Marker and Method for Determining the Composition or Purity of a Cell Culture and for Determining In Vitro the Identity of a Cartilage Cell or Synovial Cell

Abstract

The use of a marker for determining the composition or purity of a cell culture, wherein the marker is at least one marker selected from the group consisting of MMP1, ACE, POSTN, SYNPO, SULF1, ARH-GEF28, IGF2BP1, BAIAP2L1, LIMCH1, SCIN, DCLK1, PPL, CRABP2, NES, XPNPEP2 and SLIT3, and the expression level of the at least one marker is determined.

Claims

1.-15. (canceled)

16. A method for assessing the composition or purity of a cell culture, comprising the following steps: a) obtaining a cell sample from a cell culture, b) determining the expression level of at least one marker in the cell sample and c) determining the composition or purity of the cell culture based on the expression level of the at least one marker, characterized in that the at least one marker is selected from the group consisting of ACE, MMP1, POSTN, SYNPO, SULF1, ARHGEF28, IGF2BP1, BAIAP2L1, LIMCH1, SCIN, DCLK1, PPL, CRABP2, NES, XPNPEP2, SLIT3 and combinations of at least two of the aforementioned markers.

17. The method of claim 16, wherein the expression level of the at least one marker is determined in a cell sample obtained from the cell culture.

18. The method of claim 16, wherein the cell culture is a chondrocyte culture.

19. The method of claim 16, wherein the cell culture contains cells that originate from a cartilage biopsy specimen, preferably a cartilage/bone punch biopsy specimen, more particularly a cartilage tissue fragment thereof.

20. The method of claim 19, wherein the biopsy specimen is a biopsy specimen from a joint, more particularly a knee joint.

21. The method of claim 16, wherein the at least one marker is selected from the group consisting of MMP1, POSTN, SYNPO, SULF1, ARHGEF28, IGF2BP1, BAIAP2L1, LIMCH1, SCIN and combinations of at least two of the aforementioned markers, wherein an expression level of the at least one marker above a defined threshold level indicates that the cell culture contains chondrocytes.

22. The method of claim 21, wherein the threshold value is greater than or equal to the expression level of the at least one marker in a pure synovial cell culture.

23. The method of claim 16, wherein the at least one marker is selected from the group consisting of ACE, DCLK1, PPL, CRABP2, NES, XPNPEP2, SLIT3 and combinations of at least two of the aforementioned markers, wherein an expression level of the at least one marker above a defined threshold level indicates that the cell culture contains synovial cells.

24. The method of claim 23, wherein the threshold value is greater than or equal to the expression level of the at least one marker in a pure chondrocyte culture.

25. The method of claim 16, wherein the at least one marker is the ratio of ACE to MMP1 or the inverse ratio of the two.

26. The method of claim 25, wherein the value of the ratio of ACE to MMP1 above a defined threshold level indicates that the cell culture contains synovial cells, wherein the threshold value is preferably less than or equal to the value of the ratio of ACE to MMP1 in a pure synovial cell culture.

27. The method of claim 16, wherein the expression level is determined on an RNA basis, preferably an mRNA basis, or a protein basis.

28. The method of claim 16, comprising the step of using a kit, wherein the kit comprises at least one component for detection of at least one marker selected from the group consisting of ACE, MMP1, POSTN, SYNPO, SULF1, ARHGEF28, IGF2BP1, BAIAP2L1, LIMCH1, SCIN, DCLK1, PPL, CRABP2, NES, XPNPEP2 and SLIT3.

