Mass Spectrometry -Based Identification and Quantification of Posttranslational Modification Citrullination in Rheumatoid Arthritis

Downloaded from orbit.dtu.dk on: Oct 08, 2021

Mass spectrometry -based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis

Sharma, Mandvi

Publication date: 2019

Document Version Publisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA): Sharma, M. (2019). Mass spectrometry -based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis. Technical University of Denmark.

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

 Users may download and print one copy of any publication from the public portal for the purpose of private study or research.  You may not further distribute the material or use it for any profit-making activity or commercial gain  You may freely distribute the URL identifying the publication in the public portal

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Technical University of Denmark

Department of Biotechnology and Biomedicine

PhD Thesis

Mandvi Sharma

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in

rheumatoid arthritis

Academic supervisor:

Professor Birte Svensson

Technical University of Denmark, Bioengineering

Academic co-supervisor:

Associate Professor Per Hägglund

Technical University of Denmark, Bioengineering

Academic co-supervisor:

Head of Rheumatology Anne-Christine Bay-Jensen

Nordic Bioscience A/S

Submitted June 15th 2017

II PREFACE

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis III

Table of Contents

PREFACE ...... VI

ACKNOWLEDGEMENTS ...... VIII

ABBREVIATIONS ...... X

ABSTRACT ...... XIII

DANSK RESUME ...... XV

PUBLICATIONS ...... XVII

AIM ...... XIX

INTRODUCTION ...... 20 1. PROTEINS AND THE POST-TRANSLATIONAL MODIFICATION “CITRULLINATION” ...... 20 2. WHAT IS CITRULLINATION? ...... 21 2.1 PAD family of enzymes ...... 22 2.2 Role of citrullination in health and diseases ...... 25 2.3 Role of citrullination in Rheumatoid arthritis...... 27 2.4 Effect of smoking in RA ...... 28 2.5 Role of citrullination in other diseases ...... 31 3. MASS SPECTROMETRY ...... 31 3.1 Ion sources ...... 32 3.2 Mass analyzers ...... 34 3.3 Fragmentation ...... 37 3.4 Detectors...... 39 3.5 Mass spectrometry workflow ...... 40 3.6 Sample preparation ...... 40 3.7 Sample fractionation ...... 42 4. DETECTION OF CITRULLINATION ...... 43 4.1 Mass spectrometry-based identification of citrullination ...... 45 5. DATA ANALYSIS IN CITRULLINATION AND ESTIMATED RELATIVE DEGREE OF CITRULLINATION ...... 47 6. IDENTIFICATION OF CITRULLINATION SITES AND THEIR ESTIMATED RELATIVE DEGREE OF CITRULLINATION ...... 49 7. SUMMARY OF PAPERS ...... 51 7.1 PAPER 1 ...... 51 7.2 PAPER 2 ...... 52 8. CONCLUSIONS AND PERSPECTIVES ...... 53

9. REFERENCES...... 56

10 MANUSCRIPTS ...... 64

IV PREFACE

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis V

VI PREFACE

Preface

This thesis concludes my work as a Ph.D. student at DTU Bioengineering. This work was carried out from December 2014 to June 2019 under the supervision of Per Hägglund, Birte Svensson and Anne-Christine Bay-Jensen. The Ph.D. fellowship was funded by the Danish Research Foundation, Nordic Biosciences A/S and the Technical University of Denmark. The Q-Exactive Orbitrap mass spectrometers used in this study were granted by the Danish Council for Independent Research | Natural Sciences (grant number 11-106246) and the Velux foundation.

Kongens Lyngby, June 2019 Mandvi Sharma

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis VII

VIII ACKNOWLEDGEMENTS

Acknowledgements

As a Ph.D. student, I have received help and advice from a number of people. I would like to use this opportunity to express my gratitude to them. Foremost, I would like to thank Associate Professor Per Hägglund for his continuous support and great supervision through these years. I really appreciate the guidance, constant encouragement and invaluable support and patience despite all the difficult times. I am really grateful. I would also like to express my gratitude to Professor Birte Svensson for taking over the supervision and for the support and encouragement, especially during the final year. I would further like to thank Anne-Christine Bay-Jensen for being a co- supervisor. I would like to thank Professor Claus. H. Nielsen, for the fruitful collaboration, guidance, support, interesting ideas and discussions on rheumatoid arthritis. I would also like to thank Dres Damgaard, for his great help and support throughout. Dr. Ladislav Senolt is thanked for providing the samples. I would also like to thank my present and former fellow colleagues at DTU, especially Michele for all the fun times, encouragement and discussions; Nicklas, Martin, Roall, Sita, Zacharias for happy lunch hours, good company and interesting discussions. Lene Blicher, Erwin Schoof and Anne Blicher are thanked for helping me in the laboratory. A special thanks to my family, both my parents and in-laws for their endless support and always lending me a helping hand in dire moments. I would also like to thanks my sister, brother and my two wonderful sisters in law for all great advice and fun moments. I would also like to thank Dr. Geeta and her team for helping me through a very tough time. Finally, thanks to my son Mivaan. It has been wonderful to come home to a lovely little guy, who is always waiting with a big smile, full of energy and curiosity and for taking my mind off my Ph.D. Big thanks to my husband Dr. Kapil Triguna for constant support and encouragement and always believing in me. I couldn’t have done it without your support and love.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis IX

X ABBREVIATIONS

Abbreviations

Anti-citrullinated protein antibodies ACPA Antigen presenting cells APC Bronchoalveolar lavage fluid BALF Central nervous system CNS Collision-induced dissociation CID Common amino acid CAA Data-dependent acquisition mode DDA Electron transfer dissociation ETD Electrospray ionization ESI False discovery rate FDR Glial fibrillary acidic protein GFAP High collision dissociation HCD Human leucocyte antigen HLA Human serum albumin HSA Immunoglobulin gamma IgG Liquid chromatography LC Mass spectrometry MS Mass-to-charge ratios m/z Matrix-assisted laser desorption MALDI ionization Myosin basic protein MBP Neutrophil extracellular traps NETs Peptide mass fingerprinting PMF Peptide-spectrum match PSM Peptidyl arginine deiminase PAD Posttranslational modifications PTMs Rheumatoid arthritis RA Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis XI

Rheumatoid factor RF Shared epitope SE Synovial fluid SF T cell receptor TCR Time-of-flight TOF Cyclic citrullinated peptide CCP

XII

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis XIII

Abstract

Post-translational modifications (PTMs) of proteins and peptides modulate their behavior and are frequently implicated to physiological functions in a normal state, and to human diseases in aberrant states. Hence, PTMs, in general, are routinely tracked as disease markers or molecular targets for developing target-specific therapies. The PTM citrullination and rheumatoid arthritis (RA) are frequently interlinked processes. Citrullination is an inflammation-dependent phenomenon, so it is not surprising that this modification underlies many autoimmune diseases. Analysis of the biological effects of these citrullinated proteins has started to reveal the true pathological roles of this particular PTM and looks promising in revolutionizing the treatment of this autoimmune disease. Various studies have revealed the presence of many citrullinated proteins in the synovial joints of the disease-active patients. Despite being citrullinated, not all the citrullinated proteins respond similarly, but on the contrary, have distinct roles to the same stimulus. However, confident identification of citrullination is quite challenging. Hence, high- resolution mass spectrometry-based identification and analysis are required for sensitive and specific detection of PTMs such as citrullination. Here, this thesis study was done to develop a strategy for the identification and relative quantification of citrullination in complex biological samples from RA patients. In manuscript 1, identification and relative quantification of citrullinated peptides from fibrinogen have been performed in samples from four disease-active patients showing different degrees of inflammation and other disease characteristics. In order to gauge the feasibility to identify the citrullination sites in vivo, we also attempted to identify the corresponding citrullination sites in fibrinogen in vitro after treatment with human recombinant peptidylarginine deiminase 2 (PAD2). Generally, high disease-activity patients exhibited an increased number of identified citrullination sites and higher relative degree of citrullination. Twenty-three citrullination sites were identified in in vivo samples, of which 14 have not been reported in previous studies. Sites α84, α123, α129, α547, α573, α591, β334 and γ134 were identified in more than one disease-active patient, therefore these sites were regarded as hotspots. Overall, 48 citrullination sites were identified in this study following citrullination of fibrinogen in vitro using PAD2, of which six citrullination sites were found to be unreported before. Twenty-one out of the 23 citrullination sites

XIV ABSTRACT

identified in vivo were also detected in vitro, supporting the validity of the identifications. In manuscript 2, identification and quantification of citrullination sites were extended to the entire synovial fluid citrullinome of the same four RA patients. This study was done to find out which proteins are present in citrullinated form and which proteins were citrullinated at high levels and at which sites the citrullination was observed. Overall, 171 citrullinated proteins of which 48 were found to be reported in previous studies and 276 citrullination sites were identified in disease-active synovial fluid (SF) samples. Thirteen proteins were found common in all patient samples, which may be of immunological relevance and may behave as autoantigens in RA. Generally, the citrullination sites and citrullination occupancy was higher in patients with high disease activity, but some peptides displayed higher citrullination occupancy in the patient with the least disease activity than the patients with moderate to high disease activity. It was observed that we find higher citrulline occupancy in several sites of SF4 as compared to SF2 and SF3. SF4 displayed a low leukocyte count and the lowest disease activity. Nevertheless, our sample size was too small to determine if high ratios correlate with various disease parameters or not.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis XV

Dansk resume

Post-translationelle modifikationer (PTM'er) af proteiner og peptider modulerer deres egenskaber og er ofte essentielle for fysiologiske funktioner i en normal tilstand, men kan også være involveret i patogenesen af sygdomme. Derfor bliver PTM'er generelt rutinemæssigt brugt som sygdomsmarkører eller molekylære mål for udvikling af sygdomsspecifikke drugs. En af disse modifikationer er citrullinering som man mener spiller en patogenetisk rolle i leddegigt. Denne PTM, af protein-bundet arginin til protein- bundet citrullin, er katalyseret af enzym familien, Peptidylarginine deiminase. Citrullering er et betændelsesafhængigt fænomen, så det er ikke overraskende, at denne ændring kan være involveret i en autoimmun sygdom. Analyse af de biologiske effekter af disse citrullinerede proteiner er begyndt at afsløre de egentlige patologiske roller i leddegigt, og ser lovende ud i at revolutionere behandlingen af denne autoimmune sygdom. Forskellige undersøgelser har afsløret forekomsten af mange citrullinerede proteiner i leddene hos de sygdomsaktive patienter. En stor udfordring på dette forskningsområde er identifikation af citrullinerede proteiner. Derfor kræves højopløsningsmasspektrometri-baseret identifikation og analyse til følsom og specifik påvisning af PTM'er, såsom citrullinering. Denne afhandling blev udført for at udvikle en strategi for identifikation og relativ kvantificering af citrullinering i komplekse biologiske prøver fra RA patienter. I manuskript 1 er identifikation og relativ kvantificering af citrullinerede peptider fra fibrinogen blevet udført i ledvæske prøver fra fire sygdomsaktive patienter, der viser forskellige grader af inflammation og andre sygdomsparametre. For at måle muligheden for at identificere citrullineringssteder in vivo forsøgte vi også at identificere de tilsvarende citrullineringssteder i fibrinogen in vitro efter behandling med humant rekombinant peptidylarginin deiminase 2 (PAD2). Generelt udviste patienter med høj sygdomsaktivitet et forøget antal identificerede citrullineringssteder og en højere relativ grad af citrullinering. Treogtredive citrullineringssteder blev identificeret i in vivo prøver, hvoraf 14 ikke er blevet rapporteret i tidligere undersøgelser. Citrullineringsstederne α84, α123, α129, α547, α573, α591, β334 og γ134 blev identificeret i mere end en sygdomsaktiv patient, og derfor blev disse steder betragtet som hotspots. Samlet set blev 48 citrullineringssteder identificeret i dette studie efter citrullinering af fibrinogen in vitro med anvendelse af PAD2, hvoraf der blev fundet seks citrullineringssteder, der tidligere ikke

XVI DANSK RESUME

var rapporteret. Enogtyve ud af de 23 citrullineringssteder identificeret in vivo blev også påvist in vitro, hvilket understøtter identitetens gyldighed. I manuskript 2 blev identifikation og kvantificering af citrullineringssteder udvidet til hele ledvæske citrullinomet fra de samme fire leddegigt patienter. Denne undersøgelse blev udført for at finde ud af, hvilke proteiner der er til stede i citrullineret form, og hvilke proteiner der blev citrullineret i høj grad og på hvilke steder citrullineringen finder sted. Samlet set blev 171 citrullinerede proteiner, hvoraf 48 blev rapporteret i tidligere undersøgelser, og 276 citrullinerede positioner identificeret i sygdomsaktive synovialvæske (SF) prøver. Tretten citrullinerede proteiner blev fundet i alle fire patientprøver, og kan være af særlig patogenetisk relevans i RA. Generelt var citrullineringsgraden and number citrullinerings positioner højere hos patienter med høj sygdoms aktivitet i forhold til patienter med lav sygdoms aktivitet, men nogle peptider viste højere citrullineringsgrad hos patienten med laveste sygdomsaktivitet. Det blev observeret, at vi finder højere citrullineringsgrad på flere steder i SF4 sammenlignet med SF2 og SF3. SF4 viste et lavt leukocyttal og den laveste sygdomsaktivitet. Ikke desto mindre var vores prøvemateriale for lille til at bestemme, om citrullineringsgraden korrelerer med forskellige sygdomsparametre eller ej.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis XVII

Publications

Manuscripts included in this thesis:

Expanding the citrullinome of synovial fibrinogen from rheumatoid arthritis patients Mandvi Sharma, Dres Damgaard, Ladislav Senolt, Birte Svensson, Anne-Christine Bay- Jensen, Claus Henrik Nielsen, Per Hägglund (Manuscript Submitted: Journal of proteomics)

Assessment of the citrullinome and relative degree of citrullination of proteins in synovial fluid from patients with rheumatoid arthritis Mandvi Sharma, Claus Henrik Nielsen, Ladislav Senolt, Birte Svensson, Per Hägglund, Dres Damgaard (Manuscript in preparation)

Manuscripts not included in this thesis:

Identification of citrullination sites in bronchoalveolar lavage fluid from smoking and non-smoking COPD patients Mandvi Sharma, Dres Damgaard, Per Hägglund, Claus Henrik Nielsen (Manuscript in preparation)

XVIII AIM

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis XIX

Aim

The work presented in this thesis focuses on the development of mass spectrometry-based strategies for identification of citrullination sites and quantification of the estimated relative degree of citrullination at the site of modification in complex RA patient samples.

20 INTRODUCTION

Introduction

1. Proteins and the post-translational modification “citrullination”

Proteins are the most diverse macromolecules which act as biochemical machinery of life, performing a vast array of functions in necessarily all biological processes. Cells which are the structural, functional and biological unit of living organisms, function and communicate using proteins. Every single cell contains a variety of proteins, each with an exclusive function. Building blocks or monomers of proteins are called amino acids. There are 20 different types of amino acids which are commonly found in proteins however, there are more than 1,000,000 functional proteins [1]. Proteins are formed by protein synthesis (Figure 1).

Transcription is the first step in gene expression in which gene information is transcribed to make an RNA copy of a gene's DNA sequence. The information on mRNA is then decoded and translated to form a protein. This diversity in the human proteome is increase by posttranslational modifications (PTMs). These modifications can occur at any time during the lifespan of a protein and have a significant impact on biology. PTMs after protein translation considerably increase the variety of proteins present in living cells. These modifications can be either stable or temporary and may result from either targeted, enzymatically catalyzed reactions or spontaneous chemical reactions in the cell [2]. More than 300 types of PTMs have been identified which can influence the interaction between proteins carrying them [1,3]. Nearly all translated proteins undergo PTMs, which is crucial since these modifications may alter the chemical or physical properties, such as protein localization to cell compartments, folding into proper conformation, the stability of the nascent protein, activation/deactivation and subsequently the function of the proteins.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 21

Figure 1: Central dogma. Information on DNA is rewritten to make a messenger RNA copy, which is translated to build a specific protein containing a unique combination of amino acids.

2. What is citrullination?

The PTM citrullination also known as deimination is an enzymatic conversion of peptidyl-arginine to peptidyl-citrulline in the presence of Ca2+ dependent enzymes peptidyl arginine deiminase (PAD) of which five isoforms, PAD1–4 and PAD6 exist in humans [4]. PAD converts the guanidine group of peptidyl-arginine residues to an ureido group, resulting in loss of a positive charge and small increase in molecular mass by +0.98 Da (Figure 2) [3]. Citrulline takes its name from Citrullus vulgaris, the Latin word for watermelon from which citrulline was first isolated in 1914 by Koga and Othake [5]. The loss of positive charge on the conversion of arginine to citrulline can have major effects on protein structure, proteolytic degradation, intermolecular and intermolecular interactions. For example apoptotic citrullination of vimentin where the loss of a positive charge of the protein may result in failure of intramolecular and intermolecular interactions leading to structural de-stability and disruption of the cytoskeletal network. Filaggrin, an epidermal protein maintains epidermal hydration. Its highly basic nature helps it to form a tight array with negatively charged keratin IF. Citrullination of filaggrin (decrease in net positive charge) results in proteolytic processing leading to the development of moisturizing capacity [6–8]. PTM of arginine such as phosphorylation, methylation, ADP-ribosylation, glycation, carbonylation, and citrullination is unsurprisingly an important modification since it modulates several cellular processes considering its unique guanidinium group which can form hydrogen bonds with both nucleic acids and proteins [9]. It is therefore not unexpected that citrullination and its involving enzymes (PADs) can be associated with various diseases.

Figure 2. Citrullination. Peptidyl bound arginine is converted into peptidyl citrulline in the presence of calcium- dependent enzyme peptidyl arginine deiminase.

2.1 PAD family of enzymes

PADs catalyze the conversion of arginine residues in proteins to nonstandard amino acid citrulline. PAD was first identified in 1977 by Rogers and Taylor [10]. Since then five isoforms, PAD1–4 and PAD6 have been found in humans. In eukaryotic species, such as the mouse, chicken, frog, and bony fish they show extensive mutual sequence homologies (70–95% identical amino acids) but in parallel also show a very stringent tissue-specific expression [11–13]. In prokaryotes, PAD activity has only been described in Porphyromonas gingivalis, a bacterium causing periodontitis, a chronic inflammation of gums in the oral cavity [14].

The PAD enzyme family are Ca2+ dependent enzymes, requires specifically high Ca2+ concentrations for their activity. The pH optimum of PAD activity is pH 7.6 [15–17]. A concentration of 40 – 60 μM Ca2+ is required for half-maximum activity of PAD enzyme [18]. The intracellular concentration of Ca2+ is ∼1 µM in activated cells and ∼200 nM in resting cells and the concentration of Ca2+ is about 1.2 mM extracellularly considering that the plasma membrane is virtually impermeable for Ca2+ hence, this high concentration of Ca2+ seems physiologically relevant [17,19,20]. Therefore, it is suggested that cellular apoptosis and some other accompanying effects expose PAD proteins to high levels of extracellular Ca2+, which induce PAD activity [13]. Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 23

However, very little is known about the physiological trigger that initiates this calcium-dependent PAD enzyme activity.

PAD2 isoform is predicted to be the ancestral homolog of the PADs [15]. PAD2 needs at least three calcium ions for activation and uses a substrate-assisted mechanism of catalysis in which the positive charge of the substrate suppresses the pKa of Cys647 which in turn activates the enzyme. The PAD2 active site contains Cys647 and His471 in addition to Asp345 and Asp374 which position the guanidinium group for nucleophilic attack by the active site Cys647. The catalytic mechanism of PAD2 has three steps (Figure 3). Initially, the active site thiolate (Cys647) attacks the guanidinium carbon of arginine. Here, His471 acts as a general acid which protonates the guanidinium group with concomitant electrostatic stabilization by Asp351 and Asp473. As ammonia is released the newly formed S-alkyl tetrahedral intermediate collapses. Lastly, His471 activates water for the nucleophilic attack, leading to the formation of a second tetrahedral intermediate, which eventually collapses to form the citrullinated product [15].

Figure 3- Proposed catalytic mechanism of PAD2. Firstly, the active site thiolate, Cys647, attacks the guanidinum group carbon of arginine. Here, His471 acts as a general acid, protonating the guanidinium group with concomitant electrostatic stabilization by Asp351 and Asp473. The newly formed S-alkyl tetrahedral intermediate collapses when ammonia leaves. Thirdly, water is activated for nucleophilic attack by His471, which now acts as a general base. Attack of the water molecule results in the formation of the second tetrahedral intermediate, which ultimately collapses to form the citrullinated product (modified from [15].

Significant work has been done to explain the structure of PAD4 with or without Ca2+, and with substrates or with inhibitors. It has been pointed out that PAD4 has distinct N- and C-terminal domains containing five calcium binding sites in total, two binding sites in the C-terminal domain and three in the N- terminal domain. In the presence of Ca2+ residues 158–171 of the N-terminal domain show a very organized behavior and form a highly ordered α-helix however, on the contrary without Ca2+, 158–171 residues are highly disorganized. Therefore, this conformational change with change in Ca2+ controls the protein-protein interactions with the enzyme [21]. Whereas in the C- terminal domain, active site cleft is generated and Cys645 is placed in the correct position for nucleophilic attack on the bound substrate upon Ca2+ binding [22]. The cys645 residue which behaves as an active site nucleophile is present as a thiolate in the active form of the enzyme. His471 is another important active site residue, acting as a general acid/base, promotes the initial removal of ammonia from the substrate guanidinium group and eventually helping in complete substrate hydrolysis.

Each PAD family member shows a very inflexible tissue-specific expression. They play many roles in normal physiology but, on the other hand, are also found to be associated with various diseases. Their specific roles in both cases are not known. The expression sites and functions of each of the human PAD enzymes are summarized in Table 1. Among the PAD family isoforms, PAD2 is widely expressed in various tissues such as brain, uterus, skeletal muscle, pancreas, spleen, secretory glands. PAD2 substrates include histone H3 and H4, myosin basic protein (MBP), glial fibrillary acidic protein (GFAP), vimentin β, fibrinogen and γ-actins [13]. PAD1 is expressed in uterus,epidermis with filaggrin and keratin K1 as reported substrates. PAD3 is expressed in keratinocytes, hair follicles and act on THH, filaggrin. PAD4 is expressed in white blood cells and target histones H2A, H3, H4, nucleophosmin, p300/CBP, nuclear lamin C, and ING4 as substrates. Finally, PAD6 is expressed in embryos and oocytes. Generallyall PAD isoforms are found in the cell cytoplasm, except PAD4 which is localized primarily in the nucleus and has been confirmed to stimulate histone deamination. PAD2, however may also play a regulatory role in histone citrullination [11–13].

Since the PTM citrullination can only be done by enzymatic modifications after protein translation, it is important to know more about the enzymes involved in this modification. This will not only Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 25

provide us with a broader understanding of their roles in normal physiology but will also give us an insight into its role in the pathogenesis of various diseases.

Table 1- Expression sites and function of human PADs [11].

Isoform Expression site Function

PAD1 Epidermis, uterus, hair follicles, Citrullination of filaggrin and keratin, facilitating proteolysis and crosslinking of arrector pili muscles, and sweat the proteins and contributing to skin cornification. Maintains hydration of stratum glands corneum and epidermis barrier function. Differentiation of hair follicles.

PAD2 Brain astrocytes, sweat glands, Citrullination of myelin basic protein in the brain and spinal cord, promoting arrector pili muscles, skeletal electrical insulation of myelin sheaths. Citrullination of vimentin in apoptotic muscle, spleen, macrophages, monocytes and macrophages. ERα-driven transcription regulation, female monocytes, histone h3 and h4, reproduction, innate immune defense. MBP,GFAP, actin epidermis, synovial tissue, and synovial fluid

PAD3 Upper layers of epidermis and hair Citrullination of trichohyalin, contributing to directional hair growth. follicles

PAD4 Hematopoietic cells, histones Citrullination of transcriptional coactivator p300 and histones H2A, H3, and H4, H2A, H3, H4, nucleophosmin, regulating gene expression by chromatin remodelling. Citrullination of fibrin, p300/CBP, nuclear lamin C, and contributing to chronic inflammation in rheumatoid arthritis. P53-dependent ING4 and inflamed rheumatoid citrullination of proteins following DNA damage, translocation of histone synovium chaperone nucleophosmin, and p53- mediated inhibition of tumor cell growth.

PAD6 Ovary, histone H4 and testis tissue Amino acids known to be conserved in PAD enzymatic activity are not conserved and peripheral blood leukocytes in PAD6. Female fertility, oocyte cytoskeletal sheet formation, early embryo development in mouse

2.2 Role of citrullination in health and diseases

The immune system is the body’s defense system comprising a complex network of biological molecules and structures that protects the body from diseases and foreign pathogens such as bacteria, and viruses. Therefore, it is inevitable that many pathologies arise from defects in the body’s defense system. In a healthy individual the immune system manages to discriminate between self and non-self. However, if our immune system fails to recognize its own body or self and begins “attacking” itself autoimmunity may occur. When the mechanism of discriminating the self vs non- self fails and immunological tolerance is compromised, the body makes autoantibodies and attacks itself. During the immune system development, the maturing B cells in the bone marrow and the

maturing T cells in the thymus are stringently going through the positive and negative selection to help them discriminate between self and non-self or foreign substance. A wide array of peripheral self-antigens are expressed in the thymus and bone marrow assisting in the positive and negative selection. Any maturing B or T lymphocyte is negatively selected or eliminated if found to be reactive with self-antigens. Nevertheless, the maturing B and T lymphocytes that escape the first check point are filtered in the peripheral organs, where proteins may be subjected to PTMs and affecting immune recognition leading to altered physical and chemical properties. In these peripheral organs, protein modifications may not take place in an exact manner as is in the thymus hence these modified proteins are never tolerated during lymphocyte maturation [23]. This results in changes in the physical and chemical properties of proteins such as the primary and tertiary structure or proteolytic degradation that is important in generating immunogenic or tolerogenic self- peptides. In other words, loss of tolerance pushes the body to make autoantibodies and attacks itself leading to an autoimmune response. However, not all PTMs induce immune responses but just induce B and T cell activation. Protein citrullination is an inflammation-dependent process which does not always result in serious diseases however it has also been found to be involved in various normal physiological processes. Rheumatoid arthritis (RA) is an autoimmune disorder, where immune complexes are formed towards proteins in which peptidyl bound arginine is posttranslationally modified into citrulline. Anti-citrullinated protein antibodies generated in the majority of RA patients are used for the clinical diagnosis of this autoimmune condition and are used as biomarkers of RA [24,25]. The exact role of the synovial citrullinome in the pathology and physiology of RA is still uncovered. Henceforth, in the following paragraph, we will discuss this PTM.

Citrullination plays an important role in gene regulation through a decrease in the histone Arg methylation and other modifications [12,26–28]. Histone citrullination by PAD4 causes chromatin decondensation, aiding neutrophil extracellular traps (NETs) formation and contributes to antibacterial function [29]. Deimination also has an important role in skin physiology. PAD1 ̶ 3 are found to be expressed in the epidermis [30]. Keratin K1, filaggrin, and trichohyalin are the main substrates for PAD1 [31]. Citrullination of trichohyalin in the hair follicle is important for the mechanical resistance of cells in the hair follicles [32]. Filaggrin's are another major target of PADs in the epidermis. Citrullination of fillagrin loosens the keratin/fillagrin complex, generate moisturizing factors and hydrates the upper epidermis [33]. In the central nervous system, elevated Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 27

levels of deimination in MBP of the brain under 2 years of age makes it much more disorganized and less plastic. However, after 4 years of age citrullination levels of the brain MBP reduce dramatically making it more organized, tight, and plastic [34]. Citrullination of chemokines (chemoattractant biomolecules) reduces the influx of neutrophils during inflammation and causes impaired T cell activation [35].

2.3 Role of citrullination in Rheumatoid arthritis

RA is a chronic, multifactorial autoimmune disease characterized by sustained inflammation of the synovium surrounding the joints, causing bone and cartilage destruction, leading to joint deformation [36]. RA patients can be seen showing clinically positive (in most cases) and clinically negative (in few cases) characteristics. The presence of autoantibodies makes a disease autoimmune in nature. Rheumatoid factor (RF) is the one such antibody which is found associated with autoimmune disease. However, it’s presence is not only in RA patients and can also be seen in patients with other autoimmune diseases, with some infections and is also found in healthy individuals. A distinguishing characteristic of this autoimmune disorder is the production of anti- citrullinated protein antibodies (ACPA), which can be detected prior to the onset of disease in nearly all patients and are definitely much better indicator than RF [24,25,37]. The detection of ACPA in RA patient can be done today using various commercially available assays. Most commonly used is the anti-cyclic citrullinated peptide (CCP) assay which shows good sensitivity for the detection ACPA [38]. Abundant citrullinated proteins are present in the synovium of the inflamed joints [39–44]. Inciting an immune response to citrullinated proteins or formation of immune complexes plays an important role in the pathophysiology of this autoimmune disease. Also, hypercitrullination in the oral cavity caused by PAD enzymes from the bacterium Porphyromonas gingivalis during periodontitis is thought to be associated with RA [45,46]. This oral colonization was seen more frequently in untreated RA patients [47]. However, the presence of citrullinated proteins is not an exclusive feature of RA as these proteins can also be seen in various inflammation sites, perhaps the presence of certain specific structural features on these proteins make them ready for ACPA binding. When citrullinated proteins form citrullinated protein-specific immune complexes, it is only then that they become arthritogenic. However, it is still important to

have a complete overview of the citrullinated proteins in the inflamed synovium of RA patients, to have a better and deeper understanding of the disease physiology.

RA is a multifactorial disease associated with genetic risk factors and environmental factors. Genetic risk factors account for approximately 60% of the variation in liability to disease [48]. The extremely crucial and in fact the largest genetic risk factor for RA is found within the human leucocyte antigen (HLA) class II region, encoding HLA-DRB1 molecules [49]. In 1981, Klareskog et al. proposed that the presence of HLA-DR expressing antigen-presenting cells in synovial tissue and their interactions with T cells was an important characteristic of this autoimmune disorder [50]. Henceforth, several other HLA-DRB1 allelic variants associated with the pathogenesis of the disease were shown in 1989 [51]. The specific alleles of HLA-DRB1 found to be associated with RA encode a conserved amino acid sequence known as the shared epitope (SE). It is apprehended that the presence of diverse alleles in the SE impact the communication between the HLA class II molecule and the T cell receptor (TCR), and could henceforth directly affect the efficiency of antigen presentation [52].

2.4 Effect of smoking in RA

Likewise, environmental factors play an important role in the development of RA. Various factors such as silica dust [53,54], asbestos dust [55], mineral oils [56], and smoking [57–59] can lead to the onset of the disease. However, studies have shown that smoking is the most influencing environmental factor associated with this autoimmune disorder [57–59]. Hence, after knowing that RA is triggered by both genetic and environmental factors, the next question that comes to our minds is how do they come into play and trigger the onset of this autoimmune disorder? PAD2 and PAD4, expressed in lungs have been identified in synovial tissue of RA patients in close association with inflamed tissues [60].In 2006 Klareskog et al. demonstrated the connection between the presence of the HLA-DRB1 SE alleles and smoking as a potential risk factor for the generation of ACPA-positive RA. Smokers carrying two copies of the HLA-DRB1 shared epitope alleles were at a very high risk of developing ACPA-positive RA in contrast to a nonsmoker, not carrying any HLA-DRB1 shared epitope alleles [61].

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 29

Altogether, it was hypothesized [13] (Figure 3) that smoking triggers higher expression of the PAD2 in lungs and causes higher apoptosis of bronchoalveolar lavage (BAL) cells [62]. This probably results in higher Ca2+ influx, activating PAD2 and subsequently citrullination of BAL cell proteins [63]. In genetically predisposed smokers carrying SEs, citrullination of protein causes enhanced antigen presentation and T cell activation [64]. Activated T cells probably migrate to regional lymph nodes and interact with B cells for the production of ACPAs. A second trauma such as an injury or an infection will cause inflammation in the joints, accumulation of inflammatory cells and eventually activation of PADs leads to the citrullination of synovial proteins. RA autoantigens interact with ACPAs and form immune complexes at the trauma sites (joints) [65]. RA specific immune complexes will further initiate production of pro-inflammatory cytokines [66] or stimulate T and B cells leading to the production of ACPA in the synovial compartment. Hence, a vicious cycle continuously producing ACPAs, pro-inflammatory cytokines, recruiting inflammatory cells into the joint, causing PAD activation resulting in further citrullination may lead to this autoimmune disorder called RA.

However, in order to achieve a comprehensive understanding of this autoimmune disorder (RA), it is essential to have an overview of citrullinated proteins and their sites of modification in the inflamed joints of RA patients. Furthermore, it is important to know which of these are targeted by the immune system to form immune complexes. Indeed it is still important to know whether complete inflamed synovial fluid protein and citrullination site profiling can facilitate in the development of patient-specific therapies or help in predicting the onset of clinical arthritis.

