FEBS Letters 589 (2015) 1040–1047

journal homepage: www.FEBSLetters .org

MiR-193b regulates early chondrogenesis by inhibiting the TGF-beta2signaling pathway

http://dx.doi.org/10.1016/j.febslet.2015.02.0170014-5793/� 2015 Published by Elsevier B.V. on behalf of the Federation of European Biochemical Societies.

Author contributions: Conception and experiment design: Changhe Hou, ZhiqiZhang, Weiming Liao. Experiment implementation: Changhe Hou, Zibo Yang, ZhiqiZhang, Yan Kang. Collection and assembly of data: Changhe Hou, Zibo Yang, MingFu, Aishan He, Ziji Zhang. Analysis and interpretation of the data: Ming Fu, AishanHe, Ziji Zhang. Drafting of the article: Changhe Hou, Zibo Yang. Critical revision ofthe article for important intellectual content: Zhiqi Zhang, Weiming Liao. Finalapproval of the article: Changhe Hou, Zibo Yang, Zhiqi Zhang, Weiming Liao.⇑ Corresponding authors at: Department of Joint Surgery, First Affiliated Hospital

of Sun Yat-sen University, No. 58, Zhongshan No. 2 Road, Yuexiu District,Guangzhou, Guangdong, China. Fax: +86 20 87332150.

E-mail addresses: zhzhiqi@163.com (Z. Zhang), lwmjoint@163.com (W. Liao).1 These authors contributed equally to this study.

Changhe Hou 1, Zibo Yang 1, Yan Kang, Ziji Zhang, Ming Fu, Aishan He, Zhiqi Zhang ⇑, Weiming Liao ⇑Joint Department, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, Guangdong 510080, China

a r t i c l e i n f o

Article history:Received 16 December 2014Revised 30 January 2015Accepted 18 February 2015Available online 26 February 2015

Edited by Tamas Dalmay

Keywords:ChondrogenesisOsteoarthritisMiR-193bTGF-betaTNF-alpha

a b s t r a c t

Cartilage generation and degradation are regulated by miRNAs. Our previous study has shownaltered expression of miR-193b in chondrogenic human adipose-derived mesenchymal stem cells(hADSCs). In the current study, we investigated the role of miR-193b in chondrogenesis and cartilagedegradation. Luciferase reporter assays showed that miR-193b targeted seed sequences of the TGFB2and TGFBR3 30-UTRs. MiR-193b suppressed the expression of early chondrogenic markers in chon-drogenic ATDC5 cells, and TNF-alpha expression in IL-1b-induced PMCs. In conclusion, MiR-193bmay inhibit early chondrogenesis by targeting TGFB2 and TGFBR3, and may regulate inflammationby repressing TNF-alpha expression in inflamed chondrocytes.

� 2015 Published by Elsevier B.V. on behalf of the Federation of European Biochemical Societies.

1. Introduction

Cartilage tissues are degraded and destroyed during the processof osteoarthritis and cartilage is unable to recover on its own.Thus, cure or prevention of osteoarthritis requires an understand-ing of the molecular mechanisms of cartilage generation anddegradation.

Multiple factors and signaling pathways regulating cartilagegeneration and degradation have been identified [1,2]. Chondro-genesis is usually divided into two stages: early chondrogenesis,characterized by upregulated collagen 2 (col2a1) and sox9, andhypertrophy, characterized by upregulated collagen 10 (col10a1)and runx2. These two stages are regulated by multiple factorsand signaling pathways [3], and they function to repress each other[4–7]. Among these signaling pathways, the TGF-beta pathway

functions to enhance early chondrogenesis, repressing hypertro-phy and degradation of cartilage [8–10]. Many stimuli activatesTGF-beta signaling [11,12] and activate smad2 and smad3 phos-phorylation [10]. Phosphorylated smad3 induces chondrogenesisthrough activation of sox9 [13], which is the pivotal factor in earlychondrogenesis [14]. TGF-beta receptor 3 (TGFBR3) is anotherenhancer of the TGF-beta pathway. Although it does not directlyactivate downstream phosphorylation of smads after ligandbinding, it functions to present ligands to TGF-beta receptors 1and 2, and increase the affinity of TGF-beta receptors 1 and 2 tothe ligand [15].

The microRNAs (miRNAs, miRs) are a class of small, non-coding,single-stranded RNAs identified as important post-transcriptionalregulators. The miRNAs either repress the translation of mRNA oftarget genes or induce mRNA degradation, usually by thecomplementary base pairing of mature miRNAs and the 30-UTR(untranslated region) of the target gene’s mRNA. MiRNAs havebeen reported to have important roles in multiple biological pro-cesses, including differentiation, development, proliferation, andtumorigenesis.

Several miRNAs was confirmed to participate in the processes ofcartilage differentiation and degradation [16–18], and these miR-NAs were identified as either enhancing or inhibiting chondro-genesis. MiR-365 has been shown to enhance chondrogenesis bytargeting HDAC4 [19], and miR-145 has been shown to inhibit ear-ly chondrogenesis by targeting sox9 [20]. Other miRNAs regulatechondrogenesis by targeting components of the TGF-beta signaling

C. Hou et al. / FEBS Letters 589 (2015) 1040–1047 1041

pathway. For example, miR-337 regulates chondrogenesis by tar-geting TGF-beta receptor 2 [21], and miR-455 regulates chondro-genesis by targeting ACVR2B, SMAD2, and CHRDL1, all of whichbelong to the TGF-beta pathway [22]. However, the roles of addi-tional miRNAs in chondrogenesis require clarification.

