Epigenetic regulation in chondrocyte phenotype maintenance for cell-based cartilage repair

Abstract

Loss of hyaline chondrocyte phenotype during the monolayer culture in vitro is a major obstacle for cell-based articular cartilage repair. Increasing evidence implicates an important role of the epigenetic regulation in maintaining the chondrocyte phenotype. DNA methylation, histone modifications and microRNAs have all been shown to contribute to chondrocyte dedifferentiation and hypertrophy. Moreover, the interplay among epigenetic regulators forms a complicated epigenetic network in regulating chondrocyte dedifferentiation. This review provides a detailed overview of the epigenetic regulation in maintaining the chondrocyte phenotype for chondrocyte-based cartilage repair.

Introduction

Articular cartilage has very poor regeneration ability. Once damaged, it is very difficult for the cartilage to get repaired by itself. Traditional therapies (e.g. micro-fracture) have limited capacity to treat cartilage defects and the long-term outcome is often unsatisfactory [1]. Since 1994, autologous chondrocyte implantation (ACI) was established to treat cartilage defects [2], it has attracted much attention and become a golden therapy to repair focal defects except in the case of osteoarthritis (OA) or other complications [3]. However, the efficacy of ACI in repairing cartilage defects is unsatisfactory due to its variability in forming the desirable hyaline cartilage [4,5]. Monolayer expansion in vitro is a crucial step for the treatment of cartilage defects with ACI. One of the major obstacles accompanying with the monolayer culture is the loss of hyaline chondrocyte phenotype, leading to the chondrocyte dedifferentiation or chondrocyte hypertrophy, and the production of inferior matrix unsuitable for ACI [3]. Epigenetics plays an essential role in the maintenance of articular cartilage and has been implicated in the degenerative articular cartilage of OA that is characterized by the excessive chondrocyte dedifferentiation and hypertrophy [6-9]. Several lines of evidence have suggested the role of epigenetic modification in regulating chondrocyte phenotype. A detailed overview of the epigenetic regulation in gene expression during the chondrocyte dedifferentiation and hypertrophy may present a new perspective on the maintenance of chondrocyte phenotype.

In this review, we first summarize several key genes during the chondrocyte differentiation and cartilage matrix homeostasis. Secondly, we review several epigenetic mechanisms including DNA methylation, histone modifications and microRNAs (miRNAs) involved in the chondrocyte dedifferentiation and hypertrophy during in vitro chondrocyte expansion and passage. Thirdly, we also discuss recent advances in epigenetic research on cartilage defect repair including 5-hydroxymethylcytosine (5-hmC) and long non-coding RNAs (lncRNAs). Finally, we highlight the interactions between epigenetic regulators in the maintenance of chondrocyte phenotype.

Key genes involved in the chondrocyte differentiation and cartilage homeostasis

Accumulating evidence has shown that epigenetics regulates the gene expression during the chondrocyte proliferation, differentiation and function. Derived from mesenchymal stem cells (MSCs), chondrocytes migrate and are sparsely distributed in the articular cartilage. The chondrocyte is the only cell type responsible for the remodeling of the cartilage extracellular matrix (ECM). The cartilage ECM contains collagen proteins such as type II, XI, IX collagen (COL-2, COL-11, COL-9), and non-collagen proteins such as hyaluronan, aggrecan (ACAN), cartilage link protein 1 (CRTL-1) (reviewed in Table 1). Catabolic genes such as matrix metalloproteinase-13 (MMP-13) and a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTs) maintain the low turnover of cartilage matrix. In turn, cartilage ECM influences the nature of chondrocyte phenotype during the cartilage matrix remodeling [10]. Once the cartilage remodeling is disturbed, hyaline chondrocytes will dedifferentiate into fibroblast-like cells or undergo hypertrophy-like changes, and eventually result in apoptosis.

Table 1.