Figure 3. Hypothetical model showing pathogenesis of ACPA-positive RA. Citrullination of proteins in lungs of smokers with genetic risk factors increases antigen presentation, activation of T cells and B cell causing ACPA production. RA autoantigens interact with ACPAs to form immune complexes which further trigger the production of pro-inflammatory cytokines or activate T and B cells for ACPA production in the synovial compartment. Therefore, a vicious cycle of continuous production of ACPAs, pro-inflammatory cytokines, and PAD activation will lead to RA. = native protein, = citrullinated proteins, APC= antigen presenting cells

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 31

2.5 Role of citrullination in other diseases

Citrullination is found to be an important PTM in autoimmune diseases such as RA. However, there are various other diseases where citrullination results in pathological consequences. Deimidation plays an important role in the autoimmune skin disease psoriasis [67,68]. Some psoriatic patients with arthritis frequently show a polyarthritic pattern as seen in RA [69]. Several clinical studies have shown an important role of PAD (specially PAD4 and also possibly PAD2) enzymes in cancer [70,71]. The serum PAD4 levels dropped after the removal of tumor indicating high expression of PAD enzyme, thus citrullination plays a role in tumorigenesis and PAD4 may act as a clinical biomarker of this disease [72,73]. In patients with mammary tumors, loss of nuclear PAD2 expression may lead to uncontrolled multiplication, and tumor progression [74]. Multiple sclerosis[75,76] is another autoimmune neurodegenerative disease where MBP of the central nervous system (CNS) in multiple sclerosis patients had abnormally high levels of citrullination (45%) in comparison to the controls (18%) [34] leads to poor electrical conductivity of the neurons due to loss of the myelin sheath [4]. Elevated levels of PAD2 and PAD4 in myelin of the brain was seen in the multiple sclerosis patients, supporting the observed hypercitrullination in disease [77].

3. Mass spectrometry

Mass spectrometry (MS) is extensively used for the analysis of proteins and PTMs. Advances in the techniques and instrumentation have increased the possibility to do an in- depth study of proteome and various PTMs [78]. The mass spectrometer is a powerful analytical instrument which basically consists of at least three major components, each having a unique role: an ionization source, a mass analyzer, and a detector (Figure 4). First, the sample is introduced into the ion source where gas phase ions are generated. These ions are then directed towards the mass analyzer/mass filter where they are sorted depending on their mass-to-charge ratios (m/z). Finally, the ions are recorded by a detector and the instrument computer converts these signal to mass spectra that can be analyzed using suitable software. In some instruments, the mass analyzer also functions as detector.

Figure 4. Basic components of mass spectrometer.

3.1 Ion sources

Matrix-assisted laser desorption ionization (MALDI) In mass spectrometry, MALDI was introduced by Tanaka et al in 1988 [79]. It is a soft ionization technique, although some fragmentation can occur. The sample is mixed with a matrix and applied to a target plate (Figure 5A). The matrix is generally composed of a small aromatic acidic molecule that facilitates ionization by absorbing the laser energy and transferring it to the sample analytes. When the target plate is irradiated with a laser beam, the sample analytes and matrix molecules desorb and sublimates resulting in the formation of gas phase ions. The detailed mechanism of the ionization process in MALDI has not been established but it has been proposed that analyte ions are generated by the transfer of protons from the matrix to the analyte. These ions are then subjected to an electrostatic field and accelerated towards the analyzer. It is a pulsed ionization technique. It is robust, easy to use, gives a very little carryover, relatively tolerant to buffers and salts. It has been shown recently, that whole cells or tissues can be directly analyzed on the MALDI targets [80]. Nevertheless, due to its incompatibility with the on-line separation of peptides, the MALDI technique is not an ideal choice for the deep proteomics analysis.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 33

A. B.

Figure 5. Sample ionization. A. In MALDI sample embedded in the matrix is ionized when a laser beam hits the sampling spot. Energy absorbed by the matrix from the laser beam is transferred to the sample, resulting in the ionization of the analyte. B. In ESI, positively and negatively charged ions are generated when the solvent-sample mixture passes through the electrode tip. Subsequently, the solvent from the charged droplets evaporated, resulting in columbic fission if the charged droplets, forming charged gas ions. (modified from Agilent presentation).

Electrospray ionization (ESI) ESI is a soft ionization technique, developed by Fenn in 1989 [81]. In the ESI technique, high ion spray voltage is applied to the electrode tip of the emitter resulting in the formation of a taylor cone containing both positively or negatively charged ions depending of the mode in which machine is run. The taylor cone releases the charged droplets (Figure 5B). As the solvent molecules in the charged droplets evaporate, charges come closer together, until they reach the threshold of ionic repulsion resulting in columbic fission, formingcharged gas ions. Eventually, these charged gas ions enter the MS, through a heated capillary, which evaporated the remaining solvent molecules. Another ESI model assumes that during the electrospray process generated droplets contain only one analytical ion. This ion is released when the solvent evaporates [82]. ESI generate singly and multiply charged ions and works at atmospheric pressure. In contrast to MALDI, it generates ions continuously, making it a good choice for on-line coupling to chromatographic sample separation systems.

3.2 Mass analyzers

Mass analyzers are important components of mass spectrometry where ions are filtered and separated based on their m/z values. Various different types of mass analyzers are available today (Table 3) for mass spectrometry analysis.

Table 3: Different mass analyzers and their separation principle.

Mass analyzer Separation principle

Quadrupole Mass Analyzer m/z (stability of ion trajectory)

TOF (Time of Flight) Mass Kinetic energy and velocity of Analyzer ions

Ion Trap Mass Analyzers m/z (resonance frequency)

Magnetic Sector Mass Analyzer m/z (ions are separated in the magnetic field)

Electrostatic Sector Mass m/z (ions are separated in the Analyzer electrostatic field)

Ion Cyclotron Resonance (ICR) m/z (resonance frequency) Analyzer

Orbitrap m/z( Osscilating along the axis)

The time-of-flight (TOF) concept was introduced by William E. Stephens in 1946. However, significant improvements in mass spectrometry came with the introduction of reflectrons Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 35

by Mamyrin in 1972. Henceforth, from the time when first TOF instrument was available commercially, various improvements in the instrument have helped it to enhance its performance many folds. In TOF analyzer, ions of different m/z ratio are subjected to the same electric field potential and allowed to travel in a region of the constant electric field, where they travel with velocity depending on their m/z (Figure 6).

Figure 6. Time of flight mass analyzer. Ions of different m/z values are provided same kinetic energy, after which they travel in a drift region with constant electric field and vacuum. These ions travel through this region with a velocity depending on their size and hit the detector at different time points.

Alexander Makarov presented the first orbitrap mass analyzer in 1999, however, it is commercially present from 2005. Interaction of ions with an electrostatic field is the basis of ion separation in the orbitrap mass analyzer. Ions rotate around the central electrode as seen in Figure 7A and move back and forth along the axis with a frequency which is m/z dependent. Ions to be analyzed are first trapped in an ion accumulating compartment called C-trap. Ions leave the C-trap in the form of small ion packets and are directed tangentially toward the orbitrap (Figure 7B). The orbitrap consists of an inner electrode and an outer electrode. When the ion packets are injected into the obitrap the voltage of the inner electrode is ramped to induce a circular motion. However, only the oscillation of the ions along the axis is detected as an image current in the two halves of the outer electrode using an amplifier. These ions generate complex signals that are converted to mass spectra using a fourier transformation alogoritm. As the work done in this thesis was carried out using the Q-exactive orbitrap mass analyzer based mass spectrometer, only this type of instrument will be explained in detail.

The Q-exactive mass spectrometers have a stacked-ring ion guiding lens called S-lens, a quadrupole mass filter, an ion accumulation compartment called C-trap, a fragmentation cell called HCD cell and an orbitrap mass analyzer (Figure 8). Ions generated in the ESI ion source are directed by the S-lens into the quadrupole mass filter. The quadrupole is composed of two pairs of electrodes (rods), each pair receiving opposite radio frequency than the other which keeps altering very rapidly between the electrode pairs. When the ions reach this mass filter, ions only in a certain m/z range and stable trajectory are allowed to pass through the quadrupole while the ions outside this m/z range have unstable trajectories, hit the oscillating rods of the quadrupole and are neutralized. Ions that reach the other end of the quadrupole are accumulated in the C-trap and then sent to the orbitrap mass analyzer. As explained above orbitrap mass analyzer is an electrostatic ion trap where the recorded spectra are transformed into the readable signal using fourier transformation. Ramping voltage is applied to the inner electrode from 3-5 kV while the outer electrode is virtually grounded. When the ion packets reach the orbitrap mass analyzer they start rotating around the inner spile electrode in circular motions and also start oscillating along the axis back and forth. This back and forth oscillating motion is proportional to their m/z values which are recorded by the outer electrode. When this instrument is run on a data-dependent acquisition mode (DDA), it performs a full scan of peptides. To perform a full scan of peptides, ions that reach the other end of the electrode are accumulated in a C-trap and injected into the orbitrap. A full scan of peptides is carried out in this mass analyzer, where the oscillating ions are recorded to generate their m/z values. Subsequently, ions entering the quadrupole are then selected based on their abundance in the full scan and sent to the fragmentation cell called HCD cell. The ions fragmented in the HCD cell are used for the MS/MS analysis. These fragmented ions are sent to the C-trap for accumulation and injected into the orbitrap for the MS/MS scans. In the fragmentation cell, fragmentation is done using an inert gas such as nitrogen, which collides with the ions present in the cell resulting in their fragmentation. The information gathered from the precursor ion and the fragmentation spectra help in the unambiguous identification of peptides in the samples. This is a high resolution, high accuracy instrument, which can work in parallel acquisition mode where while the ions are measured in the orbitrap mass analyzer, another set of ions can be fragmented and stored. Hence, the optimized fast and sensitive acquition method can help in enhancing the Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 37

instrument capability to identify peptides, outperforming previous mass spectrometers [83].

Figure 7. A. Orbitrap consists of an inner electrode and an outer electrode. B. Ions injected in the orbitrap rotate around the inner electrode and oscillate back and forth along the axis. Modified from [84]

3.3 Fragmentation

As described above, after ions have been analysed in analysed in a MS scan, depending on their abundance the subsequent ions are selected in the mass filter and sent for fragmentation in the fragmentation cell. Fragmentation is the pre-requisite to generate an MS/MS scan. The MS/MS scan is important for correct peptide and modification site identification. There are a variety of fragmentation techniques such as collision-induced dissociation (CID), high collision dissociation (HCD), and electron transfer dissociation (ETD). Fragmentation of the ions can be done in a variety of ways. One such method is the use of inert gas for the fragmentation [85]. When the sample ions collide with the inert gas such as nitrogen, helium, during a collision transfer of kinetic energy increases the internal energy of the ions resulting in breakage of covalent bonds in the peptide. The

mobile proton model, showing the fragmentation patterns, demonstrates that the breakage of the amide bond is due to migration of the N-terminus proton down the peptide backbone [86]. In the CID fragmentation mode, ions of interest are stored and then subjected to the resonance frequency, which induces vibrations in the ions, forcing them to collide with the inert gas molecules. This type of fragmentation is relatively slow where the lowest energy pathway is taken for the fragmentation [85,87]. In some PTMs such as glycosylation and phosphorylation, CID fragmentation may result in the loss of the modification because the covalent bond between the PTM and the peptide is generally weak, hence the modification is lost during fragmentation.

Figure 8. (A) Ions in the gas phase enter the MS through stacked radio frequency lenses known as S-lens. When ions enter the MS which is at low pressure from the atmospheric pressure, drop in pressure prompts the ions to scatter, therefore S-lens focuses and compress the ions. (B) Curvature in the bent flatapole only allows the positively charged ions to move further, removes the uncharged molecules or solvent droplets. (C) All the positively charged ions are allowed to pass through the quadrupole during a full scan however, it acts as a mass filter allowing only specific ions to pass through it for the MS/MS scan. (D) Ions are then accumulated into the C-trap and are either sent into the (E) orbitrap for a full scan or into the (F) fragmentation/HCD cell for the fragmentation. In the HCD cells, selected ions are fragmented using an inert gas such as nitrogen. Fragmented ions are then stored into the C-trap and injected into the orbitrap for the MS/MS analysis. Modified from [84]

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 39

HCD fragmentation is similar to CID fragmentation in some ways such as the generation of b and y ions (Figure 10). HCD fragmentation does not have the low mass cut-off restriction, and with high mass accuracy fragmentation spectra, HCD has successfully provided more informative ion series.

Figure 9. Peptide fragmentation. a, b and c ions are generated from the N-terminus while x, y and z ions are generated from the C- terminus. The nomenclature is based on Roepstorff and Fohlman, 1984.

Electron transfer dissociation (ETD) involves a different fragmentation principle than CID and HCD. In ETD fragmentation is done by transferring electrons from an radical anion such as fluoranthene to a higher charge state (+3 and higher) multi protonated cationic peptide ions, resulting in formation of cationic radical species which undergo fragmentation mainly at the N-Cα peptide backbone bonds resulting in the formation of c and z ions (Figure 9) [88]. ETD fragmentation is mainly suitable for larger peptides carrying a high charge state, unlike CID where fragmentation readily occur in doubly charged ions [85]. This fragmentation breaks the peptide bond, retaining labile modifications such as glycosylation, hence can be used for studying such modifications [89].

3.4 Detectors

Ions generated in the full scan and after fragmentation are separated or sorted in a mass analyzer and then eventually recorded using a detector. Ions produce signals in detectors either when they hit the detector’s surface or by creating electric currents. Mass analyzers such as the orbitrap also functions as detector where the m/z ratio of the ions are determined by the frequency of oscillation

along the orbitrap axis. An Amplifier amplifies this complex signal, which with the help of fourier transformation is simplified and presented for analysis. Electron multipliers are another type of detectors which are used for example together with TOF analyzers.

3.5 Mass spectrometry workflow

Top-down and bottom-up are two main strategies for proteomic analysis using MS [90]. In the top- down strategy, intact protein ions are analyzed by fragmentation in the mass spectrometer, whereas in the bottom-up strategy of proteome analysis, proteins of interest are proteolytically digested before mass spectrometry analysis. However, in both strategies identification is dependent on the analysis of the mass spectrometry data [91,92]. The top-down strategy is used for the identification of PTMs and intact proteins however, separation of proteins in complex protein samples such as biological samples is quite challenging. In the bottom-up approach, there is a better separation of peptides and high sensitivity since peptides are more homogenous in their physiochemical properties and have less complex isotope envelops than proteins used in the top-down approach [93]. Also in the bottom-up strategy, identification of modification site is done using the information available for the modification such as mass shift and fragmentation spectra (MS/MS) [94].

3.6 Sample preparation

In proteomics, samples can be prepared in various ways depending on the aim of the study and the nature of the available samples. Sample complexity is one of the greatest concerns when it comes to biological samples. It can decrease the detection and identification ability of low abundance proteins by MS. Hence, it is essential to reduce the sample complexity in order to improve the detection of low abundance proteins. Biological samples such as body fluids generally contain a high concentration of albumin and immunoglobulin. Depletion method is often used to remove albumin and immunoglobulins from such samples to reduce their complexity. Nevertheless, the drawback with this method is that these abundant proteins often drag some low abundance proteins with them out of the sample. Protein solubilization is done after the depletion step. Hence, when the protein sample is ready for the analysis via a bottom-up approach, it is Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 41

solubilized/denaturated, reduced, alkylated and proteolytically digested before introduction to the separation system and the mass spectrometer (Figure 10).

Figure 10. Sample preparation. For the bottom-up approach, a complex protein sample is solubilized, reduced and alkylated before proteolytic digestion.

Protein solubilization Chaotropic agents or denaturants such as urea are normally used to solubilize proteins, depending on the source of protein (for instance cells, tissues, body fluids). However, many denaturants such as SDS, with good ionization capability but which can interfere with the MS analysis of peptides in the sample, need to be removed before the analysis. Hence, different approaches have been developed for the removal of denaturants such as the use of acid cleavable anionic detergents which can be degraded before the MS analysis [95].

Reduction and alkylation Followed by solubilization, the sample is reduced, to ensure efficient unfolding, as it reduces cysteine, by cleaving disulfide bonds. The resulting cysteines are then alkylated to prevent refolding of the proteins in the solution. DTT or TCEP are generally used as reducing agents, reducing the disulfide bridges. Alkylation is done using iodoacetamide to block free cysteines.

Proteolytic digestion In a bottom-up proteomics approach, proteins are digested before the analysis. Trypsin is the most widely used protease for the proteolytic digestion of protein before MS which 41

cleaves C-terminally to arginine and lysine with high specificity [96]. It produces small peptides with an average length of 14 amino acids and protonation at the N terminus and Arg/Lys side chain forming doubly and triply charged peptides making them favorable for MS analysis [97]. In some cases when the target protein does not contain a tryptic site or the tryptic site is modified several other proteases such as Glu-C, Lys-N, Asp-N, Arg-C can be used to get improved sequence coverage and increase the confidence of protein identifications [98]. Trypsin does not cleave peptides on the C-terminal side of arginine residues that have been converted to citrulline since the positive charge of the substrate needed for enzyme-catalyzed hydrolysis is lost resulting in miscleavages and formation of longer peptides. Isolated studies have reported some tryptically cleaved C-terminal citrulline peptides, but these results have not been validated by enzymatic assays [99]. Furthermore, comparison between trypsin-digested citrullinated and non-citrullinated peptides is complicated since the latter ones are shorter than the former ones. To avoid these problems the endoproteinase LysC can be used which peptides on the C-terminal side of lysine, independent of the modification on citrullinated arginine, however, the drawback of using LysC is the generation of long peptides with an average size of 20 residues [100]. Lastly, before the samples are introduced in the mass spectrometer, the samples are desalted using solid phase extraction such as stage tip [101] to remove buffers and salts.

3.7 Sample fractionation

Generally, in a bottom-up approach proteolytically digested peptides are separated using a liquid chromatography (LC) system which is coupled to MS. LC is commonly done using a reverse phase material such as C18 and the eluted peptides are directly focused on an ionization source. However, owing to the high complexity of the samples, this online sample fractionation is not sufficient to get the desired results thus arising a need to pre-fractionate the samples. Various pre-fractionation techniques have been developed such as SDS PAGE, isoelectric focusing, size exclusion chromatography which are often used to pre- fractionate the proteolytically digested samples.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 43

4. Detection of citrullination

The PTM citrullination results in the loss of one positive charge and a mass shift of +0.98 Da per converted residue. This very small change in mass makes their identification quite challenging. With the advances in mass spectrometry instruments with high resolution and sensitivity, identification of citrullinated proteins and their citrullination sites has become possible. Mass spectrometry solely or in combination with immunodetection or enrichment techniques has helped us to unveil many citrullinated proteins present at the site of interest of the RA patients. Various approaches have been developed to obtain good identification of this PTM. Tutturen et al. exhibited specific biotinylation and sensitive enrichment strategy for the enrichment of citrullinated peptides [102]. In this technique, citrullinated peptides were biotinylated using biotin-PEG-GBA (BPG), having 4-glyoxalbenzoic acid (GBA) group, which reacts specifically with the ureido group of the citrullinated peptides followed by enrichment of modified peptides using streptavidin beads.

Modified peptides were eluted off the streptavidin beads using excess free biotin and used for mass spectrometry analysis [102]. Another study presented by Kubota et al showed the identification of 18 citrullination sites using H2 O [103]. This study demonstrated a heavy isotope labeling technique 18 18 where peptides were citrullinated in 50% H2 O. Incorporation of O during citrullination in the peptide is displayed by a characteristic isotope distribution pattern (Figure 12).

Figure 11- Schematic view of the modification and enrichment of citrullinated peptides. I. Citrulline residues in a peptide mixture are modified by BPG during incubation at low pH. II. Excess BPG is removed by SCX and modified 43

citrullinated peptides are enriched using streptavidin beads. III. Modified citrullinated peptides are eluted off the streptavidin beads by using an excess of free biotin and non-modified peptides are washed off before being analyzed by MALDI-TOF/ TOF MS. Modified from [102].

18 Figure 12- Schematic view of the modification of a peptide citrullinated in 50% H2 O. During in vitro citrullination by 18 PAD H2O is incorporated while releasing NH3. Thus, the peptide citrullinated in H2 O gives a mass shift of +3 Da and 18 +1 Da when citrullined in H2O in the mass spectrum. The peptide citrullinated in 50% H2 O shows a distinct isotope distribution pattern. Modified from [103] .

The citrullinated peptides having 18O showed a mass shift of +3 Da and citrullinated peptides having 16O showed a mass shift of +1 Da in contrast to the un-modified peptide. Since this thesis work was performed using mass spectrometry-based identification, only this procedure will be discussed in detail here. An overview of the existing methods without derivatization is given in table 2.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 45

4.1 Mass spectrometry-based identification of citrullination

In 2009 Hao et al demonstrated that in CID fragmentation synthetic citrullinated peptide precursor ions exhibited abundant neutral loss of 43 Da. This was due to the exclusion of an HNCO moiety from citrulline ureido group. Therefore, for the characterization of citrullinated peptide, the sample was analyzed by tandem mass spectrometry. The generated CID spectrum was extracted for the multiply charged peptides [104].

Thereafter in another study, by Jin et al a combination of CID and HCD fragmentation was used to analyze complex samples for the identification of citrullination. The unique HNCO moiety loss feature in CID fragmentation mode was used as a signature ion which triggered another fragmentation of that peptide using HCD [105]. The HCD fragmentation spectrum contained low- mass region ions and sequence informative b and y ions, hence providing better peptide sequence coverage than seen in spectra generated by the CID fragmentation mode. However, since the signature ion was generated through CID fragmentation mode, an abundance of which was significantly reduced in HCD and better sequence coverage was the feature of the HCD fragmentation mode, both these features were taken advantage by combining these two fragmentation modes for the identification of citrullination sites both in in vitro and in vivo samples.

Hermansson et al. demonstrated another identification method based on the usage of accurate mass and retention time [106]. It was shown that after identifying the accurate mass and retention time difference of the synthetic citrullinated and non-citrullinated peptide, the same peptide can be identified in in- vivo samples. Citrullinated fibrinogen peptides were thus identified from complex RA synovial samples by using the accurate mass and retention time obtained from the in vitro non- citrullinated and citrullinated fibrinogen. However, since deamidation also results in the same mass shift as seen in citrullination (+0.98Da), peptides containing arginine and glutamine were excluded from the analysis. This approach limited the identification capacity since in vivo citrullination site identification was dependent on the in vitro citrullination site identification.

Tutturen et al. assessed the RA citrullinome from the depleted RA samples using direct LC- MS/MS analysis and an enrichment strategy using citrulline-specific biotinylation [107]. From MS

analysis, they showed frequencies of citrullinated peptides and the degree of citrullination of individual sites and also found a novel autocitrullination site in PAD4 [107].

Beers et al., separated the synovial fluid by SDS-PAGE followed by MS analysis for the identification of citrullinated peptides [39]. MS analysis of synovial fluid was done in another study by Romeo et al which used in solution digested unfractionated samples [108].

Wang et al. demonstrated the identification of citrullinated sites in another study using citrullinated proteins [44]. Characteristic neutral loss, using CID fragmentation mode was used for the identification of citrullination sites [44].

Lee et al. demonstrated mass spectrometry-based deep proteomic profiling for identification of citrullination in endogenous peptides [109]. Citrullinated and deamidated synthetic peptides were synthesized to develop a library of reference spectra. Identification of citrullination sites was done by using the manual spectrum interpretation, neutral loss of isocyanic acid, the immonium ion of citrulline, and reference spectra from synthetic citrullinated or deamidated peptides [109].

MS analysis of RA synovial fluid after removal of human serum albumin (HSA) and immunoglobulin gamma (IgG) was done for the identification of citrullinated peptides [107] (manuscript 1 and 2). Depletion of HSA and IgG was performed to identify citrullinated proteins of low abundance. In the third manuscript presented in this thesis, the MS analysis was done on BAL fluid after a simple fractionation step for the depletion of HSA, so as to identify low abundance citrullinated proteins (manuscript 3). Therefore, advancements in MS methodology have dramatically improved the instruments, giving high resolution and high data acquisition speed. MS- based analysis for the identification of citrullinated peptides and the citrullinated sites has been made possible owing to improved sensitivity and increased mass accuracy of these improved instruments.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 47

Table 2. Overview of different methods without modifications for identification of citrullination. Modified from [110]

Methods Advantages Disadvantages

MS-based identification without derivatisation

Accurate mass[39,44,107,109] Applicable on in vivo samples High-resolution equipment

Accurate mass and retention time [106] Applicable on in vivo samples Non-distinguishable from deamidation of N or Q

Time-consuming

No trypsin cleavage after citrulline [111– Applicable on in vivo samples Miscleavages 113]

Neutral loss of isocyanic acid (43 Da) Specific Specialized equipment after CID [104] Applicable on in vivo samples

Neutral loss of isocyanic acid and the Applicable on in vivo samples High-resolution equipment immonium ion [109]

5. Data analysis in citrullination and estimated relative degree of citrullination

Data analysis is a very crucial element of MS, which decodes the information in the raw spectra for the identification of proteins. Earlier mainly peptide mass fingerprinting (PMF) was used for protein identification. Basically, masses of the proteolytically digested peptides were compared with a list of in silico digested proteins. Henceforth, after the introduction of PMF, data obtained from the fragmentation of the peptides was used to decide the arrangement of residues in the peptides using methods such as de novo sequencing or sequence tags. Today thousands of MS and MS/MS spectra are generated within a very short period of time, hence manual annotation of each spectrum is very impractical, inefficient and nearly impossible. Therefore, data analysis in modern-day high- throughput proteomics depends on software tools for automatic assignment of spectra.

Identification of peptides from the MS spectra is the first step in data analysis. This is done by combining information from both MS and MS/MS spectra [114] and matching it with the database.

The various software tools available today for the protein and peptide identification can be chiefly categorized into two groups namely de novo sequencing based tools and database search based tools. De novo sequencing tools such as Lutefisk [115] encode the peptide sequence straight from the MS/MS spectrum while the database search tools such as PEAKS [116,117], Mascot [118], MaxQuant [119,120] looks into the sequence database to decode the spectrum.

When a protein sequence database is available, the database search tools are the most preferred software for peptide identification because the protein sequence database used in the database search tools narrows down the search space for the software to go through.

Accuracy and sensitivity are the two main factors in data analysis. Accuracy is generally determined by the false discovery rate (FDR). FDR is the percentage of false identifications out of all the identification obtained above the cutoff score. However, if the FDR is lowered the sensitivity suffers since the identification rate also comes down. Automated validation of database search results is another essential requirement to achieve high throughput data, which can be done using a target-decoy search strategy [121].

PEAKS [117], the software tool used for data analysis in this thesis combines the decoy and target sequences of the same protein together as a single entry of the database. The algorithm of this software tool is divided into various steps such as de novo sequencing, protein and peptide shortlisting, peptide scoring, result validation and lastly protein identification. In the de novo sequencing step a de novo sequence and sequence tag considering the user-specified modifications and mass error tolerance are generated from the input fragmentation spectrum. Each amino acid in the de novo sequence is given a confidence score in percentage. If the confidence score of an amino acid is less than 30%, it is replaced by its mass value and converted to a sequence tag, totaling the mass of the amino acids in the peptide. These sequence tags are used in the protein shortlisting step. Peptide sequence tags are matched with the database peptides, based on which a protein list is generated. Each protein is ranked based on the common amino acid (CAA) score, which is the number of common amino acids between the sequence tag and the database peptide. Proteins with Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 49

the highest CAA peptide scores are placed highest in the list. Seven thousand top proteins are selected and a protein list is generated, which is used for further analysis, reducing the analysis time. In silico digested peptides of the selected proteins in the protein shortlist are generated. Each peptide can have various modifications, hence the user-specified modifications are considered and peptides with these modifications are also generated. Peptide mass of the modified and unmodified peptides is calculated and this mass is compared with the precursor mass of the input MS/MS spectra. Each peptide-spectrum match (PSM) is given a score. Top 512 candidate sequence peptides are selected for each input spectrum. These top 512 sequence peptides are reranked to find the best peptide for each input spectrum by reranking the peptides using a precise scoring function where the similarity between the de novo sequence and the database peptide is quite crucial. To validate the results target and decoy sequences are combined together to form a decoy fusion sequence. Hence, the target and decoy sequences are not used as different entries but the protein entries are the same, just with the double length in comparison to the original length of the protein. An FDR score is generated as the ratio between the number of decoy hits and the number of target hits above the user specified score threshold. All the high confidence peptides are grouped together to form an inferred protein.

6. Identification of citrullination sites and their estimated relative degree of citrullination

Many different methods have been developed for the identification of citrullinated proteins but only a few methods have been developed to gain insight into the abundance of citrullinated peptides relative to their non-citrullinated counterparts. Quantification of the relative degree of citrullination is desirable since it may give insight into the physiological importance of specific citrullination sites. The relative degree of citrullination at a given site can be defined as the percentage of the particular arginine residue, in its citrullinated state. MS-based identification and quantification of citrullination sites is a challenging task. This PTM results in a mass shift of +0.98Da. However, the same mass shift is also seen in deamidation, which makes it much more complicated to do the correct identification. Nevertheless, before that correct and unambiguous identification of citrullination sites is of prime importance. Identification, of the citrullinated peptide mere o the

basis of accurate mass does not give us confidence in our identification, since many peptides other similar masses or a peptide with deamidation may be identified and confused with citrullination. Therefore, further confirmation using the MS/MS spectra, Ascore and the -10lgP score will give us confidence in our identification (the same approach has been used in papers 1, 2 and 3 presented in this thesis).

Most of the citrullination site measurement and quantification studies have been done on in vitro samples subjected to citrullination by PAD enzymes. In 2010 Beers et al. presented a study which demonstrated a map of citrullination sites on fibrinogen α, β, and γ chains [122]. The human fibrinogen was citrullinated using PAD and then fibrinogen α, β, and γ chains were separated and isolated from an SDS PAGE. The digested peptides were subjected to MS analysis. They observed that PAD2 (48%) citrullinated more arginine sites on fibrinogen chains than PAD4 (35%), suggesting it was more sequence specific. In another study done by Stensland et al, quantification of citrullination in citrullinated synthetic peptides (by PAD4) was done using the isotopic distribution patterns generated by using MALDI-TOF instrument. Quantitation of citrullination was done using the centroid mass, the difference in the centroid masses of the citrullinated and non-citrullinated peptides corresponded to the average degree of modification [123]. In another study, quantitation was done using a skewed isotopic distribution pattern [124]. The theoretical isotopic distribution pattern of the non-citrullinated peptide was compared with the isotopic distribution pattern of the citrullinated peptide and the changes in the pattern were used to do quantitation. Hermansson et al quantified in vivo citrullinated peptides using accurate mass and retention time [106]. The ratio of citrullinated peptides peak area to non-citrullinated peptide peak area was used to quantify the relative degree of citrullination. However, in another study done by Tutturen et al quantification was done by using the area under the curve obtained by using Maxquant software tool. Here the relative degree of citrullination was calculated as the ratio of area under the curve of the citrullinated peptide to the sum of area under the curve of all identical peptides independent of modification [107] as is also shown in paper 1 and 2 of this thesis.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 51

7. Summary of papers

7.1 PAPER 1

Expanding the citrullinome of synovial fibrinogen from rheumatoid arthritis patients

Background The post-translational modification, citrullination, has been found to be correlated with several inflammatory diseases, for instance, RA [125]. RA is a complex multifactorial, autoimmune disease characterized by chronic inflammation and joint degradation. A variety of citrullinated proteins are found in the joint synovium which may or may not be the auto-antigens of this disease [39,40,111]. Fibrinogen is one of the auto-antigens of RA, which has been studied by several investigators in association with citrullination and RA. Several studies have been done for the identification of citrullination sites in fibrinogen [126–132]. Nevertheless, very little information is available regarding the relative degree of citrullination at the sites of modification [128,133–135].

Aim In this study, an MS-based proteomics approach was used to estimate the relative degree of citrullination in fibrinogen at modification sites in SF from RA patients and compared with the modification site-specific patterns of fibrinogen citrullinated in vitro by PAD2 enzyme.

Key findings  High disease activity correlated with a high number of identified citrullination sites and a high citrulline occupancy.  Twenty-five sites were identified to be citrullinated in vivo, 14 of which have not been previously reported and these are presented as novel identifications.  Eight possible hotspots (sites identified in more than one active patient) were identified in patients of this study.  Forty-eight sites were identified to be citrullinated in vitro, 6 of which have not been previously reported and these are presented as novel identifications.  Twenty-four citrullination sites identified in vivo were also detected in vitro, supporting the validity of the identifications.

