‐1/caspase‐9 complex initiates an apoptotic protease cascade.ʺ Cell 91(4): 479‐

489.

Li, S., C. M. Armstrong, N. Bertin, H. Ge, S. Milstein, M. Boxem, P. O. Vidalain, J. D. Han, A. Chesneau, T. Hao, D. S.

Goldberg, N. Li, M. Martinez, J. F. Rual, P. Lamesch, L. Xu, M. Tewari, S. L. Wong, L. V. Zhang, G. F. Berriz, L.

Jacotot, P. Vaglio, J. Reboul, T. Hirozane‐Kishikawa, Q. Li, H. W. Gabel, A. Elewa, B. Baumgartner, D. J. Rose,

H. Yu, S. Bosak, R. Sequerra, A. Fraser, S. E. Mango, W. M. Saxton, S. Strome, S. Van Den Heuvel, F. Piano, J.

Vandenhaute, C. Sardet, M. Gerstein, L. Doucette‐Stamm, K. C. Gunsalus, J. W. Harper, M. E. Cusick, F. P.

Roth, D. E. Hill and M. Vidal (2004). ʺA map of the interactome network of the metazoan C. elegans.ʺ Science

303(5657): 540‐543.

Li, X., Z. Obradovic, C. J. Brown, E. C. Garner and A. K. Dunker (2000). ʺComparing predictors of disordered protein.ʺ

Genome Inform Ser Workshop Genome Inform 11: 172‐184.

Li, X., P. Romero, M. Rani, A. K. Dunker and Z. Obradovic (1999). ʺPredicting Protein Disorder for N‐, C‐, and Internal

Regions.ʺ Genome Inform Ser Workshop Genome Inform 10: 30‐40.

Librizzi, F., V. Fodera, V. Vetri, C. Lo Presti and M. Leone (2007). ʺEffects of confinement on insulin amyloid fibrils

formation.ʺ Eur Biophys J 36(7): 711‐715.

 154

References

Lindberg, M. O. and M. Oliveberg (2007). ʺMalleability of protein folding pathways: a simple reason for complex

behaviour.ʺ Curr Opin Struct Biol 17(1): 21‐29.

Lindholm, D., H. Wootz and L. Korhonen (2006). ʺER stress and neurodegenerative diseases.ʺ Cell Death Differ 13(3): 385‐

392.

Liu, C. W., M. J. Corboy, G. N. DeMartino and P. J. Thomas (2003). ʺEndoproteolytic activity of the proteasome.ʺ Science

299(5605): 408‐411.

Liu, L. L. and K. J. Franz (2005). ʺPhosphorylation of an alpha‐synuclein peptide fragment enhances metal binding.ʺ J Am

Chem Soc 127(27): 9662‐9663.

Liu, L. L. and K. J. Franz (2007). ʺPhosphorylation‐dependent metal binding by alpha‐synuclein peptide fragments.ʺ J Biol

Inorg Chem 12(2): 234‐247.

Liu, X., C. N. Kim, J. Yang, R. Jemmerson and X. Wang (1996). ʺInduction of apoptotic program in cell‐free extracts:

requirement for dATP and cytochrome c.ʺ Cell 86(1): 147‐157.

Llinas, R., H. Moreno, M. Sugimori, M. Mohammadi and J. Schlessinger (1997). ʺDifferential pre‐ and postsynaptic

modulation of chemical transmission in the squid giant synapse by tyrosine phosphorylation.ʺ Proc Natl Acad Sci

U S A 94(5): 1990‐1994.

Lo, T. W., M. E. Westwood, A. C. McLellan, T. Selwood and P. J. Thornalley (1994). ʺBinding and modification of proteins

by methylglyoxal under physiological conditions. A kinetic and mechanistic study with N alpha‐acetylarginine, N

alpha‐acetylcysteine, and N alpha‐acetyllysine, and bovine serum albumin.ʺ J Biol Chem 269(51): 32299‐32305.

Lobley, A., L. Whitmore and B. A. Wallace (2002). ʺDICHROWEB: an interactive website for the analysis of protein

secondary structure from circular dichroism spectra.ʺ Bioinformatics 18(1): 211‐212.

Lohmann, K. (1932). ʺBeitrag zur enzymatischen Umwandling von synthetischem Methylglyoxal in Milchsäure.ʺ

Biochemische Zeitschrift 254: 332‐354.

Lomakin, A., D. B. Teplow, D. A. Kirschner and G. B. Benedek (1997). ʺKinetic theory of fibrillogenesis of amyloid beta‐

protein.ʺ Proc Natl Acad Sci U S A 94(15): 7942‐7947.

Loske, C., A. Gerdemann, W. Schepl, M. Wycislo, R. Schinzel, D. Palm, P. Riederer and G. Munch (2000). ʺTransition

metal‐mediated glycoxidation accelerates cross‐linking of beta‐amyloid peptide.ʺ Eur J Biochem 267(13): 4171‐

4178.

Lovell, S. C., J. M. Word, J. S. Richardson and D. C. Richardson (2000). ʺThe penultimate rotamer library.ʺ Proteins 40(3):

389‐408.

Lue, L. F., Y. M. Kuo, A. E. Roher, L. Brachova, Y. Shen, L. Sue, T. Beach, J. H. Kurth, R. E. Rydel and J. Rogers (1999).

ʺSoluble amyloid beta peptide concentration as a predictor of synaptic change in Alzheimerʹs disease.ʺ Am J Pathol

155(3): 853‐862.

Luthra, M. and D. Balasubramanian (1993). ʺNonenzymatic glycation alters protein structure and stability. A study of two

eye lens crystallins.ʺ J Biol Chem 268(24): 18119‐18127.

 155

References 

Lyles, G. A. and J. Chalmers (1992). ʺThe metabolism of aminoacetone to methylglyoxal by semicarbazide‐sensitive amine

oxidase in human umbilical artery.ʺ Biochem Pharmacol 43(7): 1409‐1414.

Lynch, W. P., V. M. Riseman and A. Bretscher (1987). ʺSmooth muscle caldesmon is an extended flexible monomeric protein

in solution that can readily undergo reversible intra‐ and intermolecular sulfhydryl cross‐linking. A mechanism

for caldesmonʹs F‐actin bundling activity.ʺ J Biol Chem 262(15): 7429‐7437.

Lyubchenko, Y. L., S. Sherman, L. S. Shlyakhtenko and V. N. Uversky (2006). ʺNanoimaging for protein misfolding and

related diseases.ʺ J Cell Biochem 99(1): 52‐70.

Ma, B. and R. Nussinov (2002). ʺStabilities and conformations of Alzheimerʹs beta ‐amyloid peptide oligomers (Abeta 16‐22,

Abeta 16‐35, and Abeta 10‐35): Sequence effects.ʺ Proc Natl Acad Sci U S A 99(22): 14126‐14131.

MacArthur, M. W. and J. M. Thornton (1991). ʺInfluence of proline residues on protein conformation.ʺ J Mol Biol 218(2): 397‐

412.

MacArthur, M. W. and J. M. Thornton (1996). ʺDeviations from planarity of the peptide bond in peptides and proteins.ʺ J

Mol Biol 264(5): 1180‐1195.

Maeta, K., K. Mori, Y. Takatsume, S. Izawa and Y. Inoue (2005). ʺDiagnosis of cell death induced by methylglyoxal, a

metabolite derived from glycolysis, in Saccharomyces cerevisiae.ʺ FEMS Microbiol Lett 243(1): 87‐92.

Maigret, B., D. Perahia and B. Pullman (1970). ʺMolecular orbital calculations on the conformation of polypeptides and

proteins. IV. The conformation of the prolyl and hydroxyprolyl residues.ʺ J Theor Biol 29(2): 275‐291.

Maillard, L. (1912). ʺAction des acides amines sur les sucre: formation des melanoidine par voie methodique.ʺ C R Hebd

Seances Acad. Sci 154: 66‐68.

Manuelidis, L., W. Fritch and Y. G. Xi (1997). ʺEvolution of a strain of CJD that induces BSE‐like plaques.ʺ Science 277(5322):

94‐98.

Marchal, S., E. Shehi, M. C. Harricane, P. Fusi, F. Heitz, P. Tortora and R. Lange (2003). ʺStructural instability and fibrillar

aggregation of non‐expanded human ataxin‐3 revealed under high pressure and temperature.ʺ J Biol Chem

278(34): 31554‐31563.

Marcon, G., G. Plakoutsi and F. Chiti (2006). ʺProtein aggregation starting from the native globular state.ʺ Methods

Enzymol 413: 75‐91.

Maroteaux, L., J. T. Campanelli and R. H. Scheller (1988). ʺSynuclein: a neuron‐specific protein localized to the nucleus and

presynaptic nerve terminal.ʺ J Neurosci 8(8): 2804‐2815.

Mathes, E., E. L. OʹDea, A. Hoffmann and G. Ghosh (2008). ʺNF‐kappaB dictates the degradation pathway of

IkappaBalpha.ʺ EMBO J 27(9): 1357‐1367.

Matthews, B. W. (1993). ʺStructural and genetic analysis of protein stability.ʺ Annu Rev Biochem 62: 139‐160.

Mattson, M. P. and W. A. Pedersen (1998). ʺEffects of amyloid precursor protein derivatives and oxidative stress on basal

forebrain cholinergic systems in Alzheimerʹs disease.ʺ Int J Dev Neurosci 16(7‐8): 737‐753.

 156

References

Mayer, M. P. and B. Bukau (2005). ʺHsp70 chaperones: cellular functions and molecular mechanism.ʺ Cell Mol Life Sci 62(6):

670‐684.

Mayo, S. L. and R. L. Baldwin (1993). ʺGuanidinium chloride induction of partial unfolding in amide proton exchange in

RNase A.ʺ Science 262(5135): 873‐876.

McCutchen, S. L., Z. Lai, G. J. Miroy, J. W. Kelly and W. Colon (1995). ʺComparison of lethal and nonlethal transthyretin

variants and their relationship to amyloid disease.ʺ Biochemistry 34(41): 13527‐13536.

