Article pubs.acs.org/jpr
Identification of Autoantigens in Body Fluids by Combining PullDowns and Organic Precipitations of Intact Immune Complexes with Quantitative Label-Free Mass Spectrometry Juliane Merl,† Cornelia A. Deeg,‡ Margarete E. Swadzba,‡ Marius Ueffing,†,§,⊥ and Stefanie M. Hauck*,†,⊥ †
Research Unit Protein Science, Helmholtz Zentrum Muenchen, German Research Center for Environmental Health (GmbH), D-85764 Neuherberg, Germany ‡ Institute of Animal Physiology, Department of Veterinary Sciences, Ludwig-Maximilians University, 80539 Muenchen, Germany § Centre of Ophthalmology, University Medical Centre, University of Tuebingen, 72076 Tuebingen, Germany S Supporting Information *
ABSTRACT: Most autoimmune diseases are multifactorial diseases and are caused by the immunological reaction against a number of autoantigens. Key for understanding autoimmune pathologies is the knowledge of the targeted autoantigens, both initially and during disease progression. We present an approach for autoantigen identification based on isolation of intact autoantibody−antigen complexes from body fluids. After organic precipitation of high molecular weight proteins and free immunoglobulins, released autoantigens were identified by quantitative label-free liquid chromatography mass spectrometry. We confirmed feasibility of target enrichment and identification from highly complex body fluid proteomes by spiking of a predefined antibody−antigen complex at low level of abundance. As a proof of principle, we studied the blinding disease autoimmune uveitis, which is caused by autoreactive T-cells attacking the inner eye and is accompanied by autoantibodies. We identified three novel autoantigens in the spontaneous animal model equine recurrent uveitis (secreted acidic phosphoprotein osteopontin, extracellular matrix protein 1, and metalloproteinase inhibitor 2) and confirmed the presence of the corresponding autoantibodies in 15−25% of patient samples by enzyme-linked immunosorbent assay. Thus, this workflow led to the identification of novel autoantigens in autoimmune uveitis and may provide a versatile and useful tool to identify autoantigens in other autoimmune diseases in the future. KEYWORDS: autoantibody, autoantigen, autoimmune uveitis, label-free mass spectrometry, Progenesis LC-MS, acetonitrile precipitation
■
surface proteins,5−7 which are, however, most likely the main targets of primary autoimmune reactions. Furthermore, analysis of body fluids in close proximity to the diseased tissues might increase the probability to identify disease-associated immune targets but includes the challenge of working with highly complex samples with protein concentrations spanning several orders of magnitude, like serum/plasma,8 urine, 9 and cerebrospinal fluid.10,11 In the eye disease autoimmune uveitis, the vitreous is the closest body fluid to target tissues, containing disease-related proteins and complexes.12,13 Autoimmune uveitis is mainly characterized by a breakdown of the blood−retinal barrier, infiltration of autoreactive CD4 T-cells, and retinal tissue destruction ultimately leading to blindness.14,15 Disease progression is accompanied by the production of autoantibodies, which are thought to be not directly involved in causing this disease but can serve as a tool for the identification of eyespecific autoantigens.14,16,17 Disease progression and pathology