7.2 PAPER 2

Identification of proteins with high citrulline occupancy in synovial fluid from patients with rheumatoid arthritis

Background Citrullination is an inflammation-dependent protein modification. It is considered to play an important role in chronic inflammatory diseases such as RA. Multiple citrullinated proteins are potential autoantigens in RA. Identification of different citrullinated proteins in RA patients has been done in the past [3,39,44], however, it is unclear whether these proteins are abundantly citrullinated in the synovial joints of each patient, or if their citrullination is common among RA patients. However, very little information is available regarding the citrulline’s relative degree of modification at eligible amino acid positions.

Aim This study was done to identify citrullinated proteins in synovial fluid (SF) from four RA patients and estimated the relative degree of citrullination at individual amino acid positions using a mass spectrometry-based approach.

Key findings  One hundred and seventy-one citrullinated proteins were identified, 13 of which were present in all disease-active patients of this study.  Two hundred and seventy-six citrullination sites were identified in all patient samples of this study.  Thirty-five citrullinated peptides were quantified and compared.  High disease-active patients produced a high number of citrullination sites.  Citrullinated proteins identified are potential autoantigens in RA.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 53

8. Conclusions and perspectives

Despite the remarkably increased research interest in citrullination in relation to RA and other diseases, several intriguing and important aspects of this PTM still remain incomprehensible and is an important topic for further investigation. While citrullination is an inflammation-dependent phenomenon, the role of this PTM in RA and other diseases such as psoriasis, multiple sclerosis, and cancer are still poorly understood. Meanwhile, how this modification can carry potential immunological effects in these diseases are unknown? Thus, identifying and relatively quantifying the sites of deimination may be of immense biological significance. However, unambiguous detection of deimination is still challenging. Hence, the call for advanced techniques for comprehensive high-resolution mass spectrometry-based identification and analysis are required for sensitive and specific detection of PTMs such as citrullination.

Identification of immunologically relevant citrullinated proteins and peptides showing an interconnection to RA or other pathologies is expected. Also, quantification of the relative degree of citrullination in a disease, such RA is desirable. The main aim of this thesis was to develop a strategy for the identification and relative quantification of citrullination in complex biological samples. The applied strategy lead to the identification of citrullination sites both in auto-antigen fibrinogen and other possibly identified proteins in the synovial fluid and BAL fluid in our study. In this work, we have here applied a mass spectrometry-based proteomics approach to estimate the relative degree of citrullination in fibrinogen and other proteins in SF from RA patients and also compared the site-specific patterns to those observed for fibrinogen citrullinated in vitro by PAD2.

In manuscript 1 we studied the citrullination pattern of fibrinogen in joints of RA patients, using SF from four anti-CCP positive patients with varying degrees of disease activity. The mass spectrometric analysis revealed 23 citrullinated sites in SF samples from the four RA patients. We next compared the total number of citrullination sites in fibrinogen between the four RA patients, it was apparent that a higher number of citrullination sites were identified in the samples from the two patients with high disease activity. We also attempted to estimate the degree of citrullination at

individual sites since we found that mere identification of citrullination sites in autoantigens might not be sufficient since the extent of citrullination at a given site can vary substantially.

Several improvements can be suggested for this work. First, further studies with larger sample size are needed to show whether these citrullination sites can serve as diagnostic and prognostic biomarkers indicating disease progression and/or severity in patients with RA. Manual confirmation of citrullination sites within the mass spectrum is often required and also expected. However, the benefit of using this strategy is that the samples are not introduced to the complex chemical derivatization steps, which might show a preference for particular citrullination sites over others. Second, the large difference in the estimated relative degree of citrullination for the different sites identified here underscores the importance of quantitative assessments in future explorations of the citrullinome. Third, specific enrichment strategies can be used to improve the depth of citrullinome identification. Fourth, comparison of the citrullination patterns in ACPA-positive and ACPA- negative RA patients can help in better understanding of the immune response.

In manuscript 2 we extended our citrullination relative quantification study to proteins other than just fibrinogen that were found to be citrullinated in SF across RA patients. The characterization of citrullinated proteins which can develop autoantigenicity, can possibly unveil the pathogenic mechanism of RA. The majority of these citrullinated proteins have been reported in disease active SF before, but it was never clear that these proteins had sites that are abundantly citrullinated. Therefore, recognition of such proteins is important in relation to determining pathogenic citrullinated proteins in RA, such as proteins that can be found by ACPAs to form immune complexes or be presented by antigen presenting cells. However, a large sample size is a necessary improvement in this work to make statistically significant conclusions. Also, the enrichment strategies can more likely help in further in depth analysis of the citrullinome.

Improvements in the mass spectrometric technology have facilitated the MS-based identification of citrullination sites in the proteins. Regardless of the method employed whether with modification or without modification, identification and quantification of citrullination sites in the proteins still remain challenging in complex biological samples. Nevertheless, advancements in MS instruments, in combination with various enrichment methods and identification strategies, will most definitely progress the research in this field. Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 55

Progress in the study of protein deimination by MS has much to offer in clinical studies for the development of improved therapeutic tools used in disease diagnosis. Characterization and recognition of disease-specific biomarkers that can precisely identify the risk of developing a disease such as RA, is essential for improved diagnosis and treatment of patients at risk or at the active disease stage and improve the quality of life in such individuals.

56 9. REFERENCES

9. References

[1] O.N. Jensen, Modification-specific proteomics: Characterization of post-translational modifications by mass spectrometry, Curr. Opin. Chem. Biol. 8 (2004) 33–41. doi:10.1016/j.cbpa.2003.12.009. [2] M. Černý, J. Skalák, H. Cerna, B. Brzobohatý, Advances in purification and separation of posttranslationally modified proteins, J. Proteomics. 92 (2013) 2–27. doi:10.1016/j.jprot.2013.05.040. [3] T. Bennike, K.B. Lauridsen, M.K. Olesen, V. Andersen, S. Birkelund, A. Stensballe, Optimizing the identification of citrullinated peptides by mass spectrometry: utilizing the inability of trypsin to cleave after citrullinated amino acids, J. Proteomics Bioinform. 06 (2013) 288–295. doi:10.4172/jpb.1000293. [4] E.R. Vossenaar, A.J.W. Zendman, W.J. Van Venrooij, G.J.M. Pruijn, PAD , a growing family of citrullinating enzymes : genes , features and involvement in disease, Bioessays. 25 (2003) 1106–1118. doi:10.1002/bies.10357. [5] A.J. Creese, H.J. Cooper, The deimination of arginine to citrulline, Anal. Protein Post-Translational Modif. by Mass Spectrom. (2017) 275–306. doi:10.1002/9781119250906.ch7. [6] M. Inagaki, H. Takahara, Y. Nishi, K. Sugawara, C. Sato, Ca2+-dependent deimination-induced dissembly of intermediate filaments involves specific modification of the amino-terminal head domain, J. Biol. Chem. 264 (1989) 18119–18127. [7] E. Tarcsa, L.N. Marekov, G. Mei, G. Melino, S.C. Lee, P.M. Steinert, Protein unfolding by peptidylarginine deiminase: Substrate specificity and structural relationships of the natural substrates trichohyalin and filaggrin, J. Biol. Chem. 271 (1996) 30709–30716. doi:10.1074/jbc.271.48.30709. [8] S. Chavanas, M.C. Méchin, R. Nachat, V. Adoue, F. Coudane, G. Serre, M. Simon, Peptidylarginine deiminases and deimination in biology and pathology: Relevance to skin homeostasis, J. Dermatol. Sci. 44 (2006) 63–72. doi:10.1016/j.jdermsci.2006.07.004. [9] D.J. Slade, V. Subramanian, J. Fuhrmann, P.R. Thompson, Chemical and biological methods to detect posttranslational modifications of arginine, Biopolymers. 101 (2014) 133–143. doi:10.1002/bip.22256. [10] G.E. Rogers, L.D. Taylor, The Enzymic Derivation of Citrulline Residues from Arginine Residues In Situ during the Biosynthesis of Hair Proteins that are Cross-Linked by Isopeptide Bonds, Protein Crosslink. (1977) 283–294. [11] P. Mangat, N. Wegner, P.J. Venables, J. Potempa, Bacterial and human peptidylarginine deiminases: Targets for inhibiting the autoimmune response in rheumatoid arthritis?, Arthritis Res. Ther. 12 (2010). doi:10.1186/ar3000. [12] S. Wang, Y. Wang, Peptidylarginine deiminases in citrullination, gene regulation, health and pathogenesis, Biochim Biophys Acta. 1829 (2013) 1126–1135. doi:10.1016/j.bbagrm.2013.07.003. [13] A.P. Nicholas, S.K. Bhattacharya, Protein Deimination in Human Health and Disease, Springer, 2014. doi:10.1007/978-1-4614-8317-5. [14] W.T. McGraw, J. Potempa, D. Farley, J. Travis, Purification, characterization, and sequence analysis of a potential virulence factor from Porphyromonas gingivalis, peptidylarginine deiminase, Infect. Immun. 67 (1999) 3248–3256. [15] J.E. Jones, D.M. Lewallen, P.R. Thompson, C.J. Dreyton, B. Knuckley, Mechanistic Studies of Protein Arginine Deiminase 2: Evidence for a Substrate-Assisted Mechanism, Biochemistry. 53 (2014) 4426–4433. doi:10.1021/bi500554b. [16] S.N. Abdullah, E.A. Farmer, L. Spargo, R. Logan, N. Gully, Porphyromonas gingivalis peptidylarginine deiminase substrate specificity, Anaerobe. 23 (2013) 102–108. doi:10.1016/j.anaerobe.2013.07.001. [17] P.L. Kearney, M. Bhatia, N.G. Jones, L. Yuan, M.C. Glascock, K.L. Catchings, M. Yamada, P.R. Thompson, Kinetic characterization of protein arginine deiminase 4: A transcriptional corepressor implicated in the onset and progression of rheumatoid arthritis, Biochemistry. 44 (2005) 10570–10582. doi:10.1021/bi050292m. [18] H. Takahara, H. Okamoto, K. Sugawara, Calcium-dependent properties of peptidylarginine deiminase from Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 57

rabbit skeletal muscle, Agric. Biol. Chem. 50 (1986) 2899–2904. doi:10.1080/00021369.1986.10867830. [19] R.S. Lewis, Calcium signaling mechanisms in T lymphocytes, Annu. Rev. Immunol. 19 (2001) 497–521. doi:10.1146/annurev.immunol.19.1.497. [20] N.E.L. Saris, E. Carafoli, A historical review of cellular calcium handling, with emphasis on mitochondria, Biochem. 70 (2005) 187–194. doi:10.1007/s10541-005-0100-9. [21] K. Arita, H. Hashimoto, T. Shimizu, K. Nakashima, M. Yamada, M. Sato, Structural basis for Ca2+-induced activation of human PAD4, Nat. Struct. Mol. Biol. 11 (2004) 777–783. doi:10.1038/nsmb799. [22] K.L. Bicker, P.R. Thompson, The protein arginine deiminases (PADs): Structure, Function, Inhibition, and Disease, Biopolymers. 99 (2013) 155–163. doi:10.1002/bip.22127. [23] H.A. Doyle, M.J. Mamula, Autoantigenesis: The evolution of protein modifications in autoimmune disease, Curr Opin Immunol. 24 (2012) 112–118. doi:10.1016/j.coi.2011.12.003.Autoantigenesis. [24] G.A. Schellekens, H. Visser, B.A.W. de Jong, F.H.J. van den Hoogen, J.M.W. Hazes, F.C. Breedveld, W.J. van Venrooij, The diagnostic properties of rheumatoid arthritis antibodies recognizing a cyclic citrullinated peptide, Arthritis Rheum. 43 (2000) 155–163. doi:10.1002/1529-0131(200001)43:1<155::AID- ANR20>3.0.CO;2-3. [25] G.A. Schellekens, B.A.W. de Jong, F.H.J. van den Hoogen, L.B.A. van de Putte, W.J. Van, Venrooij, Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritis – specific autoantibodies, J. Clin. Invest. 101 (1998) 273–281. doi:10.1172/JCI1316. [26] G.L. Cuthbert, S. Daujat, A.W. Snowden, H. Erdjument-bromage, T. Hagiwara, M. Yamada, R. Schneider, P.D. Gregory, P. Tempst, A.J. Bannister, T. Kouzarides, Histone Deimination Antagonizes Arginine Methylation, Cell. 118 (2004) 545–553. doi:http://dx.doi.org/10.1016/j.cell.2004.08.020. [27] P. Li, H. Yao, Z. Zhang, M. Li, Y. Luo, P.R. Thompson, D.S. Gilmour, Y. Wang, Regulation of p53 Target Gene Expression by Peptidylarginine Deiminase 4, Mol. Cell. Biol. 28 (2008) 4745–4758. doi:10.1128/mcb.01747-07. [28] M. Leshner, S. Wang, C. Lewis, H. Zheng, X.A. Chen, L. Santy, Y. Wang, PAD4 mediated histone hypercitrullination induces heterochromatin decondensation and chromatin unfolding to form neutrophil extracellular trap-like structures, Front. Immunol. 3 (2012) 1–11. doi:10.3389/fimmu.2012.00307. [29] V. Brinkmann, U. Reichard, C. Goosmann, B. Fauler, Y. Uhlemann, D.S. Weiss, Neutrophil extracellular traps kill bacteria, Science. 303 (2004) 1532–1536. doi:10.1126/science.1092385. [30] J. Zhang, J. Dai, E. Zhao, Y. Lin, L. Zeng, J. Chen, H. Zheng, Y. Wang, X. Li, K. Ying, Y. Xie, Y. Mao, cDNA cloning, gene organization and expression analysis of human peptidylarginine deiminase type VI., Biochem. J. 370 (2003) 167–174. doi:0451041051. [31] E. Candi, R. Schmidt, G. Melino, The cornified envelope: A model of cell death in the skin, Nat. Rev. Mol. Cell Biol. 6 (2005) 328–340. doi:10.1038/nrm1619. [32] P.M. Steinert, D.A.D. Parry, L.N. Marekov, Trichohyalin Mechanically Strengthens the Hair Follicle, J. Biol. Chem. 278 (2003) 41409–41419. doi:10.1074/jbc.m302037200. [33] C.R. Harding, I.R. Scott, Histidine-rich proteins (filaggrins): Structural and functional heterogeneity during epidermal differentiation, J. Mol. Biol. 170 (1983) 651–673. doi:10.1016/S0022-2836(83)80126-0. [34] M.A. Moscarello, D.D. Wood, C. Ackerley, C. Boulias, Myelin in multiple sclerosis is developmentally immature, J. Clin. Invest. 94 (1994) 146–154. doi:10.1172/JCI117300. [35] P. Proost, T. Loos, A. Mortier, E. Schutyser, M. Gouwy, S. Noppen, C. Dillen, I. Ronsse, R. Conings, S. Struyf, G. Opdenakker, P.C. Maudgal, J. Van Damme, Citrullination of CXCL8 by peptidylarginine deiminase alters receptor usage, prevents proteolysis, and dampens tissue inflammation, J. Exp. Med. 205 (2008) 2085–2097. doi:10.1084/jem.20080305. [36] G.S. Firestein, Evolving concepts of rheumatoid arthritis, Nature. 423 (2003) 356–61. doi:10.1038/nature01661. [37] W.J. Van Venrooij, J.J.B.C. Van Beers, G.J.M. Pruijn, Anti-CCP antibodies: The past, the present and the future, Nat. Rev. Rheumatol. 7 (2011) 391–398. doi:10.1038/nrrheum.2011.76. [38] G.J. Pruijn, A. Wiik, W.J. van Venrooij, The use of citrullinated peptides and proteins for the diagnosis of rheumatoid arthritis., Arthritis Res. Ther. 12 (2010) 203. doi:10.1186/ar2903. [39] J.J.B.C. van Beers, C.M. Schwarte, J.. Stammen-Vogelzangs, E.. Oosterink, B.. Božič, G.J.M. Pruijn, The rheumatoid arthritis synovial fluid citrullinome reveals novel citrullinated epitopes in apolipoprotein E , 57

58 9. REFERENCES

myeloid nuclear differentiation antigen , and β-actin, Arthritis Rheum. 65 (2013) 69–80. doi:10.1002/art.37720. [40] C. Masson-Bessiere, M. Sebbag, E. Girbal-Neuhauser, L. Nogueira, C. Vincent, T. Senshu, G. Serre, The major synovial targets of the rheumatoid arthritis-specific antifilaggrin autoantibodies are deiminated forms of the alpha- and beta-chains of fibrin., J. Immunol. 166 (2001) 4177–4184. doi:10.4049/jimmunol.166.6.4177. [41] K. Matsuo, Y. Xiang, H. Nakamura, K. Masuko, K. Yudoh, K. Noyori, K. Nishioka, T. Saito, T. Kato, Identification of novel citrullinated autoantigens of synovium in rheumatoid arthritis using a proteomic approach, Arthritis Res. Ther. 8 (2006) 1–13. doi:10.1186/ar2085. [42] K. Tilleman, K. Van Steendam, T. Cantaert, F. De Keyser, D. Elewaut, D. Deforce, Synovial detection and autoantibody reactivity of processed citrullinated isoforms of vimentin in inflammatory arthritides, Rheumatology. 47 (2008) 597–604. doi:10.1093/rheumatology/ken077. [43] K. Van Steendam, K. Tilleman, M. De Ceuleneer, F. De Keyser, D. Elewaut, D. Deforce, Citrullinated vimentin as an important antigen in immune complexes from synovial fluid of rheumatoid arthritis patients with antibodies against citrullinated proteins, Arthritis Res. Ther. 12 (2010). doi:10.1186/ar3070. [44] F. Wang, F.-F. Chen, W.-B. Gao, H.-Y. Wang, N.-W. Zhao, M. Xu, D.-Y. Gao, W. Yu, X.-L. Yan, J.-N. Zhao, and X.-J. Li, Identification of citrullinated peptides in the synovial fluid of patients with rheumatoid arthritis using LC-MALDI-TOF/TOF, Clin. Rheumatol. 35 (2016) 2185–2194. doi:10.1007/s10067-016- 3247-4. [45] F.B. Mercado, R.I. Marshall, A.C. Klestov, P.M. Bartold, Relationship Between Rheumatoid Arthritis and Periodontitis, J. Periodontol. 72 (2001) 779–787. doi:10.1902/jop.2001.72.6.779. [46] F. Mercado, R.I. Marshall, A.C. Klestov, P.M. Bartold, Is there a relationship between rheumatoid arthritis and periodontal disease?, J Clin Periodontol. 27 (2000) 267–272. doi:10.1016/j.rbre.2014.03.007. [47] J.U. Scher, C. Ubeda, M. Equinda, R. Khanin, Y. Buischi, A. Viale, L. Lipuma, M. Attur, M.H. Pillinger, G. Weissmann, D.R. Littman, E.G. Pamer, W.A. Bretz, S.B. Abramson, Periodontal Disease and the Oral Microbiota in New-Onset Rheumatoid Arthritis, Arthritis Rheum. 64 (2012) 3083–3094. doi:10.1002/art.34539.Periodontal. [48] A.J. MacGregor, H. Snieder, A.S. Rigby, M. Koskenvuo, J. Kaprio, K. Aho, A.J. Silman, Characterizing the quantitative genetic contribution to rheumatoid arthritis using data from twins, Arthritis Rheum. 43 (2000) 30–37. doi:10.1002/1529-0131(200001)43:1<30::AID-ANR5>3.0.CO;2-B. [49] P.A. Gourraud, P. Dieudé, J.F. Boyer, L. Nogueira, A. Cambon-Thomsen, B. Mazières, F. Cornélis, G. Serre, A. Cantagrel, A. Constantin, A new classification of HLA-DRB1 alleles differentiates predisposing and protective alleles for autoantibody production in rheumatoid arthritis, Arthritis Res. Ther. 9 (2007) 1–8. doi:10.1186/ar2131. [50] L. Klareskog, U. Forsum, M. Tjernlund, D. Kabelitz, A. Wigren, Appearance of anti-HLA-DR-reactive cells in normal and rheumatoid synovial tissue, Scand. J. Immunol. 14 (1981) 183–192. doi:10.1111/j.1365- 3083.1981.tb00198.x. [51] G.T. Nepom, P. Byers, C. Seyfried, L.A. Healey, K.R. Wilske, D. Stage, B.S. Nepom, HLA Genes associated with rheumatoid Arthritis. Identification of susceptibility alleles using specific oligonucleotide probes, Arthritis Rheum. 32 (1989) 15–21. doi:10.1002/anr.1780320104. [52] P.K. Gregersen, J. Silver, R.J. Winchester, The Shared Epitope Hypothesis, Arthritis Rheum. 30 (1987) 1205–1212. doi:10.1002/art.1780301102. [53] M. Klockars, R.S. Koskela, E. Järvinen, P.J. Kolari, A. Rossi, Silica exposure and rheumatoid arthritis: a follow up study of granite workers 1940-81., Br. Med. J. 294 (1987) 997–1000. doi:10.1136/bmj.294.6578.997. [54] C. GM, R. FL, B. JM, B. J.M, S. JW, S. WT, Occupational silica exposure and risk of various diseases: An analysis using death certificates from 27 states of the United States, Occup. Environ. Med. 60 (2003) 122– 129. doi:10.1136/oem.60.2.122. [55] Å. Reckner Olsson, T. Skogh, O. Axelson, G. Wingren, Occupations and exposures in the work environment as determinants for rheumatoid arthritis, Occup. Environ. Med. 61 (2004) 233–238. doi:10.1136/oem.2003.007971. [56] B. Sverdrup, H. Källberg, C. Bengtsson, I. Lundberg, L. Padyukov, L. Alfredsson, L. Klareskog, Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 59

Epidemiological Investigation of Rheumatoid Arthritis Study Group, Association between occupational exposure to mineral oil and rheumatoid arthritis: results from the Swedish EIRA case-control study., Arthritis Res. Ther. 7 (2005) R1296-303. doi:10.1186/ar1824. [57] L.A. Criswell, L.A. Merlino, J.R. Cerhan, T.R. Mikuls, A.S. Mudano, M. Burma, A.R. Folsom, K.G. Saag, Cigarette smoking and the risk of rheumatoid arthritis among postmenopausal women:, Am. J. Med. 112 (2002) 465–471. doi:10.1016/s0002-9343(02)01051-3. [58] L. Klareskog, J. Rönnelid, K. Lundberg, L. Padyukov, L. Alfredsson, Immunity to citrullinated proteins in rheumatoid arthritis., Annu. Rev. Immunol. 26 (2008) 651–75. doi:10.1146/annurev.immunol.26.021607.090244. [59] D. Damgaard, M.F.B. Nielsen, M. Quisgaard Gaunsbaek, Y. Palarasah, V. Svane-Knudsen, C.H. Nielsen, Smoking is associated with increased levels of extracellular peptidylarginine deiminase 2 (PAD2) in the lungs, Clin. Exp. Rheumatol. 33 (2015) 405–408. doi:10.1016/S0304-3940(00)01735-3. [60] C. Foulquier, M. Sebbag, C. Clavel, S. Chapuy-Regaud, R. Al Badine, M.C. Méchin, C. Vincent, R. Nachat, M. Yamada, H. Takahara, M. Simon, M. Guerrin, G. Serre, Peptidyl arginine deiminase type 2 (PAD-2) and PAD-4 but not PAD-1, PAD-3, and PAD-6 are expressed in rheumatoid arthritis synovium in close association with tissue inflammation, Arthritis Rheum. 56 (2007) 3541–3553. doi:10.1002/art.22983. [61] L. Klareskog, P. Stolt, K. Lundberg, H. Källberg, C. Bengtsson, J. Grunewald, J. Rönnelid, H.E. Harris, A.K. Ulfgren, S. Rantapää-Dahlqvist, A. Eklund, L. Padyukov, L. Alfredsson, A new model for an etiology of rheumatoid arthritis: Smoking may trigger HLA-DR (shared epitope)-restricted immune reactions to autoantigens modified by citrullination, Arthritis Rheum. 54 (2006) 38–46. doi:10.1002/art.21575. [62] D. Makrygiannakis, M. Hermansson, A.K. Ulfgren, A.P. Nicholas, A.J.W. Zendman, A. Eklund, J. Gmnewald, C.M. Skold, L. Klareskog, A.I. Catrina, Smoking increases peptidylarginine deiminase 2 enzyme expression in human lungs and increases citrullination in BAL cells, Ann. Rheum. Dis. 67 (2008) 1488– 1492. doi:10.1136/ard.2007.075192. [63] H. Asaga, M. Yamada, T. Senshu, Selective deimination of vimentin in calcium ionophore-induced apoptosis of mouse peritoneal macrophages, Biochem. Biophys. Res. Commun. 243 (1998) 641–646. doi:10.1006/bbrc.1998.8148. [64] J.A.. Hill, S.. Southwood, A.. Sette, A.M.. Jevnikar, D.A.. Bell, E. Cairns, Cutting edge: the conversion of arginine to citrulline allows for a high-affinity peptide interaction with the rheumatoid arthritis-associated HLA-DRB1*0401 MHC class II molecule., J. Immunol. 171 (2003) 538–541. doi:10.4049/jimmunol.171.2.538. [65] L. De Rycke, A.P. Nicholas, T. Cantaert, E. Kruithof, J.D. Echols, B. Vandekerckhove, E.M. Veys, F. De Keyser, D. Baeten, Synovial intracellular citrullinated proteins colocalizing with peptidyl arginine deiminase as pathophysiologically relevant antigenic determinants of rheumatoid arthritis-specific humoral autoimmunity, Arthritis Rheum. 52 (2005) 2323–2330. doi:10.1002/art.21220. [66] C. Clavel, L. Nogueira, L. Laurent, C. Iobagiu, C. Vincent, M. Sebbag, G. Serre, Induction of macrophage secretion of tumor necrosis factor α through Fcγ receptor IIa engagement by rheumatoid arthritis-specific autoantibodies to citrullinated proteins complexed with fibrinogen, Arthritis Rheum. 58 (2008) 678–688. doi:10.1002/art.23284. [67] A. Ishida-Yamamoto, T. Senshu, H. Takahashi, K. Akiyama, K. Nomura, H. Iizuka, Decreased deiminated keratin K1 in psoriatic hyperproliferative epidermis, J. Invest. Dermatol. 114 (2000) 701–705. doi:10.1046/j.1523-1747.2000.00936.x. [68] A. Ehrlich, S. Booher, Y. Becerra, D.L. Borris, W.D. Figg, M.L. Turner, A. Blauvelt, Micellar paclitaxel improves severe psoriasis in a prospective phase II pilot study, J. Am. Acad. Dermatol. 50 (2004) 533–540. doi:10.1016/j.jaad.2003.09.018. [69] G.M. Alenius, E. Berglin, S. Rantapää Dahlqvist, Antibodies against cyclic citrullinated peptide (CCP) in psoriatic patients with or without joint inflammation, Ann. Rheum. Dis. 65 (2006) 398–400. doi:10.1136/ard.2005.040998. [70] L. Wang, X. Chang, G. Yuan, Y. Zhao, P. Wang, Expression of peptidylarginine deiminase type 4 in ovarian tumors, Int. J. Biol. Sci. 6 (2010) 454–464. doi:10.7150/ijbs.6.454. [71] B.D. Cherrington, E. Morency, A.M. Struble, S.A. Coonrod, J.J. Wakshlag, Potential role for peptidylarginine deiminase 2 (PAD2) in citrullination of canine mammary epithelial cell histones, PLoS One. 59

60 9. REFERENCES

5 (2010). doi:10.1371/journal.pone.0011768. [72] X. Chang, J. Han, Expression of Peptidylarginine Deiminase Type 4(PAD4) in Various Tumors, Mol. Carcinog. 45 (2006) 183–196. doi:10.1002/mc. [73] X. Chang, J. Han, L. Pang, Y. Zhao, Y. Yang, Z. Shen, Increased PADI4 expression in blood and tissues of patients with malignant tumors, BMC Cancer. 9 (2009) 1–11. doi:10.1186/1471-2407-9-40. [74] B.D. Cherrington, S. Mohanan, A.N. Diep, R. Fleiss, D. Sudilovsky, L.J. Anguish, S.A. Coonrod, J.J. Wakshlag, Comparative Analysis of Peptidylarginine Deiminase-2 Expression in Canine, Feline and Human Mammary Tumours, J. Comp. Pathol. 147 (2012) 139–146. doi:10.1016/j.jcpa.2012.01.021. [75] A.D. Sadovnick, G.C. Ebers, D.A. Dyment, N.J. Risch, Evidence for genetic basis of multiple sclerosis, Lancet. 347 (1996) 1728–1730. doi:10.1016/S0140-6736(96)90807-7. [76] A. Compston, A. Coles, Multiple sclerosis, Lancet. 372 (2008) 1502–17. doi:10.1007/978-1-4939-7880-9_8. [77] D.D. Wood, C.A. Ackerley, B. Van Den Brand, L. Zhang, R. Raijmakers, F.G. Mastronardi, M.A. Moscarello, Myelin localization of peptidylarginine deiminases 2 and 4: Comparison of PAD2 and PAD4 activities, Lab. Investig. 88 (2008) 354–364. doi:10.1038/labinvest.3700748. [78] A.M.N. Silva, R. Vitorino, M.R.M. Domingues, C.M. Spickett, P. Domingues, Post-translational modifications and mass spectrometry detection, Free Radic. Biol. Med. 65 (2013) 925–941. doi:10.1016/j.freeradbiomed.2013.08.184. [79] K. Tanaka, H. Waki, Y. Ido, S. Akita, Y. Yoshida, T. Yoshida, T. Matsuo, Protein and polymer analyses up to m/z 100 000 by laser ionization time‐of‐flight mass spectrometry, Rapid Commun. Mass Spectrom. 2 (1988) 151–153. doi:10.1002/rcm.1290020802. [80] R.L. Caldwell, R.M. Caprioli, Tissue Profiling by Mass Spectrometry, Mol. Cell. Proteomics. 4 (2005) 394– 401. doi:10.1074/mcp.R500006-MCP200. [81] J.B. Fenn, M. Mann, C.K. Meng, S.F. Wong, M. Craig, C.K.A.I. Meng, M. Mann, C.M. Whitehouse, Electrospray Ionization of Large for Mass Spectrometry Biomolecules, Science. 246 (1989) 64–71. [82] M. Wilm, Principles of Electrospray Ionization, Mol. Cell. Proteomics. 10 (2011) 1–8. doi:10.1074/mcp.M111.009407. [83] C.D. Kelstrup, C. Young, R. Lavallee, M.L. Nielsen, J. V. Olsen, Optimized fast and sensitive acquisition methods for shotgun proteomics on a quadrupole orbitrap mass spectrometer, J. Proteome Res. 11 (2012) 3487–3497. doi:10.1021/pr3000249. [84] S. Eliuk, A. Makarov, Evolution of Orbitrap Mass Spectrometry Instrumentation, Annu. Rev. Anal. Chem. 8 (2015) 61–80. doi:10.1146/annurev-anchem-071114-040325. [85] A.W. Jones, H.J. Cooper, Dissociation techniques in mass spectrometry-based proteomics, Analyst. 136 (2011) 3419–3429. doi:10.1039/c0an01011a. [86] B. Palzs, S. Suhal, Fragmentation pathways of protonated peptides, Mass Spectrom. Rev. 24 (2005) 508– 548. doi:10.1002/mas.20024. [87] M.W. Senko, P.M. Remes, J.D. Canterbury, R. Mathur, Q. Song, S.M. Eliuk, C. Mullen, L. Earley, M. Hardman, J.D. Blethrow, H. Bui, A. Specht, O. Lange, E. Denisov, A. Makarov, S. Horning, V. Zabrouskov, Novel parallelized quadrupole/linear ion trap/orbitrap tribrid mass spectrometer improving proteome coverage and peptide identification rates, Anal. Chem. 85 (2013) 11710–11714. doi:10.1021/ac403115c. [88] J.E.P. Syka, J.J. Coon, M.J. Schroeder, J. Shabanowitz, D.F. Hunt, Peptide and protein sequence analysis by electron transfer dissociation mass spectrometry, PNAS. 101 (2004) 9528–9533. doi:10.1016/S0167- 2789(01)00356-6. [89] L.M. Mikesh, B. Ueberheide, A. Chi, J.J. Coon, J.E.P. Syka, J. Shabanowitz, D.F. Hunt, The utility of ETD Mass Spectrometry in Proteomic Analysis: BIOCHIMICA ET BIOPHYSICA ACTA Proteins and Proteomics Posttranslational Modifications in Proteomics, Biochim Biophys Acta. 1764 (2006) 1811–1822. doi:10.1016/j.bbapap.2006.10.003. [90] K.A. Resing, N.G. Ahn, Proteomics strategies for protein identification, FEBS Lett. 579 (2005) 885–889. doi:10.1016/j.febslet.2004.12.001. [91] K. Breuker, M. Jin, X. Han, H. Jiang, F.W. McLafferty, Top-Down Identification and Characterization of Biomolecules by Mass Spectrometry, J Am Soc Mass Spectrom. 19 (2008) 1045–1053. doi:10.1016/j.jasms.2008.05.013. [92] R. Aebersold, M. Mann, Mass-spectrometric exploration of proteome structure and function, Nature. 537 Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 61

