Advertisement

The differentiation of prehypertrophic into hypertrophic chondrocytes drives an OA-remodeling program and IL-34 expression

Open ArchivePublished:December 06, 2020DOI:https://doi.org/10.1016/j.joca.2020.10.013

      Summary

      Objectives

      We hypothesize that chondrocytes from the deepest articular cartilage layer are pivotal in maintaining cartilage integrity and that the modification of their prehypertrophic phenotype to a hypertrophic phenotype will drive cartilage degradation in osteoarthritis.

      Design

      Murine immature articular chondrocytes (iMACs) were successively cultured into three different culture media to induce a progressive hypertrophic differentiation. Chondrocyte were phenotypically characterized by whole-genome microarray analysis. The expression of IL-34 and its receptors PTPRZ1 and CSF1R in chondrocytes and in human osteoarthritis tissues was assessed by RT-qPCR, ELISA and immunohistochemistry. The expression of bone remodeling and angiogenesis factors and the cell response to IL-1β and IL-34 were investigated by RT-qPCR and ELISA.

      Results

      Whole-genome microarray analysis showed that iMACs, prehypertrophic and hypertrophic chondrocytes each displayed a specific phenotype. IL-1β induced a stronger catabolic effect in prehypertrophic chondrocytes than in iMACs. Hypertrophic differentiation of prehypertrophic chondrocytes increased Bmp-2 (95%CI [0.78; 1.98]), Bmp-4 (95%CI [0.89; 1.59]), Cxcl12 (95%CI [2.19; 5.41]), CCL2 (95%CI [3.59; 11.86]), Mmp 3 (95%CI [10.29; 32.14]) and Vegf mRNA expression (95%CI [0.20; 1.74]). Microarray analysis identified IL-34, PTPRZ1 and CSFR1 as being strongly overexpressed in hypertrophic chondrocytes. IL-34 was released by human osteoarthritis cartilage; its receptors were expressed in human osteoarthritis tissues. IL-34 stimulated CCL2 and MMP13 in osteoblasts and hypertrophic chondrocytes but not in iMACs or prehypertrophic chondrocytes.

      Conclusion

      Our results identify prehypertrophic chondrocytes as being potentially pivotal in the control of cartilage and subchondral bone integrity. Their differentiation into hypertrophic chondrocytes initiates a remodeling program in which IL-34 may be involved.

      Keywords

      Introduction

      Osteoarthritis (OA) is characterized by the irreversible degradation of cartilage, which is associated with a pathological remodeling of the subchondral bone, including sclerosis and osteophyte formation. Cartilage degradation mainly results from the proteolysis of the cartilage extracellular matrix by chondrocyte-secreted proteases
      • Troeberg L.
      • Nagase H.
      Proteases involved in cartilage matrix degradation in osteoarthritis.
      . The degradation observed in the deep zone of articular cartilage is explained by an endochondral ossification-like process at the cartilage/bone interface
      • Goldring S.R.
      • Goldring M.B.
      Changes in the osteochondral unit during osteoarthritis: structure, function and cartilage-bone crosstalk.
      • Lories R.J.
      • Luyten F.P.
      The bone-cartilage unit in osteoarthritis.
      • Mahjoub M.
      • Berenbaum F.
      • Houard X.
      Why subchondral bone in osteoarthritis? The importance of the cartilage bone interface in osteoarthritis.
      , which involves the hypertrophic differentiation of chondrocytes, the calcification and the vascularization of the extracellular matrix followed by the replacement of cartilage with bone.
      While chondrocytes are the unique cell type present within cartilage, different chondrocyte phenotypes exist, depending on the type of cartilage and on the chondrocyte localization within cartilage. Articular cartilage is organized in different layers from the surface until the subchondral bone. The phenotype of chondrocytes differs upon the layer considered
      • Zuscik M.J.
      • Hilton M.J.
      • Zhang X.
      • Chen D.
      • O'Keefe R.J.
      Regulation of chondrogenesis and chondrocyte differentiation by stress.
      . Chondrocytes from the deepest articular cartilage layer of non-calcified cartilage display an intermediate phenotype between that of the chondrocytes of the surface layers and that of the chondrocytes found in the calcified cartilage, which are hypertrophic. They indeed express molecules that characterize both surface layer chondrocytes and hypertrophic chondrocytes, including type II and type X collagens
      • Aigner T.
      • Bertling W.
      • Stoss H.
      • Weseloh G.
      • von der Mark K.
      Independent expression of fibril-forming collagens I, II, and III in chondrocytes of human osteoarthritic cartilage.
      • Pullig O.
      • Weseloh G.
      • Ronneberger D.
      • Kakonen S.
      • Swoboda B.
      Chondrocyte differentiation in human osteoarthritis: expression of osteocalcin in normal and osteoarthritic cartilage and bone.
      • Wachsmuth L.
      • Soder S.
      • Fan Z.
      • Finger F.
      • Aigner T.
      Immunolocalization of matrix proteins in different human cartilage subtypes.
      . They also express Ihh and osteomodulin as prehypertrophic chondrocytes from the growth plate cartilage
      • Belluoccio D.
      • Etich J.
      • Rosenbaum S.
      • Frie C.
      • Grskovic I.
      • Stermann J.
      • et al.
      Sorting of growth plate chondrocytes allows the isolation and characterization of cells of a defined differentiation status.
      • Muller C.
      • Khabut A.
      • Dudhia J.
      • Reinholt F.P.
      • Aspberg A.
      • Heinegard D.
      • et al.
      Quantitative proteomics at different depths in human articular cartilage reveals unique patterns of protein distribution.
      • Semevolos S.A.
      • Nixon A.J.
      • Fortier L.A.
      • Strassheim M.L.
      • Haupt J.
      Age-related expression of molecular regulators of hypertrophy and maturation in articular cartilage.
      • Vortkamp A.
      • Lee K.
      • Lanske B.
      • Segre G.V.
      • Kronenberg H.M.
      • Tabin C.J.
      Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein.
      .
      Cartilage degradation in OA results in chondrocyte phenotypic modifications. We hypothesize that the chondrocytes from the deepest articular cartilage layer play a crucial role in maintaining cartilage integrity and that the modification of their prehypertrophic phenotype to a hypertrophic phenotype will drive cartilage degradation in OA. In the present study, we developed a model of progressive differentiation of murine immature articular chondrocytes (iMACs) into hypertrophic chondrocytes, and this model includes an intermediate prehypertrophic state. Here, we show that the differentiation of prehypertrophic chondrocytes into hypertrophic chondrocytes shifts chondrocytes towards an OA-inducing phenotype. This phenotype is associated with an increased expression of IL-34, a recently discovered cytokine that could be involved in both cartilage and bone integrity.

      Materials and methods

      See supplementary information for detailed Material and methods.

      Collection of osteoarthritis human samples

      Human OA knee explants (n = 33) obtained from patients undergoing total knee joint replacement surgery were dissected, as described
      • Priam S.
      • Bougault C.
      • Houard X.
      • Gosset M.
      • Salvat C.
      • Berenbaum F.
      • et al.
      Identification of soluble 14-3-3 as a novel subchondral bone mediator involved in cartilage degradation in osteoarthritis.
      .

      Immunohistochemistry

      Immunohistochemistry was performed with a mouse monoclonal antibody against PTPRZ1 (clone 12/RPTPb, BD Transduction Laboratories; dilution 1:50) and a rabbit polyclonal antibody against CSF-1R (H-300, Santa Cruz Biotechnology; dilution 1:50) as the primary antibodies. The R.T.U. Vectastain kit (Vector) was used for detection, followed by counterstaining with Mayer's hematoxylin. Irrelevant control antibodies (Dako) were incubated at the same concentration to assess nonspecific staining.