INTRODUCTION Common autoimmune diseases share characteristic immunopathological features, like immune attacks to autoantigens and inflammation contributing to the destruction of the diseaserelated target tissues. They are often caused by the immunological reaction against a number of autoantigens, both initially and during disease progression.1 In some of these diseases, immune targets have been identified. Very recently, autoantibodies to potassium channel Kir4.1 were identified in a subgroup of multiple sclerosis patients with a 1D/2D gel-based approach.2 Despite these important successes, in many other autoimmune diseases, the related autoantigens potentially participating in disease progression remain widely unidentified so far. However, comprehensive identification of autoantigens not only provides the potential for detailed diagnostics but also helps in patient stratifications (e.g., neuromyelitis optica versus multiple sclerosis3) and even prediction of relapses, disease severity, and progression rates (e.g., in type 1 diabetes4). Unfortunately, classical gel-based proteomic approaches are highly limited with regard to the analysis of membrane or cell © 2013 American Chemical Society
Received: June 24, 2013 Published: September 23, 2013 5656
dx.doi.org/10.1021/pr4005986 | J. Proteome Res. 2013, 12, 5656−5665
Journal of Proteome Research
Article
Table 1. Quality Control of the Described Workflow Using a Spiked Antibody−Antigen Complex enrichment factora controls accession
unique peptides
mascot score
protein description
ENSECAP00000009171 ENSECAP00000016524 P02754 ENSECAP00000007649 ENSECAP00000000334
67 42 10 10 2
5171 2893 534 479 106
serum albumin precursor serotransferrin precursor β-lactoglobulin apolipoprotein A-I immunoglobulin heavy
Lac-spiked
replicate 1 replicate 2 replicate 3 0.02 0.02 0.28 0.11 0.02
0.02 0.03 0.27 0.1 0.01
0.02 0.03 0.25 0.1 0
replicate 1
replicate 2
replicate 3
0.03 0.06 1.45 0.34 0
0.09 0.12 7.74 0.36 0.01
0.06 0.1 4.13 0.47 0
a
The enrichment factor of the quantified proteins was calculated by dividing the raw abundance in the ACN supernatant by the raw abundance in the crude vitrectomy sample.
samples of the control horses originated from horses euthanized because of ERU-unrelated diseases or from a local slaughterhouse. All animals were treated according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. No experimental animals were used in this study.
can be studied in the well-described spontaneous animal model, in equine recurrent uveitis (ERU). 18 The disease is characterized by changes in retinal membrane protein expression patterns,19 like increased abundances of complement components20 and differential expression of K+ channels and aquaporins.21 Furthermore, both the matrix metalloproteinase (MMP) network22 and the extracellular matrix (ECM) of uveitic horse retinas showed significant differences in comparison to control tissue from healthy animals.23,24 So far, some targets of autoantibodies in both the human and the equine disease were identified and verified using classical 1D or 2D gel-based approaches, like retinal arrestin (SAntigen),25−27 interphotoreceptor retinoid binding protein (IRBP), 2 8 , 2 9 and cellular retinoid binding protein (CRALBP).16,30 Still, not all ERU cases show antibody reactivity against any of these three described autoantigens, indicating that additional immune targets might be involved in the pathomechanisms of the disease. As autoimmune diseases are highly dynamic pathological processes, it can be assumed that the identity of associated immune targets would differ at distinct stages of disease development. Unfortunately, the commonly used methods for identification of autoantibody targets do not allow for the discrimination of currently active immunological complexes compared to memory responses from former disease states. Therefore, we developed a novel strategy for identification of acutely targeted proteins by combining enrichment of existing autoantibody−antigen complexes from diseased body fluids and subsequent precipitation of high molecular weight proteins with label-free quantitative mass spectrometry. As a proof of principle, we studied the spontaneous animal model for autoimmune uveitis, but the aim was to establish an approach that complements previous methods and, more importantly, might also access novel immune targets in other autoimmune diseases.
