Assessing the Citrullinome in Rheumatoid Arthritis Synovial Fluid with

Apr 14, 2014 - Centre for Immune Regulation, Department of Immunology, University of Oslo, Oslo University Hospital-Rikshospitalet, Oslo 0372, Norway ...
0 downloads 8 Views 1MB Size
Download PDF
Article pubs.acs.org/jpr

Assessing the Citrullinome in Rheumatoid Arthritis Synovial Fluid with and without Enrichment of Citrullinated Peptides Astrid E. V. Tutturen,*,† Burkhard Fleckenstein,‡,§ and Gustavo A. de Souza‡ †

Centre for Immune Regulation, Department of Immunology, University of Oslo, Oslo University Hospital-Rikshospitalet, Oslo 0372, Norway ‡ Centre for Immune Regulation, Department of Immunology, University of Oslo, Oslo 0372, Norway S Supporting Information *

ABSTRACT: Protein citrullination is a posttranslational modification that has attracted increased attention, especially for its involvement in rheumatoid arthritis (RA). Here, we assess the citrullinome in RA synovial fluid by direct LC−MS/MS analysis and by the use of an enrichment strategy based on citrulline specific biotinylation. RA synovial fluid was depleted for abundant proteins, and total and depleted fractions were analyzed. Frequency of citrullinated peptides and their degree of citrullination could be determined for four known RA autoantigens, as well as a novel in vivo autocitrullination site of peptidylarginine deiminase 4. From the analysis of total and depleted synovial fluid after enrichment we could estimate the numbers of citrullinated peptides to be approximately 3600 and 2100, respectively. However, identification of these biotinylated peptides by MS/MS turned out to be very difficult due to fragmentation of the biotin moiety. By direct MS analysis of the total and depleted synovial fluid without enrichment, 119 and 157 citrullinated peptides were identified, respectively. This indicates that direct analysis allows identification of only a fraction of the citrullinated proteins present in synovial fluid and that specific enrichment is still needed for a comprehensive in-depth elucidation of the citrullinome. KEYWORDS: protein citrullination, rheumatoid arthritis, synovial fluid, enrichment, mass spectrometry, LC−MS/MS, quantification

■

INTRODUCTION Protein citrullination denotes the posttranslational modification where the guanidine group of an arginine residue in a protein is converted to an ureido group, resulting in a citrulline residue. This conversion is enzymatically mediated by peptidylarginine deiminases (PADs; EC 3.5.3.15), which exist in five isoforms in mammals, PAD1−4 and PAD6.1 Citrullination results in a mass increase of 0.98 Da and the loss of one positive charge per converted residue. This posttranslational modification has received increased attention in relation to both physiological processes and the pathophysiology of several diseases (reviewed in refs 2 and 3). There has been a particular interest in rheumatoid arthritis (RA). In this disorder, the majority of the patients have autoantibodies against citrullinated proteins, socalled anti-citrullinated protein antibodies (ACPAs).4 The presence of these antibodies can be observed years before disease onset and predict more severe disease progression.5 Further, while citrullinated proteins have been found to be present in inflammation in general,6 the production of ACPAs is restricted to RA,7 indicating the importance of citrullination in autoimmune response of RA. To understand the pathophysiology of RA, there is need for a comprehensive identification of citrullinated proteins in the RA synovium. Several methods have been developed to specifically enrich and/or detect in vivo citrullination (reviewed in refs 8 and 9). © 2014 American Chemical Society

These methods includes specific chemical modifications of the ureido group of citrulline followed by either immune staining on Western blots using antibodies against citrulline residues modified by 2,3-butandione and antipyrine,10 strategies based on selective enrichment by specific immobilization11 or biotinylation of citrullinated peptides,12 or comparison of a 2,3-butandione-modified sample versus an unmodified sample,13 all followed by identification by mass spectrometry (MS). Further, methods have been proposed based on filtering specific diagnostic ions by MS in both unmodified and modified citrulline-containing peptides, through the detection of the neutral loss of isocyanic acid14 or a specific signature ion observed in CID fragmentation of citrulline modified by 2,3butaedione and antipyrine,15 respectively. To our knowledge only a few of these methods have been employed for citrulline characterization in complex biological samples such as synovial fluid. New advances in MS have led to instruments with high resolution and high acquisition speed, resulting in remarkable mass accuracy and depth of characterization. On the basis of these improvements in MS technology, citrullinated proteins in synovial fluid have been characterized by direct MS analysis Received: January 9, 2014 Published: April 14, 2014 2867