      Primary culture of murine osteoblasts and articular chondrocytes

      Osteoblasts and iMACs were isolated and cultured, as described in
      • Sanchez C.
      • Gabay O.
      • Salvat C.
      • Henrotin Y.E.
      • Berenbaum F.
      Mechanical loading highly increases IL-6 production and decreases OPG expression by osteoblasts.
      ,
      • Gosset M.
      • Berenbaum F.
      • Thirion S.
      • Jacques C.
      Primary culture and phenotyping of murine chondrocytes.
      ,
      • Salvat C.
      • Pigenet A.
      • Humbert L.
      • Berenbaum F.
      • Thirion S.
      Immature murine articular chondrocytes in primary culture: a new tool for investigating cartilage.
      and supplementary information. Prehypertrophic chondrocytes were obtained by culturing iMACs for 28 days in culture medium 2 (DMEM/HAM-F12 medium supplemented with fetal calf serum (5%), penicillin (100 U/mL), streptomycin (100 μg/mL), L-glutamine (4 mM), ascorbic acid (40 μg/mL), insulin-transferrin-sodium selenite (1%) and triiodo-L-thyronine (50 ng/mL)). Chondrocytes were further cultured for 42 days in medium 2 supplemented with β-glycerophosphate (10 mM), retinoic acid (100 nM) and 1α,25-dihydroxyvitamin D3 (10 nM) (Medium 3) to obtain hypertrophic chondrocytes. All cultures were performed in standard conditions with the exception of the last differentiation step, performed in 3% CO2/95% air. At the end of the culture, cells were serum-starved for 24 h and stimulated by recombinant human IL-1β (1 ng/mL) or murine IL-34 (3, 30 and 100 ng/mL) for 24 h. Conditioned media were kept, centrifuged and stored at −20°C. Cells were either fixed in 3.7% paraformaldehyde (PFA) or used for mRNA or protein extraction.
      The experimental design for the cell culture study is shown in Fig. S1.

      Primary culture of human articular chondrocytes

      Human chondrocytes were isolated from the less damaged areas of OA cartilage from patients who underwent total knee arthroplasty. Their hypertrophic differentiation was performed according to Yahara et al.
      • Yahara Y.
      • Takemori H.
      • Okada M.
      • Kosai A.
      • Yamashita A.
      • Kobayashi T.
      • et al.
      Pterosin B prevents chondrocyte hypertrophy and osteoarthritis in mice by inhibiting Sik3.
      .

      Microarray analysis

      mRNA expression profiling was performed using SurePrint G3 Mouse Gene Expression v2 8 × 60 K Microarray (G4852B, Agilent Technologies) and SurePrint Mouse miRNA Microarray Kit v21 8 × 60 K (G4859C, Agilent Technologies). For mRNA profiling, probe labeling and 60 mer-oligonucleotide microarray hybridization were performed according to the manufacturer's instructions
      • Hughes T.R.
      • Mao M.
      • Jones A.R.
      • Burchard J.
      • Marton M.J.
      • Shannon K.W.
      • et al.
      Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer.
      . An Agilent scanner and Feature Extraction 11.5.1.1 software (Agilent Technologies) were used to obtain the raw microarray data for both analyses.

      Statistical analysis

      We used repeated measures one-way ANOVA to compare iMACs, prehypertrophic and hypertrophic chondrocytes. Paired t-tests were used to compare prehypertrophic to hypertrophic chondrocyte gene expression observations and gene or protein comparison in OA patients. For the IL-1β and IL-34 stimulation study, we used Dunnett's post hoc test. The analyses were performed using GraphPad Prism 7 (GraphPad Software Inc., San Diego, CA, USA).

      Results

      Model of the progressive differentiation of murine immature articular chondrocytes into prehypertrophic and hypertrophic chondrocytes

      To study the consequences of the hypertrophic differentiation of chondrocytes in OA cartilage, we established an in vitro model of the progressive differentiation of iMACs into prehypertrophic and hypertrophic chondrocytes. iMACs displayed the typical phenotype of articular chondrocytes, characterized by the expression of Col2a1, Acan, Chm 1 and Sox 9 and by the almost complete absence of the expression of the hypertrophic chondrocyte markers Runx 2, Osterix, MMP-13, Col10a1 and Tnap (n = 6) (Fig. 1(B)–(E) and I-M). iMACs also showed weak alkaline phosphatase activity and no evidence of matrix calcification (Fig. 1O and P). Culturing the iMACs in Medium 2 then in Medium 3 induced progressive decreases in the mRNA expression of Acan, Col2a1, Sox9 and Chm1 [Fig. 1(B)–(E)]. Inversely, the mRNA expression of the hypertrophic markers Runx2, Osterix, Mmp 13, osteocalcin and Tnap was stimulated by culture in Medium 2 (Fig. 1I-N). The mRNA expression of Runx2, Col10a1 and osteocalcin further increased after the culture in Medium 3. Consistent with the increased mRNA expression of Tnap, the activity of alkaline phosphatase was strongly stimulated by culture in Medium 2 and Medium 3 (Fig. 1O). Significant calcification of the chondrocyte cultures was only observed when cells were cultured in Medium 3 (Fig. 1P). Cells cultured in Medium 2 expressed both iMAC and hypertrophic chondrocyte markers and did not mineralize their matrix. They also expressed the markers of prehypertrophic chondrocytes Ihh, Snorc and Osteomodulin [Fig. 1(F)–(H)]. Moreover, they expressed higher mRNA levels of Snorc and Osteomodulin than iMACs and hypertrophic chondrocytes. Thus, iMACs progressively became prehypertrophic and hypertrophic after culture in Medium 2 and Medium 3.
      Fig. 1
      Fig. 1Hypertrophic chondrocyte differentiation markers assessed in iMACs differentiated into prehypertrophic and hypertrophic chondrocytes. A) Model of progressive hypertrophic differentiation of iMACs. iMACs isolated from the femoral heads and the knees of newborn mice were cultured for 7 days in Medium 1. Confluent iMACs were then cultured for 28 days in Medium 2 to obtain prehypertrophic chondrocytes and for an additional 42 days for prehypertrophic to hypertrophic phenotype changes. B–N) Gene expression pattern of articular chondrocyte markers (BE), prehypertrophic markers (F–H) and hypertrophic markers (I–N) in iMACs, prehypertrophic chondrocytes and hypertrophic chondrocytes from six independent cell cultures. The results are expressed as fold expression compared to those of iMACs, whose mRNA expression was set to one for each culture. O) Measurement of alkaline phosphatase activity associated with iMACs, prehypertrophic chondrocytes and hypertrophic chondrocytes (n = 8). Lower panels, representative photomicrographs of the cytochemical determination of phosphatase alkaline activity in iMACs, prehypertrophic chondrocytes and hypertrophic chondrocytes. P) Quantification of chondrocyte culture mineralization by alizarin red staining (n = 12). Lower panels, representative photomicrographs of alizarin red staining of iMACs, prehypertrophic chondrocytes and hypertrophic chondrocytes. Only hypertrophic chondrocytes showed positive alizarin red staining. Bars indicate the mean expression levels.
      To further characterize our model of progressive hypertrophic differentiation of iMACs, the transcriptomic signatures of iMACs, prehypertrophic and hypertrophic chondrocytes were explored by high-throughput genomic methods (n = 8). A two-dimensional PCA of the genes expressed revealed that all displayed globally distinct gene expression patterns (Fig. S2(B)), which were also confirmed by their distinct clustering patterns (Fig. S2(C)). Interestingly, prehypertrophic and hypertrophic chondrocytes showed a homogenous clustering pattern of gene expression, indicating that they displayed a specific molecular phenotype, different from iMAC phenotype.
      The most variable genes identified by PCA (top 5,000 ranked by decreasing standard deviation) accounted for 40% and 17% of the total gene expression variability in the principal component (PC) 1 and PC2 groups, respectively. Genes positively correlated with PC1 displayed an upregulation across the differentiation process. The number of genes positively correlated with PC2 was more restricted, and those were of particular interest to characterize prehypertrophic cells since they were more specifically associated with this group of cells.
      The repeated spotted probes and the probes targeting the same gene were not averaged but were analyzed for similar expression. We identified 8,121 differentially expressed (DE) genes (9306 DE probes, at false discovery rate (FDR)-adjusted p-value ≤ 0.05 & |Fold Change (FC)|≥1.3) between iMACs and hypertrophic chondrocytes (4,311 and 3,810 genes overexpressed in hypertrophic chondrocytes and iMACs, respectively), including markers of articular and hypertrophic chondrocytes (Table S1). The specific part of this signature represented 1,467 genes (1768 DE probes, 809 and 959 overexpressed in hypertrophic chondrocytes and iMACs, respectively) [Fig. 2(A)]. We found 6,829 DE genes (7,744 probes) between prehypertrophic chondrocytes and iMACs (3,601 and 3,228 overexpressed in prehypertrophic chondrocytes and iMACs, respectively) (Table S1), while 1023 DE probes were specific to this contrast (540 and 483 overexpressed in prehypertrophic chondrocytes and iMACs, respectively) [Fig. 2(B)]. Finally, 5,308 genes (6,220 probes) were DE between the hypertrophic and prehypertrophic chondrocytes (2,970 and 2,338 genes overexpressed in hypertrophic and prehypertrophic chondrocytes, respectively) (Table S1). Eight hundred nine genes (1002 DE probes) were specific to differences between hypertrophic and prehypertrophic chondrocytes (626 and 376 probes overexpressed in hypertrophic and prehypertrophic chondrocytes, respectively) [Fig. 2(C)]. Hypertrophic chondrocytes were thus molecularly the most different from other cells.
      Fig. 2
      Fig. 2Whole-genome transcriptomic characterization across iMAC differentiation processes into hypertrophic chondrocytes. Results from differential analysis of gene expression in the chondrocytes. Signatures of differentially expressed genes (DEG) identified between A) hypertrophic vs articular chondrocytes (n = 8 in each group), B) prehypertrophic vs articular (n = 8 in each group), C) and hypertrophic vs prehypertrophic chondrocytes (n = 8 in each group). A Venn diagram compares the signatures obtained. Each comparison is illustrated by a hierarchical clustering (correlation distance on genes and Euclidean distance on samples). Volcano plots present the log2-fold changes and the significance of each gene for the three comparisons. Gene and probe numbers differ due to repeated spotting and/or missing annotations.
      Together, these results show that our culture model allows a progressive differentiation of the iMACs into prehypertrophic and hypertrophic chondrocytes, each displaying a specific molecular phenotype.