(2016) 347–355. doi:10.1038/nature19949. [93] J.R. Yates, C.I. Ruse, A. Nakorchevsky, Proteomics by mass spectrometry: approaches, advances, and applications., Annu. Rev. Biomed. Eng. 11 (2009) 49–79. doi:10.1146/annurev-bioeng-061008-124934. [94] C.R. Marshall, Fossil Record Reveals Tropics as Cradle and Museum, Science. 314 (2006) 66–67. doi:10.1126/science.1133351. [95] Y.Q. Yu, M. Gilar, P.J. Lee, E.S.P. Bouvier, J.C. Gebler, Enzyme-Friendly, Mass Spectrometry-Compatible Surfactant for In-Solution Enzymatic Digestion of Proteins, Anal. Chem. 75 (2003) 6023–6028. doi:10.1021/ac0346196. [96] J. V. Olsen, S.-E. Ong, M. Mann, Trypsin Cleaves Exclusively C-terminal to Arginine and Lysine Residues, Mol. Cell. Proteomics. 3 (2004) 608–614. doi:10.1074/mcp.T400003-MCP200. [97] J.M. Burkhart, C. Schumbrutzki, S. Wortelkamp, A. Sickmann, R.P. Zahedi, Systematic and quantitative comparison of digest efficiency and specificity reveals the impact of trypsin quality on MS-based proteomics, J. Proteomics. 75 (2012) 1454–1462. doi:10.1016/j.jprot.2011.11.016. [98] P. Giansanti, L. Tsiatsiani, T.Y. Low, A.J.R. Heck, Six alternative proteases for mass spectrometry-based proteomics beyond trypsin, Nat. Protoc. 11 (2016) 993–1006. doi:10.1038/nprot.2016.057. [99] X. Zhao, N.L. Okeke, O. Sharpe, F.M. Batliwalla, A.T. Lee, P.P. Ho, B.H. Tomooka, P.K. Gregersen, W.H. Robinson, Circulating immune complexes contain citrullinated fibrinogen in rheumatoid arthritis, Arthritis Res Ther. 10 (2008) 1–13. doi:10.1186/ar2478. [100] G. Cagney, S. Amiri, T. Premawaradena, M. Lindo, A. Emili, In silico proteome analysis to facilitate proteomics experiments using mass spectrometry, Proteome Sci. 1 (2003) 1–15. doi:10.1186/1477-5956-1-5. [101] J. Rappsilber, Y. Ishihama, M. Mann, Stop And Go Extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics, Anal. Chem. 75 (2003) 663–670. doi:10.1021/ac026117i. [102] A.E.V. Tutturen, A. Holm, B. Fleckenstein, Specific biotinylation and sensitive enrichment of citrullinated peptides, Anal. Bioanal. Chem. 405 (2013) 9321–9331. doi:10.1007/s00216-013-7376-1. [103] K. Kubota, T. Yoneyama-Takazawa, K. Ichikawa, Determination of sites citrullinated by peptidylarginine deiminase using18O stable isotope labeling and mass spectrometry, Rapid Commun. Mass Spectrom. 19 (2005) 683–688. doi:10.1002/rcm.1842. [104] G. Hao, D. Wang, J. Gu, Q. Shen, S.S. Gross, Yanming, Wang, Neutral loss of isocyanic acid in peptide CID spectra: a novel diagnostic marker for mass spectrometric identification of protein citrullination, J Am Soc Mass Spectrom. 20 (2009) 723–727. doi:10.1016/j.jasms.2008.12.012. [105] Z. Jin, Z. Fu, J. Yang, J. Troncosco, A.D. Everett, J.E. Van Eyk, Identification and Characterization of Citrulline-modified Brain Proteins by Combining HCD and CID Fragmentation, Proteomics. 13 (2013) 2682–2691. doi:10.1002/pmic.201300064.Identification. [106] M. Hermansson, K. Artemenko, E. Ossipova, H. Eriksson, J. Lengqvist, D. Makrygiannakis, A.I. Catrina, A.P. Nicholas, L. Klareskog, M. Savitski, R.A. Zubarev, P.J. Jakobsson, MS analysis of rheumatoid arthritic synovial tissue identifies specific citrullination sites on fibrinogen, Proteomics - Clin. Appl. 4 (2010) 511– 518. doi:10.1002/prca.200900088. [107] A.E. V Tutturen, B. Fleckenstein, G.A. De Souza, Assessing the citrullinome in rheumatoid arthritis synovial fluid with and without enrichment of citrullinated peptides, J. Proteome Res. 13 (2014) 2867–2873. doi:10.1021/pr500030x. [108] V. Romero, J. Fert-bober, P.A. Nigrovic, E. Darrah, U.J. Haque, D.M. Lee, J. Van Eyk, A. Rosen, F. Andrade, Immune-mediated pore-forming pathways induce cellular hypercitrullination and generate citrullinated autoantigens in rheumatoid arthritis, Sci Transl Med. 5 (2013). doi:10.1126/scitranslmed.3006869.Immune-mediated. [109] C.-Y. Lee, H. Hahne, B. Kuster, K. Schnatbaum, T. Schmidt, M. Wilhelm, D.P. Zolg, D. Wang, M. Uhlén, U. Reimer, F. Pontén, Mining the Human Tissue Proteome for Protein Citrullination, Mol. Cell. Proteomics. 17 (2018) 1378–1391. doi:10.1074/mcp.ra118.000696. [110] M. De Ceuleneer, K. Van Steendam, M. Dhaenens, D. Deforce, In vivo relevance of citrullinated proteins and the challenges in their detection, Proteomics. (2012) 752–760. doi:10.1002/pmic.201100478. [111] A. Kinloch, V. Tatzer, R. Wait, D. Peston, K. Lundberg, P. Donatien, D. Moyes, P.C. Taylor, P.J. Venables, Identification of citrullinated alpha-enolase as a candidate autoantigen in rheumatoid arthritis., Arthritis Res. 61

62 9. REFERENCES

Ther. 7 (2005) R1421-9. doi:10.1186/ar1845. [112] A.J. Kjeinnijenhuis, C. Hedegaard, D. Lundvig, S. Sundbye, O.G. Issinger, O.N. Jensen, P.H. Jensen, Identification of multiple post-translational modifications in the porcine brain specific p25α, J. Neurochem. 106 (2008) 925–933. doi:10.1111/j.1471-4159.2008.05437.x. [113] A. Ordóñez, I. Martínez-Martínez, F.J. Corrales, C. Miqueo, A. Miñano, V. Vicente, J. Corral, Effect of citrullination on the function and conformation of antithrombin, FEBS J. 276 (2009) 6763–6772. doi:10.1111/j.1742-4658.2009.07391.x. [114] K.R. Clauser, P. Baker, A.L. Burlingame, Role of accurate mass measurement (±10 ppm) in protein identification strategies employing MS or MS/MS and database searching, Anal. Chem. 71 (1999) 2871– 2882. doi:10.1021/ac9810516. [115] J.A. Taylor, R.S. Johnson, Sequence database searches via de Novo peptide sequencing by tandem mass spectrometry, Rapid Commun. Mass Spectrom. 11 (1997) 1067–1075. doi:10.1002/(SICI)1097- 0231(19970615)11:9<1067::AID-RCM953>3.0.CO;2-L. [116] B. Ma, K. Zhang, C. Hendrie, C. Liang, M. Li, A. Doherty-Kirby, G. Lajoie, PEAKS: Powerful software for peptide de novo sequencing by tandem mass spectrometry, Rapid Commun. Mass Spectrom. 17 (2003) 2337–2342. doi:10.1002/rcm.1196. [117] J. Zhang, L. Xin, B. Shan, W. Chen, M. Xie, D. Yuen, W. Zhang, Z. Zhang, G.A. Lajoie, B. Ma, PEAKS DB: De Novo Sequencing Assisted Database Search for Sensitive and Accurate Peptide Identification , Mol. Cell. Proteomics. 11 (2011) M111.010587. doi:10.1074/mcp.m111.010587. [118] D.N. Perkins, D.J.C. Pappin, D.M. Creasy, J.S. Cottrell, Probability-based protein identification by searching sequence databases using mass spectrometry data, Electrophoresis. 20 (1999) 3551±3567. doi:10.1097/MAO.0000000000000737. [119] J. Cox, M. Mann, MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification., Nat. Biotechnol. 26 (2008) 1367–72. doi:10.1038/nbt.1511. [120] J. Cox, N. Neuhauser, A. Michalski, R.A. Scheltema, J. V. Olsen, M. Mann, Andromeda: A peptide search engine integrated into the MaxQuant environment, J. Proteome Res. 10 (2011) 1794–1805. doi:10.1021/pr101065j. [121] J.E.E. And, S.P. Gygi, Target-Decoy Search Strategy for Mass Spectrometry-Based Proteomics, Methods Mol Biol. 604 (2010) 55–71. doi:10.1002/0471140864.ps2500s60. [122] J.J.B.C. van Beers, R. Raijmakers, L.-E. Alexander, J. Stammen-Vogelzangs, A.M.C. Lokate, A.J.R. Heck, R.B.M. Schasfoort, G.J.M. Pruijn, Mapping of citrullinated fibrinogen B-cell epitopes in rheumatoid arthritis by imaging surface plasmon resonance, Arthritis Res. Ther. 12 (2010) R219. doi:10.1186/ar3205. [123] M.E. Stensland, S. Pollmann, Ø. Molberg, L.M. Sollid, B. Fleckenstein, Primary sequence, together with other factors, influence peptide deimination by peptidylarginine deiminase-4, Biol. Chem. 390 (2009) 99– 107. doi:10.1515/BC.2009.019. [124] M.D.. Ceuleneer, M.. E.. D.. van Steendam, Katleen.: Dhaenens, D. Deforce, Quantification of Citrullination by Means of Skewed Isotope Distribution Pattern, J. Proteome Res. 11 (11) (2012) pp 5245–5251. doi:10.1021/pr3004453. [125] T. Cantaert, L. De Rycke, T. Bongartz, E.L. Matteson, P.P. Tak, A.P. Nicholas, D. Baeten, Citrullinated proteins in rheumatoid arthritis, Arthritis Rheum. 54 (2006) 3381–3389. doi:10.1002/art.22206. [126] O. Sanchez-Pernaute, M. Filkova, A. Gabucio, M. Klein, H. Maciejewska-Rodrigues, C. Ospelt, F. Brentano, B. a Michel, R.E. Gay, G. Herrero-Beaumont, S. Gay, M. Neidhart, A. Juengel, Citrullination enhances the pro-inflammatory response to fibrin in rheumatoid arthritis synovial fibroblasts., Ann. Rheum. Dis. 72 (2013) 1400–6. doi:10.1136/annrheumdis-2012-201906. [127] J. Sokolove, X. Zhao, P.E. Chandra, W.H. Robinson, Immune complexes containing citrullinated fibrinogen costimulate macrophages via toll-like receptor 4 and Fcγ receptor receptor, Arthritis Rheum. 63 (2011) 53– 62. doi:10.1002/art.30081. [128] R. Raijmakers, J.J. van Beers, M. El-Azzouny, N.F. Visser, B. Božič, G.J. Pruijn, A.J. Heck, Elevated levels of fibrinogen-derived endogenous citrullinated peptides in synovial fluid of rheumatoid arthritis patients, Arthritis Res Ther. 14 (2012) 1–10. doi:10.1186/ar3840. [129] A. Suzuki, R.Y.O. Yamada, K. Yamamoto, Citrullination by peptidylarginine deiminase in rheumatoid Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 63

arthritis, Ann N Y Acad Sci. 339 (2007) 323–339. doi:10.1196/annals.1422.034. [130] K.Y. Makiko Nakayama-Hamadaa, Akari Suzuki, Kazuishi Kubota, Tomoko Takazawa, Mizuko Ohsaka, Reimi Kawaida, Mitsuru Ono, Atsushi Kasuya, Hidehiko Furukawa, Ryo Yamada, Comparison of enzymatic properties between hPADI2 and hPADI4, Biochem Biophys Res Commun. 327 (2005) 192–200. doi:10.1016/j.bbrc.2004.11.152. [131] J.J.B.C. van Beers, R. Raijmakers, L.-E. Alexander, J. Stammen-Vogelzangs, A.M.C. Lokate, A.J.R. Heck, R.B.M. Schasfoort, G.J.M. Pruijn, Mapping of citrullinated fibrinogen B-cell epitopes in rheumatoid arthritis by imaging surface plasmon resonance., Arthritis Res. Ther. 12 (2010) R219. doi:10.1186/ar3205. [132] S.. Blachere, Nathale. E.; Parveen, M.O.. Frank, H.. Dill, Brian. D.; Molina, D.E. Orange, High-Titer Rheumatoid Arthritis Antibodies Preferentially Bind Fibrinogen Citrullinated by Peptidylarginine Deiminase 4, ARTHRITIS Rheumatol. 69 (2017) 986–995. doi:10.1002/art.40035. [133] S.-E. Ong, M. Mann, Mass spectrometry-based proteomics turns quantitative., Nat. Chem. Biol. 1 (2005) 252–62. doi:10.1038/nchembio736. [134] R. Raijmakers, A.J.W. Zendman, W.V. Egberts, E.R. Vossenaar, J. Raats, C. Soede-huijbregts, F.P.J.T. Rutjes, P.A. Van Veelen, J.W. Drijfhout, G.J.M. Pruijn, Methylation of Arginine Residues Interferes with Citrullination by Peptidylarginine Deiminases in vitro, J Mol Biol. 367 (2007) 1118–1129. doi:10.1016/j.jmb.2007.01.054. [135] M.L. Pérez, M.J. Gómara, G. Ercilla, R. Sanmartí, I. Haro, Antibodies to citrullinated human fibrinogen synthetic peptides in diagnosing rheumatoid arthritis, J. Med. Chem. 50 (2007) 3573–3584. doi:10.1021/jm0701932.

64 10 MANUSCRIPTS

10 Manuscripts

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 65

Manuscript 1

Expanding the citrullinome of synovial fibrinogen from rheumatoid

arthritis patients

Mandvi Sharma1, Dres Damgaard2,3, Ladislav Senolt4, Birte Svensson1,, Anne-Christine Bay- Jensen5, Claus Henrik Nielsen2,3, Per Hägglund1#*

1Department of Biotechnology and Biomedicine, Technical University of Denmark, Kgs. Lyngby, Denmark

2Institute for Inflammation Research, Center for Rheumatology and Spine Diseases, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark

3Section for Periodontology, Microbiology and Community Dentistry, Department of Odontology, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

4Institute of Rheumatology and Department of Rheumatology, 1st Faculty of Medicine, Charles University, Prague, Czech Republic

5Rheumatology, Biomarkers and Research, Nordic Biosciences, Herlev, Denmark

* Corresponding Author: [email protected] ; +45 35327900

#Present address: Department of Biomedical Sciences, Faculty of Health and Medical Sciences, Panum Institute, University of Copenhagen, Copenhagen, Denmark

Author contributions: MS, DD, CHN, and PH were involved in the study design and conceptualization. LS collected the SF samples. MS performed the mass spectrometric analysis. MS, DD, CHN, and PH wrote the paper with input from BS, ACBJ, LS.

Abstract

Citrullination is a post-translational protein modification, which is associated with inflammation in general and is thought to play an important pathogenic role in rheumatoid arthritis (RA). Here a mass spectrometry-based proteomics approach was applied to identify citrullination sites in synovial fluid fibrinogen from four RA patients. In general, high disease activity correlated with increased number of identified citrullination sites and higher relative citrulline occupancy. Altogether, 23 sites were identified, of which 14 have not been previously reported to be citrullinated in vivo. Citrullination at site α84, α123, α129, α547, α573, α591, β334 and γ134 was identified in more than one patient, and these positions were therefore regarded as hotspots. Following citrullination of fibrinogen in vitro using human recombinant peptidylarginine deiminase 2 (PAD2), a total of 48 citrullination sites were identified, including 6 hitherto unreported in vitro citrullination sites. Twenty-one out of the 23 citrullination sites identified in vivo were also detected in vitro, supporting the validity of the identifications.

Significance: This work provides information about previously uncharacterized citrullination sites in synovial fluid fibrinogen from rheumatoid arthritis patients. Detection of these novel citrullination sites may prove to have diagnostic or prognostic value in RA and enhance our understanding of the immune pathogenesis.

Keywords: Citrullination; peptidylarginine deiminase; rheumatoid arthritis; mass spectrometry; post-translational modifications; synovial fluid

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 67

1. Introduction

Peptidyl-citrulline is a deiminated form of peptidyl-arginine, modified by the calcium-dependent enzyme peptidylarginine deiminase (PAD), of which five isoforms, PAD1–4 and PAD6 exist in humans.[1] PAD converts the guanidine group of protein-bound arginine residues to an ureido group, resulting in loss of a positive charge and a monoisotopic mass shift of +0.98 Da that can be detected by mass spectrometry.[2] This post-translational modification, referred to as citrullination, occurs during inflammation in general[3] and has been associated with several inflammatory diseases, in particular, rheumatoid arthritis (RA)[4] and multiple sclerosis,[5] as well as with various cancers.[6]

RA is a multifactorial, complex autoimmune disorder characterized by chronic inflammation of the synovium surrounding the joints, causing gradual destruction of bone and cartilage and, eventually, joint deformation.[7] Around two-thirds of RA patients produce anti-citrullinated protein antibodies (ACPAs) and express HLA-DR alleles containing the so-called shared epitope, which is capable of accommodating citrullinated peptides.[8–10] PAD2 and PAD4, expressed by various cells of hematopoietic origin,[1] have been identified in synovial tissue of RA patients in close association with tissue inflammation.[11–13] The presence of a variety of citrullinated proteins including fibrinogen, vimentin, and α-enolase has been demonstrated in the synovium.[14–17]

Fibrinogen is a 340 kDa protein consisting of two pairs of three polypeptide chains (α, β, and γ). Citrullinated forms of fibrinogen are usually recognized by ACPAs with results comparable to the anti-cyclic citrullinated peptide (CCP) test.[18–20] Several investigators have studied citrullinated fibrinogen in SF and serum obtained from peripheral blood from RA patients.[21–24] Deposition of fibrin in the joints of RA patients is a distinct characteristic of RA and eventually cause pannus formation and joint destruction.[25]

A number of citrullination sites have been identified following citrullination with PAD2 and PAD4 in vitro.[26–28] So far, however, only limited information is available regarding the degree of citrullination at individual sites.[23,31,35,43] We have here applied a mass spectrometry-based proteomics approach to estimate the relative degree of fibrinogen citrullination in SF from RA

patients and compared the site-specific patterns to those observed for fibrinogen citrullinated in vitro by PAD2.

2. Materials and methods

2.1.Patient material

SF samples were collected from four patients with RA, all fulfilling the American College of Rheumatology criteria for the diagnosis of RA.[33] Collection of SF was approved by the local ethics committee of the Institute of Rheumatology in Prague, Czech Republic, and all patients gave written informed consent. Data on age, sex, swollen joint count, tender joint count, erythrocyte sedimentation rate (ESR), plasma C-reactive protein (CRP), anti-CCP and rheumatoid factor (RF) were recorded (Table 1). Disease activity score including (DAS28) was calculated on the basis of ESR (DAS28-ESR).[34] SF from four anti-CCP-positive patients with varying CRP values, leukocyte count, and disease activity was selected for inclusion in this study.

2.2.Depletion and processing of synovial fluid

The four SF samples were centrifuged (1900 g, 10 min) to remove cells and stored at −80°C. The cell-free fractions were subjected to immunoglobulin (IgG) and albumin depletion as described[30] with a few modifications. Briefly, 100 µL SF was thawed at 4°C and mixed with 100 µL 100 mM EDTA, followed by centrifugation (14,000 g, 4°C, 20 min). The supernatant was diluted in PBS (pH 7.4) to a volume of 200 µL and kept on ice, while the pellet was stored for proteolytic digestion (see below). For depletion of IgG, 200 µL protein G Sepharose beads (GE Healthcare, Uppsala, Sweden) were transferred to a reactor unit with Teflon frits (25 µm pore size, MultiSynTech, Witten, Germany) followed by two washes in PBS. Diluted supernatant was added to the reactor, incubated (2½ h, 4°C), forced through the reactor unit by pressing the plunger into the barrel, and the IgG-depleted supernatant was collected. Beads in the reactor unit were incubated 2 × 10 min with 150 µL PBS at room temperature. After each incubation, the eluent was transferred into the tube containing IgG-depleted supernatant. Ice-cold acetone containing 10 % trichloroacetic acid (1 mL) was added to the IgG-depleted supernatant (200 µL), incubated (1½ h, 20°C), and centrifuged (14,000 g, 4°C, 30 min). The (IgG- and albumin-depleted) supernatant was transferred to another tube. The pellet was the albumin fraction. Supernatant and pellet were incubated (30 min) with 3 Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 69

mL and 1 mL of ice-cold acetone, respectively, and the supernatants were discarded after centrifugation.

2.3. In vitro citrullination of fibrinogen

Fibrinogen purified from human blood plasma (Calbiochem, Darmstadt, Germany) was diluted in deimination buffer (100 mM Tris-HCl, pH 7.6, 10 mM CaCl2, 1 mM dithiothreitol (DTT; Sigma- Aldrich)) to a final protein concentration of 1 mg/mL, and incubated overnight at 37°C with 300 ng/mL recombinant human PAD2, kindly provided by Ger J. Pruijn.[35] A control without PAD2 was incubated in parallel.

2.4.Proteolytic digestion and sample preparation

Fractions from the SF depletion process and in vitro citrullinated fibrinogen samples were dried (speed-vac) and re-suspended in 20 µL 8 M urea, 50 mM Tris-HCl pH 8.0 and 2 µL 450 mM DTT. Following incubation (45 min, R.T.), 4 µL 500 mM iodoacetamide (Sigma-Aldrich) was added, and the samples were incubated in the dark (30 min, R.T.). Then, 74 µL 10 mM NH4HCO3 and 3 µL 0.1 µg/µL LysC (Promega, Madison, USA) were added and incubated overnight (37°C). Acidified samples (10 % formic acid) were desalted on 3M Empore stage-tip reversed phase discs (Fisher Scientific, Bellefonte, USA) mounted in 200 µL pipette tips. Acidified peptides loaded onto stage- tip columns were equilibrated using 0.1 % trifluoroacetic acid (TFA; Sigma), followed by washing twice with 0.1 % TFA and eluted by 50 % acetonitrile (ACN; Fluka Analytical), lyophilized and re- dissolved in 1 % formic acid.

2.5.Mass spectrometry

Proteolytic digests of fibrinogen and SF samples were analyzed by nano-liquid chromatography- mass spectrometry (LC-MS/MS) using a Q-Exactive hybrid quadrupole-orbitrap mass spectrometer (Thermo Fisher Scientific) with an EASY-spray ion source (Thermo Fisher Scientific) coupled to an EASY-nLC 1000 liquid chromatograph (Thermo Fischer Scientific). Peptides were loaded onto EASY-spray columns (50 cm length, C18 reversed-phase material, 100 Å pore-size, 2 µm particle size, 75 µm ID) and eluted using 120 min gradients (6 to 95 % solvent B (90 % ACN and 0.1 % FA in water)) with a flow rate of 250 nL/min at 45°C. Resolution of 70,000, automatic gain control

(AGC) value of 3 x 106, maximum injection time (IT) of 20 ms and a scan range of 300 - 1,750 m/z were used for MS scans. MS/MS spectra were acquired in data-dependent mode (Top 10 method) with the exclusion of singly charged ions, resolution of 17,500, AGC value of 1 x 106, maximum IT of 60 ms, normalized HCD collision energy of 25 % and dynamic exclusion of 30 s. Database searches against the human subset of UniProt (July 2013) were performed using PEAKS Studio software (Version X, Bioinformatics solutions inc.) with the following parameters: parent mass error tolerance: 4.5 ppm, fragment mass error tolerance: 0.02 Da, enzyme: Lys C (2 missed cleavages). Citrullination (R), deamidation (NQ), oxidation (M) were set as variable modification [36] while carbamidomethylation (C) was set as a fixed modification. Maximum two variable modifications per peptide and 1 % peptide level false discovery rates (FDR) were applied. Only peptides identified independently in two technical replicates with an A-score value (related to the probability of correct modification site localization) of 20 and above were taken into consideration.[37] Citrullination of peptides was further validated by comparing the retention time shift from the non-citrullinated counterpart. The citrullinated peptides are more hydrophobic than the non-citrullinated counterparts due to the loss of the positive charge,[23] and longer retention times on reversed phase columns are therefore expected. The estimated relative degree of citrullination at individual sites was calculated as the ratio of peak areas (MS1 peak area extraction algorithm in the PEAKS software (Bioinformatics Solutions Inc.) was applied) of all identified peptides citrullinated at a particular site over the sum of peak areas for all identified peptides (citrullinated and non-citrullinated) covering the same site. For these quantitative assessments, only peptides having the same length and modifications (other than citrullination) were considered. Estimation of degree of citrullination was calculated as average of two technical replicates. In some cases, retention time shifts and degree of citrullination could not be calculated, since the corresponding unmodified peptide was not observed, or the intensity of the signal for the identified peptide was very low or overlapping with isotope envelops of co-eluting peptides, such sites were annotated as not quantified (NQ). The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD006749.

3. Result and Discussion

3.1. Citrullination sites in fibrinogen isolated from SF of RA patients Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 71

To study the citrullination pattern of fibrinogen in joints of RA patients, we used SF from four anti- CCP positive patients with varying degrees of disease activity, as determined by DAS28 score, cell counts and CRP levels (Table 1). Anti-CCP antibodies are specific for RA, and associated with more severe disease.

Patients 1 and 2 had high disease activity and high leukocyte counts in the aspired SF, indicating high inflammatory activity. Patient 3 had a high DAS28 score, but a relatively low CRP level and few leukocytes in SF, indicating a moderate inflammatory activity, and patient 4 had low inflammatory activity, based on all three parameters. SF is a complex biological fluid, with abundant albumin and IgG,[38] that may obscure analysis of low abundance citrullination and therefore were depleted from SF prior to proteolytic digestion and analysis by LC-MS.

The mass spectrometric analysis revealed that SF samples from the four RA patients contained fibrinogen with 23 citrullinated sites in total; 12 in the α chain, 8 in the β chain, and 3 in the γ chain (Fig. 1 and Table S1). Among these, 11 sites have also been reported to be citrullinated in vivo previously: in an identification study by Wang et al, citrullination of α160, α84, β72, β445 and γ134 were identified, and citrullination at α573 and β72 was identified by Hermansson et al and Tutturen et al. Similarly α84, α547, α573, α591, β53, β72, β267, β436, and β445 were observed by Tutturen et al and Hermansson et al.[23,31,35,37,38] In addition, we identified 14 citrullination sites that, to our knowledge, have not been previously reported in samples from RA patients: α42, α123, α129, α425, α426, α510, α512, β44, β246, β334, γ31, and γ40. Mass spectra of citrullinated and non-citrullinated peptides representing all novel citrullination sites are displayed in Fig. S1, and an example covering site α84 is shown in Fig. 2. The citrullinated peptides were further validated by comparing the retention time shift from the non-citrullinated counterpart. The citrullinated peptides are more hydrophobic than the non-citrullinated counterparts due to the loss of the positive charge,[23] and longer retention times on reversed phase columns are therefore expected. For example, the citrullinated peptide displayed in Fig. 2B elutes approximately 10 minutes later than the corresponding non-citrullinated peptide (Fig. 2A).

Citrullination at site α84, α123,α129, α547, α573, α591, β334 and γ134 was identified in more than one patient, and these sites were therefore regarded as hotspots. The citrullinated sites α573 and α591 have been previously reported to be major target epitopes of ACPAs,[39,40,42,43], and 71

binding of ACPAs to these sites have been detected already at the disease onset stage.[40,44] Moreover, the citrullinated site α425 identified here has also been observed formerly in an ELISA screening study as an ACPA-reactive site.[39] In the β-chain, sites β53 and β72 that were found citrullinated in the more inflamed patient were also previously reported to be associated with RA.[40,43] Citrullination of β53 has been reported previously in a single RA SF sample.[2]

We next compared the citrullination profile of SF fibrinogen from four patients included in this study with their individual DAS28 scores, plasma CRP levels, and content of leukocytes in SF. Taken together, these disease parameters of inflammation correlated well with the number of citrullinated sites in fibrinogen (Fig. 3) When the total number of citrullination sites in fibrinogen was compared between the four RA patients, it was evident that a higher number of citrullination sites were identified in the samples from the two patients with high degree of inflammation, i.e. patient 1 and patient 2 (Table 2). Twenty-two out of the total 23 citrullination sites identified in vivo were found in patient 2 (≈ 88%), while 7 out of 23 (≈ 28%) were seen in patient 1.

We also attempted to estimate the relative degree of citrullination at individual sites. Quantification of citrullinated peptides is very challenging for several reasons. Firstly, the signal response of citrullinated peptides is expected to be reduced compared to the non-citrullinated counterparts due to the loss the charge on the arginine sidechain, even though this effect may be counteracted to some extent by the increase in hydrophobicity associated with conversion of arginine to citrulline. Furthermore, the loss of this charge is also likely to influence the charge state distribution for citrullinated peptides. Secondly, citrulline is not a cleavage site for trypsin, the most commonly used protease in peptide mapping experiments. Consequently, a direct comparison between tryptic citrullinated/non-citrullinated peptides can in most cases not be made since the citrullinated peptide will be longer than the non-citrullinated counterpart. Several different strategies have previously been applied to address these issues for example through independent quantification of citrullinated and non-citrullinated peptides in different samples [23,31,35,43]. Here, the estimated degree of citrullination was calculated as the ratio of mass spectrometric peak areas of LysC digested citrullinated peptides covering a particular site and the sum of peak areas of the citrullinated and non-citrullinated peptides covering the same sites. Since LysC only cleaves peptides on the C- terminal side of lysine, citrullinated and non-citrullinated peptides will have the same length. An advantage of this approach is that the abundances of citrullinated and non-citrullinated peptides Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 73

within a sample can be compared, thus avoiding inter-sample variation which is a major source of error in label-free quantification. It should, however, be emphasized that only crude estimate of the degree of citrullination can be provided with the approach used here, since the mass spectrometric response factor and charge state distribution of the citrullinated and non-citrullinated forms of peptides is likely to differ significantly, as outlined above. The values reported here should therefore be regarded as relative values of the degree of citrullination, rather than absolute values of citrullination occupancy.

When the estimated relative degree of citrullination at different sites of fibrinogen was compared between the four RA patients, it became evident that values above 50 % were seen only in the two patients with high degree of inflammation and high SF cell counts, i.e. patients 1 and 2 (Table 2) at α129 and α591 respectively. At the hotspots α84, α123,α129, α573, α591and γ134 fibrinogen from only one of the patients displayed estimated relative degree of citrullination > 15 %. Site α591 displayed one of the highest observed estimated relative degree of citrullination (up to 66 %), indicating that this site may be of high pathogenic relevance, e.g. by being recognized by ACPAs in the joints, or by being presented in the context of HLA-DR molecules containing the shared epitope. The estimated relative degree of citrullination was also considerably higher for the hotspot α573 (≈43%); higher than previously reported 19 % for this site in SF from RA patients[30]. Interestingly, it has been reported that sites α573, α591 and β72 are detected by ACPAs already at the disease onset stage, while α573 is also detected at the later stages of the disease.[40,44] All these sites were identified in the present study. PAD2 induced high citrulline occupancy on site α591 in vitro (Fig. 4).

3.2. Citrullination pattern in fibrinogen induced by PAD2 In order to benchmark the citrullination sites identified in vivo, we attempted to identify the corresponding sites in fibrinogen in vitro after treatment with recombinant human PAD2, the more promiscuous of the PAD isoforms.[26,45] Using our mass spectrometry approach, it was possible to identify 48 citrullination sites on the three polypeptide chains of fibrinogen (Fig. 1 and Table S2). Twenty-one of the 23 sites identified in patient samples were also recognized in fibrinogen treated with PAD2, which corroborates the validity of the identifications (Fig. 1 and Fig. S2). In general, PAD2 appears to be very efficient in generating citrullinated sites in vitro, but the degree of citrullination occupancy varied strikingly from site to site (Fig. 3). For example, for the site α591, 73

previously shown to be recognized by ACPA,[27] the estimated relative degree of citrullination by PAD2 was ≈ 94 %, while it was below 1 % for a number of sites such as γ134 (Fig. 4). This further highlights the importance of quantifying citrullinated peptides instead of just identifying citrullination sites.

In addition to the sites identified in patient samples, the in vitro analysis also revealed several sites that, to our knowledge, have not been reported in previous studies[23,27,46], including α178, α190, α621, β158, β267, γ301. Mass spectra supporting the identification of these sites are displayed in Fig. S3.