dx.doi.org/10.1021/pr500030x | J. Proteome Res. 2014, 13, 2867−2873

Journal of Proteome Research

Article

(HSAfraction), respectively. Both samples were incubated for 30 min on ice before being centrifuged at 14000g for 20 min at 4 °C. The acetone supernatants were discarded, and the precipitates were left to dry before being dissolved by adding 25 mM Tris-HCl, pH 8.5 with 2% sodium dodecyl sulfate (SDS; UltraPure, 10%, GIBCO, Invitrogen) and applying heat (50 °C for 10 min). To confirm protein depletion, 10 μg of each sample was analyzed by SDS-polyacrylamide gel electrophoresis (SDSPAGE; NuPAGE 4−12%, Life Technologies) followed by visualization by Coomassie staining (Collodial Blue Staining Kit, novex, Invitrogen) according to manufacturer’s instructions. The supernatant-derived samples (Sup, Sup IgG depl , SupIgG‑HSA depl, IgGfraction, HSAfraction) were further prepared by filter aided sample preparation (modified from19). Each sample (400 μg) was transferred to a filter device (Microcon, Ultracel YM-10, Millipore, Ireland) in 300 μL of 6 M urea in 50 mM Tris-HCL, pH 8.5 (UA). The samples were centrifuged at 14000g until 10−20 μL was left in the filter. This was repeated using 200 μL of UA added to each sample and mixed for 1 min at 600 rpm on a thermo mixer (Eppendorf, Germany) and centrifuged as above. DL-Dithiothreitol (DTT; Sigma), 50 μL of 10 mM in UA, was added, and the samples were mixed as above and incubated for 30 min at room temperature with gentle rotation before centrifugation. Iodoacetamide (IAA; SigmaAldrich), 50 μL 55 mM in UA, was added and treated as above. The samples were then washed 3× using 100 μL of UA and a further 3× using 200 μL of 50 mM Tris-HCl, pH 8.5. Each wash was followed by mixing and centrifugation as above. For digestion, 8 μg of lysylendopeptidase (LysC; 129−02541, Wako, Germany) dissolved in 40 μL of 50 mM Tris-HCl, pH 8.5 was added to each sample, which were then incubated overnight in a wet chamber at 37 °C. The filter devices were placed in new collection tubes and centrifuged as described, and digested peptides were collected from the flowthrough. The digestion step was repeated, but with an incubation time of 3 h. Finally, the filter devices were washed by adding 160 μL of water, mixed, and centrifuged. The washing solution was combined with the digest, and an amount corresponding to 200 μg from each fraction was analyzed by mass spectrometry. Processing of Pellet. The very viscous synovial fluid pellet was mixed with 40 μL of 0.2% ProteaseMax (Promega) in 50 mM NH4HCO3, before incubation for 30 min at room temperature with gentle agitation. An additional 187 μL of 50 mM NH4HCO3 was added. Proteins were reduced by adding 2 μL of 0.5 M DTT in 50 mM NH4HCO3 to a final concentration of 4 mM and incubated for 20 min at 56 °C. Further, for alkylation 5.4 μL of 0.55 M IAA was added and incubated for 15 min at room temperature in the dark. The protein concentration of extracted proteins was measured to be approximately 2.4 mg/mL (total volume approximately 250 μL). ProteaseMax solution (2 μL of 1%) was added together with 10 μL of 1 μg/μL LysC and incubated in a wet chamber at 37 °C overnight. Reaction was quenched by adding trifluoroacetic acid (TFA; Fluka, Sigma) to a final concentration of 0.5%. The sample was then mixed and incubated at room temperature for 5 min before being centrifuged at 14000g for 10 min. The supernatant was transferred to a new vial, and an amount corresponding to 200 μg was analyzed by mass spectrometry as the Pellet fraction.

without applying methods for citrulline detection/enrichment prior to analysis.16,17 In this study, the citrullinome of RA synovial fluid was assessed by direct LC−MS/MS analysis and by the use of an enrichment strategy based on citrulline-specific biotinylation.12 From direct MS analysis we determined frequencies of citrullinated peptides together with the degree of citrullination of individual sites and found a novel autocitrullination site in PAD4. When applying specific enrichment the number of citrullinated peptides estimated through detection of a modification-specific signature ion was more than 13 times higher than the number of citrullinated peptides identified by direct MS analysis. These results strongly indicate that only a fraction of the citrullinated peptides present in the synovial fluid was identified by direct MS analysis and that the use of specific enrichment tools will most likely provide greater insight into the synovial fluid citrullinome.

■

MATERIAL AND METHODS

Patient Material

Synovial fluid was obtained by arthrocentesis of an affected knee of a 43-year-old female patient with RA diagnosis with high titers of anti-CCP in serum at the time of sampling. The synovial fluid was immediately frozen upon sampling. The patient was under treatment at Diakonhjemmet Hospital, Oslo, Norway. All procedures were approved by the Regional Committee for Medical and Health Research Ethics, and written informed consent was obtained from the patient. Sample Preparation