Previously, we found that miR-193b expression changed sig-nificantly during chondrogenesis of human mesenchymal stemcells. We also predicted the possible target genes of miR-193busing prediction softwares and the literatures [23]. Based on ourprevious work, we hypothesized that miR-193b may contributeto chondrogenesis and osteoarthritis. In this work, we continueto investigate the specific biological functions and mechanisms ofmiR-193b.

2. Materials and methods

2.1. hADSCs culture and chondrogenesis

The hADSCs (human adipose-derived mesenchymal stem cells)were cultured for expansion, as described in Supplementarymaterials.

The hADSCs were harvested and resuspended in incompletemesenchymal stem cell chondrogenic differentiation medium(Cyagen, HUXMA-03041-194, China) at 2 � 107 cells/ml. Droplets(12.5 ll) were carefully placed in each well of a 24-well plate. Cellswere allowed to adhere at 37 �C for 90 min, followed by the addi-tion of 500 ll chondrogenic differentiation medium [23–26]. Thechondrogenic medium was replaced every 3 days.

2.2. ATDC5 cell culture and chondrogenesis

ATDC5 cells were cultured and induced to chondrogenesis asdescribed in Supplementary materials.

2.3. Primary chondrocyte isolation and culture

Mice were maintained in accordance with the guidelines of theAnimal Center of First Affiliated Hospital of Sun Yat-Sen University.Isolation of primary mouse chondrocytes (PMCs) was performed aspreviously described [27]. The method of mice chondrocyte isola-tion was in Supplementary materials.

Primary mouse chondrocytes were cultured to allow expansionin M199 (Gibco, 11150-059, USA) plus 10% FBS, 1% penicillin andstreptomycin, bFGF (Peprotech, 450-33, USA), EGF (Peprotech,315-09, USA), insulin (Sigma, 11070-73-8, USA), in a 37 �C and5% CO2 humid atmosphere.

The procedures followed were in accordance with the ethicalstandards of the ethical committee on human experimentation(First Affiliated Hospital of Sun Yat-Sen University, China) and withthe Helsinki Declaration of 1975, as revised in 2000. After informedconsent, primary human chondrocytes (PHCs) were isolated fromthe cartilage of patients suffering a fracture of the hip or knee jointduring operation. Patients with degraded cartilages, local or sys-temic immunological disorders, or tumors were excluded. Themethod of isolation was described in Supplementary materials.

Primary human chondrocytes were cultured in DMEM/F12 plus5% FBS and 1% penicillin and streptomycin, in a 37 �C and 5% CO2

humid atmosphere.

2.4. Interleukin-1beta-induced chondrocytes

Before stimulation with interleukin-1beta (IL-1b), ATDC5 cellswere incubated in chondrogenic medium with ITS+ Premix for14 days. Chondrogenic ATDC5 cells, PHCs and PMCs (after no morethan 4 passages) were treated with the indicated doses of

recombinant IL-1b (Peprotech, 200-01B, USA) for the indicatedtimes [28–30].

2.5. Staining and microscopy

The micromass was harvested at the indicated time points. Themacro-morphology was examined by photography (Leica, M205FA,Switzerland). The pathology was examined using formalin fixation,paraffin embedding method, and alcian blue staining [26].

Cultured ATDC5 cells were fixed in formalin for 4 h at roomtemperature, and then stained with alcian blue 8GX for 20 min atroom temperature. Stained cells were examined and imaged usinga microscopy (ZEISS, Axio Imager Z1, German).

2.6. RNA extraction, reverse transcription and qPCR

RNA extraction, reverse transcription and qPCR were performedas described in Supplementary materials. The fold differences ingene expression were normalized to RNU6, HPRT and GAPDH formiRNA and mRNA [31–33], and then were calculated using theDDCt method.

2.7. Protein extraction and Western blot assays

Protein extraction and Western blot were performed asdescribed in Supplementary materials.

2.8. ELISA

The supernatant of cultured cells was collected and kept frozenuntil use. The supernatant was centrifuged to remove theprecipitation. The concentration of TNF-alpha protein in the super-natant was measured by following the manufacturer’s instructions(Biotechnology, F1163).

2.9. Construction of eukaryotic expression plasmid

Mouse eukaryotic expression GV230 plasmids were purchasedfrom GeneChem (Supplementary materials, Shanghai, China).Briefly, the plasmids were constructed following standard DNAtechniques (Supplementary materials). Then, the plasmids wereverified by sequencing.

2.10. Transfection assays

The transfection assays were performed as described in Supple-mentary materials.

2.11. Luciferase reporter assay

The potential binding sites of miRNAs were predicted usingonline software (miRanda, miRDB, and TargetScan). The primersfor the gene cloning and DNA techniques were in Supplementarymaterials.

The cells were seeded into 96-well plates (1.2 � 104 cells, 100 llculture medium per well) and cultured for 24 h. Next, 5 pmolmmu-miR-193b mimic and 100 ng vector were cotransfected intoeach well. At 6 h after transfection, 100 ll culture medium wasadded to each well, and cells were cultured for a further 48 h.The luciferase reporter assay was then performed following themanufacturer’s instructions (Dual-Glo� Luciferase Assay System,E2940, Promega, USA). Luminescence was measured using a lumi-nometer (Veritas, 9100-002, USA).

1042 C. Hou et al. / FEBS Letters 589 (2015) 1040–1047

2.12. Statistical analysis

For each experiment, three completely independent trials wereanalyzed (n = 3). Error bars indicate the confidence interval (CI).Differences were analyzed using the analysis of variance (ANOVA)among groups or Student’s t-test between groups using SPSS 13.0.To test the normal distribution of data and the homogeneity ofvariances, Gaussian distribution and the homogeneity of varianceswere examined with the Shapiro–Wilk test and Levene’s testbefore applying the t-test or ANOVA, respectively. All of the datapassed the normal distribution test, but some of the data failedthe test of homogeneity of variances. Multiple comparisons inANOVA were performed by using least significant differences(LSD) if the variances were homogenous. Tamhane’s T2 was usedif the variances were not homogenous. For Student’s t test, the tvalue was adjusted if the variances were not homogenous. A P val-ue < 0.05 was regarded as statistically significant.