      Prehypertrophic to hypertrophic differentiation shifts chondrocytes towards an OA-inducing phenotype

      Inflammatory factors, such as IL-1β, alter the phenotype of chondrocytes that adopt OA-like catabolic features. As expected, IL-1β downregulated the expression of Col2a1 by iMACs (Padj = 0.0454, 95%CI [-1.41;-0.00], n = 6), whereas it strongly upregulated the mRNA expression of Il-6 (Padj = 0.0016, 95%CI [1.01; 2.60], n = 6) and Mmp 13 (Padj = 0.0483, 95%CI [0.62; 2.14], n = 6) [Fig. 3(A), (C) and (D)]. IL-1β also induced a similar catabolic phenotype in prehypertrophic and hypertrophic chondrocytes [Fig. 3(A)–(D)]. However, prehypertrophic chondrocytes appeared more sensitive to IL-1β than did iMACs since IL-1β led to a 2.0-fold, 4.1-fold and 1.8-fold lower expression of Col2a1, Acan and Chm 1 (n = 6), respectively and to a 131.7-fold higher expression of Il-6, as compared to iMACs [Fig. 3(A)–(C)]. In addition, IL-1β also stimulated the expression of Vegf and repressed those of Tsp 1 and Chm1 by prehypertrophic and hypertrophic chondrocytes (n = 6), while neither Vegf nor Tsp1 expression was regulated by IL-1β in iMACs [Fig. 3(E)–(G)].
      Fig. 3
      Fig. 3Loss of the prehypertrophic phenotype shifts chondrocytes towards an OA-inducing phenotype. A-G) iMACs, prehypertrophic chondrocytes and hypertrophic chondrocytes from six independent cell cultures were stimulated by IL-1β (1 ng/mL) for 24 h, and the mRNA expression of Col2a1 (A), Acan (B), Il-6 (C), Mmp 13 (D), Vegf (E), Tsp 1 (F) and Chm 1 (G) was determined. H–K) Phenotypic transition of prehypertrophic to hypertrophic chondrocytes induces the expression of factors involved in OA. Prehypertrophic and hypertrophic chondrocytes from six to 12 independent cell cultures were assessed for the mRNA expression of matrix proteases (H), osteoblast (I) and osteoclast activity (J) and angiogenic/angiostatic factors (K). Data are expressed as fold expression compared to those in unstimulated control cells, whose mRNA expression was set to one for each culture (AG), or to that in prehypertrophic chondrocytes (H–K). Bars indicate the mean expression levels.
      To evaluate whether the shift from prehypertrophic to hypertrophic chondrocytes mimicked OA-related osteochondral remodeling, we performed analysis focused on molecular functions involved in OA (Fig. S2(D)). Hypertrophic chondrocytes, in contrast to prehypertrophic chondrocytes, displayed activated functions related to bone differentiation, angiogenesis and deterioration/damage of connective tissues. Accordingly, the mRNA expression of the bone remodeling factors BMP-2 (2.38-fold, 95%CI [0.78; 1.98] for the difference in means), BMP-4 (2.25-fold, 95%CI [0.89; 1.59] for the difference in means), CXCL12 (4.80-fold, 95%CI [2.19; 5.41] for the difference in means), CCL2 (8.73-fold, 95%CI [3.59; 11.86] for the difference in means) and OPG (1.75-fold, 95%CI [0.13; 1.37] for the difference in means) was increased when prehypertrophic chondrocytes became hypertrophic [Fig. 3(I) and (J)]. They also tended to overexpressed Rankl mRNA, as compared to prehypertrophic chondrocytes (22.5-fold, 95%CI [-4.32; 47.43] for the difference in means) [Fig. 3(J)]. Hypertrophic chondrocytes expressed more Vegf (1.97-fold, 95%CI [0.203; 1.74] for the difference in means) and Mmp 3 mRNAs (22.21-fold, 95%CI [10.29; 32.14]) for the difference in means) and less the angiostatic factors Chm 1 (−5.99-fold, 95%CI [-0.93;-0.73]) for the difference in means) and Angptl4 than prehypertrophic chondrocytes (−1.65-fold, 95%CI [-0.68;-0.10]) for the difference in means) [Fig. 3(H) and (K)]. In contrast, neither Tgfβ1 nor Mpm13, Adamts4 and Adamts5 mRNAs were modulated by the switch from prehypertrophic to hypertrophic differentiation [Fig. 3(H) and I].