McDonald, I. K. and J. M. Thornton (1994). ʺSatisfying hydrogen bonding potential in proteins.ʺ J Mol Biol 238(5): 777‐793.

McKinney, M. and M. C. Jacksonville (2005). ʺBrain cholinergic vulnerability: relevance to behavior and disease.ʺ Biochem

Pharmacol 70(8): 1115‐1124.

McLean, C. A., R. A. Cherny, F. W. Fraser, S. J. Fuller, M. J. Smith, K. Beyreuther, A. I. Bush and C. L. Masters (1999).

ʺSoluble pool of Abeta amyloid as a determinant of severity of neurodegeneration in Alzheimerʹs disease.ʺ Ann

Neurol 46(6): 860‐866.

McLean, P. J., H. Kawamata and B. T. Hyman (2001). ʺAlpha‐synuclein‐enhanced green fluorescent protein fusion proteins

form proteasome sensitive inclusions in primary neurons.ʺ Neuroscience 104(3): 901‐912.

McLellan, A. C., S. A. Phillips and P. J. Thornalley (1992). ʺThe assay of methylglyoxal in biological systems by

derivatization with 1,2‐diamino‐4,5‐dimethoxybenzene.ʺ Anal Biochem 206(1): 17‐23.

McNulty, B. C., G. B. Young and G. J. Pielak (2006). ʺMacromolecular crowding in the Escherichia coli periplasm maintains

alpha‐synuclein disorder.ʺ J Mol Biol 355(5): 893‐897.

McPherson, J. D., B. H. Shilton and D. J. Walton (1988). ʺRole of fructose in glycation and cross‐linking of proteins.ʺ

Biochemistry 27(6): 1901‐1907.

Merlini, G. and V. Bellotti (2003). ʺMolecular mechanisms of amyloidosis.ʺ N Engl J Med 349(6): 583‐596.

Merlini, G., V. Bellotti, A. Andreola, G. Palladini, L. Obici, S. Casarini and V. Perfetti (2001). ʺProtein aggregation.ʺ Clin

Chem Lab Med 39(11): 1065‐1075.

Mezei, M. (1998). ʺChameleon sequences in the PDB.ʺ Protein Eng 11(6): 411‐414.

Miller, D. W., S. M. Hague, J. Clarimon, M. Baptista, K. Gwinn‐Hardy, M. R. Cookson and A. B. Singleton (2004). ʺAlpha‐

synuclein in blood and brain from familial Parkinson disease with SNCA locus triplication.ʺ Neurology 62(10):

1835‐1838.

Minor, D. L., Jr. and P. S. Kim (1996). ʺContext‐dependent secondary structure formation of a designed protein sequence.ʺ

Nature 380(6576): 730‐734.

Mirsky, A. E. and L. Pauling (1936). ʺOn the Structure of Native, Denatured, and Coagulated Proteins.ʺ Proc Natl Acad Sci U

S A 22(7): 439‐447.

Mirzabekov, T. A., M. C. Lin and B. L. Kagan (1996). ʺPore formation by the cytotoxic islet amyloid peptide amylin.ʺ J Biol

Chem 271(4): 1988‐1992.

 157

References 

Mitchell, M. D., R. E. Laird, R. D. Brown and C. S. Long (2007). ʺIL‐1beta stimulates rat cardiac fibroblast migration via

MAP kinase pathways.ʺ Am J Physiol Heart Circ Physiol 292(2): H1139‐1147.

Miyata, T., R. Inagi, Y. Iida, M. Sato, N. Yamada, O. Oda, K. Maeda and H. Seo (1994). ʺInvolvement of beta 2‐

microglobulin modified with advanced glycation end products in the pathogenesis of hemodialysis‐associated

amyloidosis. Induction of human monocyte chemotaxis and macrophage secretion of tumor necrosis factor‐alpha

and interleukin‐1.ʺ J Clin Invest 93(2): 521‐528.

Miyata, T., Y. Ueda, A. Saito and K. Kurokawa (2000). ʺʹCarbonyl stressʹ and dialysis‐related amyloidosis.ʺ Nephrol Dial

Transplant 15 Suppl 1: 25‐28.

Mogridge, J., P. Legault, J. Li, M. D. Van Oene, L. E. Kay and J. Greenblatt (1998). ʺIndependent ligand‐induced folding of

the RNA‐binding domain and two functionally distinct antitermination regions in the phage lambda N protein.ʺ

Mol Cell 1(2): 265‐275.

Monder, C. (1967). ʺAlpha‐keto aldehyde dehydrogenase, an enzyme that catalyzes the enzymic oxidation of methylglyoxal

to pyruvate.ʺ J Biol Chem 242(20): 4603‐4609.

Monnier, V. M. and A. Cerami (1981). ʺNonenzymatic browning in vivo: possible process for aging of long‐lived proteins.ʺ

Science 211(4481): 491‐493.

Morishima, Y., Y. Gotoh, J. Zieg, T. Barrett, H. Takano, R. Flavell, R. J. Davis, Y. Shirasaki and M. E. Greenberg (2001).

ʺBeta‐amyloid induces neuronal apoptosis via a mechanism that involves the c‐Jun N‐terminal kinase pathway

and the induction of Fas ligand.ʺ J Neurosci 21(19): 7551‐7560.

Morrison, B. M., P. R. Hof and J. H. Morrison (1998). ʺDeterminants of neuronal vulnerability in neurodegenerative

diseases.ʺ Ann Neurol 44(3 Suppl 1): S32‐44.

Muchowski, P. J. (2002). ʺProtein misfolding, amyloid formation, and neurodegeneration: a critical role for molecular

chaperones?ʺ Neuron 35(1): 9‐12.

Mulder, G. J. (1839). ʺOn the composition of some animal substances.ʺ Journal für praktische Chemie 16(129).

Munch, G., H. J. Luth, A. Wong, T. Arendt, E. Hirsch, R. Ravid and P. Riederer (2000). ʺCrosslinking of alpha‐synuclein by

advanced glycation endproducts‐‐an early pathophysiological step in Lewy body formation?ʺ J Chem Neuroanat

20(3‐4): 253‐257.

Myers, J. K., C. N. Pace and J. M. Scholtz (1995). ʺDenaturant m values and heat capacity changes: relation to changes in

accessible surface areas of protein unfolding.ʺ Protein Sci 4(10): 2138‐2148.

Myers, S. L., S. Jones, T. R. Jahn, I. J. Morten, G. A. Tennent, E. W. Hewitt and S. E. Radford (2006). ʺA systematic study of

the effect of physiological factors on beta2‐microglobulin amyloid formation at neutral pH.ʺ Biochemistry 45(7):

2311‐2321.

Myszka, D. G., R. W. Sweet, P. Hensley, M. Brigham‐Burke, P. D. Kwong, W. A. Hendrickson, R. Wyatt, J. Sodroski and

M. L. Doyle (2000). ʺEnergetics of the HIV gp120‐CD4 binding reaction.ʺ Proc Natl Acad Sci U S A 97(16): 9026‐

9031.

N. Berova, K. N., R. W. Woods (2000). Circular Dichroism: Principles and Applications. New York, Wiley‐VCH.  158

References

Nagaraj, R. H., T. Oya‐Ito, P. S. Padayatti, R. Kumar, S. Mehta, K. West, B. Levison, J. Sun, J. W. Crabb and A. K. Padival

(2003). ʺEnhancement of chaperone function of alpha‐crystallin by methylglyoxal modification.ʺ Biochemistry

42(36): 10746‐10755.

Nagaraj, R. H., I. N. Shipanova and F. M. Faust (1996). ʺProtein cross‐linking by the Maillard reaction. Isolation,

characterization, and in vivo detection of a lysine‐lysine cross‐link derived from methylglyoxal.ʺ J Biol Chem

271(32): 19338‐19345.

Naiki, H., K. Higuchi, M. Hosokawa and T. Takeda (1989). ʺFluorometric determination of amyloid fibrils in vitro using the

fluorescent dye, thioflavin T1.ʺ Anal Biochem 177(2): 244‐249.

Naiki, H., K. Higuchi, K. Matsushima, A. Shimada, W. H. Chen, M. Hosokawa and T. Takeda (1990). ʺFluorometric

examination of tissue amyloid fibrils in murine senile amyloidosis: use of the fluorescent indicator, thioflavine T.ʺ

Lab Invest 62(6): 768‐773.

Nakamura, T., H. Yamashita, T. Takahashi and S. Nakamura (2001). ʺActivated Fyn phosphorylates alpha‐synuclein at

tyrosine residue 125.ʺ Biochem Biophys Res Commun 280(4): 1085‐1092.

Nettleton, E. J., P. Tito, M. Sunde, M. Bouchard, C. M. Dobson and C. V. Robinson (2000). ʺCharacterization of the

oligomeric states of insulin in self‐assembly and amyloid fibril formation by mass spectrometry.ʺ Biophys J 79(2):

1053‐1065.

Neuberg, C. (1913). ʺUber die Zerstorung von Milchsaueraldehyd und Methylglyoxal dursch tierische Organe.ʺ Biochem. Z.

49: 573‐579.

Neuberg, C. and M. Kobel (1928). ʺÜber die Frage nach der Identität von Mutase und Ketonaldehydemutase.ʺ Biochem. Z.

203: 631‐646.

Nonaka, T., T. Iwatsubo and M. Hasegawa (2005). ʺUbiquitination of alpha‐synuclein.ʺ Biochemistry 44(1): 361‐368.

Noronha, M., J. C. Lima, P. Lamosa, H. Santos, C. Maycock, R. Ventura and A. L. Maçanita (2004). ʺIntramolecular

Fluorescence Quenching of Tyrosine by the Peptide α‐Carbonyl Group Revisited.ʺ J Phys Chem 108(12): 2155‐2166.

