Differences in the secretome of cartilage ... - Wiley Online Library

Differences in the Secretome of Cartilage Explants and Cultured Chondrocytes Unveiled by SILAC Technology Martin Polacek,1,3 Jack-Ansgar Bruun,2 Oddmund Johansen,1,3 Inigo Martinez1 1 Orthopaedic Surgery Department, Institute of Clinical Medicine, University of Tromsø, Breivika, Tromsø 9037, Norway, 2Proteomic Platform, University of Tromsø, Tromsø, Norway, 3Orthopaedic Surgery Department, University Hospital of North Norway, Tromsø, Norway

Received 3 August 2009; accepted 24 October 2009 Published online 27 January 2010 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jor.21067

ABSTRACT: The main goal of our study was to analyze and compare the profiles of secreted proteins from adult human articular chondrocytes in monolayers, and cartilage explants in culture, using a de novo protein labeling approach. Stable isotope labeling of proteins in culture was used to differentiate between chondrocyte-derived proteins and other preexisting matrix-derived components, or proteins coming from serum or synovial fluids. Proteins in culture supernatants were resolved by one-dimensional SDS-PAGE electrophoresis, and analyzed in tandem with LC/MS-MS (liquid chromatography/double mass spectrometry). Results from stable isotope labeling with amino acids in cell culture (SILAC) were validated by specific immunoblotting of four relevant proteins identified in the secretion media. After 8– 10 days of culture, over 90% of proteins secreted during monolayer growth contained 13C6-Arg and 13C6-Lys. Nonlabeled proteins corresponded mostly to plasma-associated proteins, indicating background contamination of medium with serum remnants. The majority of the secreted proteins in 2D cultures were extracellular matrix components and matrix regulators, along with some inflammatory agents and metabolic enzymes. In explants, only 25%–30% of proteins were labeled with heavy amino acids, corresponding to matrix regulators and carrier molecules. Nonlabeled proteins corresponded primarily to structural matrix components. In qualitative terms, all labeled proteins coming from cartilage explants were also found in chondrocytes supernatants. In summary, our results show differences in the labeling pattern of proteins found in supernatants from explants and monolayers. Most proteins found in the media of explants were subproducts of matrix turnover rather than newly synthesized. To our knowledge, this study is the first one so far applying SILAC technology in the context of cartilage and chondrocytes physiology. ß 2010 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 28:1040– 1049, 2010 Keywords: SILAC; chondrocytes; secretome; mass spectrometry; cartilage

The chondrocytes are the only cell type found in cartilage tissue, and account for only 1%–5% of the volume in human articular cartilage. The main function conferred to chondrocytes is the maintenance of the cartilage matrix structure, composition, and properties. The homeostasis of cartilage tissue in joints requires a constant and regulated synthesis/degradation process of the matrix components. In adults, articular cartilage poses limited or no potential of self-repair, thus cartilage defects in adults caused by trauma or by an unfavorable balance of matrix anabolic and catabolic events, would lead to erosion of the cartilage surface and eventual development of osteoarthritis (OA), a disease of steadily increasing incidence in developed countries. A proper understanding of the chondrocyte physiology would constitute a solid ground to discover potential diagnostic markers and therapeutic targets in OA patients, and would help to optimize cartilage tissue engineering strategies. Many of the protocols in use for promoting biological repair of cartilage defects, or for the development of laboratory engineered biological implants, require an initial ex vivo expansion period of the chondrocytes. During the expansion phase, the cells dedifferentiate and gradually shift to a more fibroblastic phenotype, featured by a change in the synthesis of matrix constituents.1,2 Generally, the dedifferentiation process is characterized by a downregulation of cartilage-specific Correspondence to: Inigo Martinez (T: F: þ4777645400.; E-mail: [email protected])

þ4777644686;

ß 2010 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.

1040

JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

proteoglycans and other cartilage structural proteins, such as collagen type II, cartilage oligomeric matrix protein, and collagen type IX. Concomitantly, the synthesis of matrix proteins characteristic of more fibrous tissues, such as collagen type I, are induced. Dedifferentiation of ex vivo expanded chondrocytes still represents one of the major concerns for the use of adult chondrocytes in cell-based therapies for biological repair of cartilage. The mechanisms behind in vitro cell dedifferentiation are still poorly understood. Cartilage explants, on the contrary, preserve the chondrocytes within their natural 3D environment and maintain the cartilage phenotype. Importantly, protein release into the media from cartilage explants is analogous to the bulk of proteins released in vivo into the synovium during tissue remodeling.3 Although explants supernatants lack the complexity of the intact synovial joint, cartilage cells and matrix are major contributors to the bulk of proteins encountered in synovial fluids, and thus studies with cartilage explants may contribute significantly to the understanding of the mechanism behind cartilage matrix remodeling and also arthritis. Genome and proteome-wide analyses of tissues and cells have gained enormous attention in recent years. Secretomics has emerged recently as a new discipline to study globally the proteins that are secreted by cells, tissues, or organisms at any given time or under certain conditions.4,5 Tissue remodeling is orchestrated, at least in part, by autocrine and paracrine factors released by cells during tissue regeneration, thus, study of cell and tissue secretomes represents a very attractive unbiased

CHONDROCYTES SECRETOME BY SILAC

method of searching for unknown cell-produced regulators, or eventually for exploring the effects that different culture environments exert on cells during in vitro culture. Very few studies have focused on the analysis of secreted proteins from ex vivo cartilage explants or cultured chondrocytes.67 The main drawback in those studies was to discriminate between chondrocytederived proteins or diffused proteins from plasma or synovial fluid. Stable isotope labeling by amino acids in cell culture (SILAC) represents a simple and accurate method to explore expression proteomics in cell culture.8–11 This methodology described by Ong et al.12 enables the identification of functional expression of proteins by administration to the culture medium of nonradioactive isotope labeled essential amino acids that are incorporated into all newly synthesized proteins. In our study, we have analyzed and compared the secreted proteome-profiling of cultured chondrocytes and cartilage explants by using a stable isotope labeling strategy in combination with 1D SDS-PAGE protein fractionation (shotgun), in-gel digestion, and LC/MS-MS analysis.