      Overexpression of IL-34 by hypertrophic chondrocytes

      Hypertrophic differentiation of chondrocytes may contribute to OA via the release of factors with autocrine and paracrine tissue remodeling activity. We focused the analysis on ccl, cxcl cytokines/chemokines and ILs, for which expression was upregulated with chondrocyte hypertrophy (Table S2). Seven cytokines/chemokines were overexpressed by hypertrophic chondrocytes, especially in relation to prehypertrophic chondrocytes. Among them, only CXCL12 and IL-34 have their receptors (CXCR4, PTPRZ1 and CSF1R) also upregulated, suggesting that they may act in an autocrine loop (Table S2). Interestingly, Ptrpz1 was the most overexpressed gene during the hypertrophic differentiation of chondrocytes (72.1-fold increased expression compared to iMACs), just after Mmp 3. In contrast, the levels of Cxcr7, the second receptor of CXCL12, were unchanged.
      Since the involvement of IL-34 in OA is unknown, we next focused our investigation on its expression in OA. Consistent with the microarray analysis, RT-qPCR confirmed the overexpression of Il-34, Ptprz1 and Csf1r in hypertrophic chondrocytes (n = 6) [Fig. 4(A)–(C)]. Both Il-34 and Ptprz1 mRNA levels increased progressively during the hypertrophic differentiation of chondrocytes, whereas the increase in the expression of Csf1r mRNA was observed during the conversion of iMACs to prehypertrophic chondrocytes. Concentrations of IL-34 in both cell supernatants and cell lysates also increased with iMAC hypertrophic differentiation [Fig. 4(D) and (E)]. Similar results were also observed during the hypertrophic differentiation of human chondrocytes. Their hypertrophic differentiation was associated with a decrease in the mRNA levels Sox 9 and aggrecan, an increase in those of TNAP and MMP-13, and the presence of areas of calcification within the extracellular matrix (Fig. S3). Human hypertrophic chondrocytes also showed an increase in IL-34 mRNA expression (Padj = 0.011, 95%CI [0.22; 1.17], n = 7) and in IL-34 concentration in cell supernatant (Padj = 0.0013, 95%CI [0.89; 2.08], n = 6) and cell lysates (Padj = 0.0282, 95%CI [0.27; 3.17], n = 6) [Fig. 4(F)–(H)].
      Fig. 4
      Fig. 4Increased expression of IL-34 and IL-34 receptors with chondrocyte hypertrophic differentiation. A-E) mRNA expression of Il-34 (A), Ptprz1 (B) and Csf1r (C) in iMACs, prehypertrophic chondrocytes and hypertrophic chondrocytes (n = 6) was determined. C-D) IL-34 protein levels were quantified by ELISA in cell conditioned medium (n = 7) (D) and cell lysates (n = 6) (E) of iMACs, prehypertrophic chondrocytes and hypertrophic chondrocytes. F–H) The IL-34 mRNA expression (n = 7) (F) and protein levels (n = 6) in chondrocyte conditioned medium (G) and cell lysates (H) were determined in human control and hypertrophic chondrocytes.
      IL-34 was also released by human OA cartilage, regardless of whether it originated from articular cartilage or from the thin cartilage layer covering osteophytes [Fig. 5(A)]. However, cartilage from osteophytes released higher amounts of IL-34 than OA articular cartilage (P = 0.0368, 95%CI [23.43; 506.40]). OA tissues also expressed PTPRZ1 and CSF1R [Fig. 5(B)]. Within articular cartilage, positive immunostaining for PTPRZ1 and CSF1R was mainly detected in the chondrocytes of the deeper area of the cartilage or in clusters of chondrocytes, although not all isolated chondrocyte or chondrocyte clusters were positive (Fig. 5(B), panels a–f). A more intense immunostaining was observed within the bone and was associated with osteoblasts, osteocytes and cells present in vascular channels, including vessels and mesenchymal stromal cells. A similar positive immunostaining pattern was observed within osteophytes (Fig. 5(B), panels g–l). Chondrocytes, osteoblasts, osteocytes and bone marrow cells were positive for both CSF1R and PTPRZ1. In addition, mesenchymal cells of the fibrous tissue, which often covered the osteophyte surface, were also positive for PTPRZ1 immunostaining (Fig. 5(B), panels j and k).
      Fig. 5
      Fig. 5Expression of IL-34 and IL-34 receptors in OA. A) IL-34 secreted by articular (n = 14 different donors) and osteophyte cartilages (n = 6 different donors) from OA patients was measured in tissue conditioned media by ELISA. B) Paraffin sections (5 μm) of OA cartilage bone interface (a–f) and osteophytes (g–l) (n = five to eight different donors) were stained for CSF1R (a, b, g and h), PTPRZ1 (d, e, j and k) or with irrelevant antibodies as negative controls (c, f, i and l). CSF1- and PTPRZ1-positive staining are observed at the osteochondral junction (a, b, d and e). Chondrocytes near the tidemark express both the IL-34 receptors CSF1R and PTPRZ1. Within the bone, CSF1R- and PTPRZ1-positive staining is associated with osteoblasts and vascular channels. Similar staining was observed for both CSF1R and PTPRZ1 within osteophytes (g, h, j and k). Chondrocytes, osteoblasts, osteocytes and bone marrow cells showed positive staining. In addition, mesenchymal cells of the fibrous tissue at the osteophyte surface were positive for PTPRZ1. Panels b, e, h and k show higher magnification views of the delimited areas of panels a, d, g and j, respectively. Cartilage and bone are delimited by dotted lines. Bo: bone, Cart: cartilage, CC: calcified cartilage. Bars = 200 μm (a, d, g and j) or 50 μm (b, c, e, f, h, i, k and l).

      IL-34 increases the remodeling potential of hypertrophic chondrocytes and osteoblasts

      We next determined the chondrocyte (n = 7) and osteoblast response (n = 5) to IL-34. No effect of IL-34 on iMACs and prehypertrophic chondrocytes was observed (data not shown). In contrast, both hypertrophic chondrocytes and osteoblasts were sensitive to IL-34 stimulation. A dose-dependent increase in the mRNA expression of Ccl2, Cxcl12 and Mmp 13 by hypertrophic chondrocytes was observed in response to IL-34 [Fig. 6(A)–(C)]. This was associated with a dose-dependent increase in the release of CCL2 and MMP-13 (P = 0.0360 and P = 0.0489 for CCL2 and MMP-13, respectively) [Fig. 6(E) and (F)]. No increase in MMP-3 expression and release was observed [Fig. 6(D) and (G)]. Similarly, IL-34 tended to stimulate the expression of Ccl2 (P = 0.0911), Mmp 3 (P = 0.129), Mmp13 (P = 0.141) and Tnfα (P = 0.1301), by osteoblasts [Fig. 7(A)–(D)]. Consistently, IL-34 stimulated the release of CCL2 (P = 0.0005), MMP-3 (P = 0.0085) and MMP-13 (P = 0.0052) by IL-34 in a dose-dependent manner [Fig. 7(G)–I]. IL-34 induced also a dose-dependent decrease in the mRNA expression of both Pedf (P = 0.0084) and Ptprz1 (P = 0.0917) by osteoblasts [Fig. 7(E) and (F)]. IL-34 had no observable effect on the mRNA expression of Vegf, Rankl and Csf1r in both hypertrophic chondrocytes and osteoblasts (data not shown).
      Fig. 6
      Fig. 6Increased remodeling potential of hypertrophic chondrocytes in response to IL-34. Hypertrophic chondrocytes (n = 7) were stimulated by increasing concentrations of IL-34 before RT-qPCR analysis of the mRNA expression of Cxcl12 (A), Ccl2 (B), Mmp 3 (C) and Mmp 13 (D). The release of CCL2 (E), MMP-3 (F) and MMP-13 (G) into cell conditioned medium in response to increased concentrations of IL-34 was measured by ELISAs.
      Fig. 7
      Fig. 7Increased remodeling potential of osteoblasts in response to IL-34. Osteoblasts (n = 5) were stimulated by increasing concentrations of IL-34 before RT-qPCR analysis of the mRNA expression of Cxcl12 Ccl2 (A), Mmp 3 (B), Mmp 13 (C), Tnfα (D), Pedf (E) and Ptprz1 (F). The release of CCL2 (G), MMP-3 (H) and MMP-13 (J) into cell conditioned medium in response to increased concentrations of IL-34 was measured by ELISAs.