29. A method for producing a medicinal product for novel therapies, more particularly a tissue engineering product, comprising the step of using a cell culture, wherein the composition or purity of the cell culture is determined by means of at least one marker selected from the group consisting of ACE, MMP1, POSTN, SYNPO, SULF1, ARHGEF28, IGF2BP1, BAIAP2L1, LIMCH1, SCIN, DCLK1, PPL, CRABP2, NES, XPNPEP2 and SLIT3.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0148] The figures show the following:

[0149] FIG. 1: scatterplot comparison of log 10 transformed synoviocyte/chondrocyte LQF intensities,

[0150] FIG. 2: volcano plot for visualization of the mean values of log 10 transformed synoviocyte/chondrocyte LFQ intensities (mean LFQ ratios),

[0151] FIG. 3: volcano plot for visualization of the differences in peptide count between synoviocytes and chondrocytes and the accompanying differences in the q value for all proteins not used in the mean rank test,

[0152] FIG. 4: mRNA expression of the synoviocyte marker ACE in chondrocytes and synoviocytes,

[0153] FIG. 5: mRNA expression of the chondrocyte marker MMP1 in chondrocytes and synoviocytes,

[0154] FIG. 6: ratio of mRNA expression of the synoviocyte marker ACE and chondrocyte marker MMP1 in chondrocytes and synoviocytes,

[0155] FIG. 7: ratio of mRNA expression of ACE/MMP1 in cell mixtures,

[0156] FIG. 8: mRNA expression of ACE in cell mixtures and

[0157] FIG. 9: mRNA expression of MMP1 in cell mixtures.

DETAILED DESCRIPTION OF THE FIGURES

[0158] FIG. 1 shows a scatterplot comparison of the log 10 transformed synoviocyte/chondrocyte LQF intensities of donor 1 (x axis) with the other four donors. The line through the origin with slope 1 corresponds to a 100% correlation among repeated experiments. Proteins classified in the mean rank test as showing higher expression in one cell type are shown darker.

[0159] FIG. 2 shows a volcano plot for visualization of the mean values of log 10 transformed synoviocyte/chondrocyte LFQ intensities (mean LFQ ratios) against the standard deviations of the ratios of all proteins quantified in at least 3 of 5 repeated experiments. The black vertical line indicates the border between positive (=higher expression in synoviocytes) and negative (=higher expression in chondrocytes) LFQ ratios. The respective dashed lines indicate a 2- or 10-fold difference in protein expression (corresponding to 0.3 and 1 on the logarithmic scale).

[0160] FIG. 3 shows a volcano plot for visualization of the differences in peptide count between synoviocytes and chondrocytes and the accompanying differences in q value for all proteins not used in the mean rank test. Peptide count differences greater than 1 mean higher expression in synoviocytes, and peptide count differences less than 1 mean higher expression in chondrocytes. The proteins classified in study 1 as biomarker candidates are identified by gene names.

[0161] FIG. 4 shows a graph of mRNA expression of the synoviocyte marker ACE. Chondrocytes and synoviocytes from the same donoras described below under 1. Materials and methodswere cultivated. At the end of cultivation, the mRNA expression of ACE was analyzed. The results are shown relative to expression of the reference gene GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (n=10). The boxplots used in the figure show 25/75 percentiles (boxes), 10/90 percentiles (parentheses), 5/95 percentiles (points), medians (solid lines) and mean values (dotted lines). FIG. 4 shows that the mRNA expression of ACE in synoviocytes is significantly higher than in chondrocytes. ACE is therefore particularly well-suited for use as a synoviocyte marker.

[0162] FIG. 5 shows a graph of mRNA expression of the chondrocyte marker MMP1. Chondrocytes and synoviocytes from the same donoras described below under 1. Materials and methodswere cultivated. At the end of cultivation, the mRNA expression of MMP1 was analyzed. The results are shown relative to expression of the reference gene GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (n=10). The boxplots used in the figure show 25/75 percentiles (boxes), 10/90 percentiles (parentheses), 5/95 percentiles (points), medians (solid lines) and mean values (dotted lines). FIG. 5 clearly shows that the mRNA expression of MMP1 in chondrocytes is significantly higher than in synoviocytes. MMP1 is therefore particularly well-suited for use as a chondrocyte marker.