Norris, E. H., B. I. Giasson, H. Ischiropoulos and V. M. Lee (2003). ʺEffects of oxidative and nitrative challenges on alpha‐

synuclein fibrillogenesis involve distinct mechanisms of protein modifications.ʺ J Biol Chem 278(29): 27230‐27240.

Norris, E. H., B. I. Giasson and V. M. Lee (2004). ʺAlpha‐synuclein: normal function and role in neurodegenerative

diseases.ʺ Curr Top Dev Biol 60: 17‐54.

Nunomura, A., S. Chiba, K. Kosaka, A. Takeda, R. J. Castellani, M. A. Smith and G. Perry (2002). ʺNeuronal RNA

oxidation is a prominent feature of dementia with Lewy bodies.ʺ Neuroreport 13(16): 2035‐2039.

Nystrom, F. H. and M. J. Quon (1999). ʺInsulin signalling: metabolic pathways and mechanisms for specificity.ʺ Cell Signal

11(8): 563‐574.

OʹBrien, M. (1997). Introduction to the Maillard reaction. The Glycation Hypothesis of Atherosclerosis. C. Colaco: 29‐56.

OʹDell, T. J., E. R. Kandel and S. G. Grant (1991). ʺLong‐term potentiation in the hippocampus is blocked by tyrosine kinase

inhibitors.ʺ Nature 353(6344): 558‐560.

 159

References 

Ohgushi, M. and A. Wada (1983). ʺʹMolten‐globule stateʹ: a compact form of globular proteins with mobile side‐chains.ʺ

FEBS Lett 164(1): 21‐24.

Okado, A., Y. Kawasaki, Y. Hasuike, M. Takahashi, T. Teshima, J. Fujii and N. Taniguchi (1996). ʺInduction of apoptotic

cell death by methylglyoxal and 3‐deoxyglucosone in macrophage‐derived cell lines.ʺ Biochem Biophys Res

Commun 225(1): 219‐224.

Okochi, M., J. Walter, A. Koyama, S. Nakajo, M. Baba, T. Iwatsubo, L. Meijer, P. J. Kahle and C. Haass (2000).

ʺConstitutive phosphorylation of the Parkinsonʹs disease associated alpha‐synuclein.ʺ J Biol Chem 275(1): 390‐397.

Olanow, C. W. and W. G. Tatton (1999). ʺEtiology and pathogenesis of Parkinsonʹs disease.ʺ Annu Rev Neurosci 22: 123‐144.

Olesen, L. E., M. G. Ford, E. M. Schmid, Y. Vallis, M. M. Babu, P. H. Li, I. G. Mills, H. T. McMahon and G. J. Praefcke

(2008). ʺSolitary and repetitive binding motifs for the AP2 complex alpha‐appendage in amphiphysin and other

accessory proteins.ʺ J Biol Chem 283(8): 5099‐5109.

Olsen, J. V. and M. Mann (2004). ʺImproved peptide identification in proteomics by two consecutive stages of mass

spectrometric fragmentation.ʺ Proc Natl Acad Sci U S A 101(37): 13417‐13422.

Ono, K. and M. Yamada (2006). ʺAntioxidant compounds have potent anti‐fibrillogenic and fibril‐destabilizing effects for

alpha‐synuclein fibrils in vitro.ʺ J Neurochem 97(1): 105‐115.

Onuchic, J. N., N. D. Socci, Z. Luthey‐Schulten and P. G. Wolynes (1996). ʺProtein folding funnels: the nature of the

transition state ensemble.ʺ Fold Des 1(6): 441‐450.

Ottensmeyer, F. P., D. R. Beniac, R. Z. Luo and C. C. Yip (2000). ʺMechanism of transmembrane signaling: insulin binding

and the insulin receptor.ʺ Biochemistry 39(40): 12103‐12112.

Otzen, D. E. and A. R. Fersht (1998). ʺFolding of circular and permuted chymotrypsin inhibitor 2: retention of the folding

nucleus.ʺ Biochemistry 37(22): 8139‐8146.

Otzen, D. E., L. S. Itzhaki, N. F. elMasry, S. E. Jackson and A. R. Fersht (1994). ʺStructure of the transition state for the

folding/unfolding of the barley chymotrypsin inhibitor 2 and its implications for mechanisms of protein folding.ʺ

Proc Natl Acad Sci U S A 91(22): 10422‐10425.

Outeiro, T. F. and F. Giorgini (2006). ʺYeast as a drug discovery platform in Huntingtonʹs and Parkinsonʹs diseases.ʺ

Biotechnol J 1(3): 258‐269.

Outeiro, T. F. and S. Lindquist (2003). ʺYeast cells provide insight into alpha‐synuclein biology and pathobiology.ʺ Science

302(5651): 1772‐1775.

Owen, D. J., B. M. Collins and P. R. Evans (2004). ʺAdaptors for clathrin coats: structure and function.ʺ Annu Rev Cell Dev

Biol 20: 153‐191.

Oya‐Ito, T., B. F. Liu and R. H. Nagaraj (2006). ʺEffect of methylglyoxal modification and phosphorylation on the chaperone

and anti‐apoptotic properties of heat shock protein 27.ʺ J Cell Biochem 99(1): 279‐291.

 160

References

Oya, T., N. Hattori, Y. Mizuno, S. Miyata, S. Maeda, T. Osawa and K. Uchida (1999). ʺMethylglyoxal modification of

protein. Chemical and immunochemical characterization of methylglyoxal‐arginine adducts.ʺ J Biol Chem 274(26):

18492‐18502.

Pace, C. N. (1986). ʺDetermination and analysis of urea and guanidine hydrochloride denaturation curves.ʺ Methods

Enzymol 131: 266‐280.

Pace, C. N., B. A. Shirley and J. A. Thomson (1990). Measuring the conformational stability of a protein. In protein structure:

a pratical approach. T. E. E. Creighton. Oxford UK, Oxford University Press: 311‐330.

Pal, D. and P. Chakrabarti (1999). ʺCis peptide bonds in proteins: residues involved, their conformations, interactions and

locations.ʺ J Mol Biol 294(1): 271‐288.

Pandey, A., J. S. Andersen and M. Mann (2000). ʺUse of mass spectrometry to study signaling pathways.ʺ Sci STKE

2000(37): PL1.

Pang, D. T., J. K. Wang, F. Valtorta, F. Benfenati and P. Greengard (1988). ʺProtein tyrosine phosphorylation in synaptic

vesicles.ʺ Proc Natl Acad Sci U S A 85(3): 762‐766.

Pankratz, N. and T. Foroud (2004). ʺGenetics of Parkinson disease.ʺ NeuroRx 1(2): 235‐242.

Pauling, L., R. B. Corey and H. R. Branson (1951). ʺThe Structure of Proteins: Two Hydrogen‐bonded Helical Configurations

of the Polypeptide Chain.ʺ Proc Natl Acad Sci U S A 37: 205‐211.

Pawson, T. and P. Nash (2003). ʺAssembly of cell regulatory systems through protein interaction domains.ʺ Science

300(5618): 445‐452.

Pedchenko, V. K., S. V. Chetyrkin, P. Chuang, A. J. Ham, M. A. Saleem, P. W. Mathieson, B. G. Hudson and P. A. Voziyan

(2005). ʺMechanism of perturbation of integrin‐mediated cell‐matrix interactions by reactive carbonyl compounds

and its implication for pathogenesis of diabetic nephropathy.ʺ Diabetes 54(10): 2952‐2960.

Pepys, M. B., P. N. Hawkins, D. R. Booth, D. M. Vigushin, G. A. Tennent, A. K. Soutar, N. Totty, O. Nguyen, C. C. Blake,

C. J. Terry and et al. (1993). ʺHuman lysozyme gene mutations cause hereditary systemic amyloidosis.ʺ Nature

362(6420): 553‐557.

Pertinhez, T. A., M. Bouchard, E. J. Tomlinson, R. Wain, S. J. Ferguson, C. M. Dobson and L. J. Smith (2001). ʺAmyloid

fibril formation by a helical cytochrome.ʺ FEBS Lett 495(3): 184‐186.

Pettersen, E. F., T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng and T. E. Ferrin (2004). ʺUCSF

Chimera‐‐a visualization system for exploratory research and analysis.ʺ J Comput Chem 25(13): 1605‐1612.

Piatigorsky, J., M. Kantorow, R. Gopal‐Srivastava and S. I. Tomarev (1994). ʺRecruitment of enzymes and stress proteins as

lens crystallins.ʺ EXS 71: 241‐250.

Pickart, C. M. and R. E. Cohen (2004). ʺProteasomes and their kin: proteases in the machine age.ʺ Nat Rev Mol Cell Biol 5(3):

177‐187.

Pike, C. J., A. J. Walencewicz, C. G. Glabe and C. W. Cotman (1991). ʺIn vitro aging of beta‐amyloid protein causes peptide

aggregation and neurotoxicity.ʺ Brain Res 563(1‐2): 311‐314.

 161

References 

Plakoutsi, G., F. Bemporad, M. Monti, D. Pagnozzi, P. Pucci and F. Chiti (2006). ʺExploring the mechanism of formation of

native‐like and precursor amyloid oligomers for the native acylphosphatase from Sulfolobus solfataricus.ʺ

Structure 14(6): 993‐1001.

Polymeropoulos, M. H., C. Lavedan, E. Leroy, S. E. Ide, A. Dehejia, A. Dutra, B. Pike, H. Root, J. Rubenstein, R. Boyer, E.

S. Stenroos, S. Chandrasekharappa, A. Athanassiadou, T. Papapetropoulos, W. G. Johnson, A. M. Lazzarini, R.

C. Duvoisin, G. Di Iorio, L. I. Golbe and R. L. Nussbaum (1997). ʺMutation in the alpha‐synuclein gene identified

in families with Parkinsonʹs disease.ʺ Science 276(5321): 2045‐2047.

Ponces Freire, A., A. Ferreira, R. Gomes and C. Cordeiro (2003). ʺAnti‐glycation defences in yeast.ʺ Biochem Soc Trans 31(Pt

6): 1409‐1412.