3. Results

3.1. Elevated expression of miR-193b in chondrogenic hADSCs, ATDC5cells and IL-1b-induced chondrocytes

Cultured hADSCs are thin and spindle-shaped during expansion(Fig. 1A) [26]. The micromass made from hADSCs in chondrogenicmedium became spherical, dense, hard, and non-transparent onthe 7th day (Fig. 1B). Paraffin-embedded sections of the micromasswere stained with alcian blue (Fig. 1C). Expression of miR-193band col2a1 increased in chondrogenic hADSCs (Fig. 1D).

We examined the expression of miR-193b in the chondrogenicATDC5 cells. Alcian blue staining confirmed the chondrogenic dif-ferentiation by revealing the matrix nodules (Fig. 2A and B). Weobserved an obvious elevation of miR-193b expression and theexpression of chondrocyte-specific markers in chondrogenicATDC5 cells (Fig. 2C).

We also examined the expression of miR-193b in the IL-1b-in-duced chondrocytes. Significantly increased expression of miR-193b was detected in PMCs, PHCs, and chondrogenic ATDC5 cellstreated with IL-1b (Fig. 3A–C).

Fig. 1. Cultured hADSCs, micromasses of chondrogenic hADSCs, and miR-193b expression100 u/ml penicillin, and 100 mg/ml streptomycin. The hADSCs showed a thin, spindle-likhADSCs and maintained in chondrogenic medium for 7 days. Micromasses became spextracellular matrix for 7 days (40� amplification). (C) After 7 days of chondrogenesis, mstaining (100� amplification). (D) RNA was isolated from micromasses after 7 or 14 daysusing qPCR.

3.2. MiR-193b contributed to the inhibition of early chondrogenicdifferentiation

To investigate the effect of miR-193b on chondrogenesis, wemodulated the effects of miR-193b using an miRNA mimic andan inhibitor. The transfection was highly efficient. The miR-193bmimic dramatically decreased the expression of early chondro-genic markers in a dose-dependent manner, including col2a1,comp, and sox9, and also decreased the levels of TGFB2 andTGFBR3 (Fig. 4A). Conversely, the miR-193b inhibitor dose-depen-dently increased the expression of early chondrogenic markers andof TGFB2 and TGFBR3 (Fig. 4B).

3.3. MiR-193b suppressed TNF-alpha expression in IL-1b treatedmouse primary chondrocytes but had no effect on mmp13 expression

We further examined the role of miR-193b in the IL-1b-inducedinflammatory chondrocytes. Surprisingly, we did not find a sig-nificant change in mmp13 expression in ATDC5 cells with eitherthe miR-193b mimic or the inhibitor. However, we observed sup-pressed mRNA and protein expression of TNF-alpha with themiR-193b mimic and elevated expression of TNF-alpha with themiR-193b inhibitor in mouse chondrocytes (Fig. 5A–C).

3.4. MiR-193b expression was not affected by knockout/overexpressionof sox9/runx2

We further investigated the upstream regulation of miR-193b inchondrogenic ATDC5 cells. The expression of sox9/runx2 wasmanipulated by siRNA and overexpression plasmids. We did notobserve an obvious change in miR-193b expression in ATDC5 cellswith altered sox9/runx2 expression (Supplementary materials).

3.5. TGFB2 and TGFBR3 are target genes of miR-193b in ATDC5 cellchondrogenesis

TGF-beta2 (TGFB2) and TGFBR3 were predicted to be targetgenes of miR-193b (Fig. 6A). To validate TGFB2 as a potential targetgene, we cotransfected cells with mimic or NC (mimic or negative

. (A) Third-generation hADSCs were cultured in DMEM plus 10% fetal bovine serum,e shape at 200� amplification. (B) Micromasses were made from 12.5 ll droplets ofherical, hard, dense, and non-transparent after chondrogenesis and secretion of

icromasses were embedded in paraffin and cut into sections, followed by alcian blueof chondrogenesis, and expression levels of miR-193b and col2a1 were determined

Fig. 2. Expression of miR-193b in chondrogenic ATDC5 cells. Mouse ATDC5 cellswere cultured in medium with or without ITS+ Premix for the indicated periods. (Aand B) Cells were fixed and stained with alcian blue to verify chondrogenesis. After14 days of treatment with ITS+ Premix, ATDC5 cells were highly stained comparedto the negative control group (without ITS+ Premix). (C) Total RNA was extractedand reverse-transcribed into cDNA. Expression levels of miR-193b and thechondrogenic markers col2a1, sox9 and comp were measured using qPCR andcalculated using the DDCt method. Expression of mmu-miR-193b was significantlyupregulated in ATDC5 cells with ITS+ Premix compared to the negative control(without ITS+ Premix).

Fig. 3. Expression of miR-193b in PMCs, PHCs and chondrogenic ATDC5 cells withIL-1b. Cultured PMCs, PHCs and chondrogenic ATDC5 cells (induced with ITS+Premix for 14 days) were stimulated with 1 ng/ml IL-1b. Total RNA was extractedand reverse-transcribed into cDNA. The expression of miR-193b was measuredusing qPCR and calculated using the DDCt method. (A) Expression of miR-193b wassignificantly upregulated in PMCs after 4 h of treatment. (B) miR-193b wasupregulated in PHCs after 4 h of treatment. The relative expression level ofmmp13 cannot be calculated because mmp13 could not be detected in PHCswithout IL-1b after 40 cycles. (C) Compared to untreated cells, ATDC5 cells treatedwith IL-1b showed upregulation of miR-193b in a time-dependent manner.