Related genes in the chondrocyte differentiation

Marker	Function	References
COL-2A1, COL-11A1	Cartilage-specific ECM	[22]
COL-9A1	Regulating the integrity and stability of articular cartilage	[36,76]
COL-12A1	Providing a microenvironment that supports hyaline cartilage formation	[77]
COL-1A1, COL-3A1	Collagens associated with dedifferentiation	[22]
COL-10A1	A hypertrophic marker	[78]
ALPL	A transcriptional factor regulating hypertrophy	[78]
RUNX-2	A transcriptional factor promoting chondrocyte hypertrophy	[19,78,79]
IHH	A transcriptional factor regulating chondrocyte hypertrophy
COX-2	A transcriptional factor promoting chondrocyte hypertrophy
PTHR	Inhibiting chondrocyte hypertrophy	[80-82]
SOX-9	A critical factor for chondrocyte differentiation	[83,84]
CRTL-1, MAGP2	The most differentially expressed transcripts between chondrocyte and synovial cell cultures	[85]
CD44	Hyaluronan receptor	[86,87]
ACAN	A chondrogenic marker	[22]
VCAN	A large droitin sulfate proteoglycan expressed by fibroblasts	[23,24]
MMP-13	COL-2-degrading enzyme	[79]
ADAMTs	Aggrecan-degrading enzymes	[88,89]
Matrilin-3	Inhibiting chondrocyte hypertrophy; Essential for cartilage matrix stability	[76,90-92]
COMP	A non-collagenous glycoprotein of the extracellular matrix, which enhancing cartilage ECM organization and assembly	[83,93-95]
MGP	Mineralization inhibitory protein	[96-98]
HDAC	Inhibiting chondrocyte dedifferentiation and hypertrophy	[42,43]

Along with the articular chondrocyte expansion in the monolayer culture, the expression of the sex determining region Y box gene 9 (SOX-9) and the regenerative ability of articular chondrocytes to cartilage tissue are decreased [11,12]. SOX-9 is one of the key transcriptional factors that maintain the chondrocyte phenotype and cartilage homeostasis [13,14]. SOX-9 activates the transcription of COL-9A1, which is important for the integrity and stability of articular cartilage [15]. However, the SOX-9 activity negatively regulates the MMP-13 expression [16-18]. MMP-13 is the specific collagenase that is responsible for the degradation of COL-2 and the increased expression of MMP-13 is a notable character of chondrocyte hypertrophy [19]. Interestingly, the impact of SOX-9 on the promoter activity of COL-2A1 depends on the phenotype of chondrocytes and culture conditions. In a 2-dimensional (2-D) culture system, low-level of the SOX-9 expression promotes the COL-2A1 gene transcription, while the elevated SOX9 expression inhibits the COL-2A1 gene expression in both fully differentiated and slightly phenotypically altered chondrocytes. By contrast, in the advanced stages of dedifferentiation, SOX-9, independently of its expression level, depresses the COL-2A1 transcriptional activity [20]. Under 3-D culture conditions, the overexpression of SOX-9 promotes the dedifferentiated articular chondrocytes to regain a chondrocyte phenotype and the COL-2A1 gene expression to form a cartilaginous matrix [11]. Kim et al. reported that increased DNA methylation and decreased acetylation in the SOX-9 gene promoter region resulted in the loss of phenotypes [21]. Therefore, the epigenetic modification may be involved in the differential roles of SOX-9 in the COL-2A1 transcription.

During the monolayer culture in vitro, hyaline chondrocytes are prone to lose their phenotype and dedifferentiate into fibro-chondrocytes, which are characterized by the increased expressions of COL-1, COL-3, and versican [22-24].

During the prolonged in vitro culture, a portion of articular chondrocytes may undergo the hypertrophic transformation [25,26]. A negative feedback loop between IHH and PTHrP has been reported to be responsible for the early chondrocyte differentiation and the initiation of the hypertrophic differentiation [27]. Indian hedgehog (IHH) contributes to the chondrocyte hypertrophy [19], while the parathyroid hormone (PTH)-related peptide (PTHrP) signaling pathway can inhibit this process [28,29]. A very recent study has reported that IHH can induce the runt-related transcription factor 2 (RUNX-2) and COL-10 expression in chondrocytes [30]. Bone morphogenetic protein 2 (BMP-2) is shown to be a potent inducer of chondrogenesis. BMP-2 in combination with IHH could induce the chondrogenesis from human primary MSCs without the chondrocyte hypertrophy induction. However, BMP-2 treatment alone could induce the chondrocyte hypertrophy in vitro and in vivo [31]. Moreover, IHH gene transfer is sufficient to improve the cartilage repair quality in vivo, whereas BMP-2 treatment may increase the risk of the intra-lesional bone formation [32]. Therefore, the combinational and individual application of BMP-2 and IHH lead to different phenotype changes of chondrocytes. However, the mechanisms underlying these dramatic changes need further investigation.