[0163] FIG. 6 shows a graph of the ratio of the expression of the synoviocyte marker ACE and the chondrocyte marker MMP1 in chondrocytes and synoviocytes (n=10). The boxplots used in the figure show the 25/75 percentiles (boxes), 10/90 percentiles (parentheses), 5/95 percentiles (points), medians (solid lines) and mean values (dotted lines). FIG. 6 shows that the ratio of ACE to MMP1 (ACE/MMP1 ratio) is also suitable as a marker within the meaning of the present invention, more particularly for determining the composition or purity of a cell culture and/or for determining in vitro the identity of a chondrocyte and/or a synoviocyte.

[0164] FIG. 7 shows a graph of the ratio of mRNA expression of ACE to MMP1 in cell mixtures having different contents of chondrocytes and synoviocytes (n=7). The boxplots used in the figure show the 25/75 percentiles (boxes), 10/90 percentiles (parentheses), 5/95 percentiles (points), medians (solid lines) and mean values (dotted lines). FIG. 7 shows that the value of the ratio of ACE to MMP1 correlates with the degree of purity of a chondrocyte-containing cell mixture, specifically such that a low ratio of ACE to MMP1 correlates with a high degree of purity of a chondrocyte-containing cell mixture, and a high ratio of ACE to MMP1 correlates with a low degree of purity of a chondrocyte-containing cell mixture. FIG. 7 shows in particular that the degree of purity of a chondrocyte-containing cell mixture increases with a decreasing ratio of ACE to MMP 1.

[0165] The cell mixtures were produced as follows:

[0166] Chondrocytes and synoviocytesas described below under 1. Materials and methodswere cultivated separately from each other, and on completion of culturing, trypsinized and centrifuged. The cell pellet obtained was resuspended. The cell count of the cell suspensions was determined, and the corresponding volume of each cell suspension was then used to produce mixtures containing 100%, 90%, 75%, 50%, 25%, 10% and 0% chondrocytes. The cell mixtures prepared in this manner were again centrifuged, and the resulting cell pellet was lysed and used for gene expression analysis.

[0167] FIG. 8 shows a graph of mRNA expression of ACE in cell mixtures having different contents of chondrocytes and synoviocytes (n=7). The boxplots used in the figure show 25/75 percentiles (boxes), 10/90 percentiles (parentheses), 5/95 percentiles (points), medians (solid lines) and mean values (dotted lines).

[0168] Concerning production of the cell mixtures, reference is made to the explanations in connection with FIG. 7.

[0169] FIG. 9 shows a graph of mRNA expression of MMP1 in cell mixtures having different contents of chondrocytes and synoviocytes (n=7). The boxplots used in the figure show 25/75 percentiles (boxes), 10/90 percentiles (parentheses), 5/95 percentiles (points), medians (solid lines) and mean values (dotted lines).

[0170] Concerning production of the cell mixtures, reference is also made to the explanations in connection with FIG. 7.

DETAILED DESCRIPTION OF THE EXEMPLARY EMBODIMENTS

1. Materials and Methods

1.1 Cell Culture, Inhibitor Treatment and Cell Lysis

[0171] The chondrocytes and synoviocytes used originated from the knee joints of deceased donors from the US. The tissue was finely chopped using two scalpels and digested with collagenase for 24 h, 5% CO.sub.2 and 37 C. Throughout cultivation, the synovial tissue was treated identically to the cartilage tissue. After digestion, the samples were subjected to standing incubation in chondrocyte medium for 48 h, 5% CO.sub.2 and 37 C. After this, the cells were passed through a cell sieve, centrifuged, and expanded in fresh chondrocyte medium. Depending on growth, the cells were cultivated for 19 to a maximum of 23 days, with the medium being changed 2-3 times.

[0172] After cultivation, the cells were detached from the culture surface with trypsin, washed with PBS, frozen, and sent to Evotec Munich on dry ice.