Ponder, J. W. and F. M. Richards (1987). ʺTertiary templates for proteins. Use of packing criteria in the enumeration of

allowed sequences for different structural classes.ʺ J Mol Biol 193(4): 775‐791.

Portero‐Otin, M., R. Pamplona, M. J. Bellmunt, M. C. Ruiz, J. Prat, R. Salvayre and A. Negre‐Salvayre (2002). ʺAdvanced

glycation end product precursors impair epidermal growth factor receptor signaling.ʺ Diabetes 51(5): 1535‐1542.

Pountney, D. L., F. Chegini, X. Shen, P. C. Blumbergs and W. P. Gai (2005). ʺSUMO‐1 marks subdomains within glial

cytoplasmic inclusions of multiple system atrophy.ʺ Neurosci Lett 381(1‐2): 74‐79.

Pountney, D. L., Y. Huang, R. J. Burns, E. Haan, P. D. Thompson, P. C. Blumbergs and W. P. Gai (2003). ʺSUMO‐1 marks

the nuclear inclusions in familial neuronal intranuclear inclusion disease.ʺ Exp Neurol 184(1): 436‐446.

Praefcke, G. J., M. G. Ford, E. M. Schmid, L. E. Olesen, J. L. Gallop, S. Y. Peak‐Chew, Y. Vallis, M. M. Babu, I. G. Mills

and H. T. McMahon (2004). ʺEvolving nature of the AP2 alpha‐appendage hub during clathrin‐coated vesicle

endocytosis.ʺ EMBO J 23(22): 4371‐4383.

Ptitsyn, O. B. and A. A. Rashin (1975). ʺA model of myoglobin self‐organization.ʺ Biophys Chem 3(1): 1‐20.

Quintas, A., A. Ponces Freire and M. J. Halpern (2008). Bioquímica ‐ Organização Molecular da Vida. Lisboa, Lidel, Edições

Técnica, Lda.

Quintas, A., M. J. Saraiva and R. M. Brito (1997). ʺThe amyloidogenic potential of transthyretin variants correlates with

their tendency to aggregate in solution.ʺ FEBS Lett 418(3): 297‐300.

Quintas, A., M. J. Saraiva and R. M. Brito (1999). ʺThe tetrameric protein transthyretin dissociates to a non‐native monomer

in solution. A novel model for amyloidogenesis.ʺ J Biol Chem 274(46): 32943‐32949.

Quintas, A., D. C. Vaz, I. Cardoso, M. J. Saraiva and R. M. Brito (2001). ʺTetramer dissociation and monomer partial

unfolding precedes protofibril formation in amyloidogenic transthyretin variants.ʺ J Biol Chem 276(29): 27207‐

27213.

Racker, E. (1951). ʺThe mechanism of action of glyoxalase.ʺ J Biol Chem 190(2): 685‐696.

Radford, S. E. (2000). ʺProtein folding: progress made and promises ahead.ʺ Trends Biochem Sci 25(12): 611‐618.

Radivojac, P., Z. Obradovic, D. K. Smith, G. Zhu, S. Vucetic, C. J. Brown, J. D. Lawson and A. K. Dunker (2004). ʺProtein

flexibility and intrinsic disorder.ʺ Protein Sci 13(1): 71‐80.

 162

References

Radzicka, A., L. Pedersen and R. Wolfenden (1988). ʺInfluences of solvent water on protein folding: free energies of

solvation of cis and trans peptides are nearly identical.ʺ Biochemistry 27(12): 4538‐4541.

Rakhit, R., J. P. Crow, J. R. Lepock, L. H. Kondejewski, N. R. Cashman and A. Chakrabartty (2004). ʺMonomeric Cu,Zn‐

superoxide dismutase is a common misfolding intermediate in the oxidation models of sporadic and familial

amyotrophic lateral sclerosis.ʺ J Biol Chem 279(15): 15499‐15504.

Rakhit, R., P. Cunningham, A. Furtos‐Matei, S. Dahan, X. F. Qi, J. P. Crow, N. R. Cashman, L. H. Kondejewski and A.

Chakrabartty (2002). ʺOxidation‐induced misfolding and aggregation of superoxide dismutase and its

implications for amyotrophic lateral sclerosis.ʺ J Biol Chem 277(49): 47551‐47556.

Ramachandran, G. N. (1963). ʺProtein Structure and Crystallography.ʺ Science 141(3577): 288‐291.

Ramaswamy, S., K. M. Shannon and J. H. Kordower (2007). ʺHuntingtonʹs disease: pathological mechanisms and

therapeutic strategies.ʺ Cell Transplant 16(3): 301‐312.

Rao, F., G. Settanni, E. Guarnera and A. Caflisch (2005). ʺEstimation of protein folding probability from equilibrium

simulations.ʺ J Chem Phys 122(18): 184901.

Rao, R. V. and D. E. Bredesen (2004). ʺMisfolded proteins, endoplasmic reticulum stress and neurodegeneration.ʺ Curr Opin

Cell Biol 16(6): 653‐662.

Ratnaswamy, G., E. Koepf, H. Bekele, H. Yin and J. W. Kelly (1999). ʺThe amyloidogenicity of gelsolin is controlled by

proteolysis and pH.ʺ Chem Biol 6(5): 293‐304.

Rattan, S. I. (2004). ʺMechanisms of hormesis through mild heat stress on human cells.ʺ Ann N Y Acad Sci 1019: 554‐558.

Ray, M. and S. Ray (1984). ʺPurification and partial characterization of a methylglyoxal reductase from goat liver.ʺ Biochim

Biophys Acta 802(1): 119‐127.

Ray, M. and S. Ray (1987). ʺAminoacetone oxidase from goat liver. Formation of methylglyoxal from aminoacetone.ʺ J Biol

Chem 262(13): 5974‐5977.

Reggiori, F. and D. J. Klionsky (2005). ʺAutophagosomes: biogenesis from scratch?ʺ Curr Opin Cell Biol 17(4): 415‐422.

Reimer, U., G. Scherer, M. Drewello, S. Kruber, M. Schutkowski and G. Fischer (1998). ʺSide‐chain effects on peptidyl‐

prolyl cis/trans isomerisation.ʺ J Mol Biol 279(2): 449‐460.

Reynolds, M. R., R. W. Berry and L. I. Binder (2007). ʺNitration in neurodegeneration: deciphering the ʺHowsʺ ʺnYsʺ.ʺ

Biochemistry 46(25): 7325‐7336.

Richard, J. P. (1984). ʺAcid‐Base Catalysis of the Elimination and Isomerization‐Reactions of Triose Phosphates.ʺ Journal of

the American Chemical Society 106(17): 4926‐4936.

Richard, J. P. (1993). ʺMechanism for the formation of methylglyoxal from triosephosphates.ʺ Biochem Soc Trans 21(2): 549‐

553.

Richardson, J. S. (1985). ʺSchematic drawings of protein structures.ʺ Methods Enzymol 115: 359‐380.

 163

References 

Richardson, J. S. and D. C. Richardson (1989). Principles and Patterns of Protein Conformation. Prediction of Protein

Structure and the Principles of Protein Conformation. G. D. Fasman, Plenum Press NY: 359‐389.

Richardson, J. S., D. C. Richardson, N. B. Tweedy, K. M. Gernert, T. P. Quinn, M. H. Hecht, B. W. Erickson, Y. Yan, R. D.

McClain, M. E. Donlan and et al. (1992). ʺLooking at proteins: representations, folding, packing, and design.

Biophysical Society National Lecture, 1992.ʺ Biophys J 63(5): 1185‐1209.

Richly, H., M. Rape, S. Braun, S. Rumpf, C. Hoege and S. Jentsch (2005). ʺA series of ubiquitin binding factors connects

CDC48/p97 to substrate multiubiquitylation and proteasomal targeting.ʺ Cell 120(1): 73‐84.

Richter‐Landsberg, C. and O. Goldbaum (2003). ʺStress proteins in neural cells: functional roles in health and disease.ʺ Cell

Mol Life Sci 60(2): 337‐349.

Rideout, H. J., K. E. Larsen, D. Sulzer and L. Stefanis (2001). ʺProteasomal inhibition leads to formation of ubiquitin/alpha‐

synuclein‐immunoreactive inclusions in PC12 cells.ʺ J Neurochem 78(4): 899‐908.

Rideout, H. J. and L. Stefanis (2002). ʺProteasomal inhibition‐induced inclusion formation and death in cortical neurons

require transcription and ubiquitination.ʺ Mol Cell Neurosci 21(2): 223‐238.

Riederer, P., E. Sofic, W. D. Rausch, B. Schmidt, G. P. Reynolds, K. Jellinger and M. B. Youdim (1989). ʺTransition metals,

ferritin, glutathione, and ascorbic acid in parkinsonian brains.ʺ J Neurochem 52(2): 515‐520.

Riley, B. E., H. Y. Zoghbi and H. T. Orr (2005). ʺSUMOylation of the polyglutamine repeat protein, ataxin‐1, is dependent on

a functional nuclear localization signal.ʺ J Biol Chem 280(23): 21942‐21948.

Romero, Obradovic and K. Dunker (1997). ʺSequence Data Analysis for Long Disordered Regions Prediction in the

Calcineurin Family.ʺ Genome Inform Ser Workshop Genome Inform 8: 110‐124.

Romero, P., Z. Obradovic, C. R. Kissinger, J. E. Villafranca, E. Garner, S. Guilliot and A. K. Dunker (1998). ʺThousands of

proteins likely to have long disordered regions.ʺ Pac Symp Biocomput: 437‐448.

Romero, P., Z. Obradovic, X. Li, E. C. Garner, C. J. Brown and A. K. Dunker (2001). ʺSequence complexity of disordered

protein.ʺ Proteins 42(1): 38‐48.

Ross, C. A. and M. A. Poirier (2004). ʺProtein aggregation and neurodegenerative disease.ʺ Nat Med 10 Suppl: S10‐17.