Recently, the role of nuclear factor of activated T-cells (NFAT) in chondrocyte phenotype has been studied. NFAT-1 inhibits the chondrocyte hypertrophy and maintains the metabolic balance in articular cartilage [33]. NFAT-1 deficiency in mice leads to the loss of COL-2 and ACAN, meanwhile, the up-regulation of specific matrix-degrading proteinases in young adult articular cartilage of load-bearing joints [34]. NFAT-1 has been shown to affect the function of adult articular chondrocyte via the induction of histone methylation [35].

Epigenetic regulation in the chondrocyte differentiation and the cartilage homeostasis

DNA methylation

DNA methylation is the most stable epigenetic modification that involves the addition of a methyl group to the 5′ carbon of a cytosine base pair and occurs most often in a CG dinucleotide (CG site). DNA methyltransferase (DNA MTase) family including DNMT1, DNMT2, DNMT3A, DNMT3B, and DNMT3L are involved in the DNA methylation modification. Pathological loss of DNA methylation may result in aberrant gene induction, whereas increased methylation may silence normally expressed genes [3]. Results of our study showed that DNMT3A expression level was increased in chondrocytes undergoing dedifferentiation following serial passage in monolayer culture. DNA methylation inhibitor 5-azacytidine (5-AzaC) efficiently inhibited the chondrocyte dedifferentiation and decreased the DNMT3A expression level (Figure 1). These data suggest that DNMT3A, but not DNMT3B, may be involved in regulating the chondrocyte dedifferentiation. We are currently investigating the differential role of DNMT3A and DNMT3B in regulating chondrocyte phenotype.

Figure 1.

A crucial role of DNMT3A in the chondrocyte dedifferentiation. Chondrocyte dedifferentiation always accompanied with the monolayer chondrocyte expansion in vitro. After the treatment of 5-AzaC, the cells show a typical polygonal shape (A). The dedifferentiated chondrocytes have a fibroblast-like phenotype (B). With chondrocyte passaging (from P1 to P5), the DNMT3A expression level was increased. With 5-AzaC treatment, the DNMT3A expression level was decreased while no changes of DNMT3B were observed (C).

In vitro and in vivo experiments have demonstrated that COL-9A1 gene expression levels are modulated by DNA methylation in chondrocytes. Six CpG sites of the COL-9A1 promoter have been identified to be significantly hypermethylated in chondrocytes isolated from diseased OA cartilage, leading to a significant decrease (6,000-fold) in the COL-9A1 mRNA expression. The presence of 5-AzaC prevents passaging-induced hypermethylation and up-regulates COL-9A1 gene expression. A further mechanism study indicates that DNA methylation inhibits the activity of COL-9A1 promoter via attenuating the binding of transcriptional factor SOX-9 to the COL-9A1 promoter [36].

MMP-13, a catabolic gene of cartilage matrix, has been shown to be epigenetically regulated by DNA methylation [37,38]. The depletion of DNMT1 and DNMT3A triggers a significant increase in the MMP-13 expression. On the contrary, DNMT3B depletion decreases MMP-13 mRNA in human articular chondrocytes. The -104 CpG site is found to be critical for the regulation of MMP-13 by DNA methylation because its inhibition by 5-AzaC significantly increases MMP-13 expression in human articular chondrocytes. However, CpG methylation is not engaged in the regulation of ACAN promoter activity [39].

The BMP-2 expression is suppressed during the process of dedifferentiation in human articular chondrocytes. When treated with 5-AzaC, the BMP-2 expression is up-regulated and the process of chondrocyte dedifferentiation is decelerated [3], implicating an important role of DNA methylation in the BMP-2 expression in maintaining chondrocyte phenotype. However, the underlying molecular mechanisms remain to be investigated. In a previous study, DNA methylation was shown to have no effect on the COL-2 gene expression during the chicken chondrocyte differentiation and dedifferentiation [40]. Ma et al. demonstrated that 5-AzaC had no influence on the COL-2A1 expression in human chondrocytes, suggesting that DNA methylation is unlikely to be implicated in the COL-2A1 down-regulation in chondrocyte dedifferentiation [3].