[0173] Cell lysis and enzymatic cleavage of the proteins was carried out according to a recently published protocol (Kulak, N. A., Pichler, G., Paron, I., Nagaraj, N., and Mann, M. (2014). Minimal, encapsulated proteomic-sample processing applied to copy-number estimation in eukaryotic cells. Nat Methods 11, 319-324). The cells were first stored at 80 C. and then placed on ice prior to lysis. Lysis was carried out with a cold lysis buffer (1% w/v SDC, 10 mM TCEP, 40 mM CAA, 100 mM Tris pH 8.5). Cell extracts were immediately incubated at 99 C. for 10 min and then ultrasound-treated on ice three times for 30 sec. After this, the cell debris was removed by centrifugation. The supernatant was transferred to a new reaction vessel, and protein concentration was determined (BCA Assay, Pierce). 150 g of each cell lysate was used for subsequent enzymatic protein cleavage. A total of 10 chondrocyte and 10 synoviocyte cultures were used for analysis.

1.2 Sample Preparation for Mass Spectrometry (MS)

[0174] The proteins in the cell lysates were reduced with 10 MM TCEP (tris(2-carboxyethyl)phosphine) and alkylated in the presence of 40 mM CAA (2-chloroacetamide). After centrifugation, in each case, a 150 g aliquot of the clear lysate was diluted 1:2 with H.sub.2O, a 1:1 mixture of endopeptidase Lys-C (Wako) and trypsin (Promega) was added in an enzyme-to-protein ratio of 1:50, and incubation was carried out overnight. The resulting peptide mixture was acidified by adding 99% ethyl acetate and 1% TFA, and then desalinated using a Strata-X-C column (100 mg sorbent, Phenomenex). The peptides were eluted with 80% acetonitrile and 5% v/v ammonium hydroxide, shock-frozen in liquid nitrogen and freeze-dried. After this, the peptides were fractionated at high pH by reversed-phase chromatography (Wang, Y., Yang, F., Gritsenko, M. A., Clauss, T., Liu, T., Shen, Y., Monroe, M. E., Lopez-Ferrer, D., Reno, T., Moore, R. J., et al. (2011). Reversed-phase chromatography with multiple fraction concatenation strategy for proteome profiling of human MCF10A cells. Proteomics 11, 2019-2026). For this purpose, the peptides were reconstituted in 20 mM ammonium formiate (pH 10, buffer A), loaded onto an XBridge C18 2004.6 mm analytical column (Waters) of an Akta Explorer System (GE Healthcare), and then separated by means of a segmented gradient with an increasing acetonitrile concentration of 7% to 30% buffer B (buffer A with 80% acetonitrile) for 15 min and then for 5 min with a gradient of up to 55%. The collected fractions of the eluted peptides were then combined in a non-linear manner such that a total of 12 samples were obtained. The samples were then desalinated with a self-produced C18 STAGEtip column (Rappsilber, J., Mann, M., and Ishihama, Y. (2007). Protocol for micro-purification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips. Nat Protoc 2, 1896-1906). For the mass spectrometric analysis, the peptides were reconstituted in 0.5% formic acid.

1.3 Mass Spectrometric Analysis

[0175] All LC-MS/MS analyses were conducted on a Q Exactive Plus mass spectrometer (Thermo Fisher Scientific) with the Easy n-LC 1000 UPLC System (Thermo Fisher Scientific). The samples were loaded with an autosampler onto a 40 cm quartz glass emitter (New Objective) filled with reverse phase material (Reprusil-Pur C18-AQ, 3 m, Dr. Maisch GmbH) to a maximum pressure of 800 bar. The bound peptides were eluted in a run time of 125 min. for analysis of all fractionated samples. Peptides were directly sprayed into the mass spectrometer by means of a nano-electrospray ion source (Proxeon Biosystems). The mass spectrometer was operated in a data-dependent mode in order to allow automatic switching between total MS scans in a resolution of R=70,000 (with m/z=200) with a target value of 3,000,000 counts (max. injection time=45 ms) and MS/MS fragmentation scans at R=17,500 and a target value of 100,000 ions. The 10 most intensive peptide ions were selected by means of HCD (higher-energy collisional dissociation).

