The Expression and Function of MicroRNAs inChondrogenesis and Osteoarthritis
Tracey E. Swingler,1 Guy Wheeler,1 Virginia Carmont,1 Hannah R. Elliott,2 Matthew J. Barter,2
Muhammad Abu-Elmagd,1 Simon T. Donell,3 Raymond P. Boot-Handford,4
Mohammad K. Hajihosseini,1 Andrea Munsterberg,1 Tamas Dalmay,1
David A. Young,2 and Ian M. Clark1
Objective. To use an in vitro model of chondro-genesis to identify microRNAs (miRNAs) with a func-tional role in cartilage homeostasis.
Methods. The expression of miRNAs was mea-sured in the ATDC5 cell model of chondrogenesis usingmicroarray and was verified using quantitative reversetranscription–polymerase chain reaction. MicroRNAexpression was localized by in situ hybridization. Pre-dicted miRNA target genes were validated using 3�-untranslated region-Luc reporter plasmids containingeither wild-type sequences or mutants of the miRNAtarget sequence. Signaling through the Smad pathwaywas measured using a (CAGA)12-Luc reporter.
Results. The expression of several miRNAs wasregulated during chondrogenesis. These included 39miRNAs that are coexpressed with miRNA-140 (miR-140), which is known to be involved in cartilage homeo-stasis and osteoarthritis (OA). Of these miRNAs, miR-455 resides within an intron of COL27A1 that encodes acartilage collagen. When human OA cartilage was com-
pared with cartilage obtained from patients with femo-ral neck fractures, the expression of both miR-140-5pand miR-455-3p was increased in OA cartilage. In situhybridization showed miR-455-3p expression in thedeveloping limbs of chicks and mice and in human OAcartilage. The expression of miR-455-3p was regulatedby transforming growth factor � (TGF�) ligands, andmiRNA regulated TGF� signaling. ACVR2B, SMAD2,and CHRDL1 were direct targets of miR-455-3p andmay mediate its functional impact on TGF� signaling.
Conclusion. MicroRNA-455 is expressed duringchondrogenesis and in adult articular cartilage, where itcan regulate TGF� signaling, suppressing the Smad2/3pathway. Diminished signaling through this pathwayduring the aging process and in OA chondrocytes isknown to contribute to cartilage destruction. We pro-pose that the increased expression of miR-455 in OAexacerbates this process and contributes to diseasepathology.
Osteoarthritis (OA) is a degenerative joint dis-ease characterized by degradation of articular cartilage,thickening of subchondral bone, and formation of osteo-phytes (1). The etiology of OA is complex, with thecontribution of genetic, developmental, biochemical,and biomechanical factors. Chondrocytes are the onlycells in cartilage and are responsible for the synthesisand turnover of extracellular matrix (ECM), which iscrucial to tissue function.
During development, mesenchymal cells aggre-gate and differentiate into chondrocytes, which undergoa series of differentiation events: proliferation, hypertro-phy, terminal differentiation, mineralization, and pro-grammed cell death. Blood vessels penetrate the calci-fied matrix, bringing in osteoblasts that build new bone.The cartilage model grows by rounds of chondrocyte cell
Supported by Arthritis Research UK (grants 18308 and19424).
1Tracey E. Swingler, PhD, Guy Wheeler, PhD, VirginiaCarmont, BSc, Muhammad Abu-Elmagd, PhD (current address: MiniaUniversity, Minia, Egypt), Mohammad K. Hajihosseini, PhD, AndreaMunsterberg, PhD, Tamas Dalmay, PhD, Ian M. Clark, PhD: Univer-sity of East Anglia, Norwich, UK; 2Hannah R. Elliott, PhD, MatthewJ. Barter, PhD, David A. Young, PhD: Newcastle University, New-castle upon Tyne, UK; 3Simon T. Donell, MD, FRCP(Orth): Norfolkand Norwich University Hospital, Norwich, UK; 4Raymond P. Boot-Handford, PhD: University of Manchester, Manchester, UK.
Dr. Donell has received consulting fees, speaking fees, and/orhonoraria from Tornier (less than $10,000); he is a named inventor onthe patent for the Tornier unicompartmental knee replacement deviceand receives royalties for the Tornier U Kneetec knee implant.