MATERIALS AND METHODS Patient Data and Human Material All human material used in this study was obtained from patients suffering from moderate to advanced cartilage defects in the knees. Chondrocytes for monolayer cultures were obtained from surplus cells after autologous chondrocyte implantation (ACI) operations of large cartilage defects. Cells from four different patients were included in this study, two males and two females, 45, 48, 49, and 39 years old. Cartilage explants were recovered from macroscopically healthy-looking cartilage from total knee replacement cases. Biopsies were taken from lateral femoral condyles and tibial plateaus of males 58, 56, and 51 years old. The patients participated with informed consent, and the Regional Ethical Committee at the University of Tromsø approved the study. Preparation of Cartilage Explants After operative extraction of macroscopically healthy-looking cartilage from the least affected lateral femoral and tibial condyle, the cartilage tissue was cut into small pieces of approximately 1–1.5 mm3, washed extensively in saline solution, and placed for 10 days into SILAC medium. The amount of tissue used in each experiment corresponded to approximately 500 mg wet-weight, which contained approximately 2 105 viable chondrocytes incubated in 3 ml of serumfree culture medium. Medium was changed after 5 days and replaced with new medium. Afterwards, the medium was collected and subjected to proteomic analysis. Viability of cells at the end of the experiment was above 90%, as determined by trypan-blue exclusion assay. Isolation and Culturing of Human Articular Chondrocytes Cartilage biopsies were kept in 0.9% NaCl for approximately 2 h, and then cut in 1–1.5 mm3 pieces. The cartilage pieces were kept for 18 h in 2–5 ml DMEM/HAM’s F-12 (Cat. No. T 481-50, BioChrom Labs, Terre Haute, IN) containing collagenase (Cat. No. C-9407, Sigma-Aldrich Norway AS, Oslo, Norway) at a final concentration of 0.8 mg/ml. The enzyme solution was removed after centrifugation at 200 g and by consecutive washing steps with DMEM/HAM’s F-12. There-

1041

after, the pellet was resuspended in fresh growth medium (DMEM/HAM’s F-12 supplemented with 10% human autologous serum). Cultures were further expanded by trypsinization (Cat. No. T-3924, Sigma-Aldrich, Missouri), and after repeated washing, resuspended in DMEM/HAM’s F-12 supplemented with 10% human serum from same patient. Preparation of Conditioned Medium for SILAC D-MEM and amino acids were from the SILACTM Protein Identification and Quantitation Kit purchased from Invitrogen, California (Cat. No. SM10002). Basal medium was supplemented with ascorbic acid, L-glutamine, dexamethasone, antibiotics, and ITS supplement (Sigma-Aldrich, Cat. No. I3146-5ML). Amino acids [U-13C6] L-Arginine and [U-13C6] L-Lysine were added to the medium as described in the protocol for the kit. After initial cell expansion, 2 106 chondrocytes were seeded in T-75 cell culture flasks. Culture medium was originally supplemented with 10% serum to promote cell adherence during the first 24 h. After cell attachment, the cultures were extensively washed with basal DMEM medium, and thereafter the cells were incubated in 6 ml of [U-13C6] L-Lysine and [U-13C6] L-Arginine culture medium. Supernatants were removed after 5 days and replaced by the fresh labeled medium. The new supernatants were conditioned during an additional 5 days and collected for analyses. Preparation of Protein Samples and 1D Gel Electrophoresis Six milliliters of conditioned cultured supernatants were collected from each flask (2 106 cells), and sterile-filtered with 0.22 mm syringe-filter units to eliminate potential cell debris and particulate material from fluids. Subsequently, the supernatants were concentrated by ultrafiltration (membrane cut-off 3 kDa; Vivascience, Aubagne, France) at 4,500 rpm for 50 min, until a final volume of 200–800 ml was obtained. Protein quantification was performed for each concentrated supernatant using the Bradford protein assay. Supernatants were thereafter solubilized in sample buffer (NuPAGE; Invitrogen), and sample reducing agent (NuPAGE; Invitrogen), and about 40 mg of total protein were loaded to the gel (NuPAGE Novex Bis-Tris 4%–12%; Invitrogen). The electrophoresis was performed using XCell SureLockTM Mini-Cell system (Invitrogen). Gel Fractionation (Shotgun), Trypsin Digestion, and Mass Spectrometry Following electrophoresis, the gels were fixed, washed three times for 5 min in milliQ-water, (Millipore, Massachusetts) under constant shaking and incubated in Coomassie blue for 1 h. Gels were afterwards washed extensively in milliQ-water until optimal contrast in stained bands was obtained. Afterwards, gels were processed by shotgun approach,11 and around 20 gel sections were incised from each column (Fig. 1). Gel pieces were subjected to in-gel reduction, alkylation, and tryptic digestion using 2–10 ng/ml trypsin (V511A, Promega, Madison, WI).13 Peptide mixtures containing 0.1% formic acid were loaded onto a nanoAcquityTM Ultra Performance LC (Waters, Milford, MA), containing a 3-mm Symmetry1 C18 Trap column (180 mm 22 mm) (Waters) in front of a 1.7-mm AtlantisTM C18 analytical column (100 mm 100 mm) (Waters). Peptides were separated with a gradient of 5%–95% acetonitrile, 0.1% formic acid, with a flow of 0.4 ml/ min eluted to a Q-TOF Ultima Global mass spectrometer (Micromass/Waters), and subjected to data-dependent tandem mass spectrometry analysis. Peak lists were generated by the JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