1.4 Data Processing

[0176] All of the raw data obtained in the tests were processed with the MaxQuant Software Suite (version 1.5.3.2) for peptide and protein identification and quantitation using the human SwissProt database (version 03 2014) (Cox, J., and Mann, M. (2008). MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantitation. Nat Biotechnol 26, 1367-1372; Olsen, J. V., Vermeulen, M., Santamaria, A., Kumar, C., Miller, M. L., Jensen, L. J., Gnad, F., Cox, J., Jensen, T. S., Nigg, E. A., et al. (2010). Quantitative phosphoproteomics reveals widespread full phosphorylation site occupancy during mitosis. Sci Signal 3, ra3). Label-free quantitation was activated, while the match between runs function was switched off. The maximum allowable mass deviation was 4.5 ppm for MS and 20 ppm for MS2 peaks. Carbamidomethylation of cysteines was input as a fixed modification, and oxidation of methionine and N-terminal acetylation were taken as variable modifications. All peptides had to have a minimum length of 7 amino acids, and a maximum of two uncleaved tryptic cleavage sites (missed cleavages) were allowed. The FDR (false discovery rate) for protein and peptide identification was set at 1%.

1.5 Bioinformatic Data Analysis

[0177] The protein intensities of label-free quantitation (LFQ) were used for calculating the synoviocyte/chondrocyte ratio of the same donor. This pairing resulted in a total of 5 replicates. The ratios obtained were then log 10 transformed and used for further data analysis. Proteins labeled as contaminants or reverse hits in the decoy database were removed from the modified MaxQuant protein groups table. Significant expression differences of proteins with LFQ ratios in three or more replicates were determined by the non-parametric one-sample mean rank test (Klammer M., Dybowski J. N., Hoffmann D., and Schaab C. (2014). Identification of significant features by the Global Mean Rank test. PLoS One. 2014, Aug. 13; 9(8):e104504. doi: 10.1371). The mean rank test is a global rank-based test that reliably controls the FDR (false discovery rate).

[0178] Proteins with ratios in fewer than 3 replicates were subjected to a non-parametric peptide count test, which uses the individual peptide counts for a given protein. For each protein, the mean difference in the peptide counts between chondrocytes and synoviocytes was determined. Group labels were then permuted, and the occurrence of major differences was counted over all of the proteins. This was repeated for all possible permutations using the average number of false discoveries in order to estimate the false discovery (FD) of a given protein.

[0179] For combined analysis of the two studies, a total of 10 donors with related chondrocytes and synoviocytes were used. Significant expression differences in proteins with LFQ ratios in at least 7 replicates were determined by the mean rank test, and all other proteins were subjected to the non-parametric peptide count test.

1.6 Comparison of the Two Data Sets:

[0180] In order to allow comparison of the results of the two individual studies with 5 donors each, the data were compared by either the UniProtKB identifiers and/or the gene name. Based on the quality of the comparison, the data were divided into the following four categories: (++) identical UniProtKB identifiers; (+) partially identical UniProtKB identifiers and identical gene names; (+/) partially identical UniProtKB identifiers and partially matching gene name; () not applicable.

1.7 Gene Expression Analysis

[0181] On completion of cultivation, an aliquot of 0.5 million was taken from the removed cells for analysis of gene expression, centrifuged, and the cell pellet was lysed with lysis buffer. Subsequent RNA isolation was carried out using an RNeasy mini kit 250 (Qiagen). cDNA synthesis was carried out with the Advantage RT for PCR kit (Clontech).

[0182] Gene-specific primer sequences were designed at the NCBI with PrimerBlast and synthesized by Biotez. The primers were used in a concentration of 100 nM or 200 nM/well, and expression of mRNA was analyzed by the real-time PCR method on the Roche LC480. The fluorescent dye SYBR Green was used for detection of the amplified sequences.