Address correspondence to Ian M. Clark, PhD, School ofBiological Sciences, University of East Anglia, Norwich ResearchPark, Norwich NR4 7TJ, UK. E-mail: [email protected].
Submitted for publication September 2, 2011; accepted inrevised form November 29, 2011.
1909
division accompanied by the secretion of ECM in growthplates. Chondrocytes in articular cartilage are con-strained from completing this program, allowing main-tenance of a functional cartilage layer (2).
Articular chondrocytes must enact a pattern ofgene expression in order to achieve tissue homeostasis inresponse to signals from growth factors, mechanicalload, or changes to the ECM (1). Transcription profilingdemonstrates that chondrocyte gene expression is signif-icantly altered in OA (3). One facet of this aberrant geneexpression is the replay of chondrocyte differentiationwith expression of genes associated with chondrocytehypertrophy (e.g., matrix metalloproteinase 13, type Xcollagen) (2). The mechanism of transcriptional controlof chondrogenesis is known in some detail; however,mechanisms leading to altered gene expression in OAare less well understood (1,2).
Small noncoding RNAs (19–24 nucleotides long)known as microRNAs (miRNAs) are important regula-tors of gene expression (4). MicroRNAs are first tran-scribed as primary transcripts (pri-miRNA) and pro-cessed to �70-nt stem–loop structures (pre-miRNA).Pre-miRNA is processed by the ribonuclease Dicer,forming 2 short complementary RNA molecules, one ofwhich is integrated into the RNA-induced silencingcomplex (RISC), after which miRNAs base pair withtheir complementary messenger RNA (mRNA) mole-cules, usually in the 3�-untranslated region (3�-UTR)(5). In mammals, the RISC functions to suppress trans-lation, generally leading to decreased levels of steady-state mRNA.
MicroRNAs are necessary for normal skeletaldevelopment. The conditional knockout of Dicer incartilage leads to decreased chondrocyte proliferationand accelerated hypertrophy, with consequent compro-mised skeletal growth (6). A specific miRNA, miRNA-140 (miR-140), was shown to be expressed only incartilaginous tissues in developing zebrafish (7). Weinvestigated the expression and potential targets ofmouse miR-140 (8), which is specifically expressed in thecartilage tissue of mouse embryos during the develop-ment of both long and flat bones. We identified andvalidated histone deacetylase 4 (HDAC-4), a corepres-sor of RUNX-2 and myocyte-specific enhancer factor 2(MEF-2) that is essential for chondrocyte hypertrophyand bone development, as a target of miR-140. Weproposed that miR-140 functions in the developingskeleton to promote differentiation by functionally sup-pressing HDAC-4.
We also reported that miR-140 targets Cxcl12 (9)and Smad3 (10), both of which are implicated in chon-
drocyte differentiation. The function of miR-140 in vivohas been demonstrated by Miyaki et al (11), who usedtargeted deletion to create a miR-140–null mouse. Thismouse has a mild developmental phenotype in theskeleton, potentially via reduced growth plate chondro-cyte proliferation, but displays a premature OA pheno-type driven at least in part by an increase in Adamts5expression, with Adamts5 shown as a direct target ofmiR-140. A transgenic mouse overexpressing miR-140in cartilage displayed no skeletal phenotype duringdevelopment but was protected in an antigen-inducedarthritis model. Nakamura et al (12) also showed skele-tal abnormalities in a miR-140–null mouse, with accel-erated hypertrophic chondrocyte differentiation. Dnpepwas identified as a target of miR-140, with an increase inthis aspartyl aminopeptidase in the miR-140–null mouseleading to reduced bone morphogenetic protein (BMP)signaling. MicroRNA-140 thus plays a key role in carti-lage homeostasis and OA.