4. Conclusion and remarks Using a mass-spectrometry-based proteomics approach, 48 citrullination sites in fibrinogen have been identified. Of these, 46 sites were citrullinated by PAD2 in vitro, and 23 were found in fibrinogen from RA patient synovial fluid. Several of the identified sites appear to be hotspots, being present in SF fibrinogen from more than one patient. Further studies with larger cohorts are needed to demonstrate whether these citrullination sites can serve as diagnostic and prognostic biomarkers reflecting disease progression and/or severity in patients with RA. The large difference in the estimated relative degree of citrullination for the different sites identified here underscores the importance of quantitative assessments in future explorations of the citrullinome.

Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 75

Appendix A. Supplementary material

Table S1. Citrullinated peptides identified from fibrinogen in SF of four RA patients.

Table S2. Citrullinated peptides identified from in vitro citrullinated fibrinogen.

Figure S1. Deconvoluted mass spectra of citrullinated and non-citrullinated peptides covering novel citrinullation sites in fibrinogen from RA patient synovial fluid samples.

Figure S2. Deconvoluted mass spectra of citrullinated and non-citrullinated peptides from fibrinogen citrullinated in vitro corresponding to the novel citrinullation sites identified in fibrinogen from RA patient synovial fluid samples.

Figure S3. Deconvoluted mass spectra of citrullinated and non-citrullinated peptides covering novel sites identified in fibrinogen citrullinated in vitro.

Supplementary Data 1. Detailed information about identified fibrinogen peptides derived from synovial fluid samples.

Supplementary Data 2. Detailed information about identified peptides from in vitro citrullinated fibrinogen.

Acknowledgments

The Q-Exactive Orbitrap mass spectrometers used in this study were granted by the Danish Council for Independent Research | Natural Sciences (grant number 11-106246) and the Velux foundation. CHN was supported by the Danish Council for Independent Research | Medical Sciences (grant number DFF - 7016-00233) The Danish Research Foundation, Nordic Bioscience A/S and the Technical University of Denmark are acknowledged for a joint PhD scholarship to MS. LS was supported by the Ministry of Health of the Czech Republic (MHCR) (grant #00023728). We would like to thank Lene Holberg Blicher and Anna Blicher for technical assistance.

References

[1] E.R. Vossenaar, A.J.W. Zendman, W.J. Van Venrooij, G.J.M. Pruijn, PAD , a growing family of citrullinating enzymes : genes , features and involvement in disease, Bioessays. 25 (2003) 1106–1118. doi:10.1002/bies.10357.

[2] A. Bennike, Tue.;Lauridsen, Kasper. B.;Olesen, Michael. Kruse.;Andersen, Vibeke.;Birkelund, Svend, Birkelund.;Stensballe, Optimizing the identification of citrullinated peptides by mass spectrometry: utilizing the inability of trypsin to cleave after citrullinated amino acids, J. Proteomics Bioinform. 06 (2013) 288–295. doi:10.4172/jpb.1000293.

[3] A.I. Makrygiannakis, D.; af Klint, E.; Lundberg, I.E.; Löfberg, R.; Ulfgren, A-K.; Klareskog, L.; Catrina, Citrullination is an inflammation-dependent process, Ann. Rheum. Dis. 65 (2006) 1219–1222. doi:10.1136/ard.2005.049403.

[4] T. Cantaert, L. De Rycke, T. Bongartz, E.L. Matteson, P.P. Tak, A.P. Nicholas, D. Baeten, Citrullinated proteins in rheumatoid arthritis, Arthritis Rheum. 54 (2006) 3381–3389. doi:10.1002/art.22206.

[5] M.A. Moscarello, F.G. Mastronardi, D.D. Wood, The role of citrullinated proteins suggests a novel mechanism in the pathogenesis of multiple sclerosis, Neurochem. Res. 32 (2007) 251–256. doi:10.1007/s11064-006-9144- 5.

[6] B.D. Cherrington, X. Zhang, J.L. McElwee, E. Morency, L.J. Anguish, S.A. Coonrod, Potential role for PAD2 in gene regulation in breast cancer cells, PLoS One. 7 (2012) 1–12. doi:10.1371/journal.pone.0041242.

[7] G.S. Firestein, Evolving concepts of rheumatoid arthritis, Nature. 423 (2003) 356–61. doi:10.1038/nature01661.

[8] J.A.. Hill, S.. Southwood, A.. Sette, A.M.. Jevnikar, D.A.. Bell, E. Cairns, Cutting edge: the conversion of arginine to citrulline allows for a high-affinity peptide interaction with the rheumatoid arthritis-associated HLA- DRB1*0401 MHC class II molecule., J. Immunol. 171 (2003) 538–541. doi:10.4049/jimmunol.171.2.538.

[9] E.R. Vossenaar, A.J. Zendman, W.J. Van Venrooij, Citrullination, a possible functional link between susceptibility genes and rheumatoid arthritis., Arthritis Res. Ther. 6 (2004) 1–5. doi:10.1186/ar1027.

[10] L.B.A.. van Schellekens, Gerard. A.;. de Jong, Ben A.W.; van den Hoogen, Frank. H.J.; van de Putte, W.J. Venrooij, Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritis – specific autoantibodies, J. Clin. Invest. 101 (1998) 273–281. doi:10.1172/JCI1316.

[11] C. Foulquier, M. Sebbag, C. Clavel, S. Chapuy-Regaud, R. Al Badine, M.C. Méchin, C. Vincent, R. Nachat, M. Yamada, H. Takahara, M. Simon, M. Guerrin, G. Serre, Peptidyl arginine deiminase type 2 (PAD-2) and PAD-4 but not PAD-1, PAD-3, and PAD-6 are expressed in rheumatoid arthritis synovium in close association with tissue inflammation, Arthritis Rheum. 56 (2007) 3541–3553. doi:10.1002/art.22983.

[12] D. Makrygiannakis, S. Revu, M. Engström, E. Klint, A.P. Nicholas, G.J.M. Pruijn, A.I. Catrina, Local administration of glucocorticoids decreases synovial citrullination in rheumatoid arthritis, Arthritis Res. Ther. 14 (2012) R20. doi:10.1186/ar3702.

[13] D. Damgaard, L. Senolt, C.H. Nielsen, Increased levels of peptidylarginine deiminase 2 in synovial fluid from anti-CCP-positive rheumatoid arthritis patients : Association with disease activity and inflammatory markers, Rheumatol. 55 (2016) 918–927. doi:10.1093/rheumatology/kev440.

[14] H. Bang, K. Egerer, A. Gauliard, K. Lüthke, P.E. Rudolph, G. Fredenhagen, W. Berg, E. Feist, G.R. Burmester, Mutation and citrullination modifies vimentin to a novel autoantigen for rheumatoid arthritis, Arthritis Rheum. 56 (2007) 2503–2511. doi:10.1002/art.22817. Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 77

[15] J.J.B.C. van Beers, C.M. Schwarte, J.. Stammen-Vogelzangs, E.. Oosterink, B.. Božič, G.J.M. Pruijn, The rheumatoid arthritis synovial fluid citrullinome reveals novel citrullinated epitopes in apolipoprotein E , myeloid nuclear differentiation antigen , and β-actin, Arthritis Rheum. 65 (2013) 69–80. doi:10.1002/art.37720.

[16] A. Kinloch, V. Tatzer, R. Wait, D. Peston, K. Lundberg, P. Donatien, D. Moyes, P.C. Taylor, P.J. Venables, Identification of citrullinated alpha-enolase as a candidate autoantigen in rheumatoid arthritis., Arthritis Res. Ther. 7 (2005) R1421-9. doi:10.1186/ar1845.

[17] C. Masson-Bessiere, M. Sebbag, E. Girbal-Neuhauser, L. Nogueira, C. Vincent, T. Senshu, G. Serre, The major synovial targets of the rheumatoid arthritis-specific antifilaggrin autoantibodies are deiminated forms of the alpha- and beta-chains of fibrin., J. Immunol. 166 (2001) 4177–4184. doi:10.4049/jimmunol.166.6.4177.

[18] L. Klareskog, A.I. Catrina, S. Paget, Rheumatoid arthritis, Lancet. 373 (2009) 659–672. doi:10.1016/S0140- 6736(09)60008-8.

[19] B. Vander Cruyssen, T. Cantaert, L. Nogueira, C. Clavel, L. De Rycke, A. Dendoven, M. Sebag, D. Deforce, C. Vincent, D. Elewaut, G. Serre, F. De Keyser, Diagnostic value of anti-human citrullinated fibrinogen ELISA and comparison with four other anti-citrullinated protein assays., Arthritis Res. Ther. 8 (2006) R122. doi:10.1186/ar2011.

[20] G.J. Pruijn, A. Wiik, W.J. van Venrooij, The use of citrullinated peptides and proteins for the diagnosis of rheumatoid arthritis., Arthritis Res. Ther. 12 (2010) 203. doi:10.1186/ar2903.

[21] O. Sanchez-Pernaute, M. Filkova, A. Gabucio, M. Klein, H. Maciejewska-Rodrigues, C. Ospelt, F. Brentano, B. a Michel, R.E. Gay, G. Herrero-Beaumont, S. Gay, M. Neidhart, A. Juengel, Citrullination enhances the pro- inflammatory response to fibrin in rheumatoid arthritis synovial fibroblasts., Ann. Rheum. Dis. 72 (2013) 1400– 6. doi:10.1136/annrheumdis-2012-201906.

[22] J. Sokolove, X. Zhao, P.E. Chandra, W.H. Robinson, Immune complexes containing citrullinated fibrinogen costimulate macrophages via toll-like receptor 4 and Fcγ receptor receptor, Arthritis Rheum. 63 (2011) 53–62. doi:10.1002/art.30081.

[23] R. Raijmakers, J.J. van Beers, M. El-Azzouny, N.F. Visser, B. Božič, G.J. Pruijn, A.J. Heck, Elevated levels of fibrinogen-derived endogenous citrullinated peptides in synovial fluid of rheumatoid arthritis patients, Arthritis Res Ther. 14 (2012) 1–10. doi:10.1186/ar3840.

[24] A. Suzuki, R.Y.O. Yamada, K. Yamamoto, Citrullination by peptidylarginine deiminase in rheumatoid arthritis, Ann N Y Acad Sci. 339 (2007) 323–339. doi:10.1196/annals.1422.034.

[25] G.H.-B. Sanchez-Pernaute, R Largo, E Calvo, M A Alvarez-Soria, J Egido, A fibrin based model for rheumatoid synovitis, Ann. Rheum. Dis. 62 (2003) 1135–1138. doi:10.1136/ard.2003.011767.

[26] K.Y. Makiko Nakayama-Hamadaa, Akari Suzuki, Kazuishi Kubota, Tomoko Takazawa, Mizuko Ohsaka, Reimi Kawaida, Mitsuru Ono, Atsushi Kasuya, Hidehiko Furukawa, Ryo Yamada,*, Comparison of enzymatic properties between hPADI2 and hPADI4, Biochem Biophys Res Commun. 327 (2005) 192–200. doi:10.1016/j.bbrc.2004.11.152.

[27] J.J.B.C. van Beers, R. Raijmakers, L.-E. Alexander, J. Stammen-Vogelzangs, A.M.C. Lokate, A.J.R. Heck, R.B.M. Schasfoort, G.J.M. Pruijn, Mapping of citrullinated fibrinogen B-cell epitopes in rheumatoid arthritis by imaging surface plasmon resonance., Arthritis Res. Ther. 12 (2010) R219. doi:10.1186/ar3205.

[28] S.. Blachere, Nathale. E.; Parveen, M.O.. Frank, H.. Dill, Brian. D.; Molina, D.E. Orange, High-Titer Rheumatoid Arthritis Antibodies Preferentially Bind Fibrinogen Citrullinated by Peptidylarginine Deiminase 4, ARTHRITIS Rheumatol. 69 (2017) 986–995. doi:10.1002/art.40035.

[29] M. Hermansson, K. Artemenko, E. Ossipova, H. Eriksson, J. Lengqvist, D. Makrygiannakis, A.I. Catrina, A.P. Nicholas, L. Klareskog, M. Savitski, R.A. Zubarev, P.J. Jakobsson, MS analysis of rheumatoid arthritic synovial tissue identifies specific citrullination sites on fibrinogen, Proteomics - Clin. Appl. 4 (2010) 511–518. doi:10.1002/prca.200900088.

[30] A.E. V Tutturen, B. Fleckenstein, G.A. De Souza, Assessing the citrullinome in rheumatoid arthritis synovial fluid with and without enrichment of citrullinated peptides, J. Proteome Res. 13 (2014) 2867–2873. doi:10.1021/pr500030x.

[31] S.-E. Ong, M. Mann, Mass spectrometry-based proteomics turns quantitative., Nat. Chem. Biol. 1 (2005) 252– 62. doi:10.1038/nchembio736.

[32] M.D.. Ceuleneer, M.. E.. D.. van Steendam, Katleen.: Dhaenens, D. Deforce, Quantification of Citrullination by Means of Skewed Isotope Distribution Pattern, J. Proteome Res. 11 (11) (2012) pp 5245–5251. doi:10.1021/pr3004453.

[33] F.C. Arnett, S.M. Edworthy, D.A. Bloch, D.J. Mcshane, J.F. Fries, N.S. Cooper, L.A. Healey, S.R. Kaplan, M.H. Liang, H.S. Luthra, T.A. Medsger, D.M. Mitchell, D.H. Neustadt, R.S. Pinals, J.G. Schaller, J.T. Sharp, R.L. Wilder, G.G. Hunder, The American rheumatism association 1987 revised criteria for the classification of rheumatoid arthritis, (1987).

[34] A.M. Van Gestel, C.J. Haagsma, P.L.C.M. Van Riel, Validation of rheumatoid arthritis improvement criteria that include simplified joint counts, Arthritis Rheum. (1998). doi:10.1002/1529-0131(199810)41:10<1845::AID- ART17>3.0.CO;2-K.

[35] R. Raijmakers, A.J.W. Zendman, W.V. Egberts, E.R. Vossenaar, J. Raats, C. Soede-huijbregts, F.P.J.T. Rutjes, P.A. Van Veelen, J.W. Drijfhout, G.J.M. Pruijn, Methylation of Arginine Residues Interferes with Citrullination by Peptidylarginine Deiminases in vitro, J Mol Biol. 367 (2007) 1118–1129. doi:10.1016/j.jmb.2007.01.054.

[36] C.-Y. Lee, H. Hahne, B. Kuster, K. Schnatbaum, T. Schmidt, M. Wilhelm, D.P. Zolg, D. Wang, M. Uhlén, U. Reimer, F. Pontén, Mining the Human Tissue Proteome for Protein Citrullination, Mol. Cell. Proteomics. 17 (2018) 1378–1391. doi:10.1074/mcp.ra118.000696.

[37] J.R.& S.P.G. Sean A Beausoleil, Judit Ville´n, Scott A Gerber, A probability-based approach for high-throughput protein phosphorylation analysis and site localization, 24 (2006) 1285–1292. doi:10.1038/nbt1240.

[38] D.S. Gibson, M.E. Rooney, The human synovial fluid proteome : A key factor in the pathology of joint disease, Proteomics Clin Appl. 1 (2007) 889–899. doi:10.1002/prca.200700044.

[39] M. Sebbag, N. Moinard, I. Auger, C. Clavel, J. Arnaud, L. Nogueira, J. Roudier, G. Serre, Epitopes of human fibrin recognized by the rheumatoid arthritis-specific autoantibodies to citrullinated proteins, Eur. J. Immunol. 36 (2006) 2250–2263. doi:10.1002/eji.200535790.

[40] C. Fernandes-Cerqueira, E. Ossipova, S. Gunasekera, M. Hansson, L. Mathsson, A.I. Catrina, Y. Sommarin, L. Klareskog, K. Lundberg, J. Rönnelid, U. Göransson, P.-J. Jakobsson, Targeting of anti-citrullinated protein/peptide antibodies in rheumatoid arthritis using peptides mimicking endogenously citrullinated fibrinogen antigens, Arthritis Res. Ther. 17 (2015) 155. doi:10.1186/s13075-015-0666-6.

[41] F. Wang, F.-F. Chen, W.-B. Gao, H.-Y. Wang, N.-W. Zhao, M. Xu, D.-Y. Gao, W. Yu, X.-L. Yan, J.-N. Zhao, and X.-J. Li, Identification of citrullinated peptides in the synovial fluid of patients with rheumatoid arthritis using LC- MALDI-TOF/TOF, Clin. Rheumatol. 35 (2016) 2185–2194. doi:10.1007/s10067-016-3247-4.

[42] D. Nielen, Markus. M. J .;van der Horst,A. R.;van Schaardenburg,D.;van der Horst-Bruinsma,I.E.;van de Stadt,R.J.;Aarden.L.;Dijkmans,B.A.C.;Hamann, Antibodies to citrullinated human fibrinogen (ACF) have diagnostic and prognostic value in early arthritis, Ann Rheum Dis. 64(8) (2005) 1199–1204. doi:10.1136/ard.2004.029389. Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 79

[43] J.J.B.C. van Beers, R. Raijmakers, L.-E. Alexander, J. Stammen-Vogelzangs, A.M.C. Lokate, A.J.R. Heck, R.B.M. Schasfoort, G.J.M. Pruijn, Mapping of citrullinated fibrinogen B-cell epitopes in rheumatoid arthritis by imaging surface plasmon resonance., Arthritis Res. Ther. 12 (2010) R219. doi:10.1186/ar3205.

[44] M.L. Pérez, M.J. Gómara, G. Ercilla, R. Sanmartí, I. Haro, Antibodies to citrullinated human fibrinogen synthetic peptides in diagnosing rheumatoid arthritis, J. Med. Chem. 50 (2007) 3573–3584. doi:10.1021/jm0701932.

[45] C. Assohou-Luty, R. Raijmakers, W.E. Benckhuijsen, J. Stammen-Vogelzangs, A. De Ru, P.A. Van Veelen, K.L.M.C. Franken, J.W. Drijfhout, G.J.M. Pruijn, The human peptidylarginine deiminases type 2 and type 4 have distinct substrate specificities, Biochim. Biophys. Acta - Proteins Proteomics. 1844 (2014) 829–836. doi:10.1016/j.bbapap.2014.02.019.

[46] K. Kubota, T. Yoneyama-Takazawa, K. Ichikawa, Determination of sites citrullinated by peptidylarginine deiminase using18O stable isotope labeling and mass spectrometry, Rapid Commun. Mass Spectrom. 19 (2005) 683–688. doi:10.1002/rcm.1842.

[47] A.M.. van Gestel, C.J.. Haagsma, P.L.C.M. van Riel, Validation of rheumatoid arthritis improvement criteria that include simplified joint counts, 41 (1998) 1845–1850.

Table 1. RA disease activity overview of patients used in this study.

Patient DAS28- CRP Anti-CCP Leukocytes/µl number ESR1 (mg/L) (U/mL) RF IgM (SF) Sex Age 1 6.32 86.74 2083.4 61.8 5712 M 61 2 6.33 92.49 880.4 15.4 21,624 M 66 3 5.51 15.61 891.7 400.6 36 M 53

4 3.05 2.63 874.8 300.2 149 F 69 Anti-CCP: Antibodies against the cyclic citrullinated peptide, CRP: C-reactive protein, DAS-28: Disease activity score evaluated from 28 joints, ESR: Erythrocyte sedimentation rate, Leu (SF): Leukocyte concentration in the synovial fluid, RF: IgM rheumatoid factor.

1DAS28 scores > 5.1, corresponding to moderate to severe disease activity, and scores ≤ 2.4 correspond to being in remission.[47] Mass spectrometry-based identification and quantification of posttranslational modification citrullination in rheumatoid arthritis 81

Table 2. Fibrinogen citrullination sites with an estimated relative degree of citrullination (in %) in SF samples from four patients with RA.

Peptide# Site* Patient 1∞ Patient 2∞ Patient 3∞ Patient 4∞

GLIDEVNQDFTNR(+.98)INK α84 39.70(±10.84) 6.39 (±2.11) 0.10(± 0.07) 27.55 (±1.09)

DSHSLTTNIMEILRGDFSSANNR(+.98)DNTYNRVSEDLRSRIEVLK α123 ND 25.69(± 8.41) NQ ND

DNTYNR(+.98)VSEDLRSRIEVLK α129 56.71 (± 9.75) 5.96 (± 1.34) NQ N.D

TFPGFFSPMLGEFVSETESR(+.98)GSESGIFTNTK α547 ND 8.29 (± 5.64) ND NQ

ESSSHHPGIAEFPSR(+.98)GK α573 ND 42.79 (±4.58) 0.12 (±0.01) N.D

QFTSSTSYNR(+.98)GDSTFESK α591 ND 66.13 (±2.12) 5.62 (±0.06) N.D

ISQLTR(+.98)MGPTELLIEMEDWKGDK β334 5.43 (± 4.66) NQ NQ NQ

YEASILTHDSSIR(+.98)YLQEIYNSNNQK γ134 NQ 17.33 (± 5.05) ND N.D

Lowest moderate Highest

# Detailed information about the identified peptides is available in SI Supplementary Data 2. * Hotspot sites are highlighted in grey. ∞ Average value of two technical replicates. ND = not detected. NQ = not quantified. The estimated relative degree of citrullination at some sites could not be quantified either because the corresponding unmodified peptide was not present or the isotope envelops of the identified peptides were overlapping with co-eluting peptides. Detailed information about the identified peptides is available in Supplementary Data 1.

Figures

A (α)

B (β).

C (γ).

Fig. 1. Overview of 48 citrullination sites identified in the α (A), β (B), and γ-chain (C) of fibrinogen (indicated with numbers in bold). Of these sites, 21 (green) were identified both in the in vitro citrullinated fibrinogen and in SF samples, 24 (red) were identified exclusively in the in vitro citrullinated fibrinogen, and 2 (blue) was identified only in SF samples. Detailed information about the identified peptides is available in Supplementary Data 1, 2.

Fig. 2. MS/MS spectra of the α chain fibrinogen peptide GLIDEVNQDFTNRINK in the non-citrullinated (A) and citrullinated (B) form (r=citrullinated residue α84). The b-ions and y-ions are displayed in blue and red, respectively. Inserts show the low m/z region of the spectra. 85

Fig. 3. Association of clinical and biochemical parameters with citrullination. The number of citrullinated sites in fibrinogen in synovial fluid from four patients with rheumatoid arthritis was determined by LC-MS and compared with the patients’ levels of (A) DAS28, (B) plasma CRP, and (C) leukocytes in synovial fluid. Pearson correlation coefficients and two-tailed p-values are shown.

100 90 80 70 60

40 (%) 30 20 10 0

Estimated Estimated degree of citrullination α84 α425,α426 α547 α591 γ134 Citrullination site

Fig. 4. Estimated relative degree of citrullination of selected sites in fibrinogen following incubation with PAD2 (orange). Control (grey) is the human fibrinogen incubated without PAD2. Values are based on two technical replicates and the estimated degree of citrullination at individual sites was calculated as the ratio of peak areas of citrullinated peptide covering a particular site to the sum of peak areas of the citrullinated and non-citrullinated peptides covering the same site. Detailed information about the identified peptides is available in Supplementary Data 2.

Supplementary material

Table S1. Identified citrullination sites in the three polypeptide chains of fibrinogen derived from synovial fluid of four RA patients.

Chain Peptide sequence of citrullinated peptides # Site*

α- GPRVVER(+.98)HQSAC(+57.02)K 42

α- GLIDEVNQDFTNR(+.98)INK 84

α- DSHSLTTNIMEILRGDFSSANNR(+.98)DNTYNR(+.98)VSEDLRSRIEVLK 123

α- DNTYNR(+.98)VSEDLRSRIEVLK 129

α- NVR(+.98)AQLVDMK 160

α- VSPGTR(+.98)R(+.98)EYHTEK 425,426

α- EVVTSEDGSDC(+57.02)PEAMDLGTLSGIGTLDGFR(+.98)HRHPDEAAFFDTASTGK 510

α- EVVTSEDGSDC(+57.02)PEAMDLGTLSGIGTLDGFR(+.98)HR(+.98)HPDEAAFFDTASTGK 510,512

α- TFPGFFSPMLGEFVSETESR(+.98)GSESGIFTNTK 547

α- ESSSHHPGIAEFPSR(+.98)GK 573

α- QFTSSTSYNR(+.98)GDSTFESK 591

β- VNDNEEGFFSAR(+.98)GHRPLDK 44

β- KR(+.98)EEAPSLRPAPPPISGGGYRARPAK 53

β- KREEAPSLRPAPPPISGGGYR(+.98)ARPAK 72

β- GGETSEMYLIQPDSSVKPYR(+.98)VYC(+57.02)DMNTENGGWTVIQNRQDGSVDFGRKWDPYK 267

β- PYRVYC(+57.02)DMNTENGGWTVIQNR(+.98)QDGSVDFGR(+.98)KWDPYK 294

β- ISQLTR(+.98)MGPTELLIEMEDWKGDK 334

β- EDGGGWWYNR(+.98)C(+57.02)HAANPNGR(+.98)YYWGGQYTWDMAK 436,445

β- EDGGGWWYNRC(+57.02)HAANPNGR(+.98)YYWGGQYTWDMAK 445

γ- YVATR(+.98)DNC(+57.02)C(+57.02)ILDER(+.98)FGSYC(+57.02)PTTC(+57.02)GIADFLSTYQTKVDK 31,40

γ- YEASILTHDSSIR(+.98)YLQEIYNSNNQK 134

#The mass increments of modified amino acids are indicated: Citrullination R (+0.98 Da), Carbamidomethylation C (+57.02 Da). Detailed information about the identified peptides is available in SI Supplementary Data 4. *Only peptides with Ascore ≥ 20 are displayed.

Table S2. Citrullination sites identified in the three polypeptide chains of in vitro citrullinated fibrinogen.

Chain Peptide sequence of citrullinated peptides# Site* α- ADSGEGDFLAEGGGVR(+.98)GPR(+.98)VVER(+.98)HQSAC(+57.02)K 35,38,42 α- CPSGCR(+.98)MK 69 α- GLIDEVNQDFTNR(+.98)INK 84 α DSHSLTTNIMEILR(+.98)GDFSSANNRDNTYNRVSEDLRSRIEVLK 114 α- DSHSLTTNIMEILRGDFSSANNR(+.98)DNTYNRVSEDLRSRIEVLK 123 α- NVR(+.98)AQLVDMK 160 α- I(+42.01)R(+.98)SC(+57.02)RGSC(+57.02)SRALAREVDLK 178 α- IR(+.98)SC(+57.02)R(+.98)GSC(+57.02)SR(+.98)ALAR(+.98)EVDLK 178,181,186,190 α- IRSC(+57.02)RGSC(+57.02)SR(+.98)ALAREVDLK 186 α- DLLPSR(+.98)DR(+.98)QHLPLIK 216,218 α- SGSSGPGSTGNR(+.98)NPGSSGTGGTATWK 308 α- SPGSGNAR(+.98)PNNPDWGTFEEVSGNVSPGTRREYHTEK 404 α- VSPGTR(+.98)R(+.98)EYHTEK 425,426 α- ELR(+.98)TGK 443 α- VTSGSTTTTR(+.98)R(+.98)SCSK 458,459 α- EVVTSEDGSDC(+57.02)PEAM(+15.99)DLGTLSGIGTLDGFR(+.98)HRHPDEAAFFDTASTGK 510 α- EVVTSEDGSDCPEAMDLGTLSGIGTLDGFR(+.98)HR(+.98)HPDEAAFFDTASTGK 510,512 α- EVVTSEDGSDCPEAMDLGTLSGIGTLDGFRHR(+.98)HPDEAAFFDTASTGK 512 α- TFPGFFSPMLGEFVSETESR(+.98)GSESGIFTNTK 547 α- ESSSHHPGIAEFPSR(+.98)GK 573 α- QFTSSTSYNR(+.98)GDSTFESK 591 α- MADEAGSEADHEGTHSTKR(+.98)GHAK 621 β- VNDNEEGFFSAR(+.98)GHRPLDK 44 β- VNDNEEGFFSAR(+.98)GHR(+.98)PLDK 44,47 β- KR(+.98)EEAPSLRPAPPPISGGGYRARPAK 53 β- KREEAPSLR(+.98)PAPPPISGGGYRARPAK 60 β- KREEAPSLR(+.98)PAPPPISGGGYR(+.98)ARPAK 60,72 β- KREEAPSLRPAPPPISGGGYR(+.98)ARPAK 72 β- KREEAPSLRPAPPPISGGGYR(+.98)AR(+.98)PAK 72,74 β- DLWQKR(+.98)QK 158 β- LESDVSAQMEYC(+57.02)R(+.98)TPC(+57.02)TVSC(+57.02)NIPVVSGK 224 β- ECEEIIR(+.98)KGGETSEMYLIQPDSSVK 246 β- PYR(+.98)VYCDMNTENGGWTVIQNRQDGSVDFGRKWDPYK 267 β- ISQLTR(+.98)MGPTELLIEMEDWKGDK 334 β- QC(+57.02)SKEDGGGWWYNR(+.98)C(+57.02)HAANPNGRYYWGGQYTWDMAK 436 β- EDGGGWWYNR(+.98)C(+57.02)HAANPNGR(+.98)YYWGGQYTWDM(+15.99)AK 436,445 β- EDGGGWWYNRC(+57.02)HAANPNGR(+.98)YYWGGQYTWDMAK 445 γ- YVATR(+.98)DNCCILDERFGSYCPTTCGIADFLSTYQTK 31 γ- YVATR(+.98)DNCCILDER(+.98)FGSYCPTTCGIADFLSTYQTK 31,40 γ- YEASILTHDSSIR(+.98)YLQEIYNSNNQK 134 γ- R(+.98)LDGSVDFKK 223 γ- YR(+.98)LTYAYFAGGDAGDAFDGFDFGDDPSDK 301 γ- IIPFNR(+.98)LTIGEGQQHHLGGAK 417

#The mass increments of modified amino acids are indicated: Citrullination R (+0.98 Da), Carbamidomethylation C(+57.02 Da), Oxidation M (+15.99 Da). Detailed information about the identified peptides is available in SI Supplementary Data 2). *Only peptides with Ascore ≥ 20 are displayed.

Figure S1. Deconvoluted mass spectra of citrullinated and non-citrullinated peptides covering novel citrinullation sites in fibrinogen from RA patient synovial fluid samples.

m/z= 509.2551

z= 3

RT shift= 4.3

m/z= 381.7010

z= 4

m/z= 812.0712

z= 6

RT shift= 1.8

m/z= 812.5729

z= 6

m/z= 770.0676

z= 3

RT shift = 7.1

m/z= 577.5760

z= 4

m/z= 726.3401

z= 3

RT shift = 12.3

m/z= 435.8136

z= 5

m/z= 1250.3090

z= 4

RT shift = 11.9

m/z= 1000.0593

z= 5

m/z= 730.3497

z= 3

RT shift = 8.2

m/z= 438.4160

z= 5

m/z= 789.5936

z= 5

RT = 58.2

m/z= 677.5896

z= 4

RT shift = 3.1

m/z= 897.4532

z= 3

m/z= 1341.2666

z= 3

RT shift = 7.7

m/z= 1005.7037

z= 4

Figure S2. Deconvoluted mass spectra from fibrinogen citrullinated in vitro corresponding to the novel citrinullation sites identified in fibrinogen from RA patient synovial fluid samples (see Figure S1). The y-ions are shown in red and b-ions are in blue. Selected ions are annotated in the spectra. Modified residues are shown in lower case.

m/z = 761.8562; z = 4; RT shift = 10.34

m/z = 609.2943; z = 5

100

m/z = 814.7354; z = 6; RT = 85.89

m/z = 983.7015; z = 4; RT = 50

101

m/z = 1250.3174; z = 4; RT shift = 10.4

m/z = 1003.45.31; z = 5

101

102

m/z = 548.0135; z = 4; RT shift = 7.7

m/z = 438.4157; z = 5

103

104

105

m/z = 677.5900

z =4

RT shift = 1.3

105

106

m/z = 902.7859; z =3

m/z = 1143.6246; z =2; RT shift = 5

107

m/z = 572.5637; z =4

107

108

Figure S3. Deconvoluted mass spectra covering novel sites identified in fibrinogen citrullinated in vitro. The y-ions are shown in red and b-ions are in blue. Selected ions are annotated in the spectra. Modified residues are shown in lower case.