■
Purification of Antibody−Antigen Complexes
For purification of antibody−antigen complexes, 100 μL of 50% CaptureSelect IgG-CH1 affinity matrix (Life Technologies) was used, which recognizes the CH1 domain of human IgG antibodies and allows for the purification of complete IgGs as well as Fab fragments. The beads were washed three times with 1 mL of TBS (50 mM Tris/HCl pH 7.4, 150 mM NaCl) and then incubated with equal amounts of control samples or fresh ERU vitrectomy samples (10 mL each) for 2 h at 4 °C. The supernatant was discarded, and the beads were washed three times with TBS. The final washing buffer was removed, and a mixture (3 + 4) of TBS and acetonitrile (ACN) was added to the beads. The samples were incubated for 30 min at room temperature with mixing every 5 min. After centrifugation for 15 min at 16 000g, the supernatant was transferred to a LoBind tube (Eppendorf) and air-dried. For validation of the enrichment of antibody−antigen complexes and the identification of the antigens, 10 pmol bovine β-lactoglobulin (Sigma-Aldrich) and 5 pmol of a rabbit anti-β-lactoglobulin antibody (Bethyl Laboratories) were added to 10 mL of an ERU vitrectomy sample (three replicates “Lac spiked”, Table 1), resulting in a final complex concentration of 0.5 fmol/μL. After sample preparation, both the input samples and the acetonitrile supernatants of “Lac spiked” and the three “controls” (with spiked antigen, but without spiked antibody) were analyzed by LC-MS/MS. Abundances were gained using the Progenesis LC-MS software and processed using Microsoft Excel. Sample Preparation for Mass Spectrometric Analysis
MATERIAL AND METHODS
Dried pellets from the supernatants of the ACN precipitation were dissolved in 50 μL of 50 mM ammoniumbicarbonate (ABC). For the direct digest of the vitreous input samples, 10 μg was used and diluted using 50 mM ABC. For protein reduction, 2 μL of 100 mM DTT was added to the samples and incubated for 15 min at 60 °C. After cooling the samples to room temperature, 2 μL of freshly prepared 300 mM iodoacetamide solution was added for 30 min in the dark. Two microliters of a 0.5 μg/μL trypsin solution (Promega) was added for protein in-solution digestion at 37 °C overnight. To stop the tryptic digest, 2 μL of a 0.5% trifluoroacetic acid (TFA) solution was added to the sample. The input samples were directly analyzed on the OrbitrapXL. For removal of
Collection of Vitreous Samples
For this study, a total of 167 horse vitreous specimens were sampled and analyzed. Thirty-four of them were derived from control animals, and 133 were derived from horses diagnosed with ERU, according to clinical criteria as described.31 For application of the intact antibody−antigen enrichment workflow, we needed fresh unfrozen material obtained directly from the equine clinic after therapeutic pars plana vitrectomy. The vitreous samples used for ELISA analysis do not depend on the intactness of the antibody−antigen complexes and could therefore be frozen and stored prior to analysis. The vitreous 5657
dx.doi.org/10.1021/pr4005986 | J. Proteome Res. 2013, 12, 5656−5665
Journal of Proteome Research
Article
(ELISA). PolySorb plates (Nunc) were coated with the respective autoantigen (human OPN: R&D Systems, 72.3% identity to horse OPN; human ECM1: R&D Systems, 78.4% identity; human TIMP2: PeproTech, 98.9% identity) in a concentration of 1 μg/mL in carbonate buffer (pH 9.6) at 4 °C overnight. After blocking with 0.5% gelatin in PBS for 1 h at 37 °C, the vitreous sample was added in a dilution of 1:100 in PBS and incubated for 1 h at 37 °C. A polyclonal goat anti-horse IgG Fc POD-labeled antibody (1:50000, Biozol) was used as secondary antibody and incubated for 1 h at 37 °C. The reaction was visualized using tetramethylbenzidine, stopped with 1 M sulfuric acid, and the absorbance was measured at 450 nm using a microplate reader. The differences between ERU cases and controls were statistically analyzed using a two-sided Mann−Whitney test in GraphPad Prism 5.04.
polymers and residual beads of the affinity matrix, the ACN samples were first loaded on spin columns PepClean C18 (Thermo Fisher Pierce, 30 μg binding capacity) according to the manufacturer’s protocol. The peptides were eluted using 70% ACN. The eluates containing the digested peptides were dried in a speedvac (UniEquip) and stored at −20 °C. Mass Spectrometry
Dried PepClean eluates were dissolved in 45 μL of 2% ACN/ 0.5% TFA by incubation for 30 min at room temperature under agitation. Before loading, the samples were centrifuged for 5 min at 4 °C. LC-MS/MS analysis was performed as described previously.19 Twenty microliters of every purified sample was automatically injected and loaded onto the trap column at a flow rate of 30 μL/min in 5% buffer B (98% ACN/0.1% formic acid (FA) in HPLC-grade water) and 95% buffer A (2% ACN/ 0.1% FA in HPLC-grade water). After 5 min, the peptides were eluted from the trap column and separated on the analytical column by a 170 min gradient from 5 to 31% of buffer B at 300 nL/min flow rate followed by a short gradient from 31 to 95% buffer B in 5 min. Between each sample, the gradient was set back to 5% buffer B and left to equilibrate for 20 min. From the MS prescan, the 10 most abundant peptide ions were selected for fragmentation in the linear ion trap if they exceeded an intensity of at least 200 counts and if they were at least doubly charged. During fragment analysis, a high-resolution (60 000 full width at half-maximum) MS spectrum was acquired in the Orbitrap with a mass range from 200 to 1500 Da.