C. Hou et al. / FEBS Letters 589 (2015) 1040–1047 1043

control for the mimic) mmu-miR-193b and a luciferase reportervector containing wild-type or mutated TGFB2/TGFBR3 30-UTR.Cells cotransfected with the miR-193b mimic and the vector forthe wild-type 30-UTR of either TGFB2 or TGFBR3 showed decreasesof more than 30% in cell luminosity compared to those cotransfect-ed with mimic NC and wild-type vector (Fig. 6B and C). The muta-tion of the predicted seed sequence in the 30-UTR of either TGFB2or TGFBR3 partially restored the luminosity (Fig. 6B and C). Next,we added recombinant TGFB2 to the culture medium of ATDC5cells transfected with the miR-193b mimic to restore the activationof the TGFB2 signaling pathway. Administration of TGFB2 to theculture medium restored the anti-chondrogenic effect of themiR-193b mimic in a dose-dependent manner (Fig. 6D).

3.6. MiR-193b repressed the expression of TGFB2 and TGFBR3 and thephosphorylation of smad3

We further investigated the time-dependent relationshipsamong the chondrogenic markers col2a1 and col10a1, miR-193b,and TGFB2 and TGFBR3 in chondrogenic ATDC5 cells (Fig. 7A).

Col2a1 peaked at day 14, decreased at day 21, and remained lowuntil day 28. Col10a1 peaked at day 28. miR-193b was elevatedat day 14 and continued to increase until day 28. In response tothe elevation of miR-193b, TGFBR3 decreased at day 28, and TGFB2decreased at day 21 but then increased slightly at day 28. We spec-ulate that there may be other factors regulating TGFB2 and TGFBR3transcription during chondrogenesis.

We investigated the protein expression of TGFB2 and TGFBR3and the phosphorylation of smad3 in chondrogenic ATDC5 cellswith modulated miR-193b. We found that miR-193b repressedTGFB2 and TGFBR3 expression and smad3 phosphorylation(Fig. 7B). These results show that miR-193b may negatively regu-late the TGF-beta signaling pathway by repressing the proteinexpression of TGFB2 and TGFBR3.

Fig. 4. Inhibitory effect of miR-193b on early chondrogenesis of ATDC5 cells. Themimic or inhibitor was transfected into ATDC5 cells with Lipofectamine 2000 for6 h. Next, ATDC5 cells were cultured in chondrogenic medium containing ITS+Premix for 4 days. (A) The mimic of miR-193b inhibited the mRNA expression of thechondrogenic markers col2a, comp and sox9, as well as the expression of TGFB2 andTGFBR3. (B) The inhibitor of miR-193 enhanced the mRNA expression of thesemarkers and of TGFB2 and TGFBR3.

Fig. 5. MiR-193b inhibited TNF-alpha expression in PMCs but did not affect mmp13expression. PMCs were transfected with the miR-193b mimic or inhibitor andtreated with 1 ng/ml IL-1b for 4 h. TNF-alpha and mmp13 mRNA levels weremeasured by qPCR. Additionally, the TNF-alpha protein concentration in thesupernatant was detected by ELISA. (A) miR-193b mimic inhibited the expression ofTNF-alpha mRNA but did not alter mmp13 mRNA. (B) miR-193b inhibitor enhancedthe expression of TNF-alpha mRNA but did not alter mmp13 mRNA. (C) miR-193bmimic reduced the expression of TNF-alpha protein in the supernatant. TNF-alphaprotein concentrations were undetectable in PMC culture medium with 50 nM miR-193b mimic.

1044 C. Hou et al. / FEBS Letters 589 (2015) 1040–1047

4. Discussion

According to previous studies, chondrogenic and hypertrophicphenotypes were identified at approximately 14 and 28 days inculture, respectively, in ATDC5 cells grown in chondrogenic culturemedium [34]. In this study, we found miR-193b negatively regulat-ed the early chondrogenic markers sox9, comp and col2a1, sug-gesting a negative role for miR-193b in early chondrogenesis.Hypertrophic and early chondrogenic differentiation mutuallyrepress each other, as suggested by previous studies [5–8,40,41].We speculate that miR-193b may mediate the repression of hyper-trophic differentiation on early chondrogenesis.

TGF-beta is a cytokine that is secreted into the microenviron-ments around chondrogenic ATDC5 cells and can bind to theTGF-beta receptor in an autocrine manner to regulate downstreamgene expression, including activating genes of the early chondro-genic phenotype [35]. TGFB2 has been found to positively regulatemesenchymal condensation, the early phase of chondrogenesis[35]. Phosphorylated smad3 has been shown to mediate the effectof TGF-beta on early chondrogenesis [13], and TGFBR3 acts as acoreceptor to bind and present TGF-beta ligand to the heteromericprotein kinase receptors of the TGF-beta receptor complex. There-fore, TGFBR3 can enhance the binding of TGF-beta to the receptorcomplex, thereby strengthening the TGF-beta signaling pathway[15]. In the current study, miR-193b was shown to potentially bindto the specific seed sequence of the 30-UTRs of TGFB2 and TGFBR3,inhibit the expression of TGFB2 and TGFBR3, and repress the phos-phorylation of smad3 and early chondrogenesis. Zhong Q hasshown that the TGF-beta signaling pathway was regulated bymiR-193b via the direct binding of miR-193b to smad3 mRNA[36]. These data and Zhong Q’s study suggest that miR-193b may

inhibit early chondrogenesis in ATDC5 cells via the inactivationof the TGF-beta signaling pathway, possibly by targeting the auto-crine activation of TGFB2 and TGFBR3 in the cell membrane and ofintracellular smad3.