109

110

m/z = 551.8022

z = 2

RT shift = 5.5

m/z = 551.3102

z = 2

111

m/z = 726.3403; z = 6; RT shift = 6.5

m/z = 871.2051; z = 5

111

112

113

Manuscript 2

Assessment of the citrullinome and relative degree of citrullination of proteins in synovial fluid from patients with rheumatoid arthritis

Mandvi Sharma1, Claus Henrik Nielsen2,3, Ladislav Senolt4, Birte Svensson1, Per Hägglund1#* Dres Damgaard2,3

1Department of Biotechnology and Biomedicine, Technical University of Denmark, Kgs. Lyngby, Denmark

2Institute for Inflammation Research, Center for Rheumatology and Spine Diseases, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark

3Section for Periodontology, Microbiology and Community Dentistry, Department of Odontology, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

4Institute of Rheumatology and Department of Rheumatology, 1st Faculty of Medicine, Charles University, Prague, Czech Republic

* Corresponding Author: [email protected] ; +45 35327900

#Present address: Department of Biomedical Sciences, Faculty of Health and Medical Sciences, Panum Institute, University of Copenhagen, Copenhagen, Denmark

113

114

Abstract

Citrullination is an inflammation-dependent protein modification considered to play an important role in rheumatoid arthritis (RA) and other chronic inflammatory diseases. Using a mass spectrometry-based approach, we identified citrullinated proteins in synovial fluid (SF) from four RA patients and estimated the relative degree of citrullination at individual amino acid positions. Overall, we identified 1105 proteins, 171 of which were citrullinated. Thirteen of those proteins were present in all four patients, including alpha-2-macroglobulin, antithrombin-III, apolipoprotein A-I and B-100, beta-2-microglobulin, complement factor H and C3, fibrinogen, immunoglobulin G, -A and -M, and serum albumin. A total of 276 citrullination sites were identified in all patients. Of these citrullinated peptides, 35 citrullinated peptides were eligible for calculating the estimated relative degree of citrullination ranging from 2% to 93% (at site 212 in apolipoprotein A-1). Patients with high disease activity gave a high number of citrullination sites and patients with low disease activity gave a low number of citrullination sites. In general, the number of citrullination sites identified and estimated relative degree of citrullination was higher in patients with high disease activity, but some peptides displayed the highest estimated relative degree of citrullination the patient with the least disease activity. The citrullinated proteins identified are potential autoantigens in RA.

Keywords: Citrullination; rheumatoid arthritis; mass spectrometry; post-translational modifications; synovial fluid, citrullination occupancy

115

Introduction

The post-translational conversion of peptidyl-arginine into peptidyl-citrulline (deamination) is catalyzed by peptidylarginine deiminase (PAD) of which there are five isoforms, PAD1–4 and PAD6 in humans [1]. PAD converts the guanidine group of protein-bound arginine residues to an ureido group, resulting in loss of charge and a mass shift of +0.98 Da [2]. The reaction is referred to as citrullination and may alter the function, stability, and antigenicity of a protein. The latter may lead to a breach of self-tolerance in rheumatoid arthritis (RA), where 2/3 of the patients produce autoantibodies reacting with citrullinated proteins (ACPAs) [3]. Protein citrullination has also been associated with other chronic inflammatory diseases such as Type 1 diabetes [4] and multiple sclerosis [5,6], as well as with various cancers [7]. Citrullination is further associated with inflammation in general [8], due to activation and/or release of PAD2 and PAD4 from cells of hematopoietic origin [1].

Multiple citrullinated proteins are potential autoantigens in RA, with filaggrin, fibrinogen, alpha- enolase, vimentin, and histones as most investigated candidates, but many other citrullinated proteins have been detected in the synovium of RA patients [2,9,10]. A few studies have been made for the identification of different citrullinated proteins in RA patients [2,9,11], but it is not clear if those proteins are abundantly citrullinated in the joints of individual patients, or if their citrullination is common among RA patients. Previous studies of the citrullinome have not included quantification of the citrulline occupancy at eligible amino acid positions. This information may indicate if certain proteins are major substrates for the PAD enzymes, which may be of pathogenic importance. We recently applied a mass spectrometry-based proteomics approach to estimate the degree of citrullination of fibrinogen found in synovial fluid (SF) from four RA patients and found that the relative degree of citrullination at any given site varies substantially (Sharma et al, submitted). Here, we report on the remaining citrullinome of these patients.

Methods

Patient material

SF samples were collected from four patients with RA, all fulfilling the American College of Rheumatology criteria for the diagnosis of RA [12]. All patients gave written informed consent and

115

116

collection of SF was approved by the local ethics committee of the Institute of Rheumatology in Prague, Czech Republic. Data on patient age, sex, plasma C-reactive protein (CRP), leukocyte count and Disease activity score (DAS28) scores were recorded (Table 1). DAS28 was calculated on the basis of ESR (DAS28-ESR) [13]. SF from four anti-CCP-positive patients with varying CRP values, leukocyte count, and disease activity was selected for this study.

Depletion and synovial fluid processing

The four SF patient samples were centrifuged (1900 g, 10 min) and stored at −80°C after removal of cells. The samples were subjected to albumin and immunoglobulin (IgG) depletion as described [11], with a few alterations. Briefly, 100 µL of each sample was thawed at 4°C and mixed with 100 µL 100 mM EDTA, followed by centrifugation (14,000 g, 4°C, 20 min). The supernatant was diluted in PBS (pH 7.4) to a volume of 200 µL and kept on ice for depletion, while the pellet was stored for proteolytic digestion (see below). For depletion of IgG, 200 µL protein G Sepharose beads (GE Healthcare, Uppsala, Sweden) were transferred to a reactor unit with Teflon frits (25 µm pore size, MultiSynTech, Witten, Germany) followed by two washes in PBS buffer. Diluted supernatant was added to the reactor unit, incubated (2½ h, 4°C), forced through the reactor unit, and the IgG-depleted supernatant was collected. Beads in the reactor unit were incubated (2 × 10 min, 150 µL PBS) at room temperature. After each incubation, the eluent was moved into the tube containing IgG-depleted supernatant. Ice-cold acetone with 10 % trichloroacetic acid (1 mL) was transferred to the IgG-depleted supernatant (200 µL), incubated (1½ h, 20°C), and centrifuged (14,000 g, 4°C, 30 min). The (IgG- and albumin-depleted) supernatant was moved to another tube. The pellet was the albumin fraction. Supernatant and pellet were incubated (30 min) with 3 mL and 1 mL of ice-cold acetone, respectively, and the supernatants were thrown out after centrifugation.

Sample preparation and proteolytic digestion

Fractions from the SF depletion process were dried (speed-vac) and re-dissolved in 20 µL 8 M urea, 50 mM Tris-HCl pH 8.0 and 2 µL 450 mM dithiothreitol (DTT). After incubation (45 min, R.T.), 4 µL 500 mM iodoacetamide (Sigma-Aldrich) was added, and the samples were incubated away from

light (30 min, R.T.). Then, 74 µL 10 mM NH4HCO3 and 3 µL 0.1 µg/µL LysC (Promega, Madison, 117

USA) were added and incubated overnight (37°C). Acidified samples (10 % formic acid) were desalted on 3M Empore stage-tip reversed phase discs (Fisher Scientific, Bellefonte, USA) mounted in 200 µL pipette tips. Acidified peptides loaded onto stage-tip columns. Stage-tip columns were equilibrated using 0.1 % trifluoroacetic acid (TFA; Sigma), followed by washing twice with 0.1 % TFA. Loaded peptides were eluted by 50 % acetonitrile (ACN; Fluka Analytical), lyophilized and re-dissolved in 1 % formic acid.

Mass spectrometry

Proteolytic digests of SF samples were analyzed by nano-liquid chromatography-mass spectrometry (LC-MS/MS). Q-Exactive hybrid quadrupole-orbitrap mass spectrometer (Thermo Fisher Scientific) with an EASY-spray ion source (Thermo Fisher Scientific) coupled to an EASY- nLC 1000 liquid chromatograph (Thermo Fischer Scientific) was used in this study. Digested peptides were loaded onto EASY-spray columns (50 cm length, C18 reversed-phase material, 100 Å pore-size, 2 µm particle size, 75 µm ID) and eluted using 120 min gradients (6 to 95 % solvent B (90 % ACN and 0.1 % FA in water)) with a flow rate of 250 nL/min at 45°C. For the MS scans resolution of 70,000, automatic gain control (AGC) value of 3 x 106, maximum injection time (IT) of 20 ms and a scan range of 300 - 1,750 m/z were used. MS/MS fragmentation spectra were acquired in data-dependent mode (Top 10 method) with the exclusion of singly charged ions, resolution of 17,500, AGC value of 1 x 106, maximum IT of 60 ms, normalized HCD collision energy of 25 % and dynamic exclusion of 30 s. Database searches against the human subset of UniProt (July 2013) were performed using PEAKS Studio software (Version X, Bioinformatics solutions inc.) using the following parameters: parent mass error tolerance: 15 ppm, fragment mass error tolerance: 0.02 Da, enzyme: Lys C (2 missed cleavages). Carbamidomethylation (C) was set as a fixed modification while citrullination (R), deamidation (NQ), oxidation (M) were set as variable modification. 1 % peptide level false discovery rates (FDR) and a maximum of two variable modifications per peptide were applied. Peptides showing an A-score value (related to the probability of correct modification site localization) of 20 and above were only taken into consideration [14]. The citrullinated peptides are more hydrophobic than the non-citrullinated counterparts due to the loss of the positive charge, [15] and longer retention times on reversed phase columns are therefore expected. The estimated degree of citrullination at individual sites was calculated as the ratio of peak areas (MS1 peak area was applied) of all identified peptides 117

118

citrullinated at a particular site over the sum of peak areas for all identified peptides (independent of citrullination) covering the same site and with the same length in citrullinated and native form. (see example in SI Supplementary Data 1). In some cases, degree of citrullination could not be calculated, since the corresponding unmodified peptide was not observed, or the intensity of the signal for the identified peptide was too low or overlapping with isotope envelops of co-eluting peptides. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD006749.

Results

We examined the citrullinome of SFs from four anti-CCP positive patients, who were characterized with respect to disease activity, cell count in SF and CRP levels in the blood (Table 1). Patient 1 and 2 had high DAS28 score, high leucocyte counts and high CRP. Patient 3 had mid/high DAS28, mid-CRP, and low leukocyte count. Patient 4 had low values of these three parameters.

Table 1. Disease activity overview of patients used in this study.

Patient DAS28- CRP Leukocytes/µl number ESR1 (mg/L) (SF)

1 6.32 86.74 5712

2 6.33 92.49 21,624

3 5.51 15.61 36

4 3.05 2.63 149

CRP: C-reactive protein, DAS-28: Disease activity score evaluated from 28 joints, Leu (SF): Leukocyte concentration in the synovial fluid, RF: IgM rheumatoid factor. 1DAS28 scores > 5.1, corresponds to moderate to severe disease activity, and scores ≤ 2.4 corresponds to being in remission.

We identified 1105 proteins among the four SF samples, i.e. 304 in SF 1, 926 in SF 2, 423 in SF 3 and 285 in SF4 (Figure 1A). Amongst these, we found 159 proteins shared in all four samples. Overall, 171 unique citrullinated proteins were identified (Supplementary Table S1). Out of these, 13 proteins including alpha-2-macroglobulin, antithrombin-III, apolipoprotein A-I, apolipoprotein 119

B-100, beta-2-microglobulin, complement factors H and C3, IgA, IgG, IgM and serum albumin were present in citrullinated form in all four patients (Table 2).

Table 2- List of citrullinated proteins identified in all four patients.

Accession Protein

sp|P01023|A2MG_HUMAN Alpha-2-macroglobulin

sp|P01008|ANT3_HUMAN Antithrombin-III

sp|P02647|APOA1_HUMAN Apolipoprotein A-I

sp|P04114|APOB_HUMAN Apolipoprotein B-100

sp|P61769|B2MG_HUMAN Beta-2-microglobulin

sp|P01024|CO3_HUMAN Complement C3

sp|P08603|CFAH_HUMAN Complement factor H

sp|P02671-2|FIBA_HUMAN Fibrinogen alpha chain

sp|P02675|FIBB_HUMAN Fibrinogen beta chain

sp|P06396-2|GELS_HUMAN Gelsolin

sp|P01876|IGHA1_HUMAN Ig alpha-1 chain C region

sp|P01860|IGHG3_HUMAN Ig gamma-3 chain C region

sp|P04220|MUCB_HUMAN Ig mu heavy chain disease protein

sp|P02768|ALBU_HUMAN Serum albumin

From the 171 citrullinated proteins, we identified 276 citrullination sites (Table S2). The patient with the highest disease activity (DAS28: 6.33) and leukocyte count, patient 2, displayed the highest number (167) of citrullination sites. Patient 1 with the second highest disease activity (DAS28: 6.32) and leukocyte count had the second highest number (140) of citrullination sites. The sample from patient 3 (DAS28: 5.51) contained 97 citrullination sites, while patient 4 with the lowest disease activity (DAS28: 3.05) and leukocyte count had 56 citrullination sites.

119

120

Among the 477 unique citrullinated peptides identified, only 20 - derived from proteins such as apolipoprotein A1, haptoglobin, C3, and fibrinogen - were shared among the four samples (Figure 1B). Among these citrullinated peptides, it was possible to estimate the relative degree of citrullination for 9 peptides that were present in all four SF samples (Table 3).

In addition to a site (α84) in the fibrinogen alpha chain, which we identified previously (Sharma et al.submitted), we estimated the relative degree of citrullination for 8 different peptides derived from 6 different proteins present in all four SF samples. Surprisingly, SF4 that was low on all disease parameters and on identified citrullination site count showed an estimated relative degree of citrullination at several sites comparable to that of SF1 and higher than those of SF2 and SF3. A few sites, e.g. in apolipoprotein A-I, haptoglobin, and serum albumin, showed an estimated relative degree of citrullination above 50% in more than one sample. A citrullinated peptide from fibrinogen, an established autoantigen, showed estimated relative degree of citrullination ranging from 0.1˗ 40 % at site α84 and apolipoprotein A-1 showed 2˗ 93 % occupancy at site 212.

A. B.

Figure 1 –Distribution of protein identifications (A) and citrullinated peptides (B) among the four patients in this study.

121

Table 3 – Common citrullinated peptides with an estimated relative degree of citrullination (in %) in SF samples from four patients with RA. Measurement was performed in duplicate, the mean and standard deviation of the estimated relative degree of citrullination is shown.

Highest Lowest

We extended the analysis to also include peptides where the occupancy could be calculated in 3 out of the 4 SF samples, yielding additional five peptides each containing a citrullination site from four different proteins (apolipoprotein, haptoglobin, immunoglobulin kappa chain c region, and serum amyloid A-1 protein) (Table 4). Moreover, considering peptides where the occupancy could be estimated in at least 2 out of the 4 SF samples, 15 additional citrullinated peptides, derived from 11 different proteins were included (Table 5).

SF1 showed high occupancy at several sites, whereas a more even distribution in the number of sites and citrulline occupancy were found for SF2-4.

Table 4 - Common citrullinated peptides with an estimated relative degree of citrullination (in %) in SF samples from three patients with RA. Measurement was performed in duplicate, the mean and standard deviation of the estimated relative degree of citrullination is shown.

Highest Lowest

121

122

Table 5 - Common citrullinated peptides with an estimated degree of citrullination (in %) in SF samples from two most inflamed patients. Measurement was performed in duplicate, the mean and standard deviation of the estimated relative degree of citrullination is shown.

Highest Lowest

Discussion A number of citrullinated proteins [16] and peptides have been shown to be a target of ACPAs from RA patients [2,15,17–22]. However, it is unknown which proteins are commonly citrullinated in SF of RA patients and thereby potentially constitute common self-antigens.

In this present study, we identify a total of 171 citrullinated proteins containing 477 citrullinated peptides with 276 citrullination sites in SF from four RA patients. Importantly, we also determine which sites that have a high relative degree of citrullination based on the approach we previously applied for fibrinogen from SF (Sharma et al. submitted).

Van Beers and colleagues showed that selected proteins such as apolipoproteins, complement proteins, fibrinogen, histones were found citrullinated in multiple SFs of RA patients [2]. The majority of these have been reported in the citrullinated form in SF before [2,9], but it was never clear if these proteins have sites that are abundantly citrullinated. Surprisingly, we did not detect citrullinated α-enolase or vimentin which is among the best-described autoantigens in RA [19,23]. Some proteins, which were found in citrullinated form, play roles in the immune system, including 123

immunoglobulins and complement C3. [24]. Serum amyloid A-1 protein is expressed as an acute phase response against a trauma such as tissue injury, chronic inflammatory conditions such as amyloidosis or infection [25]. The physiological and pathophysiological implications remain to be investigated, be it changes in functionality or generation of novel autoantigenic epitopes.

We extended the findings of Van Beer et al. and other groups by identifying citrullinated sites in proteins such as alpha-2-macroglobulin, apolipoprotein A-I, haptoglobin, immunoglobulins, albumin, and serum amyloid A-1 protein [2]. We detected a number of citrullination sites previously reported, with high citrulline occupancy such as Ig kappa chain C region, leucine-rich alpha-2-glycoprotein, serum amyloid A-1 protein, and transthyretin [2,9,26].

Identification of abundant citrullinated sites is important in relation to determining pathogenic citrullinated proteins in RA, e.g. proteins that can be bound by ACPAs to form immune complexes or be presented to T cells by antigen presenting cells. On the other hand, SF4 had the lowest number of identified citrullinated sites, and it appears that we find a correlation between the number of citrullination sites with disease activity and inflammation (leukocyte count). Our sample size is too small for detailed statistical analysis, but it is noteworthy that we find higher relative degree of citrullination in several sites of SF4 as compared to SF2 and SF3. SF4 was the subject with the lowest disease activity and a low leukocyte count. We cannot rule out that those sites we identify solely with high occupancy in SF1 and SF2 are more pathogenic.

Overall, by using a simple mass-spectrometry-based proteomics approach, 171 citrullinated proteins, 13 of which are common in all four patients, and 477 citrullinated peptides containing 276 citrullination sites from SF have been identified in SFs in this study. The 13 common citrullinated proteins found in all four SFs can be immunologically relevant and may be potential autoantigens in RA. We observed a general association between a number of citrullination sites identified with inflammation and disease activity. Further studies with larger sample size are needed to look further into if high citrulline occupancy at any given site is linked to high disease activity and other parameters such as bone erosion.

123

124

Appendix A. Supplementary material

Table S1. Citrullinated proteins identified in all four RA patient synovial fluid samples

Table S2. Citrullinated peptides identified in all four RA patient synovial fluid samples

Figure S3. Mass spectra of citrullinated and non-citrullinated peptides covering quantified citrinullation sites in proteins from RA patient synovial fluid samples

Acknowledgments

The Q-Exactive Orbitrap mass spectrometers used in this study were granted by the Danish Council for Independent Research | Natural Sciences (grant number 11-106246) and the Velux foundation. CHN was supported by the Danish Council for Independent Research | Medical Sciences (grant number DFF - 7016-00233) The Danish Research Foundation, Nordic Bioscience A/S and the Technical University of Denmark are acknowledged for a joint Ph.D. scholarship to MS. LS was supported by the Ministry of Health of the Czech Republic (MHCR) (grant #00023728). We would like to thank Lene Holberg Blicher and Anna Blicher for technical assistance.

References [1] E.R. Vossenaar, A.J.W. Zendman, W.J. Van Venrooij, G.J.M. Pruijn, PAD , a growing family of citrullinating enzymes : genes , features and involvement in disease, Bioessays. 25 (2003) 1106–1118. doi:10.1002/bies.10357.

[2] J.J.B.C. van Beers, C.M. Schwarte, J.. Stammen-Vogelzangs, E.. Oosterink, B.. Božič, G.J.M. Pruijn, The rheumatoid arthritis synovial fluid citrullinome reveals novel citrullinated epitopes in apolipoprotein E , myeloid nuclear differentiation antigen , and β-actin, Arthritis Rheum. 65 (2013) 69–80. doi:10.1002/art.37720.

[3] G.A. Schellekens, B.A.W. de Jong, F.H.J. van den Hoogen, L.B.A. van de Putte, W.J. Van, Venrooij, Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritis – specific autoantibodies, J. Clin. Invest. 101 (1998) 273–281. doi:10.1172/JCI1316.

[4] D. Rondas, I. Crèvecoeur, W. D’Hertog, G.B. Ferreira, A. Staes, A.D. Garg, D.L. Eizirik, P. Agostinis, K. Gevaert, L. Overbergh, C. Mathieu, Citrullinated glucose-regulated protein 78 is an autoantigen in type 1 diabetes, Diabetes. 64 (2015) 573–586. doi:10.2337/db14-0621. 125

[5] A.A. Musse, Z. Li, C.A. Ackerley, D. Bienzle, H. Lei, R. Poma, G. Harauz, M.A. Moscarello, F.G. Mastronardi, Peptidylarginine deiminase 2 (PAD2) overexpression in transgenic mice leads to myelin loss in the central nervous system, Dis. Model. Mech. 240 (2008) 229–240. doi:10.1242/dmm.00072.

[6] M.A. Moscarello, F.G. Mastronardi, D.D. Wood, The role of citrullinated proteins suggests a novel mechanism in the pathogenesis of multiple sclerosis, Neurochem. Res. 32 (2007) 251–256. doi:10.1007/s11064-006-9144-5.

[7] S. Mohanan, B.D. Cherrington, S. Horibata, J.L. McElwee, P.R. Thompson, S.A. Coonrod, Potential Role of Peptidylarginine Deiminase Enzymes and Protein Citrullination in Cancer Pathogenesis, Biochem. Res. Int. 2012 (2012) 1–11. doi:10.1155/2012/895343.

[8] A.I. Makrygiannakis, D.; af Klint, E.; Lundberg, I.E.; Löfberg, R.; Ulfgren, A-K.; Klareskog, L.; Catrina, Citrullination is an inflammation-dependent process, Ann. Rheum. Dis. 65 (2006) 1219– 1222. doi:10.1136/ard.2005.049403.

[9] F. Wang, F.-F. Chen, W.-B. Gao, H.-Y. Wang, N.-W. Zhao, M. Xu, D.-Y. Gao, W. Yu, X.-L. Yan, J.- N. Zhao, and X.-J. Li, Identification of citrullinated peptides in the synovial fluid of patients with rheumatoid arthritis using LC-MALDI-TOF/TOF, Clin. Rheumatol. 35 (2016) 2185–2194. doi:10.1007/s10067-016-3247-4.

[10] A.J. Ytterberg, V. Joshua, G. Reynisdottir, N.K. Tarasova, D. Rutishauser, E. Ossipova, A. Haj Hensvold, A. Eklund, C.M. Sköld, J. Grunewald, V. Malmström, P.J. Jakobsson, J. Rönnelid, L. Padyukov, R.A. Zubarev, L. Klareskog, A.I. Catrina, Shared immunological targets in the lungs and joints of patients with rheumatoid arthritis: Identification and validation, Ann. Rheum. Dis. 74 (2015) 1772–1777. doi:10.1136/annrheumdis-2013-204912.

[11] A.E. V Tutturen, B. Fleckenstein, G.A. De Souza, Assessing the citrullinome in rheumatoid arthritis synovial fluid with and without enrichment of citrullinated peptides, J. Proteome Res. 13 (2014) 2867–2873. doi:10.1021/pr500030x.

[12] F.C. Arnett, S.M. Edworthy, D.A. Bloch, D.J. Mcshane, J.F. Fries, N.S. Cooper, L.A. Healey, S.R. Kaplan, M.H. Liang, H.S. Luthra, T.A. Medsger, D.M. Mitchell, D.H. Neustadt, R.S. Pinals, J.G. Schaller, J.T. Sharp, R.L. Wilder, G.G. Hunder, The American rheumatism association 1987 revised criteria for the classification of rheumatoid arthritis, (1987).

[13] A.M. Van Gestel, C.J. Haagsma, P.L.C.M. Van Riel, Validation of rheumatoid arthritis improvement criteria that include simplified joint counts, Arthritis Rheum. (1998). doi:10.1002/1529- 0131(199810)41:10<1845::AID-ART17>3.0.CO;2-K. 125

126

[14] J.R.& S.P.G. Sean A Beausoleil, Judit Ville´n, Scott A Gerber, A probability-based approach for high-throughput protein phosphorylation analysis and site localization, 24 (2006) 1285–1292. doi:10.1038/nbt1240.

[15] R. Raijmakers, J.J. van Beers, M. El-Azzouny, N.F. Visser, B. Božič, G.J. Pruijn, A.J. Heck, Elevated levels of fibrinogen-derived endogenous citrullinated peptides in synovial fluid of rheumatoid arthritis patients, Arthritis Res Ther. 14 (2012) 1–10. doi:10.1186/ar3840.

[16] D. Damgaard, M. Bawadekar, L. Senolt, A. Stensballe, M.A. Shelef, C.H. Nielsen, Relative efficiencies of peptidylarginine deiminase 2 and 4 in generating target sites for anti-citrullinated protein antibodies in fibrinogen, alpha-enolase and histone H3, PLoS One. 13 (2018) 1–14. doi:10.1371/journal.pone.0203214.

[17] D. van der Woude, A.H.M. van der Helm-van Mil, The target of ACPA, Rheumatol. (United Kingdom). 55 (2016) 1711–1713. doi:10.1093/rheumatology/kew270.

[18] Y. Takizawa, A. Suzuki, T. Sawada, M. Ohsaka, T. Inoue, R. Yamada, K. Yamamoto, Citrullinated fibrinogen detected as a soluble citrullinated autoantigen in rheumatoid arthritis synovial fluids, Ann Rheum Dis. 65 (2006) 1013–1020. doi:10.1136/ard.2005.044743.

[19] A. Kinloch, V. Tatzer, R. Wait, D. Peston, K. Lundberg, P. Donatien, D. Moyes, P.C. Taylor, P.J. Venables, Identification of citrullinated alpha-enolase as a candidate autoantigen in rheumatoid arthritis., Arthritis Res. Ther. 7 (2005) R1421-9. doi:10.1186/ar1845.

[20] E.R. Vossenaar, N. Després, E. Lapointe, A. Van Der Heijden, M. Lora, T. Senshu, W.J. Van Venrooij, H.A. Ménard, Rheumatoid arthritis specific anti-Sa antibodies target citrullinated vimentin, Arthritis Res. Ther. 6 (2004) R142–R150. doi:10.1186/ar1050.

[21] H. Burkhardt, B. Sehnert, R. Bockermann, Å. Engström, J.R. Kalden, R. Holmdahl, Humoral immune response to citrullinated collagen type II determinants in early rheumatoid arthritis, Eur. J. Immunol. 35 (2005) 1643–1652. doi:10.1002/eji.200526000.

[22] X. Chang, R. Yamada, A. Suzuki, Y. Kochi, T. Sawada, K. Yamamoto, Citrullination of fibronectin in rheumatoid arthritis synovial tissue, Rheumatology. 44 (2005) 1374–1382. doi:10.1093/rheumatology/kei023.

[23] K. van Steendam, K. Tilleman, D. Deforce, The relevance of citrullinated vimentin in the production of antibodies against citrullinated proteins and the pathogenesis of rheumatoid arthritis, Rheumatology. 50 (2011) 830–837. doi:10.1093/rheumatology/keq419. 127

[24] H.W.J. Schroeder, L. Cavacini, Structure and Function of Immunoglobulins, J. Allergy Clin. Immunol. 125 (2010) S41–S52. doi:10.1016/j.jaci.2009.09.046.Structure.

[25] L. Sun, R.D. Ye, Serum amyloid A1: Structure, function and gene polymorphism, Gene. 583 (2016) 48–57. doi:10.1016/j.gene.2016.02.044.

[26] T. Bennike, K.B. Lauridsen, M.K. Olesen, V. Andersen, S. Birkelund, A. Stensballe, Optimizing the identification of citrullinated peptides by mass spectrometry: utilizing the inability of trypsin to cleave after citrullinated amino acids, J. Proteomics Bioinform. 06 (2013) 288–295. doi:10.4172/jpb.1000293.

[27] M.A. Christophorou, G. Castelo-Branco, R.P. Halley-Stott, C.S. Oliveira, R. Loos, A. Radzisheuskaya, K.A. Mowen, P. Bertone, J.C.R. Silva, M. Zernicka-Goetz, M.L. Nielsen, J.B. Gurdon, T. Kouzarides, Citrullination regulates pluripotency and histone H1 binding to chromatin, Nature. 507 (2014) 104–108. doi:10.1038/nature12942.

127

128

Supplementary material

Table S1: List of citrullinated protein in RA patient samples.