Genomatix GePS (Genomatix Pathway System) Network Analysis
The three identified autoantigens osteopontin, ECM1, and TIMP2 were fed into the Genomatix GePS software (http:// www.genomatix.de/index.html) using the human homologues of the proteins. After simple network generation, the network of these three proteins was extended with the 10 most cited and expert reviewed interaction partners. All described interactions of the three input proteins with the additional proteins were selected to be displayed, including one additional connection between ECM1 and MMP9.33 The proteins were then grouped according to their cellular localization, and the resulting network was exported.
■
Label-Free Analysis Using Progenesis LC-MS
The acquired spectra were loaded to Progenesis software (version 2.5, Nonlinear Dynamics) for label-free quantification and analyzed as previously described.19,32 Profile data of the MS scans were transformed to peak lists with respective peak m/z values, intensities, abundances (areas under the peaks), and m/ z width. MS/MS spectra were treated similarly. After selecting the most complex sample as a reference, the retention times of the other samples were aligned by manual and automatic alignment to a maximal overlay of all features. Features with only one charge or more than eight charges were excluded from further analyses. The three highest quality MS/MS spectra for each feature (rank 1−3 MS/MS spectra) were exported as Mascot generic file and used for peptide identification with Mascot (version 2.2) in the Ensembl Horse protein database (12722794 residues, 22644 sequences). Search parameters used were as follows: 10 ppm peptide mass tolerance and 0.6 Da fragment mass tolerance, one missed cleavage allowed, carbamidomethylation was set as fixed modification, methionine oxidation, and asparagine or glutamine deamidation were allowed as variable modifications. A Mascot-integrated decoy database search calculated an average false discovery of 3 (Table 2). All of these proteins showed either no or much lesser enrichment in the analyzed control samples. In order to evaluate autoantibody frequency specific for these new potential autoantigens in ERU cases, a large cohort was tested for the presence of these autoantibodies in vitreous samples with ELISA. Altogether, 32 controls and 130 ERU samples were analyzed. The recombinant proteins were immobilized on a coated plate and incubated with the diluted vitreous as primary antibody. After incubation with secondary anti-horse IgG Fc antibody and developing the ELISA, the absorbance was measured at 450 nm (Supporting Information Table 3). The normalized values are given in Figure 2. For statistical analysis, a Mann−Whitney test was used, comparing the median between two independent groups of samples,46,47 resulting in a highly significant difference between ERU samples and control samples for all three novel autoantigens. For the separation of positive and negative samples, cut-offs were calculated for the ECM1 and the TIMP2 ELISAs, by adding the 10-fold standard deviation of the control values to the average value of the controls (0.06 for ECM1, 0.075 for TIMP2). Since all control values were 0 for the OPN ELISA, the cutoff was in this case artificially set to 0.05 (Supporting Information Table 3). Thus, ELISA experiments revealed OPN
or ECM1 reactivity in approximately 25% of the investigated ERU vitreous samples (19/74 positive for OPN autoantibodies, 31/116 for ECM1) and TIMP2 reactivity in 15% of the ERU samples (8/51 for TIMP2) (Figure 2). In comparison, the control samples showed no reactivity above the respective cutoff value for all three potential autoantigens. In previous studies, 50% of the investigated ERU vitreous samples were found positive for IRBP antibodies and 60% positive for S-Antigen autoantibodies, but approximately 30% showed no reactivity against either of the two autoantigens.14 Presence of autoantibodies directed against CRALBP was also studied by ELISA analyzing sera of horses, with 30% positive ERU samples in comparison to 12% positive controls.16 Genomatix GePS Network Analysis of the Identified Autoantigens