      Discussion

      OA is characterized at the cellular level by deep phenotypic modifications of cells from the different joint tissues. Notably, there is a hypertrophic differentiation of chondrocytes leading to the accumulation of calcified depots within cartilage
      • Kirsch T.
      • Swoboda B.
      • Nah H.
      Activation of annexin II and V expression, terminal differentiation, mineralization and apoptosis in human osteoarthritic cartilage.
      and an advancement of the mineralization front in the deeper part of the cartilage. Hypertrophic differentiation of chondrocytes during OA is thought to play an important role in cartilage disappearance and subchondral bone remodeling
      • Dreier R.
      Hypertrophic differentiation of chondrocytes in osteoarthritis: the developmental aspect of degenerative joint disorders.
      ,
      • Singh P.
      • Marcu K.B.
      • Goldring M.B.
      • Otero M.
      Phenotypic instability of chondrocytes in osteoarthritis: on a path to hypertrophy.
      . Therefore, the identification of molecular factors produced by hypertrophic chondrocytes and involved in cartilage and bone damage in OA could be of therapeutic interest. However, no current model of chondrocyte hypertrophic differentiation is able to lead such investigations. None use articular chondrocytes to obtain hypertrophic chondrocytes able to calcify their matrix. Either they do not provide sufficient quantities of hypertrophic chondrocytes or are based on cell lines rather than primary cultures.
      Thus, developing such a model that combines all these features represents a real need and a great challenge. Here, we have developed an original model of progressive articular chondrocyte hypertrophic differentiation and identified the recently discovered IL-34 as a factor with putative osteochondral remodeling activity.
      Our model of chondrocyte hypertrophic differentiation was achieved starting with primary cultures of iMACs. Calcification was only observed after culturing iMACs successively in Medium 1, Medium 2 and then in Medium 3. In addition, only a faint expression of osteocalcin, Osterix and Tnap was observed in the iMAC cultures, suggesting that matrix calcification was not due to contamination of cultures by osteoblasts. Although a transdifferentiation of some chondrocytes into osteoblast-like cells during the culture time cannot be totally ruled out, we consider that the features we observed are attributable to hypertrophic chondrocytes rather than osteoblast-like cells since the expression of Col10a1, a specific marker of hypertrophic chondrocytes, was strongly increased during the culture time.
      Chondrocyte hypertrophy is generally only determined by the combination of an increased expression of some hypertrophic markers with the decreased expression of chondrocyte markers
      • Cecil D.L.
      • Johnson K.
      • Rediske J.
      • Lotz M.
      • Schmidt A.M.
      • Terkeltaub R.
      Inflammation-induced chondrocyte hypertrophy is driven by receptor for advanced glycation end products.
      • Pesesse L.
      • Sanchez C.
      • Delcour J.P.
      • Bellahcene A.
      • Baudouin C.
      • Msika P.
      • et al.
      Consequences of chondrocyte hypertrophy on osteoarthritic cartilage: potential effect on angiogenesis.
      • Wehling N.
      • Palmer G.D.
      • Pilapil C.
      • Liu F.
      • Wells J.W.
      • Muller P.E.
      • et al.
      Interleukin-1 beta and tumor necrosis factor alpha inhibit chondrogenesis by human mesenchymal stem cells through NF-kappaB-dependent pathways.
      • Zhang X.
      • Crawford R.
      • Xiao Y.
      Inhibition of vascular endothelial growth factor with shRNA in chondrocytes ameliorates osteoarthritis.
      . Nevertheless, in addition to these criteria, matrix calcification appears necessary to ascertain that chondrocytes reach hypertrophy. Indeed, chondrocytes cultured in Medium 2 never calcified, although they showed a mRNA expression of hypertrophic markers. The current published culture models of hypertrophic differentiation reaching matrix calcification include primary cultures of chondrocytes isolated from limb buds and growth plate, ATDC5 cells and mesenchymal stem cells (MSCs)
      • James C.G.
      • Ulici V.
      • Tuckermann J.
      • Underhill T.M.
      • Beier F.
      Expression profiling of Dexamethasone-treated primary chondrocytes identifies targets of glucocorticoid signalling in endochondral bone development.
      • Stanton L.A.
      • Sabari S.
      • Sampaio A.V.
      • Underhill T.M.
      • Beier F.
      p38 MAP kinase signalling is required for hypertrophic chondrocyte differentiation.
      • Kirsch T.
      • Nah H.D.
      • Shapiro I.M.
      • Pacifici M.
      Regulated production of mineralization-competent matrix vesicles in hypertrophic chondrocytes.
      • Mueller M.B.
      • Tuan R.S.
      Functional characterization of hypertrophy in chondrogenesis of human mesenchymal stem cells.
      • Shukunami C.
      • Ishizeki K.
      • Atsumi T.
      • Ohta Y.
      • Suzuki F.
      • Hiraki Y.
      Cellular hypertrophy and calcification of embryonal carcinoma-derived chondrogenic cell line ATDC5 in vitro.
      . However, neither ATDC5 cells nor MSCs are chondrocytes, and despite the similarities between chondrocytes from growth plates and articular cartilage, they have distinct molecular phenotypes
      • Chau M.
      • Lui J.C.
      • Landman E.B.
      • Spath S.S.
      • Vortkamp A.
      • Baron J.
      • et al.
      Gene expression profiling reveals similarities between the spatial architectures of postnatal articular and growth plate cartilage.
      • Hissnauer T.N.
      • Baranowsky A.
      • Pestka J.M.
      • Streichert T.
      • Wiegandt K.
      • Goepfert C.
      • et al.
      Identification of molecular markers for articular cartilage.
      • Matsusaki T.
      • Aoyama T.
      • Nishijo K.
      • Okamoto T.
      • Nakayama T.
      • Nakamura T.
      • et al.
      Expression of the cadherin-11 gene is a discriminative factor between articular and growth plate chondrocytes.
      . These models are therefore not suitable for studying hypertrophic articular chondrocytes. Yahara et al. also reported matrix calcification with cultures of articular chondrocytes
      • Yahara Y.
      • Takemori H.
      • Okada M.
      • Kosai A.
      • Yamashita A.
      • Kobayashi T.
      • et al.
      Pterosin B prevents chondrocyte hypertrophy and osteoarthritis in mice by inhibiting Sik3.
      . Pellet cultures of human OA articular chondrocytes expressed hypertrophic markers and showed calcifications
      • Yahara Y.
      • Takemori H.
      • Okada M.
      • Kosai A.
      • Yamashita A.
      • Kobayashi T.
      • et al.
      Pterosin B prevents chondrocyte hypertrophy and osteoarthritis in mice by inhibiting Sik3.
      (and this study). However, only sparse calcifications were observed, suggesting that only a subset of chondrocytes reach hypertrophy. Nevertheless, this model appears useful to validate our results obtained with iMACs.
      In addition to iMACs and hypertrophic chondrocytes, our model provides insight into a third chondrocyte phenotype, which we considered as prehypertrophic chondrocytes considering its gene expression pattern. This chondrocyte population expressed intermediate levels of chondrocyte and hypertrophic markers compared to those expressed in iMACs and hypertrophic chondrocytes, and they did not calcify their matrix. These cells also expressed known prehypertrophic markers, including snorc and osteomodulin
      • Belluoccio D.
      • Etich J.
      • Rosenbaum S.
      • Frie C.
      • Grskovic I.
      • Stermann J.
      • et al.
      Sorting of growth plate chondrocytes allows the isolation and characterization of cells of a defined differentiation status.
      ,
      • Heinonen J.
      • Taipaleenmaki H.
      • Roering P.
      • Takatalo M.
      • Harkness L.
      • Sandholm J.
      • et al.
      Snorc is a novel cartilage specific small membrane proteoglycan expressed in differentiating and articular chondrocytes.
      , whose expression was stronger than that in both iMACs and hypertrophic chondrocytes. The prehypertrophic phenotype corresponded to a specific population of chondrocytes showing a homogeneous molecular pattern, as revealed by PCA of the microarray results. The whole genome transcriptomic analysis indeed revealed that two phenotypically distinct populations of chondrocytes were obtained from iMACs. We characterized them as prehypertrophic and hypertrophic chondrocytes based on their gene expression pattern and their ability to calcify or not their extracellular matrix. Investigating other features, including the cell shape and the composition and the organization of the extracellular matrix, would be of interest to validate prehypertrophic and hypertrophic states.
      The presence of several populations of cells with molecularly distinct phenotypes within articular cartilage has recently been described in human OA cartilage
      • Jayasuriya C.T.
      • Hu N.
      • Li J.
      • Lemme N.
      • Terek R.
      • Ehrlich M.G.
      • et al.
      Molecular characterization of mesenchymal stem cells in human osteoarthritis cartilage reveals contribution to the OA phenotype.
      ,
      • Ji Q.
      • Zheng Y.
      • Zhang G.
      • Hu Y.
      • Fan X.
      • Hou Y.
      • et al.
      Single-cell RNA-seq analysis reveals the progression of human osteoarthritis.
      . Ji et al. characterized seven different chondrocyte populations, including prehypertrophic and hypertrophic chondrocytes
      • Ji Q.
      • Zheng Y.
      • Zhang G.
      • Hu Y.
      • Fan X.
      • Hou Y.
      • et al.
      Single-cell RNA-seq analysis reveals the progression of human osteoarthritis.