2. Results:

[0183] In total, with a determined FDR of <1%, 10,702 different proteins were identified. A similar number of proteins were identified for chondrocytes and synoviocytes (10,550 or 10,462). The identifications were based on more than 2,300,000 peptide detections and 162,299 identified peptides. Throughout all of the experiments, a comparable number of proteins was identified, which speaks for the basic robustness of the workflow used for proteomic analysis.

[0184] In order to show the reproducibility of protein quantitation over biological replicates of 5 donors with related chondrocyte and synoviocyte preparations, scatterplot analyses of label-free quantitation ratios (LFQ ratios) were carried out. For this purpose, the intensities of the LFQ were converted into log 10 transformed ratios that represent the ratio of the LFQ intensity of the synoviocytes to the LFQ intensity of the chondrocytes.

[0185] As an example, the results for donor 1 are shown in FIG. 1. In addition to a large cloud of log 10 transformed protein ratios around zero (=around a linear ratio of approximately 1), which means that the proteins are expressed identically in both cell types, one finds a few proteins that show a different frequency in the two cell types. The scatterplot comparisons show a high correlation of the donors, as the great majority of the proteins are on a line through the origin with slope 1. The results obtained were comparable for the two studies.

[0186] In the next step, further data analyses were conducted in order to identify proteins with significantly different expression levels between the two cell types. For this purpose, all of the proteins quantitated in at least 3 of the 5 biological replicates were subjected to a statistical analysis. These requirements were met by a total of 7,361 proteins in study 2 and 5,378 proteins in study 1. Significant and reproducible changes were identified using the mean rank test with an FDR of 10% (q values 0.1) (Klammer M., Dybowski J. N., Hoffmann D., and Schaab C. (2014). Identification of significant features by the Global Mean Rank test. PLoS One. 2014, Aug. 13; 9(8):e104504. doi: 10.1371). The FDR of 10% used is an arbitrarily selected value that appeared to be suitable for the further bioinformatic analyses of the study, such as gene ontology enrichment and interaction network analysis. The distribution of the proteins with significantly differing protein expression between the two cell types can be visualized using so-called volcano plots. For this purpose, the mean values of the log 10 transformed synoviocyte/chondrocyte LFQ intensities (mean LFQ ratio) were plotted against the standard deviation of the ratios of all proteins. All of the proteins with differential expression between the two cell types are shown in red. Proteins classified as biomarker candidates were labeled by name.

[0187] After application of the mean rank test, approx. 15% of all proteins analyzed in the test were found to differ significantly in chondrocytes and synoviocytes (total of 1,135). In the next step, a peptide count was carried out in order to identify marker proteins that were expressed as exclusively as possible in only one of the two cell types. This test compares the average number of peptides by means of which one protein was identified in the two cell types. For this test, a total of 3,340 proteins were used, 16 of which were differentially expressed (10% FDR, q values 0.1). Of these 16 proteins, 8 were expressed in only one type or expressed to a significantly greater degree in only one cell type.

[0188] Selected proteins with specific protein expression in chondrocytes or synoviocytes that were found by means of the peptide count test in both studies are shown in Table 1.