Of the 33 ERU samples analyzed in all three ELISA setups, 12 vitreous samples showed reactivity against at least one of the tested proteins. Interestingly, five of those samples showed reactivity against two candidate proteins (1× ECM1 and TIMP2, 4× ECM1 and OPN) and even six against all three investigated targets (Supporting Information Table 3). Consequently, the ELISA results not only confirmed the mass spectrometric identifications and quantifications but also indicated some functional correlation between these three novel autoantigens in the pathomechanism of ERU. In order to assess a potentially functional connection between autoantibody binding to the three novel autoantigens, we performed a network analysis. The Genomatix pathway system (GePS) was chosen for the generation and visualization of the associated protein network, as it allowed the combination of our results with verified literature data for proteins and their interactions. Feeding the three novel autoantigens into the software and searching for the 10 most cited and expert reviewed interaction partners resulted in the following protein list: five matrix metalloproteinases (MMP 2, 3, 7, 9, 14); fibroblast growth factor-2 (FGF2), interferon gamma (IFNG), 5661
dx.doi.org/10.1021/pr4005986 | J. Proteome Res. 2013, 12, 5656−5665
Journal of Proteome Research
Article
Figure 3. Genomatix GePS network analysis of the new identified autoantigens. In order to evaluate whether the potential new autoantigens identified in this study are functionally correlated, a network analysis was performed. The three identified autoantigens (white) were analyzed with Genomatix GePS software. The network was extended with the 10 most cited and expert reviewed interaction partners (gray), including five MMPs, several growth factors, the epidermal growth factor receptor, and the transcription factor JUN. The interaction between ECM1 and MMP9 was added manually based on earlier results.33
or tertiary structure. Still, a reproducible autoantibody binding may be highly dependent on the integrity of three-dimensional protein structures, which are at least partially disrupted in the described methods. In another study, serum IgGs from systemic lupus erythematosus patients were first covalently coupled to ProteinG-Sepharose and then incubated with protein lysates derived from target-associated cell lines or tissue.63 Bound antibody targets were then eluted, and 154 candidate antigens could be identified with shotgun LC-MS/MS. Altogether, 11 of these proteins were selected for verification of autoantibody presence in patients’ samples by Western blots or ELISA, among them three previously unreported autoantigens. Still, in this workflow, like in most studies described so far, the presence of autoantibodies directed against the identified autoantigens was shown but not the existence of the corresponding autoantibody−autoantigen complexes in patient samples. The identification of immune targets after a direct immunoprecipitation of these complexes from diseased body fluids is challenging, due to the very high abundance of free immunoglobulins in typical patient samples, which makes the identification of comparably low abundant autoantigens very difficult. In a study by Monach et al., a direct immunoprecipitation approach using ProteinG-Sepharose to detect autoantigens in rheumatoid arthritis was described,64 where bound antigens were eluted by low-pH steps. Still, of 53 specific peptides identified by mass spectrometry, 43 were derived from IgG (∼80%) and 3 from other contaminants. Only 7 peptides from 4 different proteins were presented as putative autoantigen sequences. Analyzing immune complexes from altogether 8 ERU samples with our workflow, we could identify more than 150 proteins with 600 unique peptides (Supporting Information Table 2), of which only ∼10% were assigned to immunoglobulin sequences. In contrast to former studies, the presented explorative workflow comprises the additional organic precipitation step using ACN,39,40 in order to remove disturbing high molecular weight proteins including especially immunoglobulins prior to LC-MS/MS identification of precipitated immune targets. Thus, the described method can deliver a list of