Histone deacetylation

Acetylation and deacetylation are two major modifications of nucleosomal histones. In comparison to acetylation, deacetylation has been better characterized. There are two types of deacetylases involved in histone deacetylation: histones deacetylases (HDACs) and sirtuins deacetylases. HDACs function through a zinc-catalyzed mechanism of deacetylation, while sirtuins deacetylases are nicotinamide adenine dinucleotide (NAD⁺) dependent [41]. The major function of HDACs is to remove acetyl groups from histones, which causes condensation of the chromatins and ultimately leads to transcriptional repression. HDAC activity in primary articular chondrocytes decreases during dedifferentiation induced by a serial monolayer culture of chondrocytes and the activity recovers during re-differentiation when cultured in a 3-D system. Inhibition of HDAC with trichostatin A or PXD101 is sufficient to block the COL-2 expression in primary chondrocytes. Also, HDAC inhibitor blocks the re-differentiation of dedifferentiated chondrocytes by suppressing the synthesis and accumulation of COL-2. The detailed mechanism behind the inhibitory effect is that HDAC inhibitor could increase the expression of WNT-5A, which is known to inhibit the COL-2 expression [42]. In addition, the induction of WNT-5A expression by HDAC inhibitors is associated with acetylation of the WNT-5A promoter. These results indicate that HDACs promote the COL-2 expression at least partially by suppressing the transcription of WNT-5A [43]. Recently, HDAC4 is reported to play a chondroprotective role under inflammatory conditions. In addition to decrease the expression of interleukin 1 beta (IL-1β), overexpression of HDAC4 also blocks IL-1β-induced activation of catabolic events in human OA chondrocytes. Moreover, HDAC4 inhibits the promoter activity of RUNX-2 and MMP-13 in a dose-dependent manner, and its down-regulation is associated with the up-regulation of RUNX-2 and other OA-related genes (MMP-13, IHH, and COL-10) in human OA cartilage [44].

Silent information regulator 2 (SIR2) is an NAD-dependent deacetylase involved in chromatin silencing, which comprises seven homologues of SIRT1-7. Down-regulation of SIRT1 expression induces cellular senescence in rabbit articular chondrocytes, which in turn promotes the OA progression. Using human chondrocytes derived from OA patients, SIRT1 is found to promote target gene expression by the recruitment of SOX-9 transcription factor to cartilage-specific gene promoters. Therefore, SIRT1 may be critical in maintaining chondrocyte phenotype. Additionally, SIRT1 signaling is involved in IL-1β-mediated chondrocyte dedifferentiation, in which both p38 kinase and extracellular regulated protein kinases (ERK) are activated to initiate SIRT1-ERK signaling [6].

Histone methylation

Besides histone acetylation, histone methylation has been extensively studied and is the best characterized modifications of nucleosomal histones. Methylation occurs on both lysine (K) and arginine (E) residues. In histone H3, different lysine residues (K4, K9, K27, K36 and K79) can be methylated. Unlike acetylation that is generally associated with transcriptional activation, histone lysine methylation is associated with either gene activation or repression, depending on the specific residue modified. Methylation of histone H3 lysine 4 (H3K4), H3K36 and H3K79 is generally associated with transcriptional gene activation, whereas methylation of H3K9 and H3K20 is associated with transcriptional gene silencing. For example, in human osteoarthritic chondrocytes, the H3K4-Methyl epigenome regulates IL-1-induced cyclooxygenase 2 (COX-2) and inducible nitric oxide synthase (iNOS) expression [45]. NFAT1 is regulated by histone methylation, in which H3K4me2 is found to promote the NFAT1 expression, whereas H3K9me2 inhibits the NFAT1 expression in chondrocytes from healthy human articular cartilage. Lysine-specific demethylase-1 (Lsd1) has an inhibitory effect on the H3K4me2 activity at the NFAT1 promoter in embryonic articular chondrocytes. In 6-month articular chondrocytes, Jumonji C (Jmjc)-containing histone demethylase-2a (Jhdm2a) inhibits H3K9me2 at the NFAT1 promoter [35]. The SOX-9 transcriptional activity is also regulated by histone demethylation. PHF2, a histone lysine demethylase, positively regulates the SOX9-mediated chondrocyte differentiation. AT-rich interactive domain 5b (Arid5b) inhibits the H3K9me2 expression and thus demethylates the target genes of SOX-9. Arid5b recruits PHF2 to the promoter region of SOX-9 target genes and stimulates the H3K9me2 activation [46].