TABLE-US-00001 TABLE 1 Selected proteins with differential expression in chondrocytes or synoviocytes determined by means of the peptide test. A: results of the first study, B: results of the second study. A ID (EVT03802) Protein name Gene name Cho. #1 Cho. #2 Cho. #3 Cho. #4 Cho. #5 9320 Adseverin SCIN 23 17 4 10 3 1669 Collagen alpha-1(XI) chain COL11A1 10 19 13 19 9 1317 Interstitial collagenase MMP1 18 11 13 4 2 398 Laminin subunit alpha-5 LAMA5 2 4 9 34 0 2995 Lysyl oxidase homolog 3 LOXL3 24 26 14 22 7 1687 Anglotensin-converting enzyme ACE 0 0 0 0 0 6504 Dedicator of cytokinesis protein 10 DOCX10 0 1 5 2 0 6968 Peroxisomal acyl-coenzyme A oxidase 2 ACOX2 0 0 0 0 0 5228 Rap1-GTP-interacting Adaptor Molecule* AP881IP 0 0 0 2 1 735 Slit homolog 3 protein SLIT3 0 0 0 2 0 589 Xaa-Pro aminopeptidase 2 XPNPEP2 1 0 0 0 0 ID Peptide Count (EVT03802) Syn #1 Syn #2 Syn #3 Syn #4 Syn #5 q-value 9320 0 0 0 0 1 0.068 1669 1 1 0 2 3 0.068 1317 1 0 0 0 0 0.068 398 0 0 0 0 0 0.069 2995 1 1 1 2 1 0.062 1687 10 7 8 13 10 0.069 6504 15 12 14 13 13 0.068 6968 6 8 5 9 10 0.098 5228 22 17 16 14 15 0.062 735 14 18 16 25 24 0.062 589 16 7 7 6 5 0.098 B ID (EVT03867) Protein name Gene name Cho. #1 Cho. #2 Cho. #3 Cho. #4 Cho. #5 9320 Adseverin SCIN 9 18 12 8 5 1669 Collagen alpha-1(XI) chain COL11A1 24 36 33 38 16 1317 Interstitial collagenase MMP1 14 6 12 12 13 398 Laminin subunit alpha-5 LAMA5 26 13 3 2 1 2995 Lysyl oxidase homolog 3 LOXL3 23 15 17 18 17 1687 Anglotensin-converting enzyme ACE 3 0 0 1 0 6504 Dedicator of cytokinesis protein 10 DOCX10 14 7 10 3 2 2604 Nestin NES 21 32 9 12 5 6988 Peroxisomal acyl-coenzyme A oxidase 2 ACOX2 0 1 0 1 0 5228 Rap1-GTP-interacting Adaptor Molecule* AP881IP 5 0 0 0 5 735 Slit homolog 3 protein SLIT3 5 0 0 0 0 589 Xaa-Pro aminopeptidase 2 XPNPEP2 0 0 0 0 1 ID Peptide Count Mean Rank (EVT03867) Syn #1 Syn #2 Syn #3 Syn #4 Syn #5 q-value q-value Match type 9320 0 0 0 0 0 0.107 + 1669 2 1 0 13 0 0.000 ++ 1317 0 0 0 0 0 0.100 ++ 398 0 0 2 0 2 0.127 ++ 2995 2 1 5 0 6 0.000 ++ 1687 33 40 33 31 23 0.000 ++ 6504 33 21 20 21 21 0.008 ++ 2604 61 60 49 57 49 0.000 ++ 6988 17 21 12 21 15 0.065 ++ 5228 18 19 15 14 19 0.050 ++ 735 37 25 32 29 27 0.000 ++ 589 22 20 16 20 11 0.060 ++ *(allas) Amyloid beta AA precursor protein-binding family B member 2-interacting protein

[0189] Proteins determined by means of bioinformatic analysis to be expressed in a significantly differing manner are shown in Table 2.