1042

POLACEK ET AL.

a 1:1,000 dilution followed by a secondary antibody (antirabbit-Osteomodulin; anti-mouse-SPARC, TIPM1, GAS6) conjugated to horseradish peroxidase (1:5,000 dilution). Bound antibody was detected by chemiluminescence.14

RESULTS

Figure 1. One dimensional gel electrophoresis of the conditioned medium from chondrocytes monolayers from three different patients. Following 10 days of incubation in stable isotope labeling with amino acids in cell culture (SILAC) medium, culture supernatants from three different individuals were collected, filtered, concentrated, and 40 mg of total protein content was loaded into 4%– 12% acrylamide gel followed by staining with Coomassie blue. The entire spectrum of proteins was processed by shotgun approach. Over 20 bands were excised from gels and the protein content of each single band was analyzed by LC/MS-MS.

ProteinLynx Global server software (version 2.1). The resulting pkl files were searched against the Swiss-Prot 55.5 protein sequence data bases using an in-house Mascot server (Matrix Sciences, London, UK). Peptide mass tolerances used in the search were 100 ppm, and fragment mass tolerance was 0.1 Da. Proteins identified in Mascot containing peptides with bold red peptides containing the Label: 13C(6) (K) were identified as secreted from the chondrocytes. Proteins identified in Mascot with bold red Lysine or Arginine containing peptides were classified as not chondrocyte secreted. Western Blotting Supernatants were solubilized in sample buffer (NuPAGE; Invitrogen) and sample reducing agent (NuPAGE; Invitrogen), and about 30 mg of total protein were loaded to the gel (NuPAGE Novex Bis-Tris 4-12%; Invitrogen). The SDS-PAGE electrophoresis was performed using XCell SureLockTM MiniCell system (Invitrogen). Subsequent to electrophoresis, the proteins were transferred to polyvinylidene difluoride (PVDF) membranes with pore size 0.2 mm (Invitrogen). Western blotting was performed using primary antibodies recognizing human Metalloproteinase inhibitor 1, SPARC, Osteomodulin (Abcam, Cambridge, UK), and Growth arrest-specific protein 6 (GAS6) (Abnova, Taiwan). The membranes were blocked with 3% dried milk in 10 mM Tris–HCl, pH 7.4, 0.15 M NaCl, and 0.2% Tween, and incubated with the primary antibodies at JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

Preparation of Conditioned Medium and Protein Separation by Electrophoresis The main goal of our study was to compare the profile of proteins secreted by cartilage explants against the proteins secreted by chondrocytes cultured in monolayers. With that goal in mind, we initiated experiments collecting culture medium from chondrocytes and cartilage explants maintained for several days in serum-free conditions. Early on, it became clear that using 40 mL of serum-free culture supernatants collected after 5 days of incubation was not enough to show enough band complexity (data not shown). Subsequently, supernatants were concentrated by ultrafiltration (molecular weight cut off ¼ 3 kDa). The yield of protein recovery from 6 ml of the culture supernatant was about 2–3 mg/ml/106 cells. In this study, the same amount of proteins was loaded into each comparison (40 mg/lane). We have made several 1-DE gels from different patients in both groups in order to verify the reproducibility of the results (Fig. 1). To enhance the protein coverage, the gel was processed by shotgun technology11 (Fig. 1). The whole gel column was sliced into 20 sections regardless banding pattern and collected for LC/MS-MS analysis. Identification of Proteins Secreted by Chondrocytes Cultured in Monolayers Despite the extensive washing and careful handling of cell cultures, some of the proteins encountered in supernatants were not chondrocyte-derived, but rather plasma/serum remnant proteins from earlier culture periods (Table 1). The exact sorting of self-generated and contaminant proteins have not been established hitherto. In our study, we took advantage of the SILAC technology to discriminate between chondrocyteproduced and contaminant proteins. Initially, after 2days incubation in SILAC medium, the medium was replaced and the cells were incubated in new SILAC medium for 3 additional days. Under these conditions, the analyses revealed that around 75% of total protein types were isotope labeled (data not shown). To maximize the amount of labeled proteins, we increased the incubation period in SILAC medium to 10 days, including one medium change after 5 days. In this way, we achieved incorporation of the isotope label to over 90% of identified proteins. Table 2 summarizes the list of the 103 proteins labeled with 13C6-Lys and 13C6-Arg identified in chondrocyte supernatants. Most of the proteins were matrix components or matrix regulating factors including proteases, protease inhibitors, different growth factors, and antiinflammatory proteins. A very reduced amount of proteins were classified as intracellular components or metabolic agents.