ERG-associated protein with SET domain (ESET), a histone methyltransferase enzyme, plays a vital role in the chondrocyte hypertrophy. Deletion of the ESET gene accelerates the chondrocyte hypertrophy in both embryos and young animals. ESET interacts with HDAC4 could inhibit the activity of RUNX-2, a hypertrophy-promoting transcription factor. The underlying mechanism by which ESET represses RUNX-2-mediated gene transactivation is dependent on its H3K9 methyltransferase activity and histone deacetylase activity. In addition, knockout of ESET is associated with the repression of IHH gene in pre- and early hypertrophic chondrocytes [8,47]. Another research found that histone methyltransferase Set7/9 could inhibit SIRT1 and potentiate the euchromatin formation on the promoter site of COL-2A1, resulting in the morphology-dependent COL-2A1 gene transactivation [48].

MiRNAs

MiRNAs are small single-stranded non-coding RNA molecules (~21-22nt) that affect both mRNA expression and protein synthesis. It is estimated that one-third of all human genes are regulated by miRNAs [49]. It is generally accepted that the main function of miRNAs is to knock down gene expression, though it has been shown to activate transcription too [50].

MiRNAs emerges as a new class of gene expression regulators in chondrogenesis. Recent miRNA expression profiling studies have identified subsets of miRNAs that are up-regulated or down-regulated during the process of human articular chondrocyte dedifferentiation. Several miRNAs, including miR-221, miR-222, miR-143, miR-145, miR-146a, miR-548e, miR-342-5p and miR-1248 are up-regulated in dedifferentiated articular chondrocytes [51,52]. Potential targets of of miR-145 include SOX-9 and COL-2A1 [51]. MiR-222 expression is higher in the weight-bearing anterior medial condyle as compared with the posterior non-weight bearing medial condyle. Therefore, the loss of mechanical stimulating factors may be one of the reasons accounting for the dedifferentiation of chondrocytes when cultured in monolayer in vitro [52]. On the other hand, miR-140, miR-491-3p, let-7d and miR-26a are down-regulated during the chondrocyte dedifferentiation [52]. MiR-140 is abundantly and relatively specifically expressed in chondrocytes, and its expression is positively correlated to the expression of chondrocyte markers [53]. The small-GTPase RalA and ADAMTS-5 have been identified as the target genes of miR-140 [54,55]. Loss of miR-140 moderately accelerates the hypertrophic differentiation of chondrocytes, suggesting miR-140 as an inhibitor of chondrocyte hypertrophy.

MiR-125b is down-regulated in OA chondrocytes compared to normal chondrocytes. Overexpression of miR-125b in human OA chondrocytes reverses the effect of IL-1β on ADAMTS-4 [56]. Thus, miR-125b may play an important role in inhibiting the cartilage matrix degradation. Another miRNA, such as miR-21 is significantly elevated in OA patients, and its overexpression could attenuate the process of chondrogenesis. The levels of MMP-1, MMP-2, MMP-3, and MMP-9 are significantly increased when CH8 cells (a human articular chondrocyte cell line) are transfected with synthetic miR-21 precursor, and the inhibition of miR-21 increases levels of MMPs [57]. Moreover, miR-21 silences the expression of growth differentiation factor 5 (GDF5) target gene through promoting its mRNA decay.