Synovial fluid was allowed to thaw on ice, and EDTA was added to a final concentration of 50 mM. The synovial fluid (200 μL) was centrifuged at 14000g for 10 min at 4 °C, and the supernatant and pellet were processed separately. Processing of Supernatant. The supernatant was analyzed as total (Sup) and in fractions where immunoglobulin gamma (IgG) and human serum albumin (HSA) had been depleted. For IgG depletion, 150 μL of protein G Sepharose 4 Fast Flow (capacity to bind 3 mg IgG; GE Healthcare) was transferred to a filter unit (Reactor with Teflon frits, 2 mL, 25 μm pore size, MultiSynTech, Germany) and prewashed with 3 × 200 μL PBS. A synovial fluid volume corresponding to 7 mg of protein was diluted in PBS to a total volume of 300 μL and added to the beads for incubation for 2 h at 4 °C with rotation. The supernatant was pressed though the filter unit applying air pressure and collected, giving the IgG-depleted supernatant (SupIgG depl). The beads were washed twice with 100 μL of PBS for 10 min with rotation at room temperature. Washing solutions were combined with the Sup IgG depl fraction. Immobilized IgG was eluted by adding 300 μL of 0.1 M glycine-HCl, pH 2.5 and incubated for 5 min with rotation at room temperature, and the eluate was collected. This was repeated once with 100 μL of elution solution. The eluates were combined (IgGfraction). Further, human serum albumin (HSA) was depleted according to the protocol from Chen et al.18 A volume corresponding to 3.5 mg (250 μL) of the sample SupIgG depl was transferred to a new vial, and 1 mL of ice-cold acetone containing 10% trichloroacetic acid (TCA; ≥99.5%, Merck, Germany) was added and mixed gently. The sample was incubated for 90 min at 20 °C before being centrifuged at 14000g for 20 min at 4 °C. The TCA/acetone supernatant was transferred to a new vial (SupIgG‑HSA depl), and 3 and 1 mL of icecold acetone was added to SupIgG‑HSA depl and the precipitate 2868

dx.doi.org/10.1021/pr500030x | J. Proteome Res. 2014, 13, 2867−2873

Journal of Proteome Research

Article

from 32% B to 45% B in 20 min before column wash in 90% B for 15 min and equilibration for 15 min for the 25 cm column and for 20 min for the 50 cm column. The MS instrument was operated in data-dependent acquisition mode with automatic switching between MS and MS/MS scans. Full MS scans were acquired in resolution of 70000, with automatic gain control target value of 3 × 106 ions or maximum injection time of 50 ms within the scan range 300−1750 m/z. Peptide fragmentation was performed by higher energy collision dissociation (HCD) with normalized collision energy set to 25. The MS/ MS spectra were acquired of the 10 most abundant ions (Top10 method) in the resolution R = 17500, automatic gain control target value of 2 × 105 ions, or maximum fragment accumulation time of 100 ms. Ion selection was performed with a 3 m/z unit window, and fixed first mass was set to 100 m/z. Underfill ratio was set to 10%, and the intensity threshold was 2 × 105.

All protein concentrations were measured by infrared spectrometry (DirectDetect, Millipore) according to the manufacturer’s instruction. Enrichment Using BPG Modification

The preparation of Biotin-PEG2-4-glyoxalbenzoic acid (BPG) and enrichment of citrullinated peptides were done as previously described12 with a slight modification of the enrichment step. In short, digests corresponding to 400 μg of each of the supernatant-derived fraction were subjected to specific citrulline modification by incubation with one “aliquot” (see definition in ref 12) BPG in 50% TFA for 3.5 h at 37 °C followed by incubation in a vacuum drier until dryness in order to remove TFA. Excess BPG was removed by strong cation exchange (SCX) chromatography; samples were dissolved in in 600 μL of ACN/0.2% FA (50/50, v/v) and applied to a SCX column prepared in a 1 mL pipet tip containing 60 mg of polySULFOETHYL A beads (see ref 12) prewashed with 200 μL of ACN/0.2% FA (50/50, v/v). The flow-through was collected and recycled twice before the column was washed 8 × 400 μL of ACN/0.2% FA (50/50, v/v). Peptides were eluted using 150 μL of ACN/1 M NaCl (25/75, v/v), the elution step was repeated once, and eluates were combined. The combined eluate was incubated with Dynabeads M-270 Streptavidin beads (250 μL; 10 mg/mL solution) in PBS containing 0.1% SDS in a final bead concentration of 5.5 mg/mL. The sample was incubated at 4 °C overnight with rotation. The beads were washed according to the following steps: 3 × 700 μL of PBS containing 0.1% SDS, 4 × 700 μL of ACN/1 M NaCl (25/75, v/v), and 2 × 700 μL of ACN/water (20/80, v/v) before bound modified peptides were eluted using 200 μL of ACN/ 10% FA/2 mM biotin (70/10/20, v/v/v) and incubation for 30 min at 37 °C with occasional resuspension.