We observed that exogenous TGFB2 could prevent the inhibi-tion of sox9 by miR-193b during chondrogenesis. Sox9 is consid-ered a pivotal transcriptional factor in chondrogenesis [9]. Theseresults further suggest that miR-193b can repress chondrogenesisby inhibiting the TGF-beta pathway. However, because the TGFB2signaling pathway was shown to positively regulate chondrogene-sis [37–39], this apparent de-repressive effect could also beachieved by the positive effect of exogenous TGFB2 on ATDC5 cellchondrogenesis, which counteracted the repression of chondro-genesis by miR-193b.

Previously, miR-193b has been shown to inhibit tumor cell pro-liferation [42–46]. One of these studies found approximately 30%

Fig. 6. MiR-193b inhibited early chondrogenesis in ATDC5 cells by targeting TGFB2and TGFBR3. (A) TGFB2 and TGFBR3 were predicted potential target genes of miR-193b. (B and C) Cells seeded into 96-well plates were cotransfected with 100 ngpmiR-RB-TGFB2/TGFBR3-30-UTRwildtype/mutation (pmiR-RB-TGFB2-30-UTRwildtype: Tgf-b2-WT; pmiR-RB-TGFB2-30-UTRmutation: Tgfb2-Mut; pmiR-RB-TGFBR3-30-UTRwild-

type: Tgfbr3-WT; pmiR-RB-TGFBR3-30-UTRmutation: Tgfbr3-Mut) and 5 pmol miR-193b mimic or NC (miR-193b mimic: mmu-miR-193b; negative control for miRmimic: NC) per well and were cultured for 24 h. Next, the luminescence wasmeasured. (B) The luminescence decreased by more than 30% in cells cotransfectedwith Tgfb2-WT and mmu-miR-193b compared to cells cotransfected with Tgfb2-WT and NC. The luminescence was significantly restored in cells cotransfected withTgfb2-Mut and mmu-miR-193b compared to cells cotransfected with Tgfb2-WTand mmu-miR-193b. (C) The luminescence decreased by more than 30% in cellscotransfected with Tgfbr3-WT and mmu-miR-193b compared to cells cotransfectedwith Tgfbr3-WT and NC. Luminescence was significantly restored in cells cotrans-fected with Tgfbr3-Mut and mmu-miR-193b compared to cells cotransfected withTgfbr3-WT and mmu-miR-193b. (D) ATDC5 cells were transfected with 50 nM miR-193b mimic, followed by 4 days in chondrogenesis induction medium (with ITS+Premix) and various concentrations of recombinant TGFB2. The expression of sox9was measured using qPCR. The miR-193b mimic inhibited the expression of sox9,but exogenous TGFB2 restored the inhibitory effect of the miR-193b mimic on sox9expression in a dose-dependent manner. These results indicate that miR-193b mayinhibit ATDC5 cell chondrogenesis by targeting TGFB2 and TGFBR3.

Fig. 7. The expression of TGFB2 and TGFBR3 and the phosphorylation of smad3were repressed by miR-193b. (A) Total RNA was isolated from chondrogenic ATDC5cells at the indicated time point and reverse-transcribed. mRNA expression levels ofmiR-193b, col2a1, col10a1, TGFB2 and TGFBR3 were measured using qPCR. Col2a1peaked at day 14, and col10a1 peaked at day 28, indicating early chondrogenicdifferentiation and hypertrophic differentiation at days 14 and 28, respectively.TGFB2 and TGFBR3 were downregulated when miR-193b was upregulated,although they were not strictly ‘‘mirror opposites’’. (B) Repression by miR-193bof the protein expression of TGFB2 and TGFBR and the phosphorylation of smad3.Four days after transfection, protein was extracted. Western blots were used toassess the protein expression levels of TGFB2, TGFBR3, smad3 and tubulin and thephosphorylation of smad3. MiR-193b markedly downregulated the protein expres-sion of TGFB2 and TGFBR3 and the phosphorylation of smad3. Total smad3 wasslightly downregulated by miR-193b, consistent with the previous literature [36].

C. Hou et al. / FEBS Letters 589 (2015) 1040–1047 1045

decreased luminosity in 3 M cells expressing miR-193b and thefirefly luciferase vector with the cyclin D1 30-UTR, and that worksuccessfully identified cyclin D1 as a target gene of miR-193b[44]. However, most previous studies of miR-193b did not identifythe target genes and possible mechanisms. In the present study,TGFB2 and TGFBR3 were identified as possible target genes ofmiR-193b. The TGF-beta signaling pathway has previously beenshown to enhance tumor cell proliferation [47,48]. We suspect thatmiR-193b may suppress tumor cell proliferation, at least partly bytargeting TGFB2 and TGFBR3.