MicroRNA profiling in human cartilage has beenperformed, and miRNA targets have been identifiedwith relevance to OA. This has led to the identificationof miR-9 impacting on interleukin-1 (IL-1)–stimulatedmatrix metalloproteinase (MMP) expression (13) andmiR-22 as a regulator of peroxisome proliferator–activated receptor � and BMP-7 signaling (14). Studieshave also shown that miR-27a (15) and miR-27b (16)regulate MMP-13 expression in human OA chondro-cytes. MicroRNA-34a has been reported to modulatechondrocyte apoptosis (17). The profile of miRNAexpression is also altered between differentiated anddedifferentiated adult articular chondrocytes (18,19) orin mesenchymal stem cells as they differentiate intochondrocytes (20,21). Key miRNAs identified in thesesystems regulate matrix genes or signaling pathwayspertinent to OA. MicroRNA-1 regulates aggrecan ex-pression in a human chondrocyte-like cell line (22);miR-29a and miR-29b directly target COL2A1 (21),while miR-675 indirectly regulates COL2A1 expressionin articular chondrocytes (23). BMP signaling regulatesthe expression of miR-199a, which targets Smad1 andregulates early chondrogenesis by reducing the expres-sion of, for example, Col2a1 and Sox9 in a BMP-drivenmodel (24). A number of miRNAs have also beenidentified as regulators of osteoblastogenesis, includingmiR-29, miR-141, miR-200a, miR-206, miR-210, andmiR-2861 (for review, see ref. 25).
ATDC5 cells are a murine embryonic carcinomaline that can differentiate in vitro through chondrogen-esis, and regulation of known markers of the stages ofdifferentiation (e.g., type II collagen and type X colla-
1910 SWINGLER ET AL
gen) has been shown to mirror the in vivo process (26).We explored whether key miRNAs regulated in chon-drogenesis were also regulated in OA cartilage. Weprofiled the expression of miRNAs in the ATDC5 cellmodel and measured expression of key miRNAs inhuman cartilage. Using this approach, we identifiedgroups of miRNAs that may function cooperatively anddemonstrated that in common with miR-140, miR-455regulates and is regulated by Smad signaling. We hy-pothesize that these miRNAs regulate the switch fromSmad2/3 signaling to Smad1/5/8 signaling in endochon-dral ossification and contribute to the alteration oftransforming growth factor � (TGF�) signaling in OAcartilage.
MATERIALS AND METHODS
Cell culture and RNA purification. SW-1353,C3H10T1/2, and 3T3 cells were cultured in Dulbecco’s modi-fied Eagle’s medium (DMEM; Invitrogen) containing 10%(volume/volume) fetal bovine serum (Sigma), 2 mM glutamine,100 IU/ml penicillin, and 100 �g/ml streptomycin. ATDC5cells were maintained at 37°C, in an atmosphere of 5% CO2, inDMEM/Ham’s F-12 medium containing 5% (v/v) fetal calfserum, 2 mM glutamine, 100 IU/ml penicillin, 100 �g/mlstreptomycin, 5 ng/ml sodium selenite, and 10 �g/ml humantransferrin. For the assays, cells were seeded at 6 � 104/well ofa 6-well plate in the above-described medium containing 10�g/ml bovine pancreatic insulin, and the medium was changedevery second day. After 21 days, the medium was changed to�-minimal essential medium with the same supplements, andthe atmosphere was changed to 3% CO2. On day 42, the cellswere fixed in methanol and stained for detection of glycosami-noglycan, using 0.1% (weight/volume) Alcian blue in 0.1M HClovernight at room temperature. At selected time points, cellswere scraped into TRIzol reagent (Invitrogen) for RNApurification.
Collection of human cartilage and RNA purification.Human articular cartilage was obtained from the femoralheads of patients undergoing total hip replacement surgery atthe Norfolk and Norwich University Hospital. Cartilage frompatients with OA (n � 10; 5 women and 5 men, ages 37–86years) was compared with cartilage from patients with afracture to the neck of the femur (n � 10; 5 women and 5 men,ages 68–94 years). OA was diagnosed using clinical history andexamination coupled with radiographic findings; the grosspathology was confirmed at the time of joint removal. Thepatients with fractures had no known history of joint disease,and their cartilage was free of lesions; 80% of these patientsunderwent surgery within 36 hours of sustaining a fracture.This study was performed with ethics committee approval, andall patients provided informed consent. Cartilage was choppedinto 2–5-mm pieces and snap-frozen in liquid nitrogen within15–30 minutes of surgery. Cartilage was ground under liquidnitrogen using the a Spex CertiPrep 6750 Freezer Mill. RNAwas purified using MirVana (Ambion) and reverse transcribed,and miRNA expression was measured using a TaqMan low-
density array (Life Technologies) or, for miR-455-3p, theindividual assay described below.