Data Analysis

The MS data acquired were subjected to the MaxQuant software20 version v1.4.0.8 for peptide and protein identification. The Andromeda search engine21 was run with a human database (downloaded from www.UniProt.org September 2013). Database search was performed with mass tolerance of 20 ppm for precursor ion in initial search and 4.5 ppm in main search. MS/MS error tolerance was set to 20 ppm. Carbamidomethyl (C) was set as fixed modification, and acetylation (Protein N-trem), deamidation (NQ), pyro-glu (E), pyro-gln (Q), hydroxylation (K/P), oxidation (M), citrullination (R) and citrullination followed by BPG-modification ((R), only for BPG-enriched samples) were set as variable modifications. LysC without proline restriction was selected as the protease, and three missed cleavages were allowed. A minimum length of 6 amino acids for peptides search was used. MaxQuant determines false discovery rate (FDR) by concatenating reversed sequences of the used database, and in this work a threshold of 1% was used. Quantitative assignment for individual peptides is done through an area under curve method, which MaxQuant calculates by a threedimensional characterization of ion peak volume using its mass accuracy width, its elution profile, and peptide ion count. For elucidation of the involvement of particular biological processes (GO Biological Process), all citrullinated proteins identified were submitted to the Database for Annotation, Visualization, and Integrated Discovery (DAVID)22,23 using their official gene symbols. Class enrichment is visualized through statistical validation, and FDR estimation is done using the Benjamin-Hochberg correction. Unique monoisotopic precursor masses were counted excluding masses with charge state lower than 1 and neutral masses below 1200 Da, allowing only peptide lengths of at least 6 amino acids including the modification mass of 516.2 Da per BPG-modified citrulline residue. Quantification of citrullination was calculated as the sum of intensities of all redundant identifications of a peptide in a certain citrullination state, divided by the sum of intensities of all identifications of the same peptide, independently of modifications.

Mass Spectrometry

The samples were desalted prior to LC−MS/MS analysis. Each fraction was diluted in 200 μL of Britton & Robinson universal buffer pH 11 (composed of 20 mM acetic acid, 20 mM phosphoric acid, and 20 mM boric acid titrated with NaOH to pH 11), before being desalted by reversed phase chromatography using micro columns prepared by placing 3 discs of C18 Empore Extraction Disk (Varian, St. Paul, MN, USA) into 200 μL pipet tips. Peptides were eluted by applying 2 × 80 μL of 80% acetonitrile (ACN; Fluka Analytical, Sigma-Aldrich) and 0.1% formic acid (FA; Fluka Analytical) in water. ACN was evaporated in a vacuum drier to a volume of approximately 8 μL and diluted in 0.1% FA in water to a final volume of 20 μL. Synovial fluid samples were then analyzed in triplicate (injection volume 6 μL) by nano-LC−MS/MS using a Q Exactive hybrid quadropole-orbitrap mass spectrometer interfaced with an EASY-spray ion source (both from Thermo Fisher Scientific) and coupled to a nano-LC HPLC (UltiMate 3000, Dionex). The peptides were loaded onto a trap column (C18, 100 μm × 2 cm, PepMap RSLC, Thermo Fisher Scientific) and separated on EASY-Spray columns (PepMapRSLC, C18, 2 μm particles, 100 Å, 75 μm ID, Thermo Scientific): 25 cm for IgGfraction, HSAfraction, and all enriched samples and 50 cm for Pellet, Sup, SupIgG depl, and SupIgG‑HSA depl. A binary gradient consisting of solvent A (0.1% FA in water) and solvent B (90% ACN and 0.1% FA in water) at a flow rate of 0.3 μL/min for the 25 cm column and 0.2 μL/min for the 50 cm column, at 35 °C, was applied. The same multistepped gradient was used on both columns: linearly increasing from 5% B to 10% B in 20 min, from 10% B to 32% B in 220 min, and 2869

dx.doi.org/10.1021/pr500030x | J. Proteome Res. 2014, 13, 2867−2873

Journal of Proteome Research

■

Article

Citrullinated Proteins in Synovial Fluid Identified by Direct MS Analysis

RESULTS AND DISCUSSION

Assessing Citrullination after Depletion of Abundant Proteins

The numbers of non-IgG and non-HSA-derived citrullinated peptides identified in the fractions Pellet, Sup, and SupIgG‑HSA depl are given in Table 1. In the total supernatant of the synovial

Due to high abundant proteins in synovial fluid, such as IgG and HSA, we implemented a depletion step to improve detection of proteins present in lower amounts. Synovial fluid from an ACPA-positive RA patient was separated into a soluble (Sup) and an insoluble fraction (Pellet). The soluble fraction was depleted for IgG by incubation with protein G Sepharose beads and further for HSA using TCA/acetone precipitation. The resulting fractions were referred to as SupIgG depl, SupIgG‑HSA depl, IgGfraction, and HSAfraction. To confirm depletion of IgG and HSA, the supernatant derived samples were analyzed by SDS-PAGE and stained by Coomassie blue (Figure 1). The gel shows that both IgG and HSA were significantly,

Table 1. Number of Identifications of Non-IgG and NonHSA Peptides Using Direct MS Analysis fraction

unique citrullinated peptidesa

all unique peptides

citrullinated peptides identified in only one fractiona

Pellet Sup SupIgG‑HSAdepl

87 110 153

3742 3882 5584

21 10 68

a

Single or multiple citrullination states of one peptide is considered as one unique observation.