Rousseau, F., H. Wilkinson, J. Villanueva, L. Serrano, J. W. Schymkowitz and L. S. Itzhaki (2006). ʺDomain swapping in

p13suc1 results in formation of native‐like, cytotoxic aggregates.ʺ J Mol Biol 363(2): 496‐505.

Rual, J. F., K. Venkatesan, T. Hao, T. Hirozane‐Kishikawa, A. Dricot, N. Li, G. F. Berriz, F. D. Gibbons, M. Dreze, N.

Ayivi‐Guedehoussou, N. Klitgord, C. Simon, M. Boxem, S. Milstein, J. Rosenberg, D. S. Goldberg, L. V. Zhang,

S. L. Wong, G. Franklin, S. Li, J. S. Albala, J. Lim, C. Fraughton, E. Llamosas, S. Cevik, C. Bex, P. Lamesch, R. S.

Sikorski, J. Vandenhaute, H. Y. Zoghbi, A. Smolyar, S. Bosak, R. Sequerra, L. Doucette‐Stamm, M. E. Cusick, D.

E. Hill, F. P. Roth and M. Vidal (2005). ʺTowards a proteome‐scale map of the human protein‐protein interaction

network.ʺ Nature 437(7062): 1173‐1178.

Sakamoto, H., T. Mashima, K. Yamamoto and T. Tsuruo (2002). ʺModulation of heat‐shock protein 27 (Hsp27) anti‐

apoptotic activity by methylglyoxal modification.ʺ J Biol Chem 277(48): 45770‐45775.

 164

References

Sampathu, D. M., B. I. Giasson, A. C. Pawlyk, J. Q. Trojanowski and V. M. Lee (2003). ʺUbiquitination of alpha‐synuclein

is not required for formation of pathological inclusions in alpha‐synucleinopathies.ʺ Am J Pathol 163(1): 91‐100.

Sander, C. and R. Schneider (1991). ʺDatabase of homology‐derived protein structures and the structural meaning of

sequence alignment.ʺ Proteins 9(1): 56‐68.

Sangar, V., D. J. Blankenberg, N. Altman and A. M. Lesk (2007). ʺQuantitative sequence‐function relationships in proteins

based on gene ontology.ʺ BMC Bioinformatics 8: 294.

Santoro, M. M. and D. W. Bolen (1988). ʺUnfolding free energy changes determined by the linear extrapolation method. 1.

Unfolding of phenylmethanesulfonyl alpha‐chymotrypsin using different denaturants.ʺ Biochemistry 27(21): 8063‐

8068.

Santoro, M. M. and D. W. Bolen (1992). ʺA test of the linear extrapolation of unfolding free energy changes over an

extended denaturant concentration range.ʺ Biochemistry 31(20): 4901‐4907.

Saraiva, M. J. (2001). ʺTransthyretin mutations in hyperthyroxinemia and amyloid diseases.ʺ Hum Mutat 17(6): 493‐503.

Saraiva, M. J., S. Birken, P. P. Costa and D. S. Goodman (1984). ʺAmyloid fibril protein in familial amyloidotic

polyneuropathy, Portuguese type. Definition of molecular abnormality in transthyretin (prealbumin).ʺ J Clin

Invest 74(1): 104‐119.

Sawaya, M. R., S. Sambashivan, R. Nelson, M. I. Ivanova, S. A. Sievers, M. I. Apostol, M. J. Thompson, M. Balbirnie, J. J.

Wiltzius, H. T. McFarlane, A. O. Madsen, C. Riekel and D. Eisenberg (2007). ʺAtomic structures of amyloid

cross‐beta spines reveal varied steric zippers.ʺ Nature 447(7143): 453‐457.

Schellman, J. A. (1955). ʺThe stability of hydrogen‐bonded peptide structures in aqueous solution.ʺ C R Trav Lab Carlsberg

Chim 29(14‐15): 230‐259.

Schellman, J. A. (1987). ʺThe thermodynamic stability of proteins.ʺ Annu Rev Biophys Biophys Chem 16: 115‐137.

Schleicher, E. and O. H. Wieland (1986). ʺKinetic analysis of glycation as a tool for assessing the half‐life of proteins.ʺ

Biochim Biophys Acta 884(1): 199‐205.

Schmid, E. M., M. G. Ford, A. Burtey, G. J. Praefcke, S. Y. Peak‐Chew, I. G. Mills, A. Benmerah and H. T. McMahon

(2006). ʺRole of the AP2 beta‐appendage hub in recruiting partners for clathrin‐coated vesicle assembly.ʺ PLoS Biol

4(9): e262.

Schmid, E. M. and H. T. McMahon (2007). ʺIntegrating molecular and network biology to decode endocytosis.ʺ Nature

448(7156): 883‐888.

Schmid, F. X. (1995). ʺProtein folding. Prolyl isomerases join the fold.ʺ Curr Biol 5(9): 993‐994.

Schrauber, H., F. Eisenhaber and P. Argos (1993). ʺRotamers: to be or not to be? An analysis of amino acid side‐chain

conformations in globular proteins.ʺ J Mol Biol 230(2): 592‐612.

Schweers, O., E. Schonbrunn‐Hanebeck, A. Marx and E. Mandelkow (1994). ʺStructural studies of tau protein and

Alzheimer paired helical filaments show no evidence for beta‐structure.ʺ J Biol Chem 269(39): 24290‐24297.

 165

References 

Seet, B. T., I. Dikic, M. M. Zhou and T. Pawson (2006). ʺReading protein modifications with interaction domains.ʺ Nat Rev

Mol Cell Biol 7(7): 473‐483.

Seidler, N. W. and C. Kowalewski (2003). ʺMethylglyoxal‐induced glycation affects protein topography.ʺ Arch Biochem

Biophys 410(1): 149‐154.

Seidler, N. W. and I. Seibel (2000). ʺGlycation of aspartate aminotransferase and conformational flexibility.ʺ Biochem

Biophys Res Commun 277(1): 47‐50.

Seldeen, K. L., C. B. McDonald, B. J. Deegan and A. Farooq (2008). ʺCoupling of folding and DNA‐binding in the bZIP

domains of Jun‐Fos heterodimeric transcription factor.ʺ Arch Biochem Biophys 473(1): 48‐60.

Selkoe, D. J. (1999). ʺTranslating cell biology into therapeutic advances in Alzheimerʹs disease.ʺ Nature 399(6738 Suppl):

A23‐31.

Sell, D. R. and V. M. Monnier (1989). ʺStructure elucidation of a senescence cross‐link from human extracellular matrix.

Implication of pentoses in the aging process.ʺ J Biol Chem 264(36): 21597‐21602.

Shepherd, P. R. and B. B. Kahn (1999). ʺGlucose transporters and insulin action‐‐implications for insulin resistance and

diabetes mellitus.ʺ N Engl J Med 341(4): 248‐257.

Shibata, N., A. Hirano, E. T. Hedley‐Whyte, M. C. Dal Canto, R. Nagai, K. Uchida, S. Horiuchi, M. Kawaguchi, T.

Yamamoto and M. Kobayashi (2002). ʺSelective formation of certain advanced glycation end products in spinal

cord astrocytes of humans and mice with superoxide dismutase‐1 mutation.ʺ Acta Neuropathol (Berl) 104(2): 171‐

178.

Shinbo, Y., T. Niki, T. Taira, H. Ooe, K. Takahashi‐Niki, C. Maita, C. Seino, S. M. Iguchi‐Ariga and H. Ariga (2006).

ʺProper SUMO‐1 conjugation is essential to DJ‐1 to exert its full activities.ʺ Cell Death Differ 13(1): 96‐108.

Shinpo, K., S. Kikuchi, H. Sasaki, A. Ogata, F. Moriwaka and K. Tashiro (2000). ʺSelective vulnerability of spinal motor

neurons to reactive dicarbonyl compounds, intermediate products of glycation, in vitro: implication of inefficient

glutathione system in spinal motor neurons.ʺ Brain Res 861(1): 151‐159.

Shipanova, I. N., M. A. Glomb and R. H. Nagaraj (1997). ʺProtein modification by methylglyoxal: chemical nature and

synthetic mechanism of a major fluorescent adduct.ʺ Arch Biochem Biophys 344(1): 29‐36.

Shoemaker, B. A., J. J. Portman and P. G. Wolynes (2000). ʺSpeeding molecular recognition by using the folding funnel: the

fly‐casting mechanism.ʺ Proc Natl Acad Sci U S A 97(16): 8868‐8873.

Shtilerman, M. D., T. T. Ding and P. T. Lansbury, Jr. (2002). ʺMolecular crowding accelerates fibrillization of alpha‐

synuclein: could an increase in the cytoplasmic protein concentration induce Parkinsonʹs disease?ʺ Biochemistry

41(12): 3855‐3860.

Sian, J., D. T. Dexter, A. J. Lees, S. Daniel, Y. Agid, F. Javoy‐Agid, P. Jenner and C. D. Marsden (1994). ʺAlterations in

glutathione levels in Parkinsonʹs disease and other neurodegenerative disorders affecting basal ganglia.ʺ Ann

Neurol 36(3): 348‐355.

Sipe, J. D. (1992). ʺAmyloidosis.ʺ Annu Rev Biochem 61: 947‐975.

 166

References

Slee, E. A., M. T. Harte, R. M. Kluck, B. B. Wolf, C. A. Casiano, D. D. Newmeyer, H. G. Wang, J. C. Reed, D. W.

Nicholson, E. S. Alnemri, D. R. Green and S. J. Martin (1999). ʺOrdering the cytochrome c‐initiated caspase

cascade: hierarchical activation of caspases‐2, ‐3, ‐6, ‐7, ‐8, and ‐10 in a caspase‐9‐dependent manner.ʺ J Cell Biol

144(2): 281‐292.

Smith, M. A., S. Taneda, P. L. Richey, S. Miyata, S. D. Yan, D. Stern, L. M. Sayre, V. M. Monnier and G. Perry (1994).