      . Their results suggest that prehypertrophic chondrocytes localized in the deeper part of articular cartilage play an important role in OA progression. Here, we show that prehypertrophic chondrocytes is the most sensitive chondrocyte population to an inflammatory stimulus. Inflammatory stress is a hallmark of OA, and IL-1β induced a more potent global response by prehypertrophic chondrocytes than by iMACs and hypertrophic chondrocytes. In addition, the prehypertrophic to hypertrophic differentiation of chondrocytes is associated with an increase in the bone remodeling and angiogenic potential of chondrocytes, as evaluated by their molecular pattern. Functional studies will be needed to ascertain the increased potential for tissue remodeling of hypertrophic chondrocytes.
      The microarray analysis for cytokines/chemokines overexpressed with chondrocyte hypertrophic differentiation and able to act in an autocrine and paracrine manner highlighted the recently discovered IL-34
      • Lin H.
      • Lee E.
      • Hestir K.
      • Leo C.
      • Huang M.
      • Bosch E.
      • et al.
      Discovery of a cytokine and its receptor by functional screening of the extracellular proteome.
      . RT-qPCR results confirmed this increased mRNA expression, and we also showed that murine and human chondrocytes produced IL-34 at higher rates when chondrocytes were hypertrophic. In OA, hypertrophic chondrocytes are localized within articular cartilage and osteophytes, where they are suspected to play a major role in the pathological remodeling of the osteochondral junctions. Both articular and osteophytic cartilages from OA patients released IL-34. A differential transcriptomic analysis of articular and osteophytic cartilage from paired OA patients revealed a higher expression of genes with functions in terminal chondrocyte differentiation by osteophytic cartilage
      • Gelse K.
      • Ekici A.B.
      • Cipa F.
      • Swoboda B.
      • Carl H.D.
      • Olk A.
      • et al.
      Molecular differentiation between osteophytic and articular cartilage--clues for a transient and permanent chondrocyte phenotype.
      . Interestingly, we found that osteophytic cartilage released higher amounts of IL-34 than articular cartilage. In addition, PTRZ1 was among the most upregulated genes in osteophytic cartilage compared to its gene expression in articular cartilage
      • Gelse K.
      • Ekici A.B.
      • Cipa F.
      • Swoboda B.
      • Carl H.D.
      • Olk A.
      • et al.
      Molecular differentiation between osteophytic and articular cartilage--clues for a transient and permanent chondrocyte phenotype.
      . Both osteophytic and articular cartilages showed positive immunostaining of PTPRZ1. In particular, OA cartilage PTPRZ1-positive chondrocytes were preferentially located in the deeper zone of joint cartilage or in the chondrocyte clusters, the two areas where hypertrophic chondrocytes are usually found
      • Pullig O.
      • Weseloh G.
      • Ronneberger D.
      • Kakonen S.
      • Swoboda B.
      Chondrocyte differentiation in human osteoarthritis: expression of osteocalcin in normal and osteoarthritic cartilage and bone.
      ,
      • Kirsch T.
      • Swoboda B.
      • Nah H.
      Activation of annexin II and V expression, terminal differentiation, mineralization and apoptosis in human osteoarthritic cartilage.
      . Osteoblasts and osteocytes of the subchondral bone also express PTPRZ1, as described
      • Kaspiris A.
      • Mikelis C.
      • Heroult M.
      • Khaldi L.
      • Grivas T.B.
      • Kouvaras I.
      • et al.
      Expression of the growth factor pleiotrophin and its receptor protein tyrosine phosphatase beta/zeta in the serum, cartilage and subchondral bone of patients with osteoarthritis.
      , as well as cells present in vascular channels, including vessels and mesenchymal stromal cells. We found a similar expression pattern for CSF1R, whose the expression by bone cells and the increased expression in OA cartilage has been already reported by others
      • Wittrant Y.
      • Gorin Y.
      • Mohan S.
      • Wagner B.
      • Abboud-Werner S.L.
      Colony-stimulating factor-1 (CSF-1) directly inhibits receptor activator of nuclear factor-{kappa}B ligand (RANKL) expression by osteoblasts.
      ,
      • Rai M.F.
      • Tycksen E.D.
      • Cai L.
      • Yu J.
      • Wright R.W.
      • Brophy R.H.
      Distinct degenerative phenotype of articular cartilage from knees with meniscus tear compared to knees with osteoarthritis.
      .
      Considering the expression of IL-34 and its receptors by cells of the cartilage/subchondral bone interface, IL-34 may act as a paracrine and autocrine factor on cartilage and bone cells in OA. IL-34 is indeed a known osteoclastogenesis factor
      • Baud'huin M.
      • Renault R.
      • Charrier C.
      • Riet A.
      • Moreau A.
      • Brion R.
      • et al.
      Interleukin-34 is expressed by giant cell tumours of bone and plays a key role in RANKL-induced osteoclastogenesis.
      ,
      • Chen Z.
      • Buki K.
      • Vaaraniemi J.
      • Gu G.
      • Vaananen H.K.
      The critical role of IL-34 in osteoclastogenesis.
      and has been reported to stimulate angiogenesis
      • Segaliny A.I.
      • Mohamadi A.
      • Dizier B.
      • Lokajczyk A.
      • Brion R.
      • Lanel R.
      • et al.
      Interleukin-34 promotes tumor progression and metastatic process in osteosarcoma through induction of angiogenesis and macrophage recruitment.
      . Its biological activity on both chondrocytes and osteoblasts has never been investigated. Neither iMACs nor prehypertrophic chondrocytes responded to IL-34 stimulation. This may be explained by the differential expression of IL-34 receptors on iMACs, prehypertrophic and hypertrophic chondrocytes. However, the involvement of other molecular partners differentially produced by the phenotypically distinct chondrocytes cannot be excluded as IL-34 displays some of its biological activities independently to Ptprz1 and Csf1r. It may also bind to other cytokines to form heteromeric cytokines
      • Segaliny A.I.
      • Brion R.
      • Brulin B.
      • Maillasson M.
      • Charrier C.
      • Teletchea S.
      • et al.
      IL-34 and M-CSF form a novel heteromeric cytokine and regulate the M-CSF receptor activation and localization.
      ,
      • Segaliny A.I.
      • Brion R.
      • Mortier E.
      • Maillasson M.
      • Cherel M.
      • Jacques Y.
      • et al.
      Syndecan-1 regulates the biological activities of interleukin-34.
      . In hypertrophic chondrocytes and osteoblasts, IL-34 stimulated the mRNA expression and the release of some tissue remodeling factors, especially the mRNA expression of Cxcl12 in hypertrophic chondrocytes, the secretion of CCL2 and MMP-13 by hypertrophic chondrocytes and osteoblasts and the release of MMP-3 by osteoblasts. Although IL-34 did not stimulate VEGF expression, it may indirectly induce angiogenesis by stimulating the expression of CXCL12 and by inhibiting that of the angiostatic factor PEDF. In addition to its reported direct action on osteoclastogenesis and angiogenesis
      • Baud'huin M.
      • Renault R.
      • Charrier C.
      • Riet A.
      • Moreau A.
      • Brion R.
      • et al.
      Interleukin-34 is expressed by giant cell tumours of bone and plays a key role in RANKL-induced osteoclastogenesis.
      • Chen Z.
      • Buki K.
      • Vaaraniemi J.
      • Gu G.
      • Vaananen H.K.
      The critical role of IL-34 in osteoclastogenesis.
      • Segaliny A.I.
      • Mohamadi A.
      • Dizier B.
      • Lokajczyk A.
      • Brion R.
      • Lanel R.
      • et al.
      Interleukin-34 promotes tumor progression and metastatic process in osteosarcoma through induction of angiogenesis and macrophage recruitment.
      , IL-34 may thus indirectly stimulate these two processes by acting on hypertrophic chondrocytes and osteoblasts. Cartilage-derived IL-34 may also explain the positive association between IL-34 concentration in synovial fluid and the radiographic and symptomatic severity of knee OA
      • Wang S.L.
      • Zhang R.
      • Hu K.Z.
      • Li M.Q.
      • Li Z.C.
      Interleukin-34 synovial fluid was associated with knee osteoarthritis severity: a cross-sectional study in knee osteoarthritis patients in different radiographic stages.
      . Further studies are needed to explore the specific role of IL-34, especially on human hypertrophic chondrocytes and osteoblasts, and more precisely that of hypertrophic-derived IL-34 in OA.
      Some limitations emerge from our study. They include the characterization of our model of iMAC hypertrophic differentiation, which would be enriched with the study of the cell shape and the extracellular matrix composition and organization. Functional studies are also needed to ascertain the increased potential for tissue remodeling of hypertrophic chondrocytes and the role of IL-34.
      To conclude, we have developed a new model of articular chondrocyte hypertrophic differentiation, which allows obtaining three molecularly distinct populations: iMACs, prehypertrophic chondrocytes and matrix calcifying hypertrophic chondrocytes. Our results support the hypothesis that the phenotypic alterations of prehypertrophic chondrocytes in articular cartilage are critical for the loss of cartilage homeostasis observed in OA. The prehypertrophic to hypertrophic differentiation of chondrocytes induced the expression of a subset of genes, which together may favor the pathological remodeling of cartilage and bone, as observed in OA. Notably, the increased production of IL-34 by hypertrophic chondrocytes could act locally on hypertrophic chondrocytes and osteoblasts to indirectly stimulate osteoclastogenesis and angiogenesis. Therefore, according to the tissue remodeling potential of the recently discovered IL-34, further investigations are needed to determine whether IL-34 could be targeted in OA and/or may be used as a synovial biomarker to determine the severity of OA.