1043


Table 1. Proteins not containing chondrocytes in monolayers

Protein name Hemopexin Ig alpha-1 chain C region Ig gamma-4 chain C region Ig kappa chain C region Ig lambda chain C regions Inter-alpha-trypsin inhibitor heavy chain H4 Serotransferrin Serum albumin Vitamin D-binding protein

13

C6 labeled peptides, identified in the secretion medium of human articular

Accession no.

Unique peptides detected

Sequence coverage (%)

Mascot Score

Unique non C peptides

HEMO_HUMAN IGHA1_HUMAN IGHG4_HUMAN KAC_HUMAN LAC_HUMAN ITIH4_HUMAN

5 7 6 5 4 6

12 33 29 80 57 8

70 174 148 134 50 137

5 7 6 5 4 6

TRFE_HUMAN ALBU_HUMAN VTDB_HUMAN

41 7 5

75 11 14

3038 132 50

41 7 5

13

Unique peptides detected, number of peptides on which the protein identification is based (the peptides are not used to identify other proteins in the same Mascot search); Sequence coverage (%), portion of the protein sequence covered by the identified peptides; Mascot score, significance of the protein identification (the higher score, the higher probability). This protein score is the sum of the unique ions scores. Unique non 13C peptides, number of unique peptides identified, not containing 13C6 label.

Identification of Proteins Recovered from the Media of Cartilage Explants In cartilage explants, as much as 71% of identified proteins were not labeled with 13C6-Lys and 13C6-Arg, indicating that they were not directly produced by the cells during the ex vivo culturing period (Table 3). Nonlabeled proteins corresponded mostly to matrix components, although several metabolic proteins and remnants from plasma/serum were also identified. In the explants’ supernatants, only 29% of proteins were carrying heavy amino acids. This group of proteins thus represents newly produced proteins from the chondrocytes integrated in the tissues (Table 4). To ascertain if the degree of labeling of cell-associated proteins was similar to the numbers obtained with released proteins, we performed SILAC and MS analysis of three randomly selected gel bands from cell lysates following cartilage tissue digestion (data not shown). The results revealed that nearly all cell-associated proteins were labeled with 13C6-Lys and 13C6-Arg. The patterns of identified proteins were consistent among different individuals. Validation of MS Analysis by Immunoblotting For validation of the MS data, specific immunoblotting of four relevant proteins was performed (Fig. 2). Central matrix regulators and growth factors were chosen for analyses. Some of these factors have not been identified in chondrocytes or cartilage secretome by proteomics hitherto. Metalloproteinase inhibitor 1 (TIMP1) and SPARC were expressed in both monolayers and explants. On the opposite, Western blot of Osteomodulin (OMD) and Growth arrest-specific protein 6 (GAS 6) showed bands only in monolayers’ supernatants. The intensity of the bands corresponded to the Mascot score of proteins identified by MS, where high intensity of the band in Western blotting matched with high Mascot scores in MS analyses. The size of the bands obtained after antibody detection also matched with the

expected size of the full-length proteins, except for the case of OMD which gave only a weak signal at a molecular weight a bit lower then expected.

DISCUSSION By the use of a last generation proteomic approach, in our study we have identified the profiles of secreted proteins produced by human chondrocytes cultured ex vivo in monolayers or as explants, with the aim of improving our understanding on cartilage cells dedifferentiation. Our data with SILAC showed different scenarios of labeled and unlabeled protein profiles encountered in chondrocytes’ monolayers and tissue explants. While in monolayers, over 90% of total proteins were labeled with heavy isotopes, in cartilage explants, only 29% of identified proteins were labeled and released to the medium. Secretion studies previously done with radiolabel S35-Met in mouse cartilage explants and 2D gel protein separation were showing a similar proportion of labeled and unlabeled proteins, however, most of the identified proteins did not match with the profile of proteins found by us using human cartilage samples and 1D electrophoresis.15 This discrepancy could be explained by the fact that only one study of De Ceuninck et al.6 was working with secreted proteins of human articular cartilage. The rest were based on animal cells as experimental models. In addition, the advanced MS analysis and SILAC technology makes our study more comprehensive than the previous ones. Comparatively, the number of newly produced protein types found in explants was significantly inferior to the protein types released in monolayers. The rationale behind this observation might be of three kinds: i) the lack of tissue structure during monolayer growth would mean that newly produced extracellular matrix components are directly released into the culture medium; ii) on the contrary, in explants newly produced, matrix constituents may be correctly processed after transJOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

Table 2. Proteins containing 13C6 labeled peptides, identified in the secretion medium of human articular chondrocytes in monolayers

Protein name Proteases 72 kDa type IV collagenase precursor Cathepsin B Cathepsin D Cathepsin Z Matrilysin Procollagen C-endopeptidase enhancer 1 Serine protease HTRA1 Stromelysin-1 Tetranectin Protease inhibitors Cystatin-C Metalloproteinase inhibitor 1 Metalloproteinase inhibitor 2 Pigment epithelium-derived factor Plasma protease C1 inhibitor ECM components and regulating agents 14-3-3 protein theta Angiopoietin-related protein 7 Apolipoprotein D Basement membrane-specific heparan sulfate proteoglycan core protein Biglycan Calsyntenin-1 Cartilage intermediate layer protein 1 Cartilage oligomeric matrix protein Chondroitin sulfate proteoglycan 4 Clusterin Coiled-coil domain-containing protein 80 Collagen alpha-1(I) chain Collagen alpha-1(III) chain Collagen alpha-1(V) chain Collagen alpha-1(VI) chain Collagen alpha-1(XI) chain Collagen alpha-1(XII) chain Collagen alpha-2(I) chain Collagen alpha-2(V) chain Collagen alpha-2(VI) chain Collagen alpha-3(VI) chain Decorin Dermatopontin EGF-containing fibulin-like extracellular matrix protein 1 EGF-containing fibulin-like extracellular matrix protein 2 Fibrillin-1 Fibronectin Hemopexin Immunoglobulin superfamily containing leucine-rich repeat protein Integrin beta-like protein 1 Inter-alpha-trypsin inhibitor heavy chain H4 Lactadherin Laminin subunit alpha-2 Laminin subunit beta-1 Laminin subunit gamma-1

Accession no.