ʺAdvanced Maillard reaction end products are associated with Alzheimer disease pathology.ʺ Proc Natl Acad Sci

U S A 91(12): 5710‐5714.

Smith, P. R. and P. J. Thornalley (1992a). ʺInfluence of pH and phosphate ions on the kinetics of enolisation and

degradation of fructosamines. Studies with the model fructosamine, N epsilon‐1‐deoxy‐D‐fructos‐1‐yl‐hippuryl‐

lysine.ʺ Biochem Int 28(3): 429‐439.

Smith, P. R. and P. J. Thornalley (1992b). ʺMechanism of the degradation of non‐enzymatically glycated proteins under

physiological conditions. Studies with the model fructosamine, N epsilon‐(1‐deoxy‐D‐fructos‐1‐yl)hippuryl‐

lysine.ʺ Eur J Biochem 210(3): 729‐739.

Soreghan, B., J. Kosmoski and C. Glabe (1994). ʺSurfactant properties of Alzheimerʹs A beta peptides and the mechanism of

amyloid aggregation.ʺ J Biol Chem 269(46): 28551‐28554.

Soto, C. (2001). ʺProtein misfolding and disease; protein refolding and therapy.ʺ FEBS Lett 498(2‐3): 204‐207.

Souillac, P. O., V. N. Uversky, I. S. Millett, R. Khurana, S. Doniach and A. L. Fink (2002). ʺEffect of association state and

conformational stability on the kinetics of immunoglobulin light chain amyloid fibril formation at physiological

pH.ʺ J Biol Chem 277(15): 12657‐12665.

Sousa, M. M., I. Cardoso, R. Fernandes, A. Guimaraes and M. J. Saraiva (2001). ʺDeposition of transthyretin in early stages

of familial amyloidotic polyneuropathy: evidence for toxicity of nonfibrillar aggregates.ʺ Am J Pathol 159(6): 1993‐

2000.

Souza, J. M., B. I. Giasson, Q. Chen, V. M. Lee and H. Ischiropoulos (2000). ʺDityrosine cross‐linking promotes formation

of stable alpha ‐synuclein polymers. Implication of nitrative and oxidative stress in the pathogenesis of

neurodegenerative synucleinopathies.ʺ J Biol Chem 275(24): 18344‐18349.

Speer, O., S. Morkunaite‐Haimi, J. Liobikas, M. Franck, L. Hensbo, M. D. Linder, P. K. Kinnunen, T. Wallimann and O.

Eriksson (2003). ʺRapid suppression of mitochondrial permeability transition by methylglyoxal. Role of reversible

arginine modification.ʺ J Biol Chem 278(37): 34757‐34763.

Spillantini, M. G., R. A. Crowther, R. Jakes, M. Hasegawa and M. Goedert (1998). ʺalpha‐Synuclein in filamentous

inclusions of Lewy bodies from Parkinsonʹs disease and dementia with lewy bodies.ʺ Proc Natl Acad Sci U S A

95(11): 6469‐6473.

Spillantini, M. G. and M. Goedert (2000). ʺThe alpha‐synucleinopathies: Parkinsonʹs disease, dementia with Lewy bodies,

and multiple system atrophy.ʺ Ann N Y Acad Sci 920: 16‐27.

Spillantini, M. G., M. L. Schmidt, V. M. Lee, J. Q. Trojanowski, R. Jakes and M. Goedert (1997). ʺAlpha‐synuclein in Lewy

bodies.ʺ Nature 388(6645): 839‐840.

 167

References 

Sreerama, N. and R. W. Woody (2004). ʺComputation and analysis of protein circular dichroism spectra.ʺ Methods Enzymol

383: 318‐351.

Srere, P. A. (1987). ʺComplexes of sequencial metabolic enzymes.ʺ Ann Rev Biochem 56: 89‐124.

Stec, B. and L. Lebioda (1990). ʺRefined structure of yeast apo‐enolase at 2.25 A resolution.ʺ J Mol Biol 211(1): 235‐248.

Stefani, M. (2004). ʺProtein misfolding and aggregation: new examples in medicine and biology of the dark side of the

protein world.ʺ Biochim Biophys Acta 1739(1): 5‐25.

Stefani, M. and C. M. Dobson (2003). ʺProtein aggregation and aggregate toxicity: new insights into protein folding,

misfolding diseases and biological evolution.ʺ J Mol Med 81(11): 678‐699.

Steffan, J. S., N. Agrawal, J. Pallos, E. Rockabrand, L. C. Trotman, N. Slepko, K. Illes, T. Lukacsovich, Y. Z. Zhu, E.

Cattaneo, P. P. Pandolfi, L. M. Thompson and J. L. Marsh (2004). ʺSUMO modification of Huntingtin and

Huntingtonʹs disease pathology.ʺ Science 304(5667): 100‐104.

Stelzl, U., U. Worm, M. Lalowski, C. Haenig, F. H. Brembeck, H. Goehler, M. Stroedicke, M. Zenkner, A. Schoenherr, S.

Koeppen, J. Timm, S. Mintzlaff, C. Abraham, N. Bock, S. Kietzmann, A. Goedde, E. Toksoz, A. Droege, S.

Krobitsch, B. Korn, W. Birchmeier, H. Lehrach and E. E. Wanker (2005). ʺA human protein‐protein interaction

network: a resource for annotating the proteome.ʺ Cell 122(6): 957‐968.

Stewart, D. E., A. Sarkar and J. E. Wampler (1990). ʺOccurrence and role of cis peptide bonds in protein structures.ʺ J Mol

Biol 214(1): 253‐260.

Storkel, S., H. M. Schneider, H. Muntefering and S. Kashiwagi (1983). ʺIatrogenic, insulin‐dependent, local amyloidosis.ʺ

Lab Invest 48(1): 108‐111.

Szent‐Gyorgyi, A. (1965). ʺCell Division and Cancer.ʺ Science 149: 34‐37.

Tainer, J. A., V. A. Roberts and E. D. Getzoff (1992). ʺProtein metal‐binding sites.ʺ Curr Opin Biotechnol 3(4): 378‐387.

Takeda, A., M. Mallory, M. Sundsmo, W. Honer, L. Hansen and E. Masliah (1998). ʺAbnormal accumulation of

NACP/alpha‐synuclein in neurodegenerative disorders.ʺ Am J Pathol 152(2): 367‐372.

Taylor, R. (1991). ʺInsulin action 1991.ʺ Clin Endocrinol (Oxf) 34(2): 159‐171.

Teilum, K., J. G. Olsen and B. B. Kragelund (2009). ʺFunctional aspects of protein flexibility.ʺ Cell Mol Life Sci 66(14): 2231‐

2247.

Thomas, P. J., B. H. Qu and P. L. Pedersen (1995). ʺDefective protein folding as a basis of human disease.ʺ Trends Biochem

Sci 20(11): 456‐459.

Thompson, A. J. and C. J. Barrow (2002). ʺProtein conformational misfolding and amyloid formation: characteristics of a

new class of disorders that include Alzheimerʹs and Prion diseases.ʺ Curr Med Chem 9(19): 1751‐1762.

Thornalley, P. J. (1990). ʺThe glyoxalase system: new developments towards functional characterization of a metabolic

pathway fundamental to biological life.ʺ Biochem J 269(1): 1‐11.

Thornalley, P. J. (1993). ʺThe glyoxalase system in health and disease.ʺ Mol Aspects Med 14(4): 287‐371.

 168

References

Thornalley, P. J. (1994). ʺMethylglyoxal, Glyoxalases and the Development of Diabetic Complications.ʺ Amino Acids 6: 15‐

23.

Thornalley, P. J. (1996). ʺPharmacology of methylglyoxal: formation, modification of proteins and nucleic acids, and

enzymatic detoxification‐‐a role in pathogenesis and antiproliferative chemotherapy.ʺ Gen Pharmacol 27(4): 565‐

573.

Thornalley, P. J., A. Langborg and H. S. Minhas (1999). ʺFormation of glyoxal, methylglyoxal and 3‐deoxyglucosone in the

glycation of proteins by glucose.ʺ Biochem J 344 Pt 1: 109‐116.

Thornton, J. M., C. A. Orengo, A. E. Todd and F. M. Pearl (1999). ʺProtein folds, functions and evolution.ʺ J Mol Biol 293(2):

333‐342.

Tofaris, G. K., R. Layfield and M. G. Spillantini (2001). ʺalpha‐synuclein metabolism and aggregation is linked to

ubiquitin‐independent degradation by the proteasome.ʺ FEBS Lett 509(1): 22‐26.

Tompa, P. (2002). ʺIntrinsically unstructured proteins.ʺ Trends Biochem Sci 27(10): 527‐533.

Tompa, P. (2005). ʺThe interplay between structure and function in intrinsically unstructured proteins.ʺ FEBS Lett 579(15):

3346‐3354.

Tompa, P., J. Prilusky, I. Silman and J. L. Sussman (2008). ʺStructural disorder serves as a weak signal for intracellular

protein degradation.ʺ Proteins 71(2): 903‐909.

Tompa, P., C. Szasz and L. Buday (2005). ʺStructural disorder throws new light on moonlighting.ʺ Trends Biochem Sci 30(9):

484‐489.

Trovato, A., F. Chiti, A. Maritan and F. Seno (2006). ʺInsight into the structure of amyloid fibrils from the analysis of

globular proteins.ʺ PLoS Comput Biol 2(12): e170.

Tuffery, P., C. Etchebest and S. Hazout (1997). ʺPrediction of protein side chain conformations: a study on the influence of

backbone accuracy on conformation stability in the rotamer space.ʺ Protein Eng 10(4): 361‐372.

Tuffery, P., C. Etchebest, S. Hazout and R. Lavery (1991). ʺA new approach to the rapid determination of protein side chain

conformations.ʺ J Biomol Struct Dyn 8(6): 1267‐1289.