      Author contributions

      • Conception and design: MHLP, PP, GRV, FDC, FB, XH
      • Analysis and interpretation of the data: SVE, MLPL, IT, DV, CB, SCG, LG, MHLP, PP, GRV, FDC, FB, XH
      • Drafting of the article: SVE, DV, FB, XH
      • Critical revision of the article for important intellectual content: DV, FB, XH
      • Final approval of the article: SVE, MLPL, IT, DV, AP, DC, CB, SCG, LG, DC, SM, GN, AS, MHLP, PP, GRV, FDC, FB, XH
      • Provision of study materials or patients: AS
      • Statistical expertise: DV
      • Obtaining of funding: FP, MHLP, FDC, FB, XH
      • Collection and assembly of data: SVE, MLPL, IT, AP, DC, CB, DC, SM, GN, AS, XH

      Conflict of interest

      None.

      Role of the funding source

      This work was supported by grants from the Société Française de Rhumatologie and the Fondation Arthritis Courtin . Sandy van Eegher was supported by a PhD grant from the Ministère de l’Enseignement Supérieur et de la Recherche . Indira Toillon was supported by a PhD grant from the region Ile de France ( ARDoC program ). Stéphanie Malbos was supported by a grant from the Société Française de Rhumatologie .

      References

        • Troeberg L.
        • Nagase H.
        Proteases involved in cartilage matrix degradation in osteoarthritis.
        Biochim Biophys Acta. 2011; 1824: 133-145
        • Goldring S.R.
        • Goldring M.B.
        Changes in the osteochondral unit during osteoarthritis: structure, function and cartilage-bone crosstalk.
        Nat Rev Rheumatol. 2016; 12: 632-644
        • Lories R.J.
        • Luyten F.P.
        The bone-cartilage unit in osteoarthritis.
        Nat Rev Rheumatol. 2011; 7: 43-49
        • Mahjoub M.
        • Berenbaum F.
        • Houard X.
        Why subchondral bone in osteoarthritis? The importance of the cartilage bone interface in osteoarthritis.
        Osteoporos Int. 2012; 23: 841-846
        • Zuscik M.J.
        • Hilton M.J.
        • Zhang X.
        • Chen D.
        • O'Keefe R.J.
        Regulation of chondrogenesis and chondrocyte differentiation by stress.
        J Clin Invest. 2008; 118: 429-438
        • Aigner T.
        • Bertling W.
        • Stoss H.
        • Weseloh G.
        • von der Mark K.
        Independent expression of fibril-forming collagens I, II, and III in chondrocytes of human osteoarthritic cartilage.
        J Clin Invest. 1993; 91: 829-837
        • Pullig O.
        • Weseloh G.
        • Ronneberger D.
        • Kakonen S.
        • Swoboda B.
        Chondrocyte differentiation in human osteoarthritis: expression of osteocalcin in normal and osteoarthritic cartilage and bone.
        Calcif Tissue Int. 2000; 67: 230-240
        • Wachsmuth L.
        • Soder S.
        • Fan Z.
        • Finger F.
        • Aigner T.
        Immunolocalization of matrix proteins in different human cartilage subtypes.
        Histol Histopathol. 2006; 21: 477-485
        • Belluoccio D.
        • Etich J.
        • Rosenbaum S.
        • Frie C.
        • Grskovic I.
        • Stermann J.
        • et al.
        Sorting of growth plate chondrocytes allows the isolation and characterization of cells of a defined differentiation status.
        J Bone Miner Res. 2010; 25: 1267-1281
        • Muller C.
        • Khabut A.
        • Dudhia J.
        • Reinholt F.P.
        • Aspberg A.
        • Heinegard D.
        • et al.
        Quantitative proteomics at different depths in human articular cartilage reveals unique patterns of protein distribution.
        Matrix Biol. 2014; 40: 34-45
        • Semevolos S.A.
        • Nixon A.J.
        • Fortier L.A.
        • Strassheim M.L.
        • Haupt J.
        Age-related expression of molecular regulators of hypertrophy and maturation in articular cartilage.
        J Orthop Res. 2006; 24: 1773-1781
        • Vortkamp A.
        • Lee K.
        • Lanske B.
        • Segre G.V.
        • Kronenberg H.M.
        • Tabin C.J.
        Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein.
        Science. 1996; 273: 613-622
        • Priam S.
        • Bougault C.
        • Houard X.
        • Gosset M.
        • Salvat C.
        • Berenbaum F.
        • et al.
        Identification of soluble 14-3-3 as a novel subchondral bone mediator involved in cartilage degradation in osteoarthritis.
        Arthritis Rheum. 2013; 65: 1831-1842
        • Sanchez C.
        • Gabay O.
        • Salvat C.
        • Henrotin Y.E.
        • Berenbaum F.
        Mechanical loading highly increases IL-6 production and decreases OPG expression by osteoblasts.
        Osteoarthritis Cartilage. 2009; 17: 473-481
        • Gosset M.
        • Berenbaum F.
        • Thirion S.
        • Jacques C.
        Primary culture and phenotyping of murine chondrocytes.
        Nat Protoc. 2008; 3: 1253-1260
        • Salvat C.
        • Pigenet A.
        • Humbert L.
        • Berenbaum F.
        • Thirion S.
        Immature murine articular chondrocytes in primary culture: a new tool for investigating cartilage.
        Osteoarthritis Cartilage. 2005; 13: 243-249
        • Yahara Y.
        • Takemori H.
        • Okada M.
        • Kosai A.
        • Yamashita A.
        • Kobayashi T.
        • et al.
        Pterosin B prevents chondrocyte hypertrophy and osteoarthritis in mice by inhibiting Sik3.
        Nat Commun. 2016; 7: 10959
        • Hughes T.R.
        • Mao M.
        • Jones A.R.
        • Burchard J.
        • Marton M.J.
        • Shannon K.W.
        • et al.
        Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer.
        Nat Biotechnol. 2001; 19: 342-347
        • Kirsch T.
        • Swoboda B.
        • Nah H.
        Activation of annexin II and V expression, terminal differentiation, mineralization and apoptosis in human osteoarthritic cartilage.
        Osteoarthritis Cartilage. 2000; 8: 294-302
        • Dreier R.
        Hypertrophic differentiation of chondrocytes in osteoarthritis: the developmental aspect of degenerative joint disorders.
        Arthritis Res Ther. 2010; 12: 216
        • Singh P.
        • Marcu K.B.
        • Goldring M.B.
        • Otero M.
        Phenotypic instability of chondrocytes in osteoarthritis: on a path to hypertrophy.
        Ann N Y Acad Sci. 2019; 1442: 17-34
        • Cecil D.L.
        • Johnson K.
        • Rediske J.
        • Lotz M.
        • Schmidt A.M.
        • Terkeltaub R.
        Inflammation-induced chondrocyte hypertrophy is driven by receptor for advanced glycation end products.
        J Immunol. 2005; 175: 8296-8302
        • Pesesse L.
        • Sanchez C.
        • Delcour J.P.
        • Bellahcene A.
        • Baudouin C.
        • Msika P.
        • et al.
        Consequences of chondrocyte hypertrophy on osteoarthritic cartilage: potential effect on angiogenesis.
        Osteoarthritis Cartilage. 2013; 21: 1913-1923