IL-1b has been broadly used to stimulate chondrocytes in stud-ies of arthritis [49,50]. And destruction of the cartilage matrixbegins when chondrocytes are stimulated by a local network ofinflammatory cytokines, including the proinflammatory cytokinesTNF-alpha and IL-1b [51]. Anti-inflammatory medicines targetingIL-1b and TNF-alpha have been considered innovative osteoarthri-tis therapies [52]. In the present study, we found that miR-193bcould be time-dependently upregulated in cells stimulated withIL-1b and could inhibit TNF-alpha expression. Based on theseresults, miR-193b may participate in the negative regulation ofthe local inflammatory web and contribute to limiting the strengthof inflammation. Wang HJ found that miR-193b expression wasnegatively correlated to inflammation and sepsis [53]. Arner Efound that miR-193b inhibited expression of the proinflammatorycytokines CCL2 and IL-6 through targeting ETS1 and MAX [54].Because ETS1 may bind to the promoter of TNF-alpha and initiatetranscription [55], we speculate that ETS1 may mediates therepression of TNF-alpha by miR-193b in inflamed chondrocytes.CCL2 is an important chemokine in symptomatic osteoarthritis[56,57], and TNF-alpha has been shown to induce CCL2 expressionin the synovial tissue of rheumatism [58,59]. Based on these rela-

1046 C. Hou et al. / FEBS Letters 589 (2015) 1040–1047

tionships among miR-193b, CCL2, TNF-alpha and mmp13, miR-193b may indirectly protect cartilage by regulating the strengthof inflammation.

Role of the funding source

This study was supported by the National Natural Science Foun-dation of China (81171709, 81201388, and 81301558), the Doctor-al Scientific Fund Project of the Ministry of Education of China (No.20130171120074), and the Natural Science Foundation of Guang-dong Province, China (No. s2013040016269), and Guangzhou EliteProject (SuiJing [2014] 7, JY201420). The study sponsors had noinvolvement in the study design; the collection, analysis and inter-pretation of data; the writing of the manuscript; or the decision tosubmit the manuscript for publication.

Conflict of interest

All the authors declare that they do not have a potential conflictof interest or the appearance of a conflict of interest with regard tothis work.

Acknowledgments

We thank Xuerong Li, Shan Li, and Shang Mei in the Departmentof Parasitology, Zhongshan School of Medicine, Sun Yat-senUniversity, Guangzhou 510080, People’s Republic of China, fortheir technical assistance with the present study.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.febslet.2015.02.017.

References

[1] Wuelling, M. and Vortkamp, A. (2011) Chondrocyte proliferation anddifferentiation. Endocr. Dev. 21, 1–11.

[2] Augello, A. and De Bari, C. (2010) The regulation of differentiation inmesenchymal stem cells. Hum. Gene Ther. 21, 1226–1238.

[3] Shimizu, H., Yokoyama, S. and Asahara, H. (2007) Growth and differentiationof the developing limb bud from the perspective of chondrogenesis. Dev.Growth Differ. 49, 449–454.

[4] Cheng, A. and Genever, P.G. (2010) SOX9 determines RUNX2 trans activity bydirecting intracellular degradation. J. Bone Miner. Res. 25, 2680–2689.

[5] Yamashita, S., Andoh, M., Ueno-Kudoh, H., Sato, T., Miyaki, S. and Asahara, H.(2009) Sox9 directly promotes Bapx1 gene expression to repress Runx2 inchondrocytes. Exp. Cell Res. 315, 2231–2240.

[6] Chen, H., Ghori-Javed, F.Y., Rashid, H., Adhami, M.D., Serra, R., Gutierrez, S.E.,et al. (2014) Runx2 regulates endochondral ossification through control ofchondrocyte proliferation and differentiation. J. Bone Miner. Res. 29, 2653–2665.

[7] Ulrich, C., Rolauffs, B., Abele, H., Bonin, M., Nieselt, K., Hart, M.L., et al. (2013)Low osteogenic differentiation potential of placenta-derived mesenchymalstromal cells correlates with low expression of the transcription factors Runx2and Twist2. Stem Cells Dev. 22, 2859–2872.

[8] van der Kraan, P.M., Blaney, D.E., Blom, A. and van den Berg, W.B. (2009) TGF-beta signaling in chondrocyte terminal differentiation and osteoarthritis:modulation and integration of signaling pathways through receptor-Smads.Osteoarthritis Cartilage 17, 1539–1545.

[9] Kawakami, Y., Rodriguez-Leon, J. and Izpisua, B.J. (2006) The role of TGFbetasand Sox9 during limb chondrogenesis. Curr. Opin. Cell Biol. 18, 723–729.

[10] Derynck, R. and Zhang, Y.E. (2003) Smad-dependent and Smad-independentpathways in TGF-beta family signalling. Nature 425, 577–584.

[11] Furumatsu, T., Matsumoto, E., Kanazawa, T., Fujii, M., Lu, Z., Kajiki, R., et al.(2013) Tensile strain increases expression of CCN2 and COL2A1 by activatingTGF-beta-Smad2/3 pathway in chondrocytic cells. J. Biomech. 46, 1508–1515.

[12] Allen, J.L., Cooke, M.E. and Alliston, T. (2012) ECM stiffness primes the TGFbetapathway to promote chondrocyte differentiation. Mol. Biol. Cell 23, 3731–3742.

[13] Furumatsu, T., Tsuda, M., Taniguchi, N., Tajima, Y. and Asahara, H. (2005)Smad3 induces chondrogenesis through the activation of SOX9 via CREB-binding protein/p300 recruitment. J. Biol. Chem. 280, 8343–8350.

[14] Akiyama, H. (2008) Control of chondrogenesis by the transcription factor Sox9.Mod. Rheumatol. 18, 213–219.

[15] Lopez-Casillas, F., Wrana, J.L. and Massague, J. (1993) Betaglycan presentsligand to the TGF beta signaling receptor. Cell 73, 1435–1444.

[16] Kobayashi, T., Lu, J., Cobb, B.S., Rodda, S.J., McMahon, A.P., Schipani, E., et al.(2008) Dicer-dependent pathways regulate chondrocyte proliferation anddifferentiation. Proc. Natl. Acad. Sci. USA 105, 1949–1954.

[17] Dong, S., Yang, B., Guo, H. and Kang, F. (2012) MicroRNAs regulateosteogenesis and chondrogenesis. Biochem. Biophys. Res. Commun. 418,587–591.