Turnell, W. G. and J. T. Finch (1992). ʺBinding of the dye congo red to the amyloid protein pig insulin reveals a novel

homology amongst amyloid‐forming peptide sequences.ʺ J Mol Biol 227(4): 1205‐1223.

Ueda, H., J. Goto, H. Hashida, X. Lin, K. Oyanagi, H. Kawano, H. Y. Zoghbi, I. Kanazawa and H. Okazawa (2002).

ʺEnhanced SUMOylation in polyglutamine diseases.ʺ Biochem Biophys Res Commun 293(1): 307‐313.

Ueda, K., H. Fukushima, E. Masliah, Y. Xia, A. Iwai, M. Yoshimoto, D. A. Otero, J. Kondo, Y. Ihara and T. Saitoh (1993).

ʺMolecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease.ʺ Proc Natl

Acad Sci U S A 90(23): 11282‐11286.

Uetz, P., L. Giot, G. Cagney, T. A. Mansfield, R. S. Judson, J. R. Knight, D. Lockshon, V. Narayan, M. Srinivasan, P.

Pochart, A. Qureshi‐Emili, Y. Li, B. Godwin, D. Conover, T. Kalbfleisch, G. Vijayadamodar, M. Yang, M.

 169

References 

Johnston, S. Fields and J. M. Rothberg (2000). ʺA comprehensive analysis of protein‐protein interactions in

Saccharomyces cerevisiae.ʺ Nature 403(6770): 623‐627.

Ulmer, T. S., A. Bax, N. B. Cole and R. L. Nussbaum (2005). ʺStructure and dynamics of micelle‐bound human alpha‐

synuclein.ʺ J Biol Chem 280(10): 9595‐9603.

Uversky, V. N. (2002a). ʺNatively unfolded proteins: a point where biology waits for physics.ʺ Protein Sci 11(4): 739‐756.

Uversky, V. N. (2002b). ʺWhat does it mean to be natively unfolded?ʺ Eur J Biochem 269(1): 2‐12.

Uversky, V. N. (2003). ʺProtein folding revisited. A polypeptide chain at the folding‐misfolding‐nonfolding cross‐roads:

which way to go?ʺ Cell Mol Life Sci 60(9): 1852‐1871.

Uversky, V. N. (2007). ʺNeuropathology, biochemistry, and biophysics of alpha‐synuclein aggregation.ʺ J Neurochem 103(1):

17‐37.

Uversky, V. N. and A. L. Fink (2004). ʺConformational constraints for amyloid fibrillation: the importance of being

unfolded.ʺ Biochim Biophys Acta 1698(2): 131‐153.

Uversky, V. N., J. R. Gillespie and A. L. Fink (2000). ʺWhy are ʺnatively unfoldedʺ proteins unstructured under physiologic

conditions?ʺ Proteins 41(3): 415‐427.

Uversky, V. N., J. Li and A. L. Fink (2001a). ʺEvidence for a partially folded intermediate in alpha‐synuclein fibril

formation.ʺ J Biol Chem 276(14): 10737‐10744.

Uversky, V. N., J. Li and A. L. Fink (2001b). ʺTrimethylamine‐N‐oxide‐induced folding of alpha‐synuclein.ʺ FEBS Lett

509(1): 31‐35.

Uversky, V. N. and O. B. Ptitsyn (1994). ʺʺPartly foldedʺ state, a new equilibrium state of protein molecules: four‐state

guanidinium chloride‐induced unfolding of beta‐lactamase at low temperature.ʺ Biochemistry 33(10): 2782‐2791.

Uversky, V. N., G. Yamin, L. A. Munishkina, M. A. Karymov, I. S. Millett, S. Doniach, Y. L. Lyubchenko and A. L. Fink

(2005). ʺEffects of nitration on the structure and aggregation of alpha‐synuclein.ʺ Brain Res Mol Brain Res 134(1):

84‐102.

Uversky, V. N., G. Yamin, P. O. Souillac, J. Goers, C. B. Glaser and A. L. Fink (2002). ʺMethionine oxidation inhibits

fibrillation of human alpha‐synuclein in vitro.ʺ FEBS Lett 517(1‐3): 239‐244.

Valente, E. M., P. M. Abou‐Sleiman, V. Caputo, M. M. Muqit, K. Harvey, S. Gispert, Z. Ali, D. Del Turco, A. R.

Bentivoglio, D. G. Healy, A. Albanese, R. Nussbaum, R. Gonzalez‐Maldonado, T. Deller, S. Salvi, P. Cortelli,

W. P. Gilks, D. S. Latchman, R. J. Harvey, B. Dallapiccola, G. Auburger and N. W. Wood (2004). ʺHereditary

early‐onset Parkinsonʹs disease caused by mutations in PINK1.ʺ Science 304(5674): 1158‐1160.

Valentine, J. S. and P. J. Hart (2003). ʺMisfolded CuZnSOD and amyotrophic lateral sclerosis.ʺ Proc Natl Acad Sci U S A

100(7): 3617‐3622.

Valleix, S., S. Drunat, J. B. Philit, D. Adoue, J. C. Piette, D. Droz, B. MacGregor, D. Canet, M. Delpech and G. Grateau

(2002). ʺHereditary renal amyloidosis caused by a new variant lysozyme W64R in a French family.ʺ Kidney Int

61(3): 907‐912.

 170

References

Vander Jagt, D. L. (1993). ʺGlyoxalase II: molecular characteristics, kinetics and mechanism.ʺ Biochem Soc Trans 21(2): 522‐

527.

Vander Jagt, D. L., E. Daub, J. A. Krohn and L. P. Han (1975). ʺEffects of pH and thiols on the kinetics of yeast glyoxalase I.

An evaluation of the random pathway mechanism.ʺ Biochemistry 14(16): 3669‐3675.

Vander Jagt, D. L. and L. A. Hunsaker (2003). ʺMethylglyoxal metabolism and diabetic complications: roles of aldose

reductase, glyoxalase‐I, betaine aldehyde dehydrogenase and 2‐oxoaldehyde dehydrogenase.ʺ Chem Biol Interact

143‐144: 341‐351.

Vander Jagt, D. L., B. Robinson, K. K. Taylor and L. A. Hunsaker (1992). ʺReduction of trioses by NADPH‐dependent aldo‐

keto reductases. Aldose reductase, methylglyoxal, and diabetic complications.ʺ J Biol Chem 267(7): 4364‐4369.

Varshavsky, A. (2005). ʺRegulated protein degradation.ʺ Trends Biochem Sci 30(6): 283‐286.

Verzijl, N., J. DeGroot, Z. C. Ben, O. Brau‐Benjamin, A. Maroudas, R. A. Bank, J. Mizrahi, C. G. Schalkwijk, S. R. Thorpe,

J. W. Baynes, J. W. Bijlsma, F. P. Lafeber and J. M. TeKoppele (2002). ʺCrosslinking by advanced glycation end

products increases the stiffness of the collagen network in human articular cartilage: a possible mechanism

through which age is a risk factor for osteoarthritis.ʺ Arthritis Rheum 46(1): 114‐123.

Vestergaard, B., M. Groenning, M. Roessle, J. S. Kastrup, M. van de Weert, J. M. Flink, S. Frokjaer, M. Gajhede and D. I.

Svergun (2007). ʺA helical structural nucleus is the primary elongating unit of insulin amyloid fibrils.ʺ PLoS Biol

5(5): e134.

Vitek, M. P., K. Bhattacharya, J. M. Glendening, E. Stopa, H. Vlassara, R. Bucala, K. Manogue and A. Cerami (1994).

ʺAdvanced glycation end products contribute to amyloidosis in Alzheimer disease.ʺ Proc Natl Acad Sci U S A

91(11): 4766‐4770.

Vlassara, H., R. Bucala and L. Striker (1994). ʺPathogenic effects of advanced glycosylation: biochemical, biologic, and

clinical implications for diabetes and aging.ʺ Lab Invest 70(2): 138‐151.

Volles, M. J. and P. T. Lansbury, Jr. (2002). ʺVesicle permeabilization by protofibrillar alpha‐synuclein is sensitive to

Parkinsonʹs disease‐linked mutations and occurs by a pore‐like mechanism.ʺ Biochemistry 41(14): 4595‐4602. von Coelln, R., V. L. Dawson and T. M. Dawson (2004). ʺParkin‐associated Parkinsonʹs disease.ʺ Cell Tissue Res 318(1): 175‐

184.

Vucetic, S., Z. Obradovic, V. Vacic, P. Radivojac, K. Peng, L. M. Iakoucheva, M. S. Cortese, J. D. Lawson, C. J. Brown, J. G.

Sikes, C. D. Newton and A. K. Dunker (2005). ʺDisProt: a database of protein disorder.ʺ Bioinformatics 21(1): 137‐

140.

Wain, R., T. A. Pertinhez, E. J. Tomlinson, L. Hong, C. M. Dobson, S. J. Ferguson and L. J. Smith (2001). ʺThe cytochrome c

fold can be attained from a compact apo state by occupancy of a nascent heme binding site.ʺ J Biol Chem 276(49):

45813‐45817.

Wakabayashi, K., K. Matsumoto, K. Takayama, M. Yoshimoto and H. Takahashi (1997). ʺNACP, a presynaptic protein,

immunoreactivity in Lewy bodies in Parkinsonʹs disease.ʺ Neurosci Lett 239(1): 45‐48.

 171

References 

Wang, J., G. Xu and D. R. Borchelt (2002). ʺHigh molecular weight complexes of mutant superoxide dismutase 1: age‐

dependent and tissue‐specific accumulation.ʺ Neurobiol Dis 9(2): 139‐148.

Ward, J. J., J. S. Sodhi, L. J. McGuffin, B. F. Buxton and D. T. Jones (2004). ʺPrediction and functional analysis of native

disorder in proteins from the three kingdoms of life.ʺ J Mol Biol 337(3): 635‐645.