fluid (fraction Sup), 110 citrullinated peptides deriving from 67 proteins were identified, whereas after depletion of IgG and HSA the number of identified citrullinated peptides increased to 153 and the number of proteins to 91. The doubly depleted fraction contained the highest number of citrullinated peptides that were identified in only a single fraction, namely, 68 peptides, whereas within the Pellet and the Sup fractions, 21 and 10 citrullinated peptides were exclusively identified, respectively. This further demonstrates that depletion of abundant proteins improves identification of citrullinated peptides. Representation of Identified Citrullinated Proteins in Known Biological Processes

Having identified a list of citrullinated proteins from synovial fluid of a patient with ongoing RA, a bioinformatics analysis of these proteins could give insight into the biology of this inflammatory condition. Although synovial fluid from only one patient was analyzed in this study, we tested the potential of such a bioinformatics approach. The genes of the citrullinated proteins identified in the total synovial fluid by the direct approach were categorized by doing a gene ontology (GO) classification using the bioinformatics tool DAVID. Biological processes significantly enriched for were compared between the citrullinome and the total proteome of synovial fluid. The biological process “acute inflammatory response” was most enriched for in both groups (Supporting Information, Supplementary Table S2). Interestingly, while 19% of the citrullinated proteins belonged to this group, only 3% of the total proteome did. Further, among the top 20 biological processes most enriched for in each group, 6 were immunerelated for the total proteome, whereas for the citrullinome 17 of these were immune-related. This might indicate that proteins participating in immune-related processes are likely candidates to be citrullinated. Although this approach seems promising, analysis of synovial fluid samples from more patients should be analyzed before drawing general conclusions.

Figure 1. Depletion of IgG and HSA from synovial fluid. Supernatantderived fractions of synovial fluid were analyzed by SDS-PAGE and visualized by Coomassie blue staining: (1) Sup, (2) SupIgG depl, (3) IgGfraction, (4) SupIgG‑HSA depl, and (5) HSAfraction. IgG and HSA in the synovial fluid were significantly reduced but not completely removed.

but not completely, removed from the synovial fluid. Next, the synovial fluid fractions were digested by LysC and directly analyzed by LC−MS/MS. Peptides were identified using MaxQuant. Detailed information is provided as Supporting Information (Supplementary File S1). Evaluation of all peptide identifications shows that depletion of IgG and HSA from the synovial fluid resulted in improved detection of lower abundant proteins. With respect to all peptide identifications, depletion of both abundant proteins led to an increase of identified non-IgG and non-HSA-derived peptides from 17113 (no depletion) to 30874 (IgG and HSA depletion), whereas the number of identified citrullinated nonIgG and non-HSA-derived peptides rose from 493 to 641, respectively. This increase was achieved even though IgG and HSA was not fully removed, which was expected from the original publication of the depletion method. Notably, more efficient depletion including other abundant proteins could further improve detection of low abundant proteins. Since both IgG and HSA are known to be citrullinated, the fractions IgGfraction and HSAfraction, as well as fraction SupIgG depl, were analyzed by LC−MS/MS. Results are given in Supporting Information (Supplementary Table S1).

Quantification of Frequency of Citrullinated Peptides and Degree of Citrullination

On the basis of peptide identification from direct LC−MS/MS analysis of all synovial fluid fractions, we quantified the frequency of citrullinated peptides and the degree of citrullination in individual sites in four of the best established autoantigens in RA: fibrinogen alpha chain (Fibα), fibrinogen beta chain, (Fibβ), vimentin, and alpha enolase (reviewed in ref 24). Notably, the degree of citrullination is expected to vary 2870

dx.doi.org/10.1021/pr500030x | J. Proteome Res. 2014, 13, 2867−2873

Journal of Proteome Research

Article

Table 2. Peptides Detected with Citrulline Modification and Frequencies of Their Citrullinated form peptide

amino acids

frequency of citrullinated form (%)

Fibrinogen α-chain GLIDEVNQDFTNRINK DLLPSRa DRa QHLPLIK TFPGFFSPMLGEFVSETESRGSESGIFTNTK ESSSHHPGIAEFPSRGK QFTSSTSYNRGDSTFESK (K)REEAPSLRPAPPPISGGGYRARPAK LESDVSAQMEYCRTPCTVSCNIPVVSGK PYRVYCDMNTENGGWTVIQNRQDGSVDFGRK EDGGGWWYNRCHAANPNGRYYWGGQYTWDMAK VELQELNDRFANYIDK SRLGDLYEEEMRELRRQVDQLTNDK PDLTAALRDVRQQYESVAAK FADLSEAANRNNDALRQAK LAQANGWGVMVSHRSGETEDTFIADLVVGLCTGQIK TGAPCRSERLAK a

72−87 211−225 528−558 559−575 582−599

0.8 1.1 3.0 18.9 26.6

Fibrinogen β-chain (52)53−77 212−238 265−295 427−459 Vimentin 105−120 144−168 263−281 295−313 α-Enolase 359−394 395−406

13.5 0.2 24.5 38.5 1.0 0.4 9.9 4.1 39.6 2.4

MaxQuant could not determine which of the arginine residues that had been citrullinated based on MS/MS data.