[18] Hong, E. and Reddi, A.H. (2012) MicroRNAs in chondrogenesis, articularcartilage, and osteoarthritis: implications for tissue engineering. Tissue Eng.Part B Rev. 18, 445–453.

[19] Guan, Y.J., Yang, X., Wei, L. and Chen, Q. (2011) MiR-365: a mechanosensitivemicroRNA stimulates chondrocyte differentiation through targeting histonedeacetylase 4. FASEB J. 25, 4457–4466.

[20] Yang, B., Guo, H., Zhang, Y., Chen, L., Ying, D. and Dong, S. (2011) MicroRNA-145 regulates chondrogenic differentiation of mesenchymal stem cells bytargeting Sox9. PLoS One 6, e21679.

[21] Zhong, N., Sun, J., Min, Z., Zhao, W., Zhang, R., Wang, W., et al. (2012)MicroRNA-337 is associated with chondrogenesis through regulating TGFBR2expression. Osteoarthritis Cartilage 20, 593–602.

[22] Swingler, T.E., Wheeler, G., Carmont, V., Elliott, H.R., Barter, M.J., Abu-Elmagd,M., et al. (2012) The expression and function of microRNAs in chondrogenesisand osteoarthritis. Arthritis Rheum. 64, 1909–1919.

[23] Zhang, Z., Kang, Y., Zhang, Z., Zhang, H., Duan, X., Liu, J., et al. (2012) Expressionof microRNAs during chondrogenesis of human adipose-derived stem cells.Osteoarthritis Cartilage 20, 1638–1646.

[24] Zhang, Z.J., Zhang, H., Kang, Y., Sheng, P.Y., Ma, Y.C., Yang, Z.B., et al. (2012)miRNA expression profile during osteogenic differentiation of human adipose-derived stem cells. J. Cell. Biochem. 113, 888–898.

[25] Zhang, L., Su, P., Xu, C., Yang, J., Yu, W. and Huang, D. (2010) Chondrogenicdifferentiation of human mesenchymal stem cells: a comparison betweenmicromass and pellet culture systems. Biotechnol. Lett. 32, 1339–1346.

[26] Estes, B.T., Diekman, B.O., Gimble, J.M. and Guilak, F. (2010) Isolation ofadipose-derived stem cells and their induction to a chondrogenic phenotype.Nat. Protoc. 5, 1294–1311.

[27] Thirion, S. and Berenbaum, F. (2004) Culture and phenotyping of chondrocytesin primary culture. Methods Mol. Med. 100, 1–14.

[28] Miyaki, S., Nakasa, T., Otsuki, S., Grogan, S.P., Higashiyama, R., Inoue, A., et al.(2009) MicroRNA-140 is expressed in differentiated human articularchondrocytes and modulates interleukin-1 responses. Arthritis Rheum. 60,2723–2730.

[29] Simsa-Maziel, S. and Monsonego-Ornan, E. (2012) Interleukin-1beta promotesproliferation and inhibits differentiation of chondrocytes through amechanism involving down-regulation of FGFR-3 and p21. Endocrinology153, 2296–2310.

[30] MacRae, V.E., Farquharson, C. and Ahmed, S.F. (2006) The restricted potentialfor recovery of growth plate chondrogenesis and longitudinal bone growthfollowing exposure to pro-inflammatory cytokines. J. Endocrinol. 189, 319–328.

[31] de Araujo, M.A., Marques, T.E., Taniele-Silva, J., Souza, F.M., de Andrade, T.G.,Garcia-Cairasco, N., et al. (2014) Identification of endogenous reference genesfor the analysis of microRNA expression in the hippocampus of thepilocarpine-induced model of mesial temporal lobe epilepsy. PLoS One 9,e100529.

[32] Lin, E.A., Kong, L., Bai, X.H., Luan, Y. and Liu, C.J. (2009) MiR-199a, a bonemorphogenic protein 2-responsive MicroRNA, regulates chondrogenesis viadirect targeting to Smad1. J. Biol. Chem. 284, 11326–11335.

[33] Zhai, Z., Yao, Y. and Wang, Y. (2013) Importance of suitable reference geneselection for quantitative RT-PCR during ATDC5 cells chondrocytedifferentiation. PLoS One 8, e64786.

[34] Newton, P.T., Staines, K.A., Spevak, L., Boskey, A.L., Teixeira, C.C., Macrae, V.E.,et al. (2012) Chondrogenic ATDC5 cells: an optimised model for rapid andphysiological matrix mineralisation. Int. J. Mol. Med. 30, 1187–1193.

[35] Kawai, J., Akiyama, H., Shigeno, C., Ito, H., Konishi, J. and Nakamura, T. (1999)Effects of transforming growth factor-beta signaling on chondrogenesis inmouse chondrogenic EC cells, ATDC5. Eur. J. Cell Biol. 78, 707–714.

[36] Zhong, Q., Wang, T., Lu, P., Zhang, R., Zou, J. and Yuan, S. (2014) MiR-193bpromotes cell proliferation by targeting Smad3 in human glioma. J. Neurosci.Res. 92, 619–626.

[37] Jin, X., Sun, Y., Zhang, K., Wang, J., Shi, T., Ju, X., et al. (2007) Ectopicneocartilage formation from predifferentiated human adipose derived stemcells induced by adenoviral-mediated transfer of hTGF beta2. Biomaterials 28,2994–3003.

[38] Jin, X.B., Sun, Y.S., Zhang, K., Wang, J., Shi, T.P., Ju, X.D., et al. (2008) Tissueengineered cartilage from hTGF beta2 transduced human adipose derivedstem cells seeded in PLGA/alginate compound in vitro and in vivo. J. Biomed.Mater. Res. A 86, 1077–1087.