Unique peptides Sequence detected coverage (%)

Mascot Score


13

MMP2_HUMAN CATB_HUMAN CATD_HUMAN CATZ_HUMAN MMP7_HUMAN PCOC1_HUMAN HTRA1_HUMAN MMP3_HUMAN TETN_HUMAN

25 4 4 5 4 19 6 5 11

48 12 10 15 16 67 13 9 60

837 58 65 136 67 599 211 103 546

0 0 0 0 0 0 0 0 1

CYTC_HUMAN TIMP1_HUMAN TIMP2_HUMAN PEDF_HUMAN IC1_HUMAN

5 7 5 9 7

44 53 32 32 19

607 420 200 256 290

0 0 0 0 0

1433T_HUMAN ANGL7_HUMAN APOD_HUMAN PGBM_HUMAN

5 7 4 7

24 23 20 2

130 255 224 324

1 0 0 0

PGS1_HUMAN CSTN1_HUMAN CILP1_HUMAN COMP_HUMAN CSPG4_HUMAN CLUS_HUMAN CCD80_HUMAN CO1A1_HUMAN CO3A1_HUMAN CO5A1_HUMAN CO6A1_HUMAN COBA1_HUMAN COCA1_HUMAN CO1A2_HUMAN CO5A2_HUMAN CO6A2_HUMAN CO6A3_HUMAN PGS2_HUMAN DERM_HUMAN FBLN3_HUMAN

7 6 4 33 5 9 2 34 12 4 29 7 25 45 7 10 68 12 7 17

30 7 3 62 3 25 2 40 12 5 42 6 1023 53 8 16 30 45 60 43

121 113 56 2010 141 331 122 1147 451 95 1399 110 26 2366 148 192 1599 436 491 497

0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 1 0

FBLN4_HUMAN

6

18

127

0

FBN1_HUMAN FINC_HUMAN HEMO_HUMAN ISLR_HUMAN

14 61 6 8

6 42 20 31

236 2312 110 354

0 0 6 0

ITGBL_HUMAN ITIH4_HUMAN

3 7

8 10

74 200

0 7

MFGM_HUMAN LAMA2_HUMAN LAMB1_HUMAN LAMC1_HUMAN

9 9 13 4

30 3 9 3

105 129 210 143

0 0 0 0 (Continued)


Table 2. (Continued)

Protein name Lumican Lysyl oxidase homolog 1 Nidogen-1 Olfactomedin-like protein 3 Plasma retinol-binding protein Proteoglycan-4 SPARC Stanniocalcin-2 Sulfhydryl oxidase 1 Tenascin Thrombospondin-1 Vascular cell adhesion protein 1 Growth factors Chitinase-3-like protein 1 Chitinase-3-like protein 2 Connective tissue growth factor Growth arrest-specific protein 6 Insulin-like growth factor-binding protein 2 Insulin-like growth factor-binding protein 4 Insulin-like growth factor-binding protein 6 Insulin-like growth factor-binding protein 7 Osteomodulin Plasminogen activator inhibitor 1 Sushi, von Willebrand factor type A, EGF and pentraxin domain-containing protein 1 Transforming growth factorbeta- induced protein ig-h3 Antiinflammatory/antioxidant proteins Beta-2-microglobulin CD109 antigen Complement C1r subcomponent Complement C1s subcomponent Complement C3 Complement factor D Complement factor H Epididymal secretory protein E1 Follistatin-related protein 1 Glutathione peroxidase 3 Pentraxin-related protein PTX3 Peptidyl-prolyl cis-trans isomerase B Intracellular components Annexin A2 Gelsolin Profilin-1 Target of Nesh-SH3 Metabolic enzymes Alpha-1-antichymotrypsin L-lactate dehydrogenase A chain Mannosyl-oligosaccharide 1,2alpha-mannosidase IA Ribonuclease 4

Accession no.

Unique peptides Sequence detected coverage (%)

Mascot Score


13

LUM_HUMAN LOXL1_HUMAN NID1_HUMAN OLFL3_HUMAN RETBP_HUMAN PRG4_HUMAN SPRC_HUMAN STC2_HUMAN QSOX1_HUMAN TENA_HUMAN TSP1_HUMAN VCAM1_HUMAN