Table 3. PAD4 Peptides with Potential Citrullination Sites sequence

amino acids

a

192−210 211−220 363−377 383−421 486−499 534−554 601−615

DFFTNHTLVLHVARSEMDK VRVFQATRGKa TLPVVFDSPRNRGLKa RVMGPDFGYVTRGPQTGGISGLDSFGNLEVSPPVTVRGKa GFRLLLASPRSCYK TLREHNSFVERCIDWNRELLK PFGPVINGRCCLEEKb

citrullination sites R205 R212, R372, R383, R495 R536, R609

R218 R374 R394 R544

a

Peptides were identified only in their noncitrullinated form in the present study. bThis peptide was identified in a citrullinated and noncitrullinated form. Citrullination sites described before are given in bold R. The citrullination site identified in the present study is given as underlined R.

observed both as citrullinated and noncitrullinated in the position R609. The degree of citrullination for this site was calculated as 35%. This citrullination site has not been previously described and may indicate that the sites described by in vitro investigation are less favored for in vivo autocitrullination. However, more synovial fluids should be investigated for significant determination. PAD2 was also identified, but no peptides were detected as citrullinated.

between patients due to individual differences in citrullination patterns.16 In the synovial fluid analyzed in this study a pronounced variation in frequency was observed between citrullinated peptides derived from the same protein (Table 2). For example, for Fibα five citrullinated peptides were identified. Two of these peptides corresponding to the amino acids 559− 575 and 582−599 had a frequency of 19% and 27%, respectively, whereas the remaining three peptides had frequencies below 3%. Further, a citrullinated peptide harboring multiple citrullination sites may exist in multiple citrullination states due to differences in degree of citrullination for each targeted arginine residue. Frequencies of individual citrullination states for peptides derived from Fibα, Fibβ, vimentin, and α-enolase were quantified and are given in Supporting Information (Supplementary Figure S1).

LC−MS/MS Analysis after Enrichment of Citrullinated Peptides

Our group has recently developed a strategy for enrichment of citrullinated peptides based on specific biotinylation of citrulline side chains allowing binding of citrullinated peptides to streptavidin beads followed by MS analysis.12 The LysC digests of the fractions Sup and SupIgG‑HSA depl were subjected to that enrichment procedure followed by LC−MS/MS analysis. First the specificity of the enrichment technique when applied to such a complex sample as synovial fluid was proven. From earlier studies, using both MALDI and ESI, fragmentation of BPG is known to generate a specific signature ion at m/z 270.1 (Figure 2A).12 Therefore, this signature ion was taken as an indicator for the presence of BPG-modified citrullinated peptides. In the MS analysis of the fractions Sup and SupIgG‑HSA depl 44206 and 22993 MS/MS spectra were recorded, from which 42344 and 21920 contained the signature ion, respectively, showing a specificity of more than 95% (Figure

In Vivo Citrullination of PAD

PAD2 and PAD4 have been previously described to be present in inflamed RA synovium;25 however, in vivo autocitrullination has not been examined and is still unclear. By in vitro experiments 10 autocitrullination sites have been described for PAD4 (given in Table 3).26 In our data set from direct analysis of all synovial fluid fractions, PAD4 was identified and peptides harboring seven of the described autocitrullination sites were detected in their noncitrullinated forms only (Table 3). However, we identified a peptide specific for PAD4 corresponding to the amino acids 601−615, which was 2871

dx.doi.org/10.1021/pr500030x | J. Proteome Res. 2014, 13, 2867−2873

Journal of Proteome Research

Article

However, from the MS/MS spectra acquired in the analysis of the BPG-enriched fractions only13 and 10 citrullinated peptides derived from in total 17 proteins could be identified. Detailed information is provided as Supporting Information (Supplementary File S2). The low identification rate might derive from poor fragmentation efficiency of BPG-modified peptides. We observed that fragmentation of the BPG-structure during HCD fragmentation resulted in several BPG-derived ions in addition to the fragment observed at m/z 270.1; among these were m/z 227.1 and m/z 286.1 (Figure 2A). The extensive BPG fragmentation seems to reduce the fragmentation yield of the peptide backbone, resulting in poor sequence coverage of the analyzed peptide and a low quality of MS/MS data (data not shown). Nevertheless, our findings strongly indicate that despite highly increased performance of MS instrumentation there is still a substantial portion of the synovial fluid citrullinome that is currently unreachable for direct MS identification. Certainly, optimization of chromatographical and MS conditions together with improved sample preparation could increase the number of identifications of citrullinated peptides by the direct approach to some extent. However, this increase is most likely still far less than 13-fold, which our enrichment data indicate. Therefore, we expect that applying specific enrichment tools would most probably provide a greater insight into the citrullinome of samples with a complexity as that of synovial fluid. The problem of BPG fragmentation during MS/MS analysis might be overcome by implementing a cleavable site in the modification structure, e.g., a base labile site cleaved under basic conditions or an S−S linker cleaved under reducing conditions. As a result a significantly smaller modification structure of the citrulline side chain would be generated, and MS/MS spectra of the enriched peptides would be more informative leading to their identification.