        • Wehling N.
        • Palmer G.D.
        • Pilapil C.
        • Liu F.
        • Wells J.W.
        • Muller P.E.
        • et al.
        Interleukin-1 beta and tumor necrosis factor alpha inhibit chondrogenesis by human mesenchymal stem cells through NF-kappaB-dependent pathways.
        Arthritis Rheum. 2009; 60: 801-812
        • Zhang X.
        • Crawford R.
        • Xiao Y.
        Inhibition of vascular endothelial growth factor with shRNA in chondrocytes ameliorates osteoarthritis.
        J Mol Med (Berl). 2016; 94: 787-798
        • James C.G.
        • Ulici V.
        • Tuckermann J.
        • Underhill T.M.
        • Beier F.
        Expression profiling of Dexamethasone-treated primary chondrocytes identifies targets of glucocorticoid signalling in endochondral bone development.
        BMC Genom. 2007; 8: 205
        • Stanton L.A.
        • Sabari S.
        • Sampaio A.V.
        • Underhill T.M.
        • Beier F.
        p38 MAP kinase signalling is required for hypertrophic chondrocyte differentiation.
        Biochem J. 2004; 378: 53-62
        • Kirsch T.
        • Nah H.D.
        • Shapiro I.M.
        • Pacifici M.
        Regulated production of mineralization-competent matrix vesicles in hypertrophic chondrocytes.
        J Cell Biol. 1997; 137: 1149-1160
        • Mueller M.B.
        • Tuan R.S.
        Functional characterization of hypertrophy in chondrogenesis of human mesenchymal stem cells.
        Arthritis Rheum. 2008; 58: 1377-1388
        • Shukunami C.
        • Ishizeki K.
        • Atsumi T.
        • Ohta Y.
        • Suzuki F.
        • Hiraki Y.
        Cellular hypertrophy and calcification of embryonal carcinoma-derived chondrogenic cell line ATDC5 in vitro.
        J Bone Miner Res. 1997; 12: 1174-1188
        • Chau M.
        • Lui J.C.
        • Landman E.B.
        • Spath S.S.
        • Vortkamp A.
        • Baron J.
        • et al.
        Gene expression profiling reveals similarities between the spatial architectures of postnatal articular and growth plate cartilage.
        PLoS One. 2014; 9e103061
        • Hissnauer T.N.
        • Baranowsky A.
        • Pestka J.M.
        • Streichert T.
        • Wiegandt K.
        • Goepfert C.
        • et al.
        Identification of molecular markers for articular cartilage.
        Osteoarthritis Cartilage. 2010; 18: 1630-1638
        • Matsusaki T.
        • Aoyama T.
        • Nishijo K.
        • Okamoto T.
        • Nakayama T.
        • Nakamura T.
        • et al.
        Expression of the cadherin-11 gene is a discriminative factor between articular and growth plate chondrocytes.
        Osteoarthritis Cartilage. 2006; 14: 353-366
        • Heinonen J.
        • Taipaleenmaki H.
        • Roering P.
        • Takatalo M.
        • Harkness L.
        • Sandholm J.
        • et al.
        Snorc is a novel cartilage specific small membrane proteoglycan expressed in differentiating and articular chondrocytes.
        Osteoarthritis Cartilage. 2011; 19: 1026-1035
        • Jayasuriya C.T.
        • Hu N.
        • Li J.
        • Lemme N.
        • Terek R.
        • Ehrlich M.G.
        • et al.
        Molecular characterization of mesenchymal stem cells in human osteoarthritis cartilage reveals contribution to the OA phenotype.
        Sci Rep. 2018; 8: 7044
        • Ji Q.
        • Zheng Y.
        • Zhang G.
        • Hu Y.
        • Fan X.
        • Hou Y.
        • et al.
        Single-cell RNA-seq analysis reveals the progression of human osteoarthritis.
        Ann Rheum Dis. 2018; 78: 100-110
        • Lin H.
        • Lee E.
        • Hestir K.
        • Leo C.
        • Huang M.
        • Bosch E.
        • et al.
        Discovery of a cytokine and its receptor by functional screening of the extracellular proteome.
        Science. 2008; 320: 807-811
        • Gelse K.
        • Ekici A.B.
        • Cipa F.
        • Swoboda B.
        • Carl H.D.
        • Olk A.
        • et al.
        Molecular differentiation between osteophytic and articular cartilage--clues for a transient and permanent chondrocyte phenotype.
        Osteoarthritis Cartilage. 2012; 20: 162-171
        • Kaspiris A.
        • Mikelis C.
        • Heroult M.
        • Khaldi L.
        • Grivas T.B.
        • Kouvaras I.
        • et al.
        Expression of the growth factor pleiotrophin and its receptor protein tyrosine phosphatase beta/zeta in the serum, cartilage and subchondral bone of patients with osteoarthritis.
        Joint Bone Spine. 2013; 80: 407-413
        • Wittrant Y.
        • Gorin Y.
        • Mohan S.
        • Wagner B.
        • Abboud-Werner S.L.
        Colony-stimulating factor-1 (CSF-1) directly inhibits receptor activator of nuclear factor-{kappa}B ligand (RANKL) expression by osteoblasts.
        Endocrinology. 2009; 150: 4977-4988
        • Rai M.F.
        • Tycksen E.D.
        • Cai L.
        • Yu J.
        • Wright R.W.
        • Brophy R.H.
        Distinct degenerative phenotype of articular cartilage from knees with meniscus tear compared to knees with osteoarthritis.
        Osteoarthritis Cartilage. 2019; 27: 945-955
        • Baud'huin M.
        • Renault R.
        • Charrier C.
        • Riet A.
        • Moreau A.
        • Brion R.
        • et al.
        Interleukin-34 is expressed by giant cell tumours of bone and plays a key role in RANKL-induced osteoclastogenesis.
        J Pathol. 2010; 221: 77-86
        • Chen Z.
        • Buki K.
        • Vaaraniemi J.
        • Gu G.
        • Vaananen H.K.
        The critical role of IL-34 in osteoclastogenesis.
        PLoS One. 2011; 6e18689
        • Segaliny A.I.
        • Mohamadi A.
        • Dizier B.
        • Lokajczyk A.
        • Brion R.
        • Lanel R.
        • et al.
        Interleukin-34 promotes tumor progression and metastatic process in osteosarcoma through induction of angiogenesis and macrophage recruitment.
        Int J Canc. 2015; 137: 73-85
        • Segaliny A.I.
        • Brion R.
        • Brulin B.
        • Maillasson M.
        • Charrier C.
        • Teletchea S.
        • et al.
        IL-34 and M-CSF form a novel heteromeric cytokine and regulate the M-CSF receptor activation and localization.
        Cytokine. 2015; 76: 170-181
        • Segaliny A.I.
        • Brion R.
        • Mortier E.
        • Maillasson M.
        • Cherel M.
        • Jacques Y.
        • et al.
        Syndecan-1 regulates the biological activities of interleukin-34.
        Biochim Biophys Acta. 2015; 1853: 1010-1021
        • Wang S.L.
        • Zhang R.
        • Hu K.Z.
        • Li M.Q.
        • Li Z.C.
        Interleukin-34 synovial fluid was associated with knee osteoarthritis severity: a cross-sectional study in knee osteoarthritis patients in different radiographic stages.
        Dis Markers. 2018; 2018: 2095480