Since the identification of differentially expressed microRNA in differentiated and dedifferentiated chondrocytes, it is an attractive possibility that manipulate the miRNA expression could provide new strategies for improving cartilage tissue engineering. A thorough understanding of miRNA expression profile in chondrocytes and the manipulation of individual miRNA via introducing miRNA precursor molecules or antagonists, may open a new window for developing therapeutic strategies to inhibit the chondrocyte dedifferentiation and enhance the ECM synthesis.

Recent findings in epigenetics involved in chondrocyte dedifferentiation

Recent studies have identified 5-hydroxymethylcytosine (5-hmC) as a new epigenetic marker widely distributed in all types of tissues with varying degrees of abundance [58]. 5-hmC is an intermediate derived from the ten-eleven translocation (TET) enzyme-mediated DNA demethylation [59]. The expression levels of 5-hmC and TET-1 activity in human chondrocytes are suppressed by IL-1β and tumor necrosis factor alpha (TNF-α). Increased 5-hmC expression levels or its activities have been shown to counteract the pro-inflammatory effect to regulate the chondrocyte dedifferentiation [60].

Emerging as a new field, the lncRNAs have attracted much attention. The lncRNAs have been tentatively defined as noncoding RNAs of more than 200 nucleotides in length and are characterized by their complexity and diversity of sequences as well as mechanisms of action [61]. Accumulating evidence has shown that altered lncRNAs expression can result in aberrant gene expression, which contributes to a variety of disease states and biological functions [62-64]. OA cartilage and OA chondrocytes express higher levels of cartilage injury-related lncRNA (lncRNA-CIR). LncRNA-CIR is a negative regulator for collagen and ACAN while server as a positive regulator for matrix-degrading enzymes, such as MMP-13 and ADAMTS-5 [65]. HOTTIP is a novel lncRNA that is significantly up-regulated and suppresses the cartilage integrity factor integrin-α1 expression in OA progression [66]. Another lncRNA, growth arrest-specific 5 (GAS5) negatively regulates the miR-21 expression and thus contributes to the pathogenesis of OA through regulating the cell survival [67]. Since lncRNAs exert their functions in chondrocytes, it will be interesting to identify more IncRNAs related to the chondrocyte phenotype, which may provide new targets for intervention in treating OA patients.

Interactions between different epigenetic regulations

Recent studies have suggested that epigenetic regulators interact with each other to regulate chondrocyte phenotype (Figures 2 and 3). In this review, the interactions among miRNAs, DNA methylation and histone deacetylation will be discussed.

Figure 2.

Epigenetics regulation of miRNA expression during the chondrocyte dedifferentiation. CpG islands are located on the binding sites where transcription factors actively regulate promoters of individual miRNAs. HDAC, MBDs and DNMTs compete with transcriptional factors to bind the promoters of miRNAs, which leads to the formation of compact inactive chromatin. In addition, DNA demethylation exerts its regulatory effect on the promoters via TETs. Under the regulation of HDAC, DNMTs, and TETs, miR-140 inhibits the catabolic genes of the cartilage matrix metabolism, such as ADAMTS-5. MiR-145 negatively regulates anabolic genes of the chondrocyte differentiation, including SOX-9 and COL-2A1. MiR-21 degrades the chondrocyte hypertrophy marker COL-10 and hyaline chondrocyte marker COL-2.

Figure 3.

MiRNA in the regulation of DNA methylation and histone deacetylation during the chondrocyte dedifferentiation. Mature miRNAs regulate the DNMT and HDAC expression. MiR-152 directly targets DNMT1 by binding a recognition site located not in the usual 3’UTR but in the coding region. Both miR-1 and miR-140 directly modulate the HDAC4 expression.

The interplay between miRNAs and DNA methylation

Some miRNAs can affect the methylation pattern of the whole genome by direct target DNA methyltransferases to inhibit their activities. For instance, miRNA-152 modulates the canonical WNT pathway via targeting DNMT-1 in an arthritic rat model [68]. Conversely, DNA methylation can inhibit the transcription of miRNAs. The methylation of miRNAs in mammals was demonstrated by Yuan et al., in which they verified a tRNA methyltransferase NSun2 mediated methylation of the primary, precursor and mature miR-125b in vitro and in vivo [69]. NSun2 inhibits MiR-125b expression via inhibiting miR-125b maturation and RISC recruitment. Methyl CpG binding protein 2 (MeCP2) binds to methylated DNA and recruits the HDAC complex, and then regulates gene expression post-transcriptionally by suppressing nuclear miRNA processing [70].