Figure 2. (A) Chemical structure of BPG with fragmentation sites indicated as lines. (B,C) Analysis of citrullinated peptides present in synovial fluid supernatant before (Sup, upper panel) and after depletion of IgG and HSA (SupIgG‑HSA depl, bottom panel) using BPG-enrichment (B) or direct MS analysis (C). In panel B the pie charts are based on all MS/MS spectra recorded, with (red) and without (blue) the presence of the modification-specific signature ion, at m/z 270.1. The inserted numbers (3673 for Sup and 2146 for SupIgG‑HSA depl) represent estimated numbers of unique monoisotopic masses selected for MS/MS containing the signature ion. In panel C the pie charts are based on all identification events of citrullinated peptides (red) and noncitrullinated peptides (blue). The inserted numbers (119 for Sup and 157 for SupIgG‑HSA depl) represent the number of unique citrullinated peptides identified.

■

ASSOCIATED CONTENT

* Supporting Information S

Supplementary information, tables, and figure as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

2B). In contrast, m/z 270.1 was found in only less than 1% of the MS/MS spectra acquired for the same fractions analyzed by the direct approach, demonstrating a very low rate of false positives caused by other fragment ions with isobaric masses. This finding demonstrates the remarkable specificity of the enrichment strategy. Since data dependent MS/MS acquisition can detect the same peptide multiple times, e.g., in different charge states, the number of MS/MS spectra collected does not correspond to the number of unique peptides present in the sample. Therefore, we investigated the number of nonredundant monoisotopic precursor ion masses from the recorded MS/ MS scans in the analysis of the BPG-enriched fractions Sup and SupIgG‑HSA depl. We found 3835 (fraction Sup) and 2251 (fraction SupIgG‑HSA depl) unique monoisotopic precursor ion masses in the 44206 and 22993 MS/MS scans recorded. As the signature ion m/z 270.1 was present in 95% of the MS/MS scans, this finding indicates the presence of approximately 3600 and 2100 citrullinated peptides present in these fractions. From the direct analysis of the same two fractions 119 and 157 citrullinated peptides were identified, which is only a small portion of the number of citrullinated peptides estimated by the enrichment analysis (Figure 2C).

AUTHOR INFORMATION

Corresponding Author

*Tel: +47 230 73014. Fax: +47 230 73510. E-mail: astrid. [email protected]. Present Address §

Protagen Protein Services GmbH, Inselwiesenstraße 10, 74076 Heilbronn, Germany. Author Contributions

A.E.V.T. and G.A.dS. were involved in the study conception and design of analysis. A.E.V.T. did the sample preparation and data acquisition. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We want to thank Dr. Guro Goll Løvik and Dr. Tore Kvien at Diakonhjemmet Hospital, Oslo, Norway for providing synovial fluid to this study and Prof. Ludvig M. Sollid for reading the manuscript. Further we want to thank Dr. Magnus Arntzen (The Biotechnology Centre of Oslo, UiO, Norway) and Dr. 2872

dx.doi.org/10.1021/pr500030x | J. Proteome Res. 2014, 13, 2867−2873

Journal of Proteome Research

Article

and generate citrullinated autoantigens in rheumatoid arthritis. Sci. Transl. Med. 2013, 5 (209), 209ra150. (18) Chen, Y. Y.; Lin, S. Y.; Yeh, Y. Y.; Hsiao, H. H.; Wu, C. Y.; Chen, S. T.; Wang, A. H. A modified protein precipitation procedure for efficient removal of albumin from serum. Electrophoresis 2005, 26 (11), 2117−27. (19) Wisniewski, J. R.; Zougman, A.; Nagaraj, N.; Mann, M. Universal sample preparation method for proteome analysis. Nat. Methods 2009, 6 (5), 359−62. (20) Cox, J.; Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 2008, 26 (12), 1367−1372. (21) Cox, J.; Neuhauser, N.; Michalski, A.; Scheltema, R. A.; Olsen, J. V.; Mann, M. Andromeda: a peptide search engine integrated into the MaxQuant environment. J. Proteome Res. 2011, 10 (4), 1794−805. (22) Huang da, W.; Sherman, B. T.; Lempicki, R. A. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 2009, 4 (1), 44−57. (23) Huang da, W.; Sherman, B. T.; Lempicki, R. A. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009, 37 (1), 1−13. (24) Wegner, N.; Lundberg, K.; Kinloch, A.; Fisher, B.; Malmstrom, V.; Feldmann, M.; Venables, P. J. Autoimmunity to specific citrullinated proteins gives the first clues to the etiology of rheumatoid arthritis. Immunol.Rev. 2010, 233 (1), 34−54. (25) Foulquier, C.; Sebbag, M.; Clavel, C.; Chapuy-Regaud, S.; Al Badine, R.; Mechin, M. C.; Vincent, C.; Nachat, R.; Yamada, M.; Takahara, H.; Simon, M.; Guerrin, M.; Serre, G. 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. 2007, 56 (11), 3541−53. (26) Andrade, F.; Darrah, E.; Gucek, M.; Cole, R. N.; Rosen, A.; Zhu, X. Autocitrullination of human peptidylarginine deiminase 4 regulates protein citrullination during cell activation. Arthritis Rheum. 2010, 62, 1630−40.