transforming growth factor beta 1 (TGFB1), epidermal growth factor receptor (EGFR), and the transcription factor AP-1 (JUN) (Figure 3). All three novel autoantigens showed a direct interaction with at least one of the five depicted matrix metalloproteinases.33,48−50 In general, MMPs are known to play a major role in extracellular matrix remodeling by degradation of ECM components.51 Because of this direct influence on the cellular surface of target tissues, MMPs and their inhibitors (e.g., TIMP2) are thought to be involved in various immunological52 and autoimmune processes,22,53 including autoimmune uveitis. The other five interaction partners (FGF2, IFNG, TGFB1, EGFR, JUN) showed a direct connection with OPN and/or TIMP2. As matrix metalloproteinases are also involved in processing of growth factors and cytokines,54 this clearly proposed a connection of the novel autoantigens and the MMP network with cytokine and growthfactor-dependent cell signaling in the retina. Interestingly, OPN was previously found to be involved in several inflammatory and immune-modulating diseases: multiple sclerosis,55 type 1 diabetes,56 and also autoimmune uveitis.23,57 Additionally, the abundance of both TIMP2 and OPN was described to be significantly reduced in uveitic retina and vitreous samples, and both the MMP network and the ECM undergo severe changes in ERU-affected eyes.22,23 The binding of autoantibodies to OPN, ECM1, and/or TIMP2 at the surface of retinal cells might lead to a degradation of these proteins and therefore influence the described processes and potentially contribute directly to the destruction of retinal cells in ERU. Alternatively, the targeting of those antigens could be a side effect of and therefore an indication of occurring damages in the ECM of the retina in autoimmune uveitis. Comparison to Previously Used Methods
So far, autoantigens in autoimmune diseases were mainly identified by Western blots,2,58 through the binding of autoantibodies from body fluids to extracted, denatured, and immobilized proteins derived from the disease-associated tissue. Since autoantibodies also target conformational epitopes, protein microarrays59−62 are suited to detect such reactions because attached proteins at least partly retain their secondary 5662
dx.doi.org/10.1021/pr4005986 | J. Proteome Res. 2013, 12, 5656−5665
Journal of Proteome Research
■
interesting autoantigen candidates, which can be the basis for further validation and characterization of the respective disease processes.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail:
[email protected]. Tel: +49-89-3187 3941.
CONCLUSIONS In summary, the developed workflow proved to be highly sensitive and, depending on complex concentrations in patient samples, could be capable of detecting novel, low abundant autoantigen candidates in body fluids in the context of a variety of autoimmune diseases. In contrast to former studies, it is possible to investigate in detail autoantibody binding at different stages of the disease (e.g., relapse vs remission), as not only the presence of autoantibodies but the integrity of a corresponding active immunological complex is studied. Therefore, it might be possible to distinguish memory responses of former disease states from currently targeted proteins and pathways in order to improve understanding of disease initiation and progression and potentially even offer new ways of disease treatment.
■
Article
Author Contributions ⊥
Equal senior authors.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We like to thank Fabian Gruhn, Sandra Helm, and Silke Becker for their excellent technical assistance, and Christine von Toerne for constructive discussions. This work was supported by grants from the Deutsche Forschungsgemeinschaft, SFB 571/A5 Deeg and DFG DE 719/4-1.