Accession Description E9PDH4_HUMAN 1-phosphatidylinositol 3-phosphate 5-kinase (Fragment) FYV1_HUMAN 1-phosphatidylinositol 3-phosphate 5-kinase B7ZKJ8_HUMAN 35 kDa inter-alpha-trypsin inhibitor heavy chain H4 Q5T8M8_HUMAN Actin alpha skeletal muscle E7EVS6_HUMAN Actin cytoplasmic 1 (Fragment) ACTB_HUMAN Actin cytoplasmic 1 I3L3R2_HUMAN Actin cytoplasmic 2 N-terminally processed (Fragment) ACTG_HUMAN Actin cytoplasmic 2 ABRA_HUMAN Actin-binding Rho-activating protein A1AG1_HUMAN Alpha-1-acid glycoprotein 1 A1AG2_HUMAN Alpha-1-acid glycoprotein 2 G3V595_HUMAN Alpha-1-antichymotrypsin (Fragment) AACT_HUMAN Alpha-1-antichymotrypsin A1AT_HUMAN Alpha-1-antitrypsin A1BG_HUMAN Alpha-1B-glycoprotein A2MG_HUMAN Alpha-2-macroglobulin ANT3_HUMAN Antithrombin-III APOA1_HUMAN Apolipoprotein A-I APOA4_HUMAN Apolipoprotein A-IV APOB_HUMAN Apolipoprotein B-100 APOC1_HUMAN Apolipoprotein C-I APOC3_HUMAN Apolipoprotein C-III H0Y8C9_HUMAN ATP-binding cassette sub-family A member 2 (Fragment) ATRN_HUMAN Attractin APOH_HUMAN Beta-2-glycoprotein 1 F5H6I0_HUMAN Beta-2-microglobulin form pI 5.3 B2MG_HUMAN Beta-2-microglobulin BRCA2_HUMAN Breast cancer type 2 susceptibility protein F5GXS0_HUMAN C4b-B C4BPA_HUMAN C4b-binding protein alpha chain CATA_HUMAN Catalase H0YD13_HUMAN CD44 antigen (Fragment) 129

E7EPC6_HUMAN CD44 antigen CERU_HUMAN Ceruloplasmin CLUS_HUMAN Clusterin Q5FBE1_HUMAN Coagulation factor IX COF2_HUMAN Cofilin-2 CO3_HUMAN Complement C3 B0UZ83_HUMAN Complement C4 beta chain CO4A_HUMAN Complement C4-A CO4B_HUMAN Complement C4-B CO5_HUMAN Complement C5 CO8A_HUMAN Complement component C8 alpha chain F5H7G1_HUMAN Complement component C8 beta chain CO9_HUMAN Complement component C9 B4E1Z4_HUMAN Complement factor B CFAD_HUMAN Complement factor D CFAH_HUMAN Complement factor H FHR4_HUMAN Complement factor H-related protein 4 FHR5_HUMAN Complement factor H-related protein 5 G3XAM2_HUMAN Complement factor I light chain CFAI_HUMAN Complement factor I COR1A_HUMAN Coronin-1A CRP_HUMAN C-reactive protein E9PN83_HUMAN Disks large homolog 2 DYH10_HUMAN Dynein heavy chain 10 axonemal EF1A1_HUMAN Elongation factor 1-alpha 1 FRPD4_HUMAN FERM and PDZ domain-containing protein 4 FIBA_HUMAN Fibrinogen alpha chain FIBB_HUMAN Fibrinogen beta chain FIBG_HUMAN Fibrinogen gamma chain FINC_HUMAN Fibronectin B4DTT5_HUMAN Follistatin-related protein 1 LG3BP_HUMAN Galectin-3-binding protein GELS_HUMAN Gelsolin J3QR68_HUMAN Haptoglobin (Fragment) HPT_HUMAN Haptoglobin HBB_HUMAN Hemoglobin subunit beta HEMO_HUMAN Hemopexin HEP2_HUMAN Heparin cofactor 2 H12_HUMAN Histone H1.2 H13_HUMAN Histone H1.3 H14_HUMAN Histone H1.4 H31T_HUMAN Histone H3.1t

129

130

Q0VAC5_HUMAN HP protein IGHA1_HUMAN Ig alpha-1 chain C region IGHA2_HUMAN Ig alpha-2 chain C region IGHG1_HUMAN Ig gamma-1 chain C region IGHG2_HUMAN Ig gamma-2 chain C region IGHG3_HUMAN Ig gamma-3 chain C region IGHG4_HUMAN Ig gamma-4 chain C region HV320_HUMAN Ig heavy chain V-III region GAL IGKC_HUMAN Ig kappa chain C region LV202_HUMAN Ig lambda chain V-II region NEI LAC2_HUMAN Ig lambda-2 chain C regions LAC6_HUMAN Ig lambda-6 chain C region LAC7_HUMAN Ig lambda-7 chain C region IGHM_HUMAN Ig mu chain C region MUCB_HUMAN Ig mu heavy chain disease protein IGLL5_HUMAN Immunoglobulin lambda-like polypeptide 5 F5H7E1_HUMAN Inter-alpha-trypsin inhibitor heavy chain H1 ITIH2_HUMAN Inter-alpha-trypsin inhibitor heavy chain H2 E7ET33_HUMAN Inter-alpha-trypsin inhibitor heavy chain H3 ITIH4_HUMAN Inter-alpha-trypsin inhibitor heavy chain H4 RTEL1_HUMAN Isoform 1 of Regulator of telomere elongation helicase 1 CD44_HUMAN Isoform 10 of CD44 antigen FINC_HUMAN Isoform 10 of Fibronectin CD44_HUMAN Isoform 11 of CD44 antigen FINC_HUMAN Isoform 11 of Fibronectin CD44_HUMAN Isoform 13 of CD44 antigen FINC_HUMAN Isoform 13 of Fibronectin CD44_HUMAN Isoform 14 of CD44 antigen FINC_HUMAN Isoform 15 of Fibronectin CD44_HUMAN Isoform 16 of CD44 antigen CD44_HUMAN Isoform 17 of CD44 antigen CD44_HUMAN Isoform 18 of CD44 antigen APOL1_HUMAN Isoform 2 of Apolipoprotein L1 ATRN_HUMAN Isoform 2 of Attractin CLUS_HUMAN Isoform 2 of Clusterin CO4A_HUMAN Isoform 2 of Complement C4-A FHR4_HUMAN Isoform 2 of Complement factor H-related protein 4 DYH10_HUMAN Isoform 2 of Dynein heavy chain 10 axonemal FIBA_HUMAN Isoform 2 of Fibrinogen alpha chain GELS_HUMAN Isoform 2 of Gelsolin IGHM_HUMAN Isoform 2 of Ig mu chain C region ITIH3_HUMAN Isoform 2 of Inter-alpha-trypsin inhibitor heavy chain H3 LSP1_HUMAN Isoform 2 of Lymphocyte-specific protein 1 131

SREC2_HUMAN Isoform 2 of Scavenger receptor class F member 2 TBCD1_HUMAN Isoform 2 of TBC1 domain family member 1 VTDB_HUMAN Isoform 2 of Vitamin D-binding protein APOL1_HUMAN Isoform 3 of Apolipoprotein L1 ATRN_HUMAN Isoform 3 of Attractin FINC_HUMAN Isoform 3 of Fibronectin GELS_HUMAN Isoform 3 of Gelsolin LSP1_HUMAN Isoform 3 of Lymphocyte-specific protein 1 NCOR2_HUMAN Isoform 3 of Nuclear receptor corepressor 2 Isoform 3 of Structural maintenance of chromosomes protein SMC1B_HUMAN 1B CLUS_HUMAN Isoform 4 of Clusterin FINC_HUMAN Isoform 4 of Fibronectin CLUS_HUMAN Isoform 5 of Clusterin FINC_HUMAN Isoform 5 of Fibronectin CD44_HUMAN Isoform 6 of CD44 antigen FINC_HUMAN Isoform 6 of Fibronectin CD44_HUMAN Isoform 7 of CD44 antigen FINC_HUMAN Isoform 7 of Fibronectin CD44_HUMAN Isoform 8 of CD44 antigen FBLN1_HUMAN Isoform A of Fibulin-1 FBLN1_HUMAN Isoform C of Fibulin-1 FIBG_HUMAN Isoform Gamma-A of Fibrinogen gamma chain MYL6_HUMAN Isoform Smooth muscle of Myosin light polypeptide 6 KNG1_HUMAN Kininogen-1 A2GL_HUMAN Leucine-rich alpha-2-glycoprotein LUM_HUMAN Lumican LSP1_HUMAN Lymphocyte-specific protein 1 MNDA_HUMAN Myeloid cell nuclear differentiation antigen F8W1R7_HUMAN Myosin light polypeptide 6 PI4KA_HUMAN Phosphatidylinositol 4-kinase alpha IC1_HUMAN Plasma protease C1 inhibitor S10A8_HUMAN Protein S100-A8 S10A9_HUMAN Protein S100-A9 PRG4_HUMAN Proteoglycan 4 THRB_HUMAN Prothrombin EF1A3_HUMAN Putative elongation factor 1-alpha-like 3 SREC2_HUMAN Scavenger receptor class F member 2 E9PFD9_HUMAN Semaphorin-4D NEK5_HUMAN Serine/threonine-protein kinase Nek5 TRFE_HUMAN Serotransferrin H0YA55_HUMAN Serum albumin (Fragment) ALBU_HUMAN Serum albumin 131

132

SAA1_HUMAN Serum amyloid A-1 protein SAA2_HUMAN Serum amyloid A-2 protein SAA4_HUMAN Serum amyloid A-4 protein SAMP_HUMAN Serum amyloid P-component E9PAX1_HUMAN SKI/DACH domain-containing protein 1 C9JIE2_HUMAN TBC1 domain family member 1 (Fragment) E9PHK0_HUMAN Tetranectin TTHY_HUMAN Transthyretin E9PE77_HUMAN Ugl-Y3 VTNC_HUMAN Vitronectin ZA2G_HUMAN Zinc-alpha-2-glycoprotein 133

Table S2. Citrullinated peptides identified in all four RA patient synovial fluid samples

Accession Peptide sp|P02768|ALBU_HUMAN AAC(+57.02)LLPKLDELR(+.98)DEGK sp|P02768|ALBU_HUMAN AAC(+57.02)LLPKLDELR(+.98)DEGKASSAK sp|P01011|AACT_HUMAN ADLSGITGAR(+.98)NLAVSQVVHK sp|P02787|TRFE_HUMAN ADRDQYELLC(+57.02)LDNTR(+.98)KPVDEYK sp|P00738|HPT_HUMAN AEVGRVGYVSGWGR(+.98)NANFK sp|P02768|ALBU_HUMAN AEVSKLVTDLTKVHTEC(+57.02)C(+57.02)HGDLLEC(+57.02)ADDR(+.98)ADLAK sp|P01024|CO3_HUMAN AFSDR(+.98)NTLIIYLDK sp|P01024|CO3_HUMAN AGDFLEANYMNLQR(+.98)SYTVAIAGYALAQMGRLK sp|Q92954-3|PRG4_HUMAN AIGPSQTHTIR(+.98)IQYSPAR(+.98)LAYQDK sp|P01023|A2MG_HUMAN AIGYLNTGYQR(+.98)QLNYK sp|P01861|IGHG4_HUMAN AKGQPR(+.98)EPQVYTLPPSQEEMTK sp|P02647|APOA1_HUMAN AKPALEDLR(+.98)QGLLPVLESFK sp|P02647|APOA1_HUMAN AKPALEDLR(+.98)QGLLPVLESFKVSFLSALEEYTK sp|P01011|AACT_HUMAN AKWEMPFDPQDTHQSR(+.98)FYLSK sp|P04003|C4BPA_HUMAN ALC(+57.02)RKPELVNGR(+.98)LSVDK sp|P01023|A2MG_HUMAN ALLAYAFALAGNQDKR(+.98)K sp|P06727|APOA4_HUMAN ALVQQMEQLR(+.98)QK sp|P00450|CERU_HUMAN ALYLQYTDETFR(+.98)TTIEK APDAGGC(+57.02)LHADPDLGVLC(+57.02)PTGC(+57.02)QLQEALLQQER(+.98)PIR(+.98)NSVDELNNNVEAVSQTSSSSFQY sp|P02675|FIBB_HUMAN MYLLK sp|P02765|FETUA_HUMAN APHGPGLIYR(+.98)QPNC(+57.02)DDPETEEAALVAIDYINQNLPWGYK sp|P04114|APOB_HUMAN AQIPILR(+.98)M(+15.99)NFKQELNGNTK 133

134

sp|P01011|AACT_HUMAN AR(+.98)NLAVSQVVHK sp|P01042|KNG1_HUMAN AR(+.98)VQVVAGKK sp|P02654|APOC1_HUMAN ARELISR(+.98)IKQSELSAK sp|P01834|IGKC_HUMAN ASVVC(+57.02)LLNNFYPR(+.98)EAK sp|P06396|GELS_HUMAN ATASR(+.98)GASQAGAPQGRVPEAR(+.98)PNSMVVEHPEFLK sp|P06396|GELS_HUMAN ATASR(+.98)GASQAGAPQGRVPEARPNSMVVEHPEFLK sp|P04114|APOB_HUMAN ATFQTPDFIVPLTDLR(+.98)IPSVQINFK tr|F5GY56|F5GY56_HUMAN ATVLTTER(+.98)K sp|P02749|APOH_HUMAN ATVVYQGER(+.98)VK sp|P02749|APOH_HUMAN ATVVYQGER(+.98)VKIQEK sp|P02768|ALBU_HUMAN C(+57.02)ASLQKFGER(+.98)AFKAWAVARLSQRFPK sp|P02768|ALBU_HUMAN C(+57.02)C(+57.02)TESLVNR(+.98)R(+.98)PC(+57.02)FSALEVDETYVPKEFNAETFTFHADIC(+57.02)TLSEK sp|P02787|TRFE_HUMAN C(+57.02)HLAR(+.98)APNHAVVTRK sp|P02749|APOH_HUMAN C(+57.02)PFPSR(+.98)PDNGFVNYPAKPTLYYK sp|P02749|APOH_HUMAN C(+57.02)PFPSR(+.98)PDNGFVNYPAKPTLYYKDK sp|P00738|HPT_HUMAN C(+57.02)PKPPEIAHGYVEHSVR(+.98)YQC(+57.02)K sp|P02787|TRFE_HUMAN C(+57.02)QSFR(+.98)DHMK sp|P01024|CO3_HUMAN C(+57.02)R(+.98)EALKLEEKK sp|P10909-5|CLUS_HUMAN C(+57.02)R(+.98)EILSVDC(+57.02)STNNPSQAK tr|F8WDX4|F8WDX4_HUMAN C(+57.02)YFPYLENGYNQNHGR(+.98)K sp|P02768|ALBU_HUMAN DAHKSEVAHR(+.98)FKDLGEENFK sp|P02656|APOC3_HUMAN DALSSVQESQVAQQAR(+.98)GWVTDGFSSLK sp|P02656|APOC3_HUMAN DALSSVQESQVAQQAR(+.98)GWVTDGFSSLKDYWSTVK sp|P02787|TRFE_HUMAN DC(+57.02)HLAQVPSHTVVAR(+.98)SM(+15.99)GGKEDLIWELLNQAQEHFGKDK sp|P02787|TRFE_HUMAN DC(+57.02)HLAQVPSHTVVAR(+.98)SMGGKEDLIWELLNQAQEHFGK sp|P02787|TRFE_HUMAN DC(+57.02)HLAQVPSHTVVAR(+.98)SMGGKEDLIWELLNQAQEHFGKDK sp|P02788-2|TRFL_HUMAN DC(+57.02)HLARVPSHAVVAR(+.98)SVNGKEDAIWNLLRQAQEK sp|P02647|APOA1_HUMAN DEPPQSPWDR(+.98)VK 135

sp|P02647|APOA1_HUMAN DEPPQSPWDR(+.98)VKDLATVYVDVLK sp|P08254|MMP3_HUMAN DGAAR(+.98)GEDTSMNLVQK tr|B4E1Z4|B4E1Z4_HUMAN DISEVVTPR(+.98)FLC(+57.02)TGGVSPYADPNTC(+57.02)RGDSGGPLIVHK sp|P06312|KV401_HUMAN DIVMTQSPDSLAVSLGER(+.98)ATINC(+57.02)K sp|P01024|CO3_HUMAN DKNR(+.98)WEDPGK sp|P02647|APOA1_HUMAN DLATVYVDVLKDSGR(+.98)DYVSQFEGSALGK sp|P02647|APOA1_HUMAN DLATVYVDVLKDSGR(+.98)DYVSQFEGSALGKQLNLK sp|P02787|TRFE_HUMAN DLLFR(+.98)DDTVC(+57.02)LAK sp|P02750|A2GL_HUMAN DLLLPQPDLR(+.98)YLFLNGNK sp|P06727|APOA4_HUMAN DLR(+.98)DKVNSFFSTFK sp|P06727|APOA4_HUMAN DLR(+.98)DKVNSFFSTFKEK sp|P01023|A2MG_HUMAN DLTGFPGPLNDQDNEDC(+57.02)INR(+.98)HNVYINGITYTPVSSTNEK sp|P08603|CFAH_HUMAN DNPYIPNGDYSPLR(+.98)IK sp|P02751-10|FINC_HUMAN DNRGNLLQC(+57.02)IC(+57.02)TGNGR(+.98)GEWK sp|P01023|A2MG_HUMAN DNSVHWER(+.98)PQKPK sp|P02671|FIBA_HUMAN DNTYNR(+.98)VSEDLRSRIEVLK sp|P0DJI8|SAA1_HUMAN DPNHFR(+.98)PAGLPEKY sp|P07737|PROF1_HUMAN DR(+.98)SSFYVNGLTLGGQK sp|P02787|TRFE_HUMAN DSGFQMNQLR(+.98)GK sp|P02787|TRFE_HUMAN DSGFQMNQLR(+.98)GKK sp|P02647|APOA1_HUMAN DSGR(+.98)DYVSQFEGSALGK sp|P02647|APOA1_HUMAN DSGR(+.98)DYVSQFEGSALGKQLNLK sp|P02671|FIBA_HUMAN DSHSLTTNIMEILRGDFSSANNR(+.98)DNTYNRVSEDLRSRIEVLK sp|P01860|IGHG3_HUMAN DTLM(+15.99)ISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEVQFK sp|P01857|IGHG1_HUMAN DTLMISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEVK sp|P01857|IGHG1_HUMAN DTLMISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEVKFNWYVDGVEVHNAK

135

136

sp|P01860|IGHG3_HUMAN DTLMISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEVQFK sp|P01859|IGHG2_HUMAN DTLMISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEVQFNWYVDGVEVHNAK sp|O43866|CD5L_HUMAN DVAVLC(+57.02)R(+.98)ELGC(+57.02)GAASGTPSGILYEPPAEKEQK sp|P01876|IGHA1_HUMAN DVLVR(+.98)WLQGSQELPR(+.98)EK sp|P01876|IGHA1_HUMAN DVLVR(+.98)WLQGSQELPREK sp|P01876|IGHA1_HUMAN DVLVRWLQGSQELPR(+.98)EK sp|P00738|HPT_HUMAN DYAEVGR(+.98)VGYVSGWGR(+.98)NANFK sp|P00738|HPT_HUMAN DYAEVGR(+.98)VGYVSGWGR(+.98)NANFKFTDHLK sp|P00738|HPT_HUMAN DYAEVGRVGYVSGWGR(+.98)NANFK sp|Q8N0Z2|ABRA_HUMAN DYELAMSTR(+.98)LHK sp|P02787|TRFE_HUMAN DYELLC(+57.02)LDGTR(+.98)KPVEEYANC(+57.02)HLARAPNHAVVTRK sp|P02760|AMBP_HUMAN EC(+57.02)LQTC(+57.02)R(+.98)TVAAC(+57.02)NLPIVRGPC(+57.02)RAFIQLWAFDAVKGK sp|P02675|FIBB_HUMAN EDGGGWWYNR(+.98)C(+57.02)HAANPNGR(+.98)YYWGGQYTWDMAK sp|P02675|FIBB_HUMAN EDGGGWWYNRC(+57.02)HAANPNGR(+.98)YYWGGQYTWDMAK sp|P01024|CO3_HUMAN EDIPPADLSDQVPDTESETR(+.98)ILLQGTPVAQMTEDAVDAERLK sp|P02774|VTDB_HUMAN EFSHLGKEDFTSLSLVLYSR(+.98)KFPSGTFEQVSQLVK tr|F8W1R7|F8W1R7_HUMAN EGNGTVMGAEIR(+.98)HVLVTLGEK sp|P02787|TRFE_HUMAN EGYYGYTGAFR(+.98)C(+57.02)LVEKGDVAFVK sp|P01009|A1AT_HUMAN ELDR(+.98)DTVFALVNYIFFK sp|P01009|A1AT_HUMAN ELDR(+.98)DTVFALVNYIFFKGKWERPFEVK sp|P02768|ALBU_HUMAN ELFEQLGEYKFQNALLVR(+.98)YTK sp|P04114|APOB_HUMAN ENFAGEATLQR(+.98)IYSLWEHSTK sp|P02647|APOA1_HUMAN ENGGAR(+.98)LAEYHAK sp|P02647|APOA1_HUMAN ENGGAR(+.98)LAEYHAKATEHLSTLSEK sp|P01023|A2MG_HUMAN EQAPHC(+57.02)IC(+57.02)ANGR(+.98)QTVSWAVTPK sp|P19652|A1AG2_HUMAN EQLGEFYEALDC(+57.02)LC(+57.02)IPR(+.98)SDVMYTDWKK sp|P02763|A1AG1_HUMAN EQLGEFYEALDC(+57.02)LR(+.98)IPKSDVVYTDWK sp|P02763|A1AG1_HUMAN EQLGEFYEALDC(+57.02)LR(+.98)IPKSDVVYTDWKK 137

sp|P01024|CO3_HUMAN EQR(+.98)ISLPESLK sp|P16401|H15_HUMAN ER(+.98)NGLSLAALK sp|P02768|ALBU_HUMAN ER(+.98)QIKKQTALVELVK sp|P16403|H12_HUMAN ER(+.98)SGVSLAALK sp|P16403|H12_HUMAN ER(+.98)SGVSLAALKK sp|P00450|CERU_HUMAN ERGPEEEHLGILGPVIWAEVGDTIR(+.98)VTFHNK sp|P02671-2|FIBA_HUMAN ESSSHHPGIAEFPSR(+.98)GK sp|P02647|APOA1_HUMAN ETEGLR(+.98)QEM(+15.99)SK sp|P02647|APOA1_HUMAN ETEGLR(+.98)QEM(+15.99)SKDLEEVK sp|P02647|APOA1_HUMAN ETEGLR(+.98)QEMSK sp|P02647|APOA1_HUMAN ETEGLR(+.98)QEMSKDLEEVK sp|P02671|FIBA_HUMAN EVVTSEDGSDC(+57.02)PEAMDLGTLSGIGTLDGFR(+.98)HR(+.98)HPDEAAFFDTASTGK sp|P02671|FIBA_HUMAN EVVTSEDGSDC(+57.02)PEAMDLGTLSGIGTLDGFR(+.98)HRHPDEAAFFDTASTGK sp|P08670|VIME_HUMAN FADLSEAANR(+.98)NNDALR(+.98)QAK sp|P08670|VIME_HUMAN FADLSEAANR(+.98)NNDALRQAK sp|Q92496-2|FHR4_HUMAN FC(+57.02)DMPVFENSR(+.98)AK sp|P02751-14|FINC_HUMAN FGFC(+57.02)PMAAHEEIC(+57.02)TTNEGVMYR(+.98)IGDQWDK sp|P02788-2|TRFL_HUMAN FGR(+.98)NGSDC(+57.02)PDK sp|P02749|APOH_HUMAN FPSR(+.98)PDNGFVNYPAKPTLYYK sp|P02749|APOH_HUMAN FPSR(+.98)PDNGFVNYPAKPTLYYKDK sp|P02768|ALBU_HUMAN FQNALLVR(+.98)YTK sp|P02768|ALBU_HUMAN FQNALLVR(+.98)YTKK sp|P01023|A2MG_HUMAN FR(+.98)VVSMDENFHPLNELIPLVYIQDPK sp|P01023|A2MG_HUMAN FRVVSMDENFHPLNELIPLVYIQDPKGNR(+.98)IAQWQSFQLEGGLK sp|P01011|AACT_HUMAN FSISR(+.98)DYNLNDILLQLGIEEAFTSK sp|P01011|AACT_HUMAN FSISR(+.98)DYNLNDILLQLGIEEAFTSKADLSGITGARNLAVSQVVHK

137

138

sp|P04114|APOB_HUMAN FSR(+.98)NYQLYK sp|P00738|HPT_HUMAN FTDHLKYVMLPVADQDQC(+57.02)IR(+.98)HYEGSTVPEKK tr|E9PE77|E9PE77_HUMAN FTQVTPTSLSAQWTPPNVQLTGYR(+.98)VR(+.98)VTPK sp|P05156|CFAI_HUMAN FYGNR(+.98)FYEK sp|P01011|AACT_HUMAN GAR(+.98)NLAVSQVVHK sp|P00450|CERU_HUMAN GAYPLSIEPIGVR(+.98)FNK sp|P00738|HPT_HUMAN GC(+57.02)PKPPEIAHGYVEHSVR(+.98)YQC(+57.02)K sp|P00738|HPT_HUMAN GC(+57.02)PKPPEIAHGYVEHSVR(+.98)YQC(+57.02)KNYYK tr|B4E1Z4|B4E1Z4_HUMAN GGSFR(+.98)LLQEGQALEYVC(+57.02)PSGFYPYPVQTRTC(+57.02)RSTGSWSTLK sp|P06396|GELS_HUMAN GGTPVQSR(+.98)VVQGK sp|P51587|BRCA2_HUMAN GIQLADGGWLIPSNDGKAGKEEFYR(+.98)ALC(+57.02)DTPGVDPK sp|P01011|AACT_HUMAN GITGAR(+.98)NLAVSQVVHK sp|P01009|A1AT_HUMAN GKWER(+.98)PFEVKDTEEEDFHVDQVTTVK sp|P02671|FIBA_HUMAN GLIDEVNQDFTNR(+.98)INK sp|P02671|FIBA_HUMAN GLIDEVNQDFTNR(+.98)INKLK sp|P01023|A2MG_HUMAN GNR(+.98)IAQWQSFQLEGGLK sp|P01876|IGHA1_HUMAN GNTFR(+.98)PEVHLLPPPSEELALNELVTLTC(+57.02)LARGFSPK sp|P02741|CRP_HUMAN GPFSPNVLNWR(+.98)ALK sp|P02671|FIBA_HUMAN GPRVVER(+.98)HQSAC(+57.02)K sp|P01861|IGHG4_HUMAN GQPR(+.98)EPQVYTLPPSQEEMTK sp|P01860|IGHG3_HUMAN GQPR(+.98)EPQVYTLPPSREEM(+15.99)TK sp|P01860|IGHG3_HUMAN GQPR(+.98)EPQVYTLPPSREEMTK sp|P01860|IGHG3_HUMAN GQPR(+.98)EPQVYTLPPSREEMTKNQVSLTC(+57.02)LVK sp|P02751-3|FINC_HUMAN GQPR(+.98)QYNVGPSVSK sp|P01860|IGHG3_HUMAN GQPREPQVYTLPPSR(+.98)EEMTKNQVSLTC(+57.02)LVK sp|P02647|APOA1_HUMAN GR(+.98)DYVSQFEGSALGK sp|P01023|A2MG_HUMAN GR(+.98)QTVSWAVTPK sp|P01781|HV320_HUMAN GRFTISR(+.98)DNAK 139

sp|P01762|HV301_HUMAN GRFTISR(+.98)DNAQK sp|P00738|HPT_HUMAN GRVGYVSGWGR(+.98)NANFK sp|P01024|CO3_HUMAN GSPMYSIITPNILR(+.98)LESEETMVLEAHDAQGDVPVTVTVHDFPGK sp|P01024|CO3_HUMAN GSPMYSIITPNILR(+.98)LESEETMVLEAHDAQGDVPVTVTVHDFPGKKLVLSSEK sp|P01834|IGKC_HUMAN GTASVVC(+57.02)LLNNFYPR(+.98)EAK sp|P02763|A1AG1_HUMAN GTISR(+.98)YVGGQEHFAHLLILRDTK sp|P06396|GELS_HUMAN GTPVQSR(+.98)VVQGK sp|P01009|A1AT_HUMAN GTQGKIVDLVKELDR(+.98)DTVFALVNYIFFK sp|P01871|IGHM_HUMAN GVALHRPDVYLLPPAR(+.98)EQLNLRESATITC(+57.02)LVTGFSPADVFVQWMQRGQPLSPEK sp|P00738|HPT_HUMAN GWGR(+.98)NANFK sp|P02675|FIBB_HUMAN GWTVIQNR(+.98)QDGSVDFGRKWDPYK sp|P01024|CO3_HUMAN GYTQQLAFR(+.98)QPSSAFAAFVK sp|P00738|HPT_HUMAN GYVEHSVR(+.98)YQC(+57.02)K sp|P00738|HPT_HUMAN GYVEHSVR(+.98)YQC(+57.02)KNYYK sp|P00738|HPT_HUMAN GYVSGWGR(+.98)NANFK sp|P02774|VTDB_HUMAN HLSLLTTLSNR(+.98)VC(+57.02)SQYAAYGEK sp|P07225|PROS_HUMAN HNDIR(+.98)AHSC(+57.02)PSVWKK sp|P02768|ALBU_HUMAN HPEAKR(+.98)M(+15.99)PC(+57.02)AEDYLSVVLNQLC(+57.02)VLHEKTPVSDRVTK sp|P02768|ALBU_HUMAN HPEAKR(+.98)MPC(+57.02)AEDYLSVVLNQLC(+57.02)VLHEK sp|P02768|ALBU_HUMAN HPEAKR(+.98)MPC(+57.02)AEDYLSVVLNQLC(+57.02)VLHEKTPVSDRVTK sp|P01011|AACT_HUMAN HPNSPLDEENLTQENQDR(+.98)GTHVDLGLASANVDFAFSLYK sp|P02774|VTDB_HUMAN HQPQEFPTYVEPTNDEIC(+57.02)EAFR(+.98)K sp|P05155|IC1_HUMAN HR(+.98)LEDMEQALSPSVFK sp|A0M8Q6|LAC7_HUMAN HR(+.98)SYSC(+57.02)R(+.98)VTHEGSTVEK sp|P08603|CFAH_HUMAN HRTGDEITYQC(+57.02)R(+.98)NGFYPATR(+.98)GNTAK sp|P08603|CFAH_HUMAN HRTGDEITYQC(+57.02)R(+.98)NGFYPATRGNTAK

139

140

sp|P02787|TRFE_HUMAN HSTIFENLANKADR(+.98)DQYELLC(+57.02)LDNTRKPVDEYK sp|P25311|ZA2G_HUMAN HVEDVPAFQALGSLNDLQFFR(+.98)YNSK sp|P01023|A2MG_HUMAN HVSR(+.98)TEVSSNHVLIYLDK sp|P06396|GELS_HUMAN HVVPNEVVVQR(+.98)LFQVK sp|P02751-14|FINC_HUMAN HYQINQQWER(+.98)TYLGNALVC(+57.02)TC(+57.02)YGGSR(+.98)GFNC(+57.02)ESKPEAEETC(+57.02)FDK sp|P02751-10|FINC_HUMAN HYQINQQWERTYLGNALVC(+57.02)TC(+57.02)YGGSR(+.98)GFNC(+57.02)ESKPEAEETC(+57.02)FDK sp|P08603|CFAH_HUMAN IIYKENER(+.98)FQYK sp|P02787|TRFE_HUMAN INHC(+57.02)R(+.98)FDEFFSEGC(+57.02)APGSK sp|P02787|TRFE_HUMAN INHC(+57.02)R(+.98)FDEFFSEGC(+57.02)APGSKK sp|P02787|TRFE_HUMAN INHC(+57.02)R(+.98)FDEFFSEGC(+57.02)APGSKKDSSLC(+57.02)K sp|P04114|APOB_HUMAN INSR(+.98)FFGEGTK sp|P02675|FIBB_HUMAN IPTNLRVLRSILENLR(+.98)SK sp|P61769|B2MG_HUMAN IQRTPKIQVYSR(+.98)HPAENGK sp|P61769|B2MG_HUMAN IQVYSR(+.98)HPAENGK sp|P01024|CO3_HUMAN IR(+.98)AYYENSPQQVFSTEFEVK sp|P02675|FIBB_HUMAN ISQLTR(+.98)M(+15.99)GPTELLIEMEDWKGDK sp|P02675|FIBB_HUMAN ISQLTR(+.98)MGPTELLIEMEDWKGDK sp|P02675|FIBB_HUMAN ISQLTR(+.98)MGPTELLIEMEDWKGDKVK sp|P01011|AACT_HUMAN ITGAR(+.98)NLAVSQVVHK sp|P60709|ACTB_HUMAN IWHHTFYNELR(+.98)VAPEEHPVLLTEAPLNPK sp|P02749|APOH_HUMAN KATVVYQGER(+.98)VK sp|P02751-10|FINC_HUMAN KDNRGNLLQC(+57.02)IC(+57.02)TGNGR(+.98)GEWK sp|P01023|A2MG_HUMAN KDNSVHWER(+.98)PQKPK sp|P02787|TRFE_HUMAN KDSGFQMNQLR(+.98)GKK sp|P19827|ITIH1_HUMAN KGHVLFR(+.98)PTVSQQQSC(+57.02)PTC(+57.02)STSLLNGHFK sp|P02743|SAMP_HUMAN KGLR(+.98)QGYFVEAQPK tr|B4E1Z4|B4E1Z4_HUMAN KGSC(+57.02)ER(+.98)DAQYAPGYDK sp|P01024|CO3_HUMAN KGYTQQLAFR(+.98)QPSSAFAAFVK 141

sp|P02760|AMBP_HUMAN KIMDR(+.98)MTVSTLVLGEGATEAEISMTSTRWRK sp|P41218|MNDA_HUMAN KINQEEVGLAAPAPTAR(+.98)NK sp|P00450|CERU_HUMAN KLISVDTEHSNIYLQNGPDR(+.98)IGR(+.98)LYKK sp|P05109|S10A8_HUMAN KLLETEC(+57.02)PQYIR(+.98)KK sp|P00450|CERU_HUMAN KLVYREYTDASFTNR(+.98)K sp|P0C0L4|CO4A_HUMAN KPR(+.98)LLLFSPSVVHLGVPLSVGVQLQDVPRGQVVK sp|P00734|THRB_HUMAN KPVAFSDYIHPVC(+57.02)LPDR(+.98)ETAASLLQAGYK sp|P02675|FIBB_HUMAN KREEAPSLRPAPPPISGGGYR(+.98)ARPAK sp|P02768|ALBU_HUMAN KVPQVSTPTLVEVSR(+.98)NLGK sp|P02768|ALBU_HUMAN KVPQVSTPTLVEVSR(+.98)NLGKVGSK sp|P02647|APOA1_HUMAN KWQEEM(+15.99)ELYR(+.98)QK sp|P02647|APOA1_HUMAN KWQEEMELYR(+.98)QK sp|P02787|TRFE_HUMAN LC(+57.02)MGSGLNLC(+57.02)EPNNKEGYYGYTGAFR(+.98)C(+57.02)LVEKGDVAFVK sp|P02768|ALBU_HUMAN LC(+57.02)TVATLR(+.98)ETYGEM(+15.99)ADC(+57.02)C(+57.02)AK sp|P02768|ALBU_HUMAN LC(+57.02)TVATLR(+.98)ETYGEMADC(+57.02)C(+57.02)AK sp|P02768|ALBU_HUMAN LC(+57.02)TVATLRETYGEMADC(+57.02)C(+57.02)AKQEPER(+.98)NEC(+57.02)FLQHK sp|P08254|MMP3_HUMAN LDGAAR(+.98)GEDTSM(+15.99)NLVQK sp|P08254|MMP3_HUMAN LDGAAR(+.98)GEDTSMNLVQK sp|P06702|S10A9_HUMAN LGHPDTLNQGEFKELVR(+.98)K sp|P19823|ITIH2_HUMAN LGSYEHRIYLQPGR(+.98)LAK sp|P68871|HBB_HUMAN LHVDPENFR(+.98)LLGNVLVC(+57.02)VLAHHFGKEFTPPVQAAYQK sp|P04220|MUCB_HUMAN LIC(+57.02)QATGFSPR(+.98)QIEVSWLREGK sp|P02751-14|FINC_HUMAN LLC(+57.02)QC(+57.02)LGFGSGHFR(+.98)C(+57.02)DSSR(+.98)WC(+57.02)HDNGVNYK sp|P02647|APOA1_HUMAN LLDNWDSVTSTFSKLR(+.98)EQLGPVTQEFWDNLEK sp|P08670|VIME_HUMAN LLEGEESRISLPLPNFSSLNLR(+.98)ETNLDSLPLVDTHSK sp|P04209|LV211_HUMAN LLIYDVNSR(+.98)PSGISNR(+.98)FSGSK