10 3 12 10 5 5 11 5 7 38 2 2

37 6 16 28 32 4 34 22 11 24 2 2

264 85 219 245 164 141 548 261 70 1287 62 54

0 0 0 0 0 0 0 0 0 0 0 0

CH3L1_HUMAN CH3L2_HUMAN CTGF_HUMAN GAS6_HUMAN IBP2_HUMAN

14 14 6 10 7

46 66 28 18 26

568 380 144 236 193

0 0 0 0 0

IBP4_HUMAN

4

18

62

0

IBP6_HUMAN

4

19

122

1

IBP7_HUMAN

3

10

61

0

OMD_HUMAN PAI1_HUMAN SVEP1_HUMAN

6 18 13

20 62 4

132 690 295

0 1 0

BGH3_HUMAN

8

15

226

0

B2MG_HUMAN CD109_HUMAN C1R_HUMAN C1S_HUMAN CO3_HUMAN CFAD_HUMAN CFAH_HUMAN NPC2_HUMAN FSTL1_HUMAN GPX3_HUMAN PTX3_HUMAN PPIB_HUMAN

2 2 14 16 25 5 13 3 13 6 13 2

26 2 30 32 19 33 15 17 43 28 43 9

96 50 406 387 443 158 180 63 344 141 669 35

0 0 0 0 0 0 0 0 0 0 1 2

ANXA2_HUMAN GELS_HUMAN PROF1_HUMAN TARSH_HUMAN

3 5 4 4

9 7 35 5

37 109 39 62

0 0 4 0

AACT_HUMAN LDHA_HUMAN MA1A1_HUMAN

15 3 16

45 14 33

544 36 664

0 1 0

RNAS4_HUMAN

5

32

236

1

Unique peptides detected, number of peptides on which the protein identification is based (the peptides are not used to identify other proteins in the same Mascot search); Sequence coverage (%), portion of the protein sequence covered by the identified peptides; Mascot score, significance of the protein identification (the higher score, the higher probability). This protein score is the sum of the unique ions scores. Unique non 13C peptides, number of unique peptides identified, not containing 13C6 label. JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

1046

POLACEK ET AL.

Table 3. Proteins not containing chondrocytes in explants

13

Protein name Serum/plasma associated proteins Ig kappa chain C region Leukocyte cell-derived chemotaxin 2 Serotransferrin Serum amyloid P-component Matrix components/regulating agents Aggrecan core protein Angiogenin Biglycan Cartilage intermediate layer protein 1 Cartilage intermediate layer protein 2 Cartilage oligomeric matrix protein Chondroadherin Collagen alpha-2(XI) chain Complement C3 C-type lectin domain family 3 member A Decorin Dermatopontin Fibrinogen alpha chain Fibromodulin Hyaluronan and proteoglycan link protein 1 Lumican Mimecan Prolargin SPARC-related modular calcium- binding protein 2 Tenascin Thrombospondin-4 Transforming growth factor-betainduced protein ig-h3 Metabolic proteins Annexin A1 Annexin A2 Annexin A5 L-lactate dehydrogenase A chain Lysozyme C Phospholipase A2, membrane associated Ribonuclease 4 Triosephosphate isomerase

C6 labeled peptides, identified in the secretion medium of human articular

Accession no.



Mascot Score

Unique 13C peptides

KAC_HUMAN LECT2_HUMAN

2 4

28 32

66 79

0 0

TRFE_HUMAN SAMP_HUMAN

31 6

61 47

1803 121

0 0

PGCA_HUMAN ANGI_HUMAN PGS1_HUMAN CILP1_HUMAN

21 7 8 7

10 57 29 10

1702 211 734 422

0 0 0 0

CILP2_HUMAN

9

13

310

0

COMP_HUMAN

25

46

2463

0

CHAD_HUMAN COBA2_HUMAN CO3_HUMAN CLC3A_HUMAN

13 5 2 7

52 6 1 43

1147 135 133 352

0 0 0 0

PGS2_HUMAN DERM_HUMAN FIBA_HUMAN FMOD_HUMAN HPLN1_HUMAN

11 3 2 3 17

40 13 2 15 62

546 70 82 308 743

0 0 0 0 0

LUM_HUMAN MIME_HUMAN PRELP_HUMAN SMOC2_HUMAN

11 12 3 2

27 46 8 6

721 439 114 48

0 0 0 0

TENA_HUMAN TSP4_HUMAN BGH3_HUMAN

20 2 13

12 5 29

651 103 663

0 0 0

ANXA1_HUMAN ANXA2_HUMAN ANXA5_HUMAN LDHA_HUMAN LYSC_HUMAN PA2GA_HUMAN

3 3 5 4 8 6

10 15 17 18 56 36

126 112 242 97 362 297

0 0 0 0 0 0

RNAS4_HUMAN TPIS_HUMAN

5 3

37 22

218 69

0 0


lation, and then directly incorporated or retained in the preexisting tissue matrix, therefore not released into the culture medium; iii) alternatively, chondrocytes in explants may keep a low metabolic activity and, JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

consequently, synthesis of new proteins may occur at considerably lower speed than in chondrocyte monolayers. To shed some light on these questions, we performed SILAC analysis of proteins in cell lysates of


1047

Table 4. Proteins containing 13C6 labeled peptides, identified in the secretion medium of human articular chondrocytes in explants

Protein name Alpha-1-antitrypsin Alpha-1-antichymotrypsin Apolipoprotein D Chitinase-3-like protein 1 Chitinase-3-like protein 2 Clusterin Collagen alpha-1(III) chain Fibronectin Metalloproteinase inhibitor 1 Pentraxin-related protein PTX3 Peptidyl-prolyl cis-trans isomerase B Proteoglycan-4 SPARC Stromelysin-1

Accession no.