Profiling of miRNA and mRNA expression. ForATDC5 cells, RNA samples were analyzed on an Agilent 2100bioanalyzer and a NanoDrop spectrophotometer (ThermoScientific). For each time point, RNA from 6 culture replicateswas pooled for array. Samples were labeled using themiRCURY Hy3/Hy5 power labeling kit and hybridized on amiRCURY LNA Array v.10.0 (Exiqon). Signal was correctedfor background and normalized using the global lowess regres-sion algorithm. For transcriptomic analysis, samples werehybridized on an Illumina MouseWG-6 whole genome array(Cambridge Genomic Services). Signal was corrected for back-ground and normalized by quantile normalization using the Rpackage lumi.
Quantitative reverse transcription–polymerase chainreaction (RT-PCR). Complementary DNA was synthesizedfrom RNA using SuperScript II reverse transcriptase (Invitro-gen) and either random hexamers or miRNA-specific primersaccording to the manufacturer’s instructions. ComplementaryDNA was stored at �20°C. The relative quantitation of geneexpression was performed using an ABI Prism 7700 SequenceDetection System (Applied Biosystems), following the manu-facturer’s protocol.
In situ hybridization. Whole-mount in situ hybridiza-tion of mouse embryos and isolated tissues was performed aspreviously described in (8). Embryos were treated with protei-nase K, and endogenous alkaline phosphatase activity wasblocked by pretreatment of tissues with 2 mM levamisole.Hybridizations were performed at 50°C overnight in hybridiza-tion mix containing 100 pmoles of double-labeled lockednucleic acid (LNA) oligonucleotides (Exiqon). The nitrobluetetrazolium (NBT)/BCIP staining reaction was carried out atroom temperature, after which the embryos were fixed in 4%paraformaldehyde and stored in phosphate buffered saline(PBS) at 4°C. Embryos were then blocked in 3% agar andserially sectioned (100 �m) using a Lancer series 1000 vi-bratome. Long bones from mouse embryos (stage E18) werestained in an identical manner, paraffin-embedded, and sec-tioned (10 �m). Sections were then counterstained with hema-toxylin and eosin.
Whole-mount in situ hybridization for chick embryoswas performed as previously described (27). Embryos werefixed in 4% paraformaldehyde, dehydrated in methanol, rehy-drated, and treated with proteinase K. Hybridization withdouble-labeled LNA probes (Exiqon) was performed at 50°Covernight. After NBT/BCIP color development, embryos wereembedded in OCT compound and sectioned on a cryostat.
Transient transfection. The 3�-UTR of potential targetmRNAs was amplified by PCR and subcloned into AmbionpMIR-Report vector. Mutation of the miRNA seed sequencewas achieved using QuikChange (Agilent). The positive con-trol contains a concatamer of 3 copies of the reverse comple-ment of the mature miRNA sequence downstream of theluciferase gene in pMIR-Report. SW-1353 or 3T3 cells wereplated at 2 � 104/well in a 24-well plate and grown overnight to�80% confluency. Cells were transiently transfected with 200ng of luciferase reporter plasmid, 50 ng �-galactosidase expres-sion plasmid (Promega), and 30–50 nM miRNA mimic orcontrol (AllStars; Qiagen) using Lipofectamine 2000 accordingto the manufacturer’s instructions (Invitrogen), and incubated
MicroRNAs IN CHONDROGENESIS AND OSTEOARTHRITIS 1911
for 48 hours. For growth factor induction, use of thep(CAGA)12-luc plasmid was as previously described (10). Cellswere serum starved for 24 hours posttransfection and treatedwith TGF�1 (4 ng/ml)/TGF�3 (4 ng/ml) or activin A 20 ng/ml(R&D Systems) for 3 hours. For luciferase assay, cells werewashed with ice-cold PBS, lysed in 1� Reporter Lysis Buffer,and assayed according to the manufacturer’s instructions (Pro-mega). �-galatosidase assays were performed using a Beta-Gloassay kit according to the manufacturer’s instructions (Pro-mega). Data are presented as relative light units normalized to�-galactosidase.