Watanabe, M., M. Dykes‐Hoberg, V. C. Culotta, D. L. Price, P. C. Wong and J. D. Rothstein (2001). ʺHistological evidence

of protein aggregation in mutant SOD1 transgenic mice and in amyotrophic lateral sclerosis neural tissues.ʺ

Neurobiol Dis 8(6): 933‐941.

Watkins, N. G., C. I. Neglia‐Fisher, D. G. Dyer, S. R. Thorpe and J. W. Baynes (1987). ʺEffect of phosphate on the kinetics

and specificity of glycation of protein.ʺ J Biol Chem 262(15): 7207‐7212.

Wei, G. and J. E. Shea (2006). ʺEffects of solvent on the structure of the Alzheimer amyloid‐beta(25‐35) peptide.ʺ Biophys J

91(5): 1638‐1647.

Wei, Y., S. Kim, D. Fela, J. Baum and M. H. Hecht (2003). ʺSolution structure of a de novo protein from a designed

combinatorial library.ʺ Proc Natl Acad Sci U S A 100(23): 13270‐13273.

Weinreb, P. H., W. Zhen, A. W. Poon, K. A. Conway and P. T. Lansbury, Jr. (1996). ʺNACP, a protein implicated in

Alzheimerʹs disease and learning, is natively unfolded.ʺ Biochemistry 35(43): 13709‐13715.

Wells‐Knecht, K. J., D. V. Zyzak, J. E. Litchfield, S. R. Thorpe and J. W. Baynes (1995). ʺMechanism of autoxidative

glycosylation: identification of glyoxal and arabinose as intermediates in the autoxidative modification of proteins

by glucose.ʺ Biochemistry 34(11): 3702‐3709.

Westermark, P., M. D. Benson, J. N. Buxbaum, A. S. Cohen, B. Frangione, S. Ikeda, C. L. Masters, G. Merlini, M. J. Saraiva

and J. D. Sipe (2005). ʺAmyloid: toward terminology clarification. Report from the Nomenclature Committee of

the International Society of Amyloidosis.ʺ Amyloid 12(1): 1‐4.

Westermark, P., K. Sletten, B. Johansson and G. G. Cornwell, 3rd (1990). ʺFibril in senile systemic amyloidosis is derived

from normal transthyretin.ʺ Proc Natl Acad Sci U S A 87(7): 2843‐2845.

Westwood, M. E. and P. J. Thornalley (1997). Glycation and advanced glycation endproducts. The Glycation Hypothesis of

Atherosclerosis. C. Colaco: 57‐87.

Wetlaufer, D. B. (1973). ʺNucleation, rapid folding, and globular intrachain regions in proteins.ʺ Proc Natl Acad Sci U S A

70(3): 697‐701.

Whisstock, J. C. and A. M. Lesk (2003). ʺPrediction of protein function from protein sequence and structure.ʺ Q Rev Biophys

36(3): 307‐340.

Whitmore, L. and B. A. Wallace (2004). ʺDICHROWEB, an online server for protein secondary structure analyses from

circular dichroism spectroscopic data.ʺ Nucleic Acids Res 32(Web Server issue): W668‐673.

Wilhelm, K. R., K. Yanamandra, M. A. Gruden, V. Zamotin, M. Malisauskas, V. Casaite, A. Darinskas, L. Forsgren and L.

A. Morozova‐Roche (2007). ʺImmune reactivity towards insulin, its amyloid and protein S100B in blood sera of

Parkinsonʹs disease patients.ʺ Eur J Neurol 14(3): 327‐334.

 172

References

Williams, M. D. and J. L. Nadler (2007). ʺInflammatory mechanisms of diabetic complications.ʺ Curr Diab Rep 7(3): 242‐248.

Wilson, C. M. (1979). ʺStudies and critique of Amido Black 10B, Coomassie Blue R, and Fast Green FCF as stains for proteins

after polyacrylamide gel electrophoresis.ʺ Anal Biochem 96(2): 263‐278.

Wolfenden, R. (2007). ʺExperimental measures of amino acid hydrophobicity and the topology of transmembrane and

globular proteins.ʺ J Gen Physiol 129(5): 357‐362.

Wolynes, P. G., J. N. Onuchic and D. Thirumalai (1995). ʺNavigating the folding routes.ʺ Science 267(5204): 1619‐1620.

Wood, J. D., T. P. Beaujeux and P. J. Shaw (2003). ʺProtein aggregation in motor neurone disorders.ʺ Neuropathol Appl

Neurobiol 29(6): 529‐545.

Wood, S. J., J. Wypych, S. Steavenson, J. C. Louis, M. Citron and A. L. Biere (1999). ʺalpha‐synuclein fibrillogenesis is

nucleation‐dependent. Implications for the pathogenesis of Parkinsonʹs disease.ʺ J Biol Chem 274(28): 19509‐19512.

Wright, C. F., K. Lindorff‐Larsen, L. G. Randles and J. Clarke (2003). ʺParallel protein‐unfolding pathways revealed and

mapped.ʺ Nat Struct Biol 10(8): 658‐662.

Wright, P. E. and H. J. Dyson (1999). ʺIntrinsically unstructured proteins: re‐assessing the protein structure‐function

paradigm.ʺ J Mol Biol 293(2): 321‐331.

Xu, S. (2007). ʺAggregation drives ʺmisfoldingʺ in protein amyloid fiber formation.ʺ Amyloid 14(2): 119‐131.

Yamamoto, S., K. Hasegawa, I. Yamaguchi, S. Tsutsumi, J. Kardos, Y. Goto, F. Gejyo and H. Naiki (2004). ʺLow

concentrations of sodium dodecyl sulfate induce the extension of beta 2‐microglobulin‐related amyloid fibrils at a

neutral pH.ʺ Biochemistry 43(34): 11075‐11082.

Yamin, G., C. B. Glaser, V. N. Uversky and A. L. Fink (2003a). ʺCertain metals trigger fibrillation of methionine‐oxidized

alpha‐synuclein.ʺ J Biol Chem 278(30): 27630‐27635.

Yamin, G., V. N. Uversky and A. L. Fink (2003b). ʺNitration inhibits fibrillation of human alpha‐synuclein in vitro by

formation of soluble oligomers.ʺ FEBS Lett 542(1‐3): 147‐152.

Yan, S. D., X. Chen, A. M. Schmidt, J. Brett, G. Godman, Y. S. Zou, C. W. Scott, C. Caputo, T. Frappier, M. A. Smith and et

al. (1994). ʺGlycated tau protein in Alzheimer disease: a mechanism for induction of oxidant stress.ʺ Proc Natl

Acad Sci U S A 91(16): 7787‐7791.

Yan, S. D., S. F. Yan, X. Chen, J. Fu, M. Chen, P. Kuppusamy, M. A. Smith, G. Perry, G. C. Godman, P. Nawroth and et al.

(1995). ʺNon‐enzymatically glycated tau in Alzheimerʹs disease induces neuronal oxidant stress resulting in

cytokine gene expression and release of amyloid beta‐peptide.ʺ Nat Med 1(7): 693‐699.

Yang, J., X. Liu, K. Bhalla, C. N. Kim, A. M. Ibrado, J. Cai, T. I. Peng, D. P. Jones and X. Wang (1997). ʺPrevention of

apoptosis by Bcl‐2: release of cytochrome c from mitochondria blocked.ʺ Science 275(5303): 1129‐1132.

Yang, J. S., W. W. Chen, J. Skolnick and E. I. Shakhnovich (2007). ʺAll‐atom ab initio folding of a diverse set of proteins.ʺ

Structure 15(1): 53‐63.

Yazaki, M., S. A. Farrell and M. D. Benson (2003). ʺA novel lysozyme mutation Phe57Ile associated with hereditary renal

amyloidosis.ʺ Kidney Int 63(5): 1652‐1657.  173

References 

Yoritaka, A., N. Hattori, K. Uchida, M. Tanaka, E. R. Stadtman and Y. Mizuno (1996). ʺImmunohistochemical detection of

4‐hydroxynonenal protein adducts in Parkinson disease.ʺ Proc Natl Acad Sci U S A 93(7): 2696‐2701.

Yu, N. T., B. H. Jo, R. C. Chang and J. D. Huber (1974). ʺSingle‐crystal Raman spectra of native insulin. Structures of insulin

fibrils, glucagon fibrils, and intact calf lens.ʺ Arch Biochem Biophys 160(2): 614‐622.

Zarranz, J. J., J. Alegre, J. C. Gomez‐Esteban, E. Lezcano, R. Ros, I. Ampuero, L. Vidal, J. Hoenicka, O. Rodriguez, B.

Atares, V. Llorens, E. Gomez Tortosa, T. del Ser, D. G. Munoz and J. G. de Yebenes (2004). ʺThe new mutation,

E46K, of alpha‐synuclein causes Parkinson and Lewy body dementia.ʺ Ann Neurol 55(2): 164‐173.

Zerovnik, E. (2002). ʺAmyloid‐fibril formation. Proposed mechanisms and relevance to conformational disease.ʺ Eur J

Biochem 269(14): 3362‐3371.

Zhou, N. E., C. M. Kay and R. S. Hodges (1992). ʺSynthetic model proteins: the relative contribution of leucine residues at

the nonequivalent positions of the 3‐4 hydrophobic repeat to the stability of the two‐stranded alpha‐helical coiled‐

coil.ʺ Biochemistry 31(25): 5739‐5746.

Zhou, Y., Q. Wang, B. Mark Evers and D. H. Chung (2006). ʺOxidative stress‐induced intestinal epithelial cell apoptosis is

mediated by p38 MAPK.ʺ Biochem Biophys Res Commun 350(4): 860‐865.

Zhu, X., B. Su, X. Wang, M. A. Smith and G. Perry (2007). ʺCauses of oxidative stress in Alzheimer disease.ʺ Cell Mol Life

Sci 64(17): 2202‐2210.

Zierath, J. R., A. Krook and H. Wallberg‐Henriksson (2000). ʺInsulin action and insulin resistance in human skeletal

muscle.ʺ Diabetologia 43(7): 821‐835.

 174