■
REFERENCES
(1) Mackay, I. R.; Rowley, M. J. Autoimmune epitopes: autoepitopes. Autoimmun. Rev. 2004, 3, 487−492. (2) Srivastava, R.; Aslam, M.; Kalluri, S. R.; Schirmer, L.; Buck, D.; Tackenberg, B.; Rothhammer, V.; Chan, A.; Gold, R.; Berthele, A.; Bennett, J. L.; Korn, T.; Hemmer, B. Potassium channel KIR4.1 as an immune target in multiple sclerosis. N. Engl. J. Med. 2012, 367, 115− 123. (3) Roemer, S. F.; Parisi, J. E.; Lennon, V. A.; Benarroch, E. E.; Lassmann, H.; Bruck, W.; Mandler, R. N.; Weinshenker, B. G.; Pittock, S. J.; Wingerchuk, D. M.; Lucchinetti, C. F. Pattern-specific loss of aquaporin-4 immunoreactivity distinguishes neuromyelitis optica from multiple sclerosis. Brain 2007, 130, 1194−1205. (4) Batstra, M. R.; Aanstoot, H. J.; Herbrink, P. Prediction and diagnosis of type 1 diabetes using beta-cell autoantibodies. Clin. Lab. 2001, 47, 497−507. (5) Braun, R. J.; Kinkl, N.; Beer, M.; Ueffing, M. Two-dimensional electrophoresis of membrane proteins. Anal. Bioanal. Chem. 2007, 389, 1033−1045. (6) Zahedi, R. P.; Moebius, J.; Sickmann, A. Two-dimensional BAC/ SDS-PAGE for membrane proteomics. Subcell. Biochem. 2007, 43, 13− 20. (7) Elschenbroich, S.; Kim, Y.; Medin, J. A.; Kislinger, T. Isolation of cell surface proteins for mass spectrometry-based proteomics. Expert Rev. Proteomics 2010, 7, 141−154. (8) Anderson, N. L.; Anderson, N. G. The human plasma proteome: history, character, and diagnostic prospects. Mol. Cell. Proteomics 2002, 1, 845−867. (9) Thongboonkerd, V. Practical points in urinary proteomics. J. Proteome Res. 2007, 6, 3881−3890. (10) Monsarrat, B.; Gonzalez de Peredo, A.; Mouton-Barbosa, E.; Roux-Dalvai, F.; Bouyssié, D.; Berger, F.; Schmidt, E.; Righetti, P. G.; Guerrier, L.; Boschetti, E.; Burlet-Schiltz, O. In depth exploration of cerebrospinal fluid by combining peptide ligand library treatment and label free protein quantification. Mol. Cell. Proteomics 2010, 9, 1006− 1021. (11) Stoop, M. P.; Dekker, L. J.; Titulaer, M. K.; Burgers, P. C.; Smitt, P. A. E. S.; Luider, T. M.; Hintzen, R. Q. Multiple sclerosisrelated proteins identified in cerebrospinal fluid by advanced mass spectrometry. Proteomics 2008, 8, 1576−1585. (12) Angi, M.; Kalirai, H.; Coupland, S. E.; Damato, B. E.; Semeraro, F.; Romano, M. R. Proteomic analyses of the vitreous humour. Mediators Inflammation 2012, 2012, 148039. (13) Hauck, S. M.; Hofmaier, F.; Dietter, J.; Swadzba, M. E.; Blindert, M.; Amann, B.; Behler, J.; Kremmer, E.; Ueffing, M.; Deeg, C. A. Label-free LC-MS/MS analysis of vitreous from autoimmune uveitis reveals a significant decrease in secreted Wnt signalling inhibitors DKK3 and SFRP2. J. Proteomics 2012, 75, 4545−4554.
ASSOCIATED CONTENT
S Supporting Information *
Supplementary Table 1: Raw abundances and enrichment factors of all identified proteins in the Lac-IP control data set. Supplementary Table 2: Raw abundances and enrichment factors of all identified proteins in the antibody−antigen complex enrichment data set. Supplementary Table 3: Summary of samples analyzed in the three presented ELISA data sets, with their respective status regarding antibody reactivity against the three putative autoantigens. Supplementary Table 3.1: Measured ELISA values with osteopontin as putative autoantigen. Supplementary Table 3.2: Measured ELISA values with ECM1 as putative autoantigen. Supplementary Table 3.3: Measured ELISA values with TIMP2 as putative autoantigen. Supplementary Figure 1: Molecular weight distribution of proteins after acetonitrile precipitation. (A) Frequency of molecular weights within the data set comprising 151 proteins from ERU IP after acetonitrile precipitation is plotted against their respective molecular weights; 76% of the proteins have a predicted molecular weight below the assumed precipitation cutoff of 70 kDa. (B) ERU vitrectomy sample (10 mL) was processed following the described workflow. Samples were taken from the different fractions and analyzed on a silver-stained 12% SDS-PAGE. Both the vitreous input and the flow-through contain high amounts of serum albumin (∼70 kDa); the pellet of the ACN precipitation shows strong protein signals at ∼50 and ∼25 kDa representing the heavy and the light chain of precipitated immunoglobulins. The supernatant of the precipitation still contains low amounts of the IgG chains and shows additionally weak signals at