The interactions between miRNAs and histone deacetylation

MiR-1 plays an important role in the regulation of chondrocyte phenotype during the development of growth plate via interacting with HDAC4 [71]. Evidence has demonstrated that miR-1 down-regulates HDAC4 by directly targeting its three prime untranslated region (3’-UTR), while inhibition of miR-1 attenuates its repression effect on HDAC4. Conversely, overexpression of HDAC4 inhibits the effect of miR-1 on the chondrocyte phenotype.

Recently, HDAC4 has been shown to be a target of miR-140. MiR-140 inhibits the mRNA translation via direct binding to the 3’-UTR of HDAC4 mRNA [72]. Another study has shown that, miR-140 reduces the myocyte-specific enhancer factor 2C expression, which results in enhancing the PTHrP-HDAC4 pathway to suppress the chondrocyte hypertrophy [73]. These results support an important role of the interactions between miRNAs and HDAC in maintaining the chondrocyte phenotype.

Future perspectives

The interactions between DNMT and histone deacetylase in many cell types suggest a possible crosstalk between these epigenetic regulators [74]. Given the important functions of epigenetic regulation, inhibition of DNA methylation could provide a promise approach in maintaining the chondrocyte phenotype, while site-specific promoter methylation or demethylation must be optimized. Epigenetic profiling and subsequent bioinformatics analysis may represent powerful strategies to identify region-specific DNA methylation that is critical for chondrocyte phenotype. In addition, recent studies have shown that RNA methylation is involved in the cell fate decision [75]. It is speculated that RNA methylation may play an important role in the chondrocyte phenotype maintenance. It may also open a new avenue to decelerate the chondrocyte dedifferentiation and hypertrophy. To design effective miRNA-based treatment modalities that have minimal or no off-target side effects, future work should be directed toward identifying key miRNA targets that functionally impact on the chondrocyte dedifferentiation and hypertrophy. More information regarding new epigenetic regulators for chondrocyte phenotype are needed to explore. With the increasing knowledge of the molecular mechanisms underlying the chondrocyte dedifferentiation and hypertrophy, better treatment strategies may be on the horizon to maintain the chondrocyte phenotype and simultaneously enhance the hyaline cartilage matrix to improve the outcome of ACI.

Disclosure of conflict of interest

None.

Abbreviations

ACI

autologous chondrocyte implantation

ACAN

aggrecan

ADAMTs

a disintegrin and metalloproteinase with thrombospondin motifs

Arid5b

AT-rich interactive domain 5b

BMP-2

bone morphogenetic protein 2

COL-1

type I collagen

COL-2

type II collagen

COX-2

cyclooxygenase 2

CRTL-1

cartilage link protein 1

DNMTs

DNA methyltransferases

ECM

extracellular matrix

ERK

extracellular regulated protein kinases

ESET

an ERG-associated protein with a SET domain

5-AzaC

5-azacytidine

5-hmC

5-hydroxymethylcytosine

HDACs

histones deacetylases

H3K4

histone H3 lysine 4

IHH

Indian hedgehog

IL-1β

interleukin-1 beta

iNOS

inducible nitric oxide synthase

Jmjc

Jumonji C

Jhdm2a

Jumonji C containing histone demethylase-2a

LncRNAs

long noncoding RNAs

LncRNA-CIR

cartilage injury-related lncRNA

Lsd1

Lysine-specific demethylase-1

MMP-13

matrix metalloproteinase-13

MeCP2

methyl CpG binding protein 2

miRNAs

microRNAs

MSCs

mesenchymal stem cells

NAD

nicotinamide adenine dinucleotide

NFAT

nuclear factor of activated T-cells

OA

osteoarthritis

PTHrP

parathyroid hormone related peptide

RUNX-2

runt-related transcription factor 2

SIR2

silent information regulator 2

SOX-9

sex determining region Y box gene 9

2-D

2-dimension

TET

ten-eleven translocation

3’-UTR

three prime untranslated region

TNF-α

tumor necrosis factor alpha