Mascot Score

Unique 13C peptides

A1AT_HUMAN AACT_HUMAN APOD_HUMAN CH3L1_HUMAN CH3L2_HUMAN CLUS_HUMAN CO3A1_HUMAN FINC_HUMAN TIMP1_HUMAN PTX3_HUMAN PPIB_HUMAN PRG4_HUMAN SPRC_HUMAN MMP3_HUMAN

2 6 6 7 7 4 3 33 6 2 2 3 4 15

4 18 30 21 24 10 3 18 38 7 11 2 14 29

59 310 139 209 246 143 64 2519 207 98 36 111 80 670

2 6 6 7 7 3 2 19 6 2 2 3 4 15


Figure 2. Validation of protein expression and identification by Western blot. Specific immunodetection of four relevant proteins identified by proteomics was performed. Supernatants from one of the chondrocytes in monolayers and one cartilage explant were analyzed. Immunoblotting was performed using primary antibodies against TIMP1, SPARC, OMD, and GAS6 at a 1:1,000 dilution followed by a secondary antibody conjugated to horseradish peroxidase (1:5,000 dilution). Bound antibody was detected by chemiluminescence.

chondrocytes isolated from the cartilage explants. The shift of the overall protein cargo in chondrocytes from cartilage explants was nearly completed after the incubation period, indicating that chondrocytes in tissues are metabolically active, and that probably most of the released proteins from chondrocytes in explants are indeed retained or incorporated in the existing matrix. In explants, most of the unlabeled proteins found in supernatants corresponded to extracellular matrix components. These proteins may possibly represent catalytic products resulting from the high turn-over of connective tissue during the ex vivo culture phase. The chances of getting proteins from the mechanical excision of the tissue are negligible, because the supernatants from the initial days of incubation were systematically collected and discarded. On the other hand, most of the labeled proteins found in supernatants of explants corresponded to matrix regulating agents, including factors such as a-1-antichymotrypsin, Chitinase-3-like proteins, Clusterin, TIMP1, MMP3, Proteoglycan 4. All together, these results reflect the high rate of matrix remodeling events occurring during the ex vivo culture. Qualitatively, all newly produced protein types in cartilage explants were also found in supernatants of chondrocytes monolayers, corresponding mostly to matrix components, and matrix regulating agents such as MMP3, TIMP1, Clusterin, Proteoglycan-4, Chitinase3-like proteins, and a-1-antitrypsin. These data indicate that after several weeks in monolayer culture, adult chondrocytes are still partially equipped with the tools to regulate matrix synthesis and degradation. Proteomic studies of chondrocytes supernatants from monolayer cultures are scarce. Previous attempts based on 2D gel protein separation showed that proteoglycans and JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

1048

POLACEK ET AL.

GAGs secreted by chondrocytes significantly interfered with the 2D gel focusing.7,15,16 Our approach based on 1D gel protein separation combined with SILAC and LC/ MS-MS has enabled us to identify over 100 different types of proteins produced by chondrocytes during the culture period. To our knowledge, no previous report has shown such a long spectrum of proteins produced by chondrocytes in culture. Our data reveal that most of the proteins produced by chondrocytes in monolayers are known cartilage matrix components. Some matrix regulatory factors, growth hormones, and antiinflammatory agents such as Dermatopontin, Hemopexin, Integrin beta-like protein, Proteoglycan-4, Vascular cell adhesion protein, Growth arrest-specific protein, Osteomodulin, Sushi, Epididymal secretory protein, and Follistatin-related protein have not been identified before in chondrocyte supernatants by proteomics. Among the regulators of matrix degradation/remodeling, we identified nine different types of proteases and only five types of protease inhibitors, showing the potential dominance of catabolic substances over anabolic ones in secretion media of chondrocytes in monolayers. The antiinflammatory and antioxidant proteins that we have detected in monolayer supernatants may be part of the local defense mechanisms developed by chondrocytes. Among these, the Complement subcomponents and Pentraxin-related protein are associated with acute phase response. Subcomponent C1r, factor C3, and factor H have an important role in innate immunity where they are a part of an antigen-independent alternative complement activation pathway.7,17,18 Factor H was detected in both normal and osteoarthritic cartilage already by Kumar et al.,19 based on a detection of expressed sequence tags, and our study confirms their findings by proteomic analysis of the secretome. Presently, stable isotope labeling technology may represent one of the most promising strategies for approaching comparative proteomic analyses in in vitro established systems. The method has been successfully applied in comparative and quantitative proteomics of different cell types, especially while pursuing identification of biomarkers of several cancer types.20–22 Other studies have applied SILAC to study cell function or tissue characterization.23–25 Our study is the first one applying SILAC in proteomic studies of cartilage or chondrocytes. The validation experiments by Western blot included in our study were in full accordance with the results achieved by SILAC proteomics, thus further confirming the robustness and reproducibility of the method. Overall, our results show the differences in the profile of proteins that are newly produced and secreted by chondrocytes during monolayer culture or as explants. Stable 13C isotope labeling of amino acids Arginine and Lysine during cell culture has shown to be an appropriate method to make a distinction between selfgenerated and exogenous proteins. Although chondrocytes maintained in 2D cultures are known to be phenotypically instable, our data shows that all newly JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2010

produced proteins identified in cartilage explants were also produced by chondrocytes in monolayers, even though some typical dedifferentiation markers were also found in 2D culture supernatants. Most proteins found in the media of cartilage explants corresponded to degradation products of the existing extracellular matrix. Additionally, in monolayer cultures, we were able to identify a bunch of proteins and factors not previously described by proteomics. Although promising, our results are not able to show the differences in quantitative terms. Further studies using quantitative SILAC, or isotope-coded affinity tags (ICAT), in tandem with real-time PCR or Western blots for validation, may help to more accurately reveal the global changes in chodrocyte metabolism and phenotype associated with de- and redifferentiation during ex vivo culture of cartilage cells.