141

142

sp|P04208|LV106_HUMAN LLIYKDNQR(+.98)PSGVPDRFSGSK sp|P00738|HPT_HUMAN LPEC(+57.02)EADDGC(+57.02)PKPPEIAHGYVEHSVR(+.98)YQC(+57.02)K sp|P01008|ANT3_HUMAN LQPLDFKENAEQSR(+.98)AAINK sp|P02750|A2GL_HUMAN LQVLGKDLLLPQPDLR(+.98)YLFLNGNK sp|P02647|APOA1_HUMAN LR(+.98)EQLGPVTQEFWDNLEK sp|P02647|APOA1_HUMAN LR(+.98)QGLLPVLESFK sp|P04114|APOB_HUMAN LR(+.98)R(+.98)NLQNNAEWVYQGAIR(+.98)QIDDIDVRFQK sp|P04114|APOB_HUMAN LR(+.98)TSSFALNLPTLPEVK sp|P02647|APOA1_HUMAN LREQLGPVTQEFWDNLEKETEGLR(+.98)QEMSK sp|P01024|CO3_HUMAN LSINTHPSQKPLSITVR(+.98)TK sp|P02768|ALBU_HUMAN LVEVSR(+.98)NLGK sp|P01008|ANT3_HUMAN LVSANR(+.98)LFGDK sp|P00450|CERU_HUMAN LVYREYTDASFTNR(+.98)K sp|P19823|ITIH2_HUMAN LWAYLTINQLLAER(+.98)SLAPTAAAK tr|B7WNR0|B7WNR0_HUMAN LYLYEIAR(+.98)R(+.98)HPYFYAPELLFFAK sp|P02679|FIBG_HUMAN M(+15.99)LEEIMKYEASILTHDSSIR(+.98)YLQEIYNSNNQK sp|P06702|S10A9_HUMAN M(+15.99)SQLER(+.98)NIETIINTFHQYSVK sp|P02679|FIBG_HUMAN MLEEIM(+15.99)KYEASILTHDSSIR(+.98)YLQEIYNSNNQK sp|P02679|FIBG_HUMAN MLEEIMKYEASILTHDSSIR(+.98)YLQEIYNSNNQK sp|P06702|S10A9_HUMAN MSQLER(+.98)NIETIINTFHQYSVK sp|P00450|CERU_HUMAN NC(+57.02)VTR(+.98)IYHSHIDAPK sp|P00738|HPT_HUMAN NDVTDIADDGC(+57.02)PKPPEIAHGYVEHSVR(+.98)YQC(+57.02)K sp|P01023|A2MG_HUMAN NEDC(+57.02)INR(+.98)HNVYINGITYTPVSSTNEK sp|Q12841|FSTL1_HUMAN NFDNGDSR(+.98)LDSSEFLK sp|P02790|HEMO_HUMAN NFPSPVDAAFR(+.98)QGHNSVFLIK sp|P02790|HEMO_HUMAN NFPSPVDAAFR(+.98)QGHNSVFLIKGDK sp|P02748|CO9_HUMAN NFR(+.98)TEHYEEQIEAFK tr|B4E1Z4|B4E1Z4_HUMAN NGDKKGSC(+57.02)ER(+.98)DAQYAPGYDK 143

sp|P61769|B2MG_HUMAN NGER(+.98)IEKVEHSDLSFSK sp|P25311|ZA2G_HUMAN NILDR(+.98)QDPPSVVVTSHQAPGEK sp|P00747|PLMN_HUMAN NLDENYC(+57.02)R(+.98)NPDGK sp|P04003|C4BPA_HUMAN NLR(+.98)WTPYQGC(+57.02)EALC(+57.02)C(+57.02)PEPKLNNGEITQHR(+.98)K sp|P04003|C4BPA_HUMAN NLR(+.98)WTPYQGC(+57.02)EALC(+57.02)C(+57.02)PEPKLNNGEITQHRK sp|P04003|C4BPA_HUMAN NLRWTPYQGC(+57.02)EALC(+57.02)C(+57.02)PEPKLNNGEITQHR(+.98)K sp|P00450|CERU_HUMAN NNEGTYYSPNYNPQSR(+.98)SVPPSASHVAPTETFTYEWTVPK tr|B4E1Z4|B4E1Z4_HUMAN NPR(+.98)EDYLDVYVFGVGPLVNQVNINALASKK sp|P00738|HPT_HUMAN NPVQR(+.98)ILGGHLDAK sp|P02671|FIBA_HUMAN NR(+.98)GDSTFESK sp|P04114|APOB_HUMAN NR(+.98)NNALDFVTK sp|P01024|CO3_HUMAN NR(+.98)WEDPGK sp|Q92954|PRG4_HUMAN NSAANR(+.98)ELQK sp|P01011|AACT_HUMAN NSPLDEENLTQENQDR(+.98)GTHVDLGLASANVDFAFSLYK sp|P01876|IGHA1_HUMAN NTFR(+.98)PEVHLLPPPSEELALNELVTLTC(+57.02)LARGFSPK sp|P02671|FIBA_HUMAN NVR(+.98)AQLVDMK sp|P02763|A1AG1_HUMAN NWGLSVYADKPETTKEQLGEFYEALDC(+57.02)LR(+.98)IPK sp|P02675|FIBB_HUMAN NYC(+57.02)GLPGEYWLGNDKISQLTR(+.98)MGPTELLIEMEDWKGDK sp|P00747|PLMN_HUMAN NYR(+.98)GTMSK sp|P00738|HPT_HUMAN NYYKLR(+.98)TEGDGVYTLNDK sp|P00738|HPT_HUMAN NYYKLR(+.98)TEGDGVYTLNNEK sp|P00738|HPT_HUMAN NYYKLR(+.98)TEGDGVYTLNNEKQWINK sp|P02647|APOA1_HUMAN PALEDLR(+.98)QGLLPVLESFK sp|P02741|CRP_HUMAN PFSPNVLNWR(+.98)ALK sp|P00738|HPT_HUMAN PIC(+57.02)LPSKDYAEVGRVGYVSGWGR(+.98)NANFKFTDHLK sp|P01023|A2MG_HUMAN PLNDQDNEDC(+57.02)INR(+.98)HNVYINGITYTPVSSTNEK

143

144

sp|P02675|FIBB_HUMAN PNGR(+.98)YYWGGQYTWDMAK sp|P00450|CERU_HUMAN PQSR(+.98)SVPPSASHVAPTETFTYEWTVPK sp|P02768|ALBU_HUMAN PTLVEVSR(+.98)NLGK sp|P02768|ALBU_HUMAN PYFYAPELLFFAKR(+.98)YKAAFTEC(+57.02)C(+57.02)QAADK sp|P02675|FIBB_HUMAN QC(+57.02)SKEDGGGWWYNR(+.98)C(+57.02)HAANPNGR(+.98)YYWGGQYTWDMAK sp|P02675|FIBB_HUMAN QC(+57.02)SKEDGGGWWYNRC(+57.02)HAANPNGR(+.98)YYWGGQYTWDMAK sp|P02768|ALBU_HUMAN QEPER(+.98)NEC(+57.02)FLQHK sp|P02768|ALBU_HUMAN QEPER(+.98)NEC(+57.02)FLQHKDDNPNLPRLVRPEVDVMC(+57.02)TAFHDNEETFLK sp|P02671|FIBA_HUMAN QFTSSTSYNR(+.98)GDSTFESK sp|P02671|FIBA_HUMAN QFTSSTSYNR(+.98)GDSTFESKSY sp|P00738|HPT_HUMAN QKVSVNER(+.98)VMPIC(+57.02)LPSK sp|P00738|HPT_HUMAN QKVSVNER(+.98)VMPIC(+57.02)LPSKDYAEVGRVGYVSGWGRNANFK sp|P00738|HPT_HUMAN QKVSVNERVMPIC(+57.02)LPSKDYAEVGRVGYVSGWGR(+.98)NANFK sp|P25311|ZA2G_HUMAN QKWEAEPVYVQR(+.98)AK sp|P01024|CO3_HUMAN QLANGVDR(+.98)YISK sp|P01024|CO3_HUMAN QLANGVDR(+.98)YISKYELDK sp|P02647|APOA1_HUMAN QLNLKLLDNWDSVTSTFSKLR(+.98)EQLGPVTQEFWDNLEK sp|P02768|ALBU_HUMAN QNC(+57.02)ELFEQLGEYKFQNALLVR(+.98)YTK sp|P01024|CO3_HUMAN QQR(+.98)TFISPIK sp|P02768|ALBU_HUMAN QR(+.98)LKC(+57.02)ASLQK sp|P0C0L4|CO4A_HUMAN QR(+.98)VEASISK tr|B4E1Z4|B4E1Z4_HUMAN QVPAHAR(+.98)DFHINLFQVLPWLK sp|P0C0L5|CO4B_HUMAN QYR(+.98)NGESVK sp|P0C0L4|CO4A_HUMAN QYR(+.98)NGESVKLHLETDSLALVALGALDTALYAAGSK sp|P05156|CFAI_HUMAN R(+.98)AQLGDLPWQVAIK sp|Q99700-4|ATX2_HUMAN R(+.98)DAFTDSAISAK sp|P00747|PLMN_HUMAN R(+.98)DVVLFEK sp|P02647|APOA1_HUMAN R(+.98)DYVSQFEGSALGK 145

sp|P0DJI8|SAA1_HUMAN R(+.98)EANYIGSDK sp|P63267|ACTH_HUMAN R(+.98)GILTLK sp|P07357|CO8A_HUMAN R(+.98)HLVC(+57.02)NGDQDC(+57.02)LDGSDEDDC(+57.02)EDVR(+.98)AIDEDC(+57.02)SQYEPIPGSQK sp|P01009|A1AT_HUMAN R(+.98)LGM(+15.99)FNIQHC(+57.02)K sp|P01009|A1AT_HUMAN R(+.98)LGMFNIQHC(+57.02)K sp|P19823|ITIH2_HUMAN R(+.98)LSNENHGIAQRIYGNQDTSSQLK sp|P01011|AACT_HUMAN R(+.98)LYGSEAFATDFQDSAAAK sp|P02768|ALBU_HUMAN R(+.98)M(+15.99)PC(+57.02)AEDYLSVVLNQLC(+57.02)VLHEK sp|P02768|ALBU_HUMAN R(+.98)MPC(+57.02)AEDYLSVVLNQLC(+57.02)VLHEK sp|P01031|CO5_HUMAN R(+.98)MPITYDNGFLFIHTDK sp|P01031|CO5_HUMAN R(+.98)MPITYDNGFLFIHTDKPVYTPDQSVK sp|P08254|MMP3_HUMAN R(+.98)NSMEPGFPK sp|P19827|ITIH1_HUMAN R(+.98)QAVDTAVDGVFIRSLK sp|P01024|CO3_HUMAN R(+.98)QGALELIK sp|P01024|CO3_HUMAN R(+.98)QGALELIKK sp|P01023|A2MG_HUMAN R(+.98)QSSEITRTITK sp|P01023|A2MG_HUMAN R(+.98)QTVSWAVTPK sp|P02774|VTDB_HUMAN R(+.98)SDFASNC(+57.02)C(+57.02)SINSPPLYC(+57.02)DSEIDAELK sp|P13671|CO6_HUMAN R(+.98)SENINHNSAFK sp|Q8WWI1|LMO7_HUMAN R(+.98)TILIK sp|P08670|VIME_HUMAN R(+.98)TLLIK sp|P01861|IGHG4_HUMAN R(+.98)VESKYGPPC(+57.02)PSC(+57.02)PAPEFLGGPSVFLFPPKPK sp|P00747|PLMN_HUMAN R(+.98)YDYC(+57.02)DILEC(+57.02)EEEC(+57.02)MHC(+57.02)SGENYDGKISK sp|O75882-3|ATRN_HUMAN R(+.98)YENC(+57.02)PK sp|P00747|PLMN_HUMAN RAPWC(+57.02)HTTNSQVR(+.98)WEYC(+57.02)K sp|P02748|CO9_HUMAN RC(+57.02)TDAVGDRRQC(+57.02)VPTEPC(+57.02)EDAEDDC(+57.02)GNDFQC(+57.02)STGR(+.98)C(+57.02)IK

145

146

sp|P01024|CO3_HUMAN RQGALELIKKGYTQQLAFR(+.98)QPSSAFAAFVK sp|P00738|HPT_HUMAN RVGYVSGWGR(+.98)NANFK sp|P01876|IGHA1_HUMAN SAVQGPPER(+.98)DLC(+57.02)GC(+57.02)YSVSSVLPGC(+57.02)AEPWNHGK sp|P01876|IGHA1_HUMAN SAVQGPPER(+.98)DLC(+57.02)GC(+57.02)YSVSSVLPGC(+57.02)AEPWNHGKTFTC(+57.02)TAAYPESK sp|P01876|IGHA1_HUMAN SAVQGPPER(+.98)DLC(+57.02)GC(+57.02)YSVSSVLPGC(+57.02)AEPWNHGKTFTC(+57.02)TAAYPESKTPLTATLSK sp|P01877|IGHA2_HUMAN SAVQGPPER(+.98)DLC(+57.02)GC(+57.02)YSVSSVLPGC(+57.02)AQPWNHGETFTC(+57.02)TAAHPELK sp|P08603|CFAH_HUMAN SC(+57.02)DNPYIPNGDYSPLR(+.98)IK SC(+57.02)DTPPPC(+57.02)PR(+.98)C(+57.02)PAPELLGGPSVFLFPPKPKDTLMISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEV sp|P01860|IGHG3_HUMAN QFK sp|P01860|IGHG3_HUMAN SC(+57.02)DTPPPC(+57.02)PR(+.98)C(+57.02)PEPK sp|P02787|TRFE_HUMAN SC(+57.02)HTGLGR(+.98)SAGWNIPIGLLYC(+57.02)DLPEPRKPLEK sp|P02787|TRFE_HUMAN SC(+57.02)HTGLGRSAGWNIPIGLLYC(+57.02)DLPEPR(+.98)KPLEK sp|P0DJI8|SAA1_HUMAN SDMR(+.98)EANYIGSDK sp|P02671|FIBA_HUMAN SETESR(+.98)GSESGIFTNTK sp|P02768|ALBU_HUMAN SEVAHR(+.98)FKDLGEENFK sp|P02768|ALBU_HUMAN SEVAHR(+.98)FKDLGEENFKALVLIAFAQYLQQC(+57.02)PFEDHVK sp|P01023|A2MG_HUMAN SGGR(+.98)TEHPFTVEEFVLPKFEVQVTVPK sp|P01876|IGHA1_HUMAN SGNTFR(+.98)PEVHLLPPPSEELALNELVTLTC(+57.02)LAR(+.98)GFSPK sp|P01876|IGHA1_HUMAN SGNTFR(+.98)PEVHLLPPPSEELALNELVTLTC(+57.02)LARGFSPK sp|P01876|IGHA1_HUMAN SGNTFR(+.98)PEVHLLPPPSEELALNELVTLTC(+57.02)LARGFSPKDVLVRWLQGSQELPREK sp|P01876|IGHA1_HUMAN SGNTFRPEVHLLPPPSEELALNELVTLTC(+57.02)LAR(+.98)GFSPK sp|P01024|CO3_HUMAN SGQSEDR(+.98)QPVPGQQMTLK sp|P02647|APOA1_HUMAN SGR(+.98)DYVSQFEGSALGK sp|P01024|CO3_HUMAN SGSDEVQVGQQR(+.98)TFISPIK sp|P01834|IGKC_HUMAN SGTASVVC(+57.02)LLNNFYPR(+.98)EAK sp|P01834|IGKC_HUMAN SGTASVVC(+57.02)LLNNFYPR(+.98)EAKVQWK sp|P01834|IGKC_HUMAN SGTASVVC(+57.02)LLNNFYPR(+.98)EAKVQWKVDNALQSGNSQESVTEQDSK tr|F5GXS0|F5GXS0_HUMAN SHALQLNNR(+.98)QIR(+.98)GLEEELQFSLGSK sp|B9A064|IGLL5_HUMAN SHR(+.98)SYSC(+57.02)QVTHEGSTVEK 147

sp|P36980|FHR2_HUMAN SHSFR(+.98)AMC(+57.02)QNGK sp|P01008|ANT3_HUMAN SKLPGIVAEGR(+.98)DDLYVSDAFHK sp|P01008|ANT3_HUMAN SKLPGIVAEGR(+.98)DDLYVSDAFHKAFLEVNEEGSEAAASTAVVIAGRSLNPNRVTFK sp|P02768|ALBU_HUMAN SLHTLFGDKLC(+57.02)TVATLR(+.98)ETYGEM(+15.99)ADC(+57.02)C(+57.02)AK sp|P02768|ALBU_HUMAN SLHTLFGDKLC(+57.02)TVATLR(+.98)ETYGEMADC(+57.02)C(+57.02)AK sp|P01024|CO3_HUMAN SNLDEDIIAEENIVSR(+.98)SEFPESWLWNVEDLKEPPK sp|P01024|CO3_HUMAN SNLDEDIIAEENIVSR(+.98)SEFPESWLWNVEDLKEPPKNGISTK sp|P04114|APOB_HUMAN SPAFTDLHLR(+.98)YQK sp|P01011|AACT_HUMAN SPLDEENLTQENQDR(+.98)GTHVDLGLASANVDFAFSLYK sp|P01024|CO3_HUMAN SPMYSIITPNILR(+.98)LESEETMVLEAHDAQGDVPVTVTVHDFPGKK sp|P05452|TETN_HUMAN SR(+.98)LDTLAQEVALLK sp|P02749|APOH_HUMAN SR(+.98)PDNGFVNYPAKPTLYYK sp|P02749|APOH_HUMAN SR(+.98)PDNGFVNYPAKPTLYYKDK sp|P33241|LSP1_HUMAN SR(+.98)WETGEVQAQSAAK sp|P01861|IGHG4_HUMAN SR(+.98)WQEGNVFSC(+57.02)SVMHEALHNHYTQK sp|P01860|IGHG3_HUMAN SR(+.98)WQQGNIFSC(+57.02)SVMHEALHNRFTQK sp|P02679|FIBG_HUMAN SRKMLEEIMKYEASILTHDSSIR(+.98)YLQEIYNSNNQK tr|B4E1Z4|B4E1Z4_HUMAN SSGQWQTPGATR(+.98)SLSK sp|P01042-2|KNG1_HUMAN SSR(+.98)IGEIKEETTSHLRSC(+57.02)EYK :sp|P02671|FIBA_HUMAN SSSYSKQFTSSTSYNR(+.98)GDSTFESK sp|P02768|ALBU_HUMAN STPTLVEVSR(+.98)NLGK sp|P01024|CO3_HUMAN STVLYR(+.98)IFTVNHK sp|P04114|APOB_HUMAN SVGFHLPSR(+.98)EFQVPTFTIPK sp|P00738|HPT_HUMAN SVR(+.98)YQC(+57.02)KNYYK sp|P01834|IGKC_HUMAN SVVC(+57.02)LLNNFYPR(+.98)EAK sp|P02671|FIBA_HUMAN SYNR(+.98)GDSTFESK

147

148

sp|P01834|IGKC_HUMAN TASVVC(+57.02)LLNNFYPR(+.98)EAK sp|P02751-14|FINC_HUMAN TC(+57.02)YGGSR(+.98)GFNC(+57.02)ESKPEAEETC(+57.02)FDK sp|P08603|CFAH_HUMAN TESGWR(+.98)PLPSC(+57.02)EEKSC(+57.02)DNPYIPNGDYSPLR(+.98)IK sp|P0C0L4|CO4A_HUMAN TFFR(+.98)GLESQTK sp|P02671|FIBA_HUMAN TFPGFFSPM(+15.99)LGEFVSETESR(+.98)GSESGIFTNTK sp|P02671|FIBA_HUMAN TFPGFFSPMLGEFVSETESR(+.98)GSESGIFTNTK sp|P01876|IGHA1_HUMAN TFRPEVHLLPPPSEELALNELVTLTC(+57.02)LAR(+.98)GFSPK sp|P01023|A2MG_HUMAN TGYQR(+.98)QLNYK sp|P01857|IGHG1_HUMAN THTC(+57.02)PPC(+57.02)PAPELLGGPSVFLFPPKPKDTLM(+15.99)ISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEVK sp|P01857|IGHG1_HUMAN THTC(+57.02)PPC(+57.02)PAPELLGGPSVFLFPPKPKDTLMISR(+.98)TPEVTC(+57.02)VVVDVSHEDPEVK tr|H0Y8C9|H0Y8C9_HUMAN TIQHIVAEKEHR(+.98)LKEVM(+15.99)K sp|P02763|A1AG1_HUMAN TISR(+.98)YVGGQEHFAHLLILRDTK sp|P01024|CO3_HUMAN TIYTPGSTVLYR(+.98)IFTVNHK sp|P01860|IGHG3_HUMAN TKGQPR(+.98)EPQVYTLPPSREEMTK sp|P01860|IGHG3_HUMAN TKGQPREPQVYTLPPSR(+.98)EEMTK sp|P00736|C1R_HUMAN TLDEFTIIQNLQPQYQFR(+.98)DYFIATC(+57.02)K sp|P10909-2|CLUS_HUMAN TLIEKTNEER(+.98)K sp|P02768|ALBU_HUMAN TLVEVSR(+.98)NLGK sp|P00747|PLMN_HUMAN TPENYPNAGLTM(+15.99)NYC(+57.02)R(+.98)NPDADK sp|P02768|ALBU_HUMAN TPTLVEVSR(+.98)NLGK sp|Q96QV6|H2A1A_HUMAN TRIIPRHLQLAIR(+.98)NDEELNK sp|P01876|IGHA1_HUMAN TTTFAVTSILR(+.98)VAAEDWK sp|P02787|TRFE_HUMAN TVR(+.98)WC(+57.02)AVSEHEATK sp|P02671-2|FIBA_HUMAN TYNR(+.98)VSEDLRSRIEVLK sp|Q16695|H31T_HUMAN VAR(+.98)KSAPATGGVK sp|P02679|FIBG_HUMAN VATR(+.98)DNC(+57.02)C(+57.02)ILDERFGSYC(+57.02)PTTC(+57.02)GIADFLSTYQTK sp|P00738|HPT_HUMAN VDSGNDVTDIADDGC(+57.02)PKPPEIAHGYVEHSVR(+.98)YQC(+57.02)K sp|P01023|A2MG_HUMAN VDSHFR(+.98)QGIPFFGQVRLVDGK 149

sp|P08670|VIME_HUMAN VELQELNDR(+.98)FANYIDK sp|P02647|APOA1_HUMAN VEPLRAELQEGAR(+.98)QK sp|P02679|FIBG_HUMAN VGPEADKYR(+.98)LTYAYFAGGDAGDAFDGFDFGDDPSDK sp|P02768|ALBU_HUMAN VHTEC(+57.02)C(+57.02)HGDLLEC(+57.02)ADDR(+.98)ADLAK sp|P00747|PLMN_HUMAN VIPAC(+57.02)LPSPNYVVADR(+.98)TEC(+57.02)FITGWGETQGTFGAGLLK sp|P02766|TTHY_HUMAN VLDAVR(+.98)GSPAINVAVHVFRK sp|P01024|CO3_HUMAN VLLDGVQNPR(+.98)AEDLVGK sp|P02675|FIBB_HUMAN VNDNEEGFFSAR(+.98)GHRPLDK sp|P02768|ALBU_HUMAN VPQVSTPTLVEVSR(+.98)NLGK sp|P02768|ALBU_HUMAN VPQVSTPTLVEVSR(+.98)NLGKVG sp|P02768|ALBU_HUMAN VPQVSTPTLVEVSR(+.98)NLGKVGSK sp|P01023|A2MG_HUMAN VSNQTLSLFFTVLQDVPVR(+.98)DLKPAIVK sp|P06310|KV206_HUMAN VSNR(+.98)DSGVPDRFSGSGSGTDFTLK sp|P02768|ALBU_HUMAN VSTPTLVEVSR(+.98)NLGK sp|P00738|HPT_HUMAN VSVNER(+.98)VMPIC(+57.02)LPSK sp|P00738|HPT_HUMAN VSVNERVMPIC(+57.02)LPSKDYAEVGRVGYVSGWGR(+.98)NANFK sp|P19827|ITIH1_HUMAN VTAWKQYR(+.98)K sp|P0C0L4|CO4A_HUMAN VVEEQESRVHYTVC(+57.02)IWR(+.98)NGK sp|P01024|CO3_HUMAN VYAYYNLEESC(+57.02)TR(+.98)FYHPEKEDGKLNK sp|P00450|CERU_HUMAN VYPGEQYTYMLLATEEQSPGEGDGNC(+57.02)VTR(+.98)IYHSHIDAPK sp|P02751-10|FINC_HUMAN WDR(+.98)QGENGQMMSC(+57.02)TC(+57.02)LGNGKGEFKC(+57.02)DPHEATC(+57.02)YDDGK sp|P00738|HPT_HUMAN WGR(+.98)NANFK sp|P02751-14|FINC_HUMAN WLPSSSPVTGYR(+.98)VTTTPK sp|P02751-3|FINC_HUMAN WLPSSSPVTGYR(+.98)VTTTPKNGPGPTK sp|P02751|FINC_HUMAN WLPSSSPVTGYR(+.98)VTTTPKNGPGPTKTK sp|P01011|AACT_HUMAN WVMVPMMSLHHLTIPYFR(+.98)DEELSC(+57.02)TVVELK

149

150

sp|B9A064|IGLL5_HUMAN YAASSYLSLTPEQWKSHR(+.98)SYSC(+57.02)QVTHEGSTVEK sp|P00738|HPT_HUMAN YAEVGRVGYVSGWGR(+.98)NANFK sp|P02679|FIBG_HUMAN YEASILTHDSSIR(+.98)YLQEIYNSNNQK sp|P0DJI8|SAA1_HUMAN YFHAR(+.98)GNYDAAK sp|P01024|CO3_HUMAN YFKPGMPFDLMVFVTNPDGSPAYR(+.98)VPVAVQGEDTVQSLTQGDGVAK tr|Q5TFM2|Q5TFM2_HUMAN YFPYLENGYNQNHGR(+.98)KFVQGK sp|P08603|CFAH_HUMAN YFPYLENGYNQNYGR(+.98)KFVQGK sp|P01876|IGHA1_HUMAN YLTWASR(+.98)QEPSQGTTTFAVTSILR(+.98)VAAEDWK sp|P01876|IGHA1_HUMAN YLTWASR(+.98)QEPSQGTTTFAVTSILRVAAEDWK sp|P02768|ALBU_HUMAN YLYEIARRHPYFYAPELLFFAKR(+.98)YKAAFTEC(+57.02)C(+57.02)QAADK sp|P51884|LUM_HUMAN YLYLR(+.98)NNQIDHIDEK sp|P01042|KNG1_HUMAN YNSQNQSNNQFVLYR(+.98)ITEATK sp|P08254|MMP3_HUMAN YPLDGAAR(+.98)GEDTSMNLVQK sp|P02679|FIBG_HUMAN YR(+.98)LTYAYFAGGDAGDAFDGFDFGDDPSDK sp|P02751|FINC_HUMAN YTGNTYR(+.98)VGDTYERPK sp|P02679|FIBG_HUMAN YVATR(+.98)DNC(+57.02)C(+57.02)ILDER(+.98)FGSYC(+57.02)PTTC(+57.02)GIADFLSTYQTKVDK sp|P00738|HPT_HUMAN YVEHSVR(+.98)YQC(+57.02)K sp|P00738|HPT_HUMAN YVEHSVR(+.98)YQC(+57.02)KNYYK sp|P00738|HPT_HUMAN YVMLPVADQDQC(+57.02)IR(+.98)HYEGSTVPEK sp|P00738|HPT_HUMAN YVMLPVADQDQC(+57.02)IR(+.98)HYEGSTVPEKK sp|P00738|HPT_HUMAN YVMLPVADQDQC(+57.02)IR(+.98)HYEGSTVPEKKTPK sp|P00738|HPT_HUMAN YVSGWGR(+.98)NANFK sp|P00738|HPT_HUMAN YVSGWGR(+.98)NANFKFTDHLK sp|P04220|MUCB_HUMAN YVTSAPMPEPQAPGR(+.98)YFAHSILTVSEEEWNTGETYTC(+57.02)VVAHEALPNRVTERTVDK

151

Figure S3. Mass spectra of citrullinated and non-citrullinated peptides covering quantified citrinullation sites in proteins from RA patient synovial fluid samples

Apolipoprotein A-I

Site 212 m/z = 709.3460; z= 2; RT shift = 6.12

151

152

Apolipoprotein A-I

Site 212 m/z = 472.5712; z= 3

153

Haptoglobin

Site 286 m/z = 783.0455; z= 3; RT shift = 3.4

153

154

Haptoglobin

Site 286 m/z = 782.7192; z= 3

155

Ceruloplasmin

Site 481 m/z = 587.9883; z= 3; RT shift = 3.48

155

156

Ceruloplasmin

Site 481 m/z = 881.4877; z= 2

157

Ig alpha-1 chain C region

Site263 m/z = 685.3715; z= 3; RT shift = 2.7

157

158

Ig alpha-1 chain C region

Site263 m/z = 685.0042; z= 3

159

Serum albumin

Site 545 m/z = 446.7711; z= 4; RT shift = 3.6

159

160

Serum albumin

Site 545 m/z = 446.5236; z= 4

161

Serum albumin

Site 452 m/z = 963.5392; z= 2; RT shift = 3.4

161

162

Serum albumin

Site 452 m/z = 963.0462; z= 2

163

Serum albumin

Site 434 m/z = 741.4257; z= 2; RT shift = 4.5

163

164

Serum albumin

Site 434 m/z = 740.9337; z= 2

165

Apolipoprotein A-I

Site 51 m/z = 908.9180; z= 2; RT shift = 2.7

165

166

Apolipoprotein A-I

Site 51 m/z = 908.4271; z= 2

167

Haptoglobin

Site 261 m/z = 865.4501; z= 2; RT shift = 5.7

167

168

Haptoglobin

Site 261 m/z = 576.9750; z= 3

169

Ig kappa chain C region

Site 34 m/z = 1064.5331; z= 2; RT shift = 3.81

169

170

Ig kappa chain C region

Site 34 m/z = 1064.0372; z= 2

171

Serum amyloid A-1 protein

Site 57 m/z = 707.3304; z= 2; RT shift = 7

171

172

Serum amyloid A-1 protein

Site 57 m/z = 706.8372; z= 2

173

Alpha-1-acid glycoprotein 1

Site 108 m/z = 654.1042; z= 4; RT shift = 2.9

173

174

Alpha-1-acid glycoprotein 1

Site 108 m/z = 653.8573; z= 4

175

Alpha-1-antitrypsin

Site 220 m/z = 1051.1785; z= 3; RT shift = 1

175

176

Alpha-1-antitrypsin

Site 220 m/z = 788.1354; z= 4

177

Alpha-2-macroglobulin

Site 1014 m/z = 951.9889; z= 2; RT shift = 1.3

177

178

Alpha-2-macroglobulin

Site 1014 m/z = 951.4946; z= 2

179

Apolipoprotein A-I

Site 140 m/z = 556.9376; z= 3; RT shift = 1.81

179

180

Apolipoprotein A-I

Site 140 m/z = 556.6114; z= 3

181

Apolipoprotein B-100

Site 3989 m/z = 788.9072; z= 2; RT shift = 3.7

181

182

Apolipoprotein B-100

Site 3989 m/z = 788.4174; z= 2

183

Complement C3

Site 148 m/z = 553.8006; z= 4; RT shift = 4

183

184

Complement C3

Site 148 m/z = 553.5544; z= 4

185

Fibrinogen alpha chain

Site 129 m/z = 770.0672; z= 3; RT shift = 7.1

185

186

Fibrinogen alpha chain

Site 129 m/z = 577.8064; z= 4

187

Fibrinogen alpha chain

Site 573 m/z = 912.9402; z= 2; RT shift = 6.7

187

188

Fibrinogen alpha chain

Site 573 m/z = 912.4455; z= 2

189

Fibrinogen alpha chain

Site 591 m/z = 1021.9486; z= 2; RT shift = 11.9

189

190

Fibrinogen alpha chain

Site 591 m/z = 1021.4583; z= 2

191

Ig kappa chain c region

Site 34 m/z = 1064.5331; z= 2; RT shift = 3.8

191

192

Ig kappa chain c region

Site 34 m/z = 1064.0372; z= 2

193

Leucine-rich alpha-2-glycoprotein

Site 239 m/z = 1065.5820; z= 2; RT shift = 2.6

193

194

Leucine-rich alpha-2-glycoprotein

Site 239 m/z = 1065.0891; z= 2

195

Serum albumin

Site 434 m/z = 749.3751; z= 4; RT shift = 6.3

195

196

Serum albumin

Site 434 m/z = 749.1299; z= 4

197

Serum albumin

Site 105 m/z = 1176.0153; z= 2; RT shift = 13.1

197

198

Serum albumin

Site 105 m/z = 1175.5195; z= 2

199

Transthyretin

Site 41 m/z = 538.0617; z= 4; RT shift = 3

199

200

Transthyretin

Site 41 m/z = 717.0848; z= 3