Sandro de Souza (Brain Institute, UFRN, Brazil) for valuable help in bioinformatics. The Proteomics Core Facility is supported by Infrastructure grants from The Norwegian South-East Health Authority.

■

REFERENCES

(1) Vossenaar, E. R.; Zendman, A. J.; van Venrooij, W. J.; Pruijn, G. J. PAD, a growing family of citrullinating enzymes: genes, features and involvement in disease. Bioessays 2003, 25 (11), 1106−1118. (2) Baka, Z.; Gyorgy, B.; Geher, P.; Buzas, E. I.; Falus, A.; Nagy, G. Citrullination under physiological and pathological conditions. Joint Bone Spine 2012, 79 (5), 431−6. (3) Gyorgy, B.; Toth, E.; Tarcsa, E.; Falus, A.; Buzas, E. I. Citrullination: a posttranslational modification in health and disease. Int. J. Biochem. Cell Biol. 2006, 38 (10), 1662−1677. (4) Schellekens, G. A.; de Jong, B. A.; van den Hoogen, F. H.; van de Putte, L. B.; van Venrooij, W. J. Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritis-specific autoantibodies. J. Clin. Invest. 1998, 101 (1), 273−281. (5) Schellekens, G. A.; Visser, H.; de Jong, B. A.; van den Hoogen, F. H.; Hazes, J. M.; Breedveld, F. C.; van Venrooij, W. J. The diagnostic properties of rheumatoid arthritis antibodies recognizing a cyclic citrullinated peptide. Arthritis Rheum. 2000, 43 (1), 155−163. (6) Makrygiannakis, D.; af Klint, E.; Lundberg, I. E.; Lofberg, R.; Ulfgren, A. K.; Klareskog, L.; Catrina, A. I. Citrullination is an inflammation-dependent process. Ann. Rheum. Dis. 2006, 65 (9), 1219−1222. (7) Kinloch, A.; Lundberg, K.; Wait, R.; Wegner, N.; Lim, N. H.; Zendman, A. J.; Saxne, T.; Malmstrom, V.; Venables, P. J. Synovial fluid is a site of citrullination of autoantigens in inflammatory arthritis. Arthritis Rheum. 2008, 58 (8), 2287−2295. (8) Hensen, S. M.; Pruijn, G. J. Methods for the detection of peptidylarginine deiminase activity and protein citrullination. Mol. Cell. Proteomics 2013, 13, 388−96. (9) De Ceuleneer, M.; van Steendam, K.; Dhaenens, M.; Deforce, D. In vivo relevance of citrullinated proteins and the challenges in their detection. Proteomics 2012, 12 (6), 752−60. (10) Senshu, T.; Sato, T.; Inoue, T.; Akiyama, K.; Asaga, H. Detection of citrulline residues in deiminated proteins on polyvinylidene difluoride membrane. Anal. Biochem. 1992, 203 (1), 94− 100. (11) Tutturen, A. E.; Holm, A.; Jorgensen, M.; Stadtmuller, P.; Rise, F.; Fleckenstein, B. A technique for the specific enrichment of citrulline-containing peptides. Anal. Biochem. 2010, 403 (1−2), 43−51. (12) Tutturen, A. E.; Holm, A.; Fleckenstein, B. Specific biotinylation and sensitive enrichment of citrullinated peptides. Anal. Bioanal. Chem. 2013, 405 (29), 9321−31. (13) De Ceuleneer, M.; De Wit, V.; van Steendam, K.; van Nieuwerburgh, F.; Tilleman, K.; Deforce, D. Modification of citrulline residues with 2,3-butanedione facilitates their detection by liquid chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 2011, 25 (11), 1536−1542. (14) Hao, G.; Wang, D.; Gu, J.; Shen, Q.; Gross, S. S.; Wang, Y. 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. 2008, 20 (4), 723−27. (15) Stensland, M.; Holm, A.; Kiehne, A.; Fleckenstein, B. Targeted analysis of protein citrullination using chemical modification and tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2009, 23 (17), 2754−2762. (16) van Beers, J. J.; Schwarte, C. M.; Stammen-Vogelzangs, J.; Oosterink, E.; Bozic, B.; Pruijn, G. J. The rheumatoid arthritis synovial fluid citrullinome reveals novel citrullinated epitopes in apolipoprotein E, myeloid nuclear differentiation antigen, and beta-actin. Arthritis Rheum. 2013, 65 (1), 69−80. (17) Romero, V.; Fert-Bober, J.; Nigrovic, P. A.; Darrah, E.; Haque, U. J.; Lee, D. M.; van Eyk, J.; Rosen, A.; Andrade, F. Immunemediated pore-forming pathways induce cellular hypercitrullination 2873

dx.doi.org/10.1021/pr500030x | J. Proteome Res. 2014, 13, 2867−2873