Cluster analysis. Hierarchical cluster analysis and vi-sualization were performed using Cluster and TreeView.
RESULTS
MicroRNA profiling during chondrogenic differ-entiation in the ATDC5 cell model. ATDC5 cells weredifferentiated over a 42-day time course, followed byAlcian blue staining for the detection of glycosaminogly-can accumulation (data not shown). The expression ofall mRNAs was profiled using an Illumina microarray inpooled RNA samples from 6 replicate wells at each timepoint. Known markers of chondrocyte differentiationwere appropriately regulated with an early increase inCol2a1 expression and a later increase in Col10a1 ex-pression (data not shown).
Our group previously demonstrated that miR-140-5p, which is selectively expressed in cartilage, targetsat least HDAC-4, CXCL12, and Smad3, all of which areimplicated in chondrocyte differentiation (8–10). Wethus profiled the expression of all miRNAs in the samepooled RNA samples as those described above. Figure 1shows the average expression of miRNAs in 7 individualgroups, based on two-way unsupervised hierarchicalclustering of the data, which demonstrated regulation ofmiRNAs during chondrocyte differentiation. Groups 1aand 1b are subclades but clearly have different expres-sion patterns; the results in groups 3a and 3b weresimilar.
Thirty-nine miRNAs grouped with a pattern ofexpression similar to that of miR-140, although severalof these were within genomic clusters and were poten-tially coregulated, collapsing this group to include 23miRNA loci (Figure 1, group 3b). Expression of thesemiRNAs increased across the time course of differenti-ation. The expression of miR-140-5p and miR-455-3p(miR-455*, but shown to be the guide strand on www.miRBase.org) was validated using quantitative RT-PCRin triplicate samples from each time point (Figures 2Aand B).
Figure 1. Expression of microRNAs in the ATDC5 cell model of chondrogenesis. For each time point, RNA from 6 culture replicates was pooled,labeled, and hybridized. The experiment used a dual-label approach comparing each test sample with a common reference sample. Signal wascorrected for background and normalized using the global lowess regression algorithm. Hierarchical cluster analysis and visualization wereperformed to generate heatmaps. The average expression of microRNAs in each cluster is plotted.
1912 SWINGLER ET AL
For the miR-140–containing group, 7 miRNAsare located within the introns of protein-coding genes(miR-99a, ENSMUST0000114231; miR-140, Wwp2;miR-149, Gpc1; miR-338, Aatk; miR-455, Col27a1; miR-466/467, Sfmbt2; miR-676, Eda), with the remainder inintergenic regions. Of these genes, the expression ofCol27a1 (containing miR-455), Wwp2 (containing miR-140), and Gpc1 (containing miR-149) was clearly regu-lated across the ATDC5 cell model in the parallelmRNA microarray experiment, while Aatk was ex-pressed but not regulated (data not shown). Sfmtb2 andEda were not detected by the probes on the array, andENSMUST00000114231 was not on the array. The ex-pression of Wwp2 and Col27a1 was validated by quanti-tative RT-PCR (Figures 2C and D), showing the generalcoregulation of the genes and miRNA. The earlierdecrease in mRNA expression compared with that ofmiRNA may reflect differences in RNA stability be-tween the the 2 RNA species.
Localization of miR-455-3p in chick and mousedevelopment. We have shown that the expression ofmiR-140 is restricted to the developing mouse skeleton(8). Because miR-455 resides in an intron of COL27A1,a collagen expressed in cartilage, we examined expres-sion of miR-455-3p in the development of chick and
mouse embryos (Figures 3A and B). In chick embryo,miR-455 was expressed in the skeleton of the developinglimbs. Expression was first detected on day 6.5(Hamburger-Hamilton stage 30; approximately equiva-lent to stage E12.5 in the mouse) with strong expressionin developing long bones and developing digits (Figure3A, parts iii–vi). Later in the course of development(days 7.5–8, approximately equivalent to stages E14.5–16.5 in the mouse), expression became more restrictedto developing joints (Figure 3A, parts vii–xii