[39] Kim, H.J. and Im, G.I. (2009) Combination of transforming growth factor-beta2and bone morphogenetic protein 7 enhances chondrogenesis from adiposetissue-derived mesenchymal stem cells. Tissue Eng. Part A 15, 1543–1551.

[40] Chen, C.G., Thuillier, D., Chin, E.N. and Alliston, T. (2012) Chondrocyte-intrinsicSmad3 represses Runx2-inducible matrix metalloproteinase 13 expression tomaintain articular cartilage and prevent osteoarthritis. Arthritis Rheum. 64,3278–3289.

C. Hou et al. / FEBS Letters 589 (2015) 1040–1047 1047

[41] Zhou, G., Zheng, Q., Engin, F., Munivez, E., Chen, Y., Sebald, E., et al. (2006)Dominance of SOX9 function over RUNX2 during skeletogenesis. Proc. Natl.Acad. Sci. USA 103, 19004–19009.

[42] Hu, H., Li, S., Liu, J. and Ni, B. (2012) MicroRNA-193b modulates proliferation,migration, and invasion of non-small cell lung cancer cells. Acta Biochim.Biophys. Sin. (Shanghai) 44, 424–430.

[43] Gao, X.N., Lin, J., Gao, L., Li, Y.H., Wang, L.L. and Yu, L. (2011) MicroRNA-193bregulates c-Kit proto-oncogene and represses cell proliferation in acutemyeloid leukemia. Leuk. Res. 35, 1226–1232.

[44] Chen, J., Feilotter, H.E., Pare, G.C., Zhang, X., Pemberton, J.G., Garady, C., et al.(2010) MicroRNA-193b represses cell proliferation and regulates cyclin D1 inmelanoma. Am. J. Pathol. 176, 2520–2529.

[45] Rauhala, H.E., Jalava, S.E., Isotalo, J., Bracken, H., Lehmusvaara, S., Tammela,T.L., et al. (2010) MiR-193b is an epigenetically regulated putative tumorsuppressor in prostate cancer. Int. J. Cancer 127, 1363–1372.

[46] Unno, K., Zhou, Y., Zimmerman, T., Platanias, L.C. and Wickrema, A. (2009)Identification of a novel microRNA cluster miR-193b-365 in multiplemyeloma. Leuk. Lymphoma 50, 1865–1871.

[47] Hilbig, A. and Oettle, H. (2011) Transforming growth factor beta in pancreaticcancer. Curr. Pharm. Biotechnol. 12, 2158–2164.

[48] Govinden, R. and Bhoola, K.D. (2003) Genealogy, expression, and cellularfunction of transforming growth factor-beta. Pharmacol. Ther. 98, 257–265.

[49] Goldring, M.B. and Otero, M. (2011) Inflammation in osteoarthritis. Curr. Opin.Rheumatol. 23, 471–478.

[50] Wang, M., Shen, J., Jin, H., Im, H.J., Sandy, J. and Chen, D. (2011) Recent progressin understanding molecular mechanisms of cartilage degeneration duringosteoarthritis. Ann. NY Acad. Sci. 1240, 61–69.

[51] Bian, Q., Wang, Y.J., Liu, S.F. and Li, Y.P. (2012) Osteoarthritis: genetic factors,animal models, mechanisms, and therapies. Front. Biosci. (Elite Ed) 4, 74–100.

[52] Calich, A.L., Domiciano, D.S. and Fuller, R. (2010) Osteoarthritis: can anti-cytokine therapy play a role in treatment? Clin. Rheumatol. 29, 451–455.

[53] Wang, H.J., Zhang, P.J., Chen, W.J., Feng, D., Jia, Y.H. and Xie, L.X. (2012) Fourserum microRNAs identified as diagnostic biomarkers of sepsis. J. TraumaAcute Care Surg. 73, 850–854.

[54] Arner, E., Mejhert, N., Kulyte, A., Balwierz, P.J., Pachkov, M., Cormont, M., et al.(2012) Adipose tissue microRNAs as regulators of CCL2 production in humanobesity. Diabetes 61, 1986–1993.

[55] Suriano, A.R., Sanford, A.N., Kim, N., Oh, M., Kennedy, S., Henderson, M.J., et al.(2005) GCF2/LRRFIP1 represses tumor necrosis factor alpha expression. Mol.Cell. Biol. 25, 9073–9081.

[56] Li, L. and Jiang, B.E. (2014) Serum and synovial fluid chemokine ligand2/monocyte chemoattractant protein 1 concentrations correlates withsymptomatic severity in patients with knee osteoarthritis. Ann. Clin. Biochem.

[57] Miller, R.E., Tran, P.B., Das, R., Ghoreishi-Haack, N., Ren, D., Miller, R.J., et al.(2012) CCR2 chemokine receptor signaling mediates pain in experimentalosteoarthritis. Proc. Natl. Acad. Sci. USA 109, 20602–20607.

[58] Harigai, M., Hara, M., Yoshimura, T., Leonard, E.J., Inoue, K. and Kashiwazaki, S.(1993) Monocyte chemoattractant protein-1 (MCP-1) in inflammatory jointdiseases and its involvement in the cytokine network of rheumatoidsynovium. Clin. Immunol. Immunopathol. 69, 83–91.

[59] Tsou, H.K., Chen, H.T., Chang, C.H., Yang, W.Y. and Tang, C.H. (2012) Apoptosissignal-regulating kinase 1 is mediated in TNF-alpha-induced CCL2 expressionin human synovial fibroblasts. J. Cell. Biochem. 113, 3509–3519.