ACKNOWLEDGMENTS The authors are thankful for the excellent technical assistance of Toril Anne Grønset in running the samples for the MS analyses, and Kirsti Rønne in running the Western blots. This project was partially supported by funds from the University of Tromsø (Norway). We are thankful for the financial support from MyJoint project (EU project) and Orthogenics AS (Tromsø, Norway). The authors have no conflicts of interest.

REFERENCES 1. Holtzer H, Abbott J, Lash J, et al. 1960. The loss of phenotypic traits by differentiated cells in vitro, I. Dedifferentiation of cartilage cells. Proc Natl Acad Sci USA 46:1533–1542. 2. Watt FM. 1988. Effect of seeding density on stability of the differentiated phenotype of pig articular chondrocytes in culture. J Cell Sci 89(Pt 3):373–378. 3. Wilson R, Whitelock JM, Bateman JF. 2009. Proteomics makes progress in cartilage and arthritis research. Matrix Biol 28:121–128. 4. Aebersold R. 2003. A mass spectrometric journey into protein and proteome research. J Am Soc Mass Spectrom 14:685–695. 5. Aebersold R, Mann M. 2003. Mass spectrometry-based proteomics. Nature 422:198–207. 6. De Ceuninck F, Marcheteau E, Berger S, et al. 2005. Assessment of some tools for the characterization of the human osteoarthritic cartilage proteome. J Biomol Tech 16: 256–265. 7. Haglund L, Bernier SM, Onnerfjord P, et al. 2008. Proteomic analysis of the LPS-induced stress response in rat chondrocytes reveals induction of innate immune response components in articular cartilage. Matrix Biol 27:107–118. 8. Mann M. 2006. Functional and quantitative proteomics using SILAC. Nat Rev Mol Cell Biol 7:952–958. 9. Ong SE, Foster LJ, Mann M. 2003. Mass spectrometric-based approaches in quantitative proteomics. Methods 29:124– 130. 10. Zhu H, Pan S, Gu S, et al. 2002. Amino acid residue specific stable isotope labeling for quantitative proteomics. Rapid Commun Mass Spectrom 16:2115–2123. 11. Hathout Y. 2007. Approaches to the study of the cell secretome. Expert Rev Proteomics 4:239–248.

CHONDROCYTES SECRETOME BY SILAC 12. Ong SE, Blagoev B, Kratchmarova I, et al. 2002. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics 1:376–386. 13. Shevchenko A, Wilm M, Vorm O, et al. 1996. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal Chem 68:850–858. 14. Yakunin AF, Hallenbeck PC. 1998. A luminol/iodophenol chemiluminescent detection system for Western immunoblots. Anal Biochem 258:146–149. 15. Hermansson M, Sawaji Y, Bolton M, et al. 2004. Proteomic analysis of articular cartilage shows increased type II collagen synthesis in osteoarthritis and expression of inhibin betaA (activin A), a regulatory molecule for chondrocytes. J Biol Chem 279:43514–43521. 16. Catterall JB, Rowan AD, Sarsfield S, et al. 2006. Development of a novel 2D proteomics approach for the identification of proteins secreted by primary chondrocytes after stimulation by IL-1 and oncostatin M. Rheumatology (Oxford) 45:1101– 1109. 17. Bradley K, North J, Saunders D, et al. 1996. Synthesis of classical pathway complement components by chondrocytes. Immunology 88:648–656. 18. Solomon S, Kassahn D, Illges H. 2005. The role of the complement and the Fc gamma R system in the pathogenesis of arthritis. Arthritis Res Ther 7:129–135.

1049

19. Kumar S, Connor JR, Dodds RA, et al. 2001. Identification and initial characterization of 5000 expressed sequenced tags (ESTs) each from adult human normal and osteoarthritic cartilage cDNA libraries. Osteoarthritis Cartilage 9:641–653. 20. Everley PA, Krijgsveld J, Zetter BR, et al. 2004. Quantitative cancer proteomics: stable isotope labeling with amino acids in cell culture (SILAC) as a tool for prostate cancer research. Mol Cell Proteomics 3:729–735. 21. Pan C, Kumar C, Bohl S, et al. 2008. Comparative proteomic phenotyping of cell lines and primary cells to assess preservation of cell type specific functions. Mol Cell Proteomics 8:443–450. 22. Gronborg M, Kristiansen TZ, Iwahori A, et al. 2006. Biomarker discovery from pancreatic cancer secretome using a differential proteomic approach. Mol Cell Proteomics 5:157– 171. 23. Molina H, Yang Y, Ruch T, et al. 2009. Temporal profiling of the adipocyte proteome during differentiation using a fiveplex SILAC based strategy. J Proteome Res 8:48–58. 24. Graumann J, Hubner NC, Kim JB, et al. 2008. Stable isotope labeling by amino acids in cell culture (SILAC) and proteome quantitation of mouse embryonic stem cells to a depth of 5,111 proteins. Mol Cell Proteomics 7:672–683. 25. Alvarez-Llamas G, Szalowska E, de Vries MP, et al. 2007. Characterization of the human visceral adipose tissue secretome. Mol Cell Proteomics 6:589–600.