TABLE-US-00002 TABLE 2 Markers for chondrocytes and synoviocytes. ID ID (Protein (Protein group, group, Match Test used q-value Test used q-value Cell type EVT03802) EVT03867) Protein name Gene name type (EVT03802) (EVT03802) (EVT03867) (EVT03867) Chondrocytes 4057 4524 Periostin POSTN ++ Mean Rank 0.009 Mean Rank 0.009 5679 6480 Synaptopodin SYNPO + Peptide Count 0.068 Mean Rank 0.023 1317 1522 Interstitial MMP1 ++ Peptide Count 0.069 Peptide Count 0.100 collagenase 5512 6273 Extracellular SULF1 ++ Mean Rank 0.009 Peptide Count 0.162 sulfatase Sulf-1 5645 6432 Rho guanine ARHGEF28 + Peptide Count 0.098 Peptide Count 0.003 nucleotide exchange factor 28 8360 9665 Insulin-like IGF2BP1 + Peptide Count 0.098 Peptide Count 0.134 growth factor-2 mRNA-binding protein 1 8638 9976 Brain-specific BAIAP2L1 ++ Peptide Count 0.098 Peptide Count 0.091 angiogenesis inhibitor 1- associated protein 2-like protein 1 8901 10270 LIM and calponin LIMCH1 + Peptide Count 0.068 Peptide Count 0.014 homology domains- containing protein 1 9320 10748 Adseverin SCIN + Peptide Count 0.069 Peptide Count 0.107 Synoviocytes 367 440 Serine/threonine- DCLK1 + Mean Rank 0.000 Mean Rank 0.000 protein kinase DCLK1 640 763 Periplakin PPL + Mean Rank 0.000 Mean Rank 0.004 2164 2455 Cellular retinoic CRABP2 ++ Mean Rank 0.000 Mean Rank 0.000 acid- binding protein 2 2604 2927 Nestin NES ++ Peptide Count 0.000 Mean Rank 0.000 589 703 Xaa-Pro XPNPEP2 ++ Peptide Count 0.098 Peptide Count 0.000 aminopeptidase 2 735 876 Slit homolog 3 SLIT3 ++ Peptide Count 0.062 Peptide Count 0.000 protein 1687 1928 Angiotensin- ACE ++ Peptide Count 0.069 Peptide Count 0.000 converting enzyme

[0190] In addition, for the proteins classified as biomarker candidates, the difference in protein expression between the two cell types (-fold difference) and the mean peptide counts are summarized below (Table 3).

TABLE-US-00003 TABLE 3 Differences in expression of the markers between chondrocytes and synoviocytes and mean value of peptide counts (shown separately according to study 1 and study 2 respectively). EVT3802_Fold EVT3867_Fold difference in difference in Test used Test used protein protein Cell type Gene name (EVT03802) (EVT03867) expression expression Chondrocytes POSTN Mean Rank Mean Rank 46.7 15.2 SYNPO Peptide Count Mean Rank 17.3 6.0 MMP1 Peptide Count Peptide Count SULF1 Mean Rank Peptide Count 46.1 2.8 ARHGEF28 Peptide Count Peptide Count IGF2BP1 Peptide Count Peptide Count BAIAP2L1 Peptide Count Peptide Count 3.7 7.5 LIMCH1 Peptide Count Peptide Count 6.7 SCIN Peptide Count Peptide Count Synoviocytes DCLK1 Mean Rank Mean Rank 67.7 19.0 PPL Mean Rank Mean Rank 116.5 4.7 CRABP Mean Rank Mean Rank 71.7 269.8 NES Peptide Count Mean Rank 26.6 114.3 XPNPEP2 Peptide Count Peptide Count SLIT3 Peptide Count Peptide Count 23.2 81.4 ACE Peptide Count Peptide Count EVT3802_Mean EVT3802_Mean EVT3867_Mean EVT3867_Mean value peptide value peptide value peptide value peptide count, count, count, count, synoviocytes chondrocytes synoviocytes chondrocytes Cell type Gene name (#1-#5) (#1-#5) (#1-#5) (#1-#5) Chondrocytes POSTN 7 54.6 12.4 30.6 SYNPO 0.6 13.2 4.6 17 MMP1 0.2 9.6 0 11.2 SULF1 1.8 17.8 3.2 9.2 ARHGEF28 0 7.8 0.8 25.4 IGF2BP1 0 7.6 0.6 7.8 BAIAP2L1 0.6 8.4 1.4 14.2 LIMCH1 0.2 12.8 1.2 22.6 SCIN 0.2 10.4 0 10.4 Synoviocytes DCLK1 29.8 2.6 23.2 10.6 PPL 131.6 10.8 89.6 56.6 CRABP 14.8 2.8 8 2 NES 60.6 5.8 55.2 15.8 XPNPEP2 8.2 0.2 17.8 0.2 SLIT3 19.4 0.4 30 1.2 ACE 0 0.8 indicates data missing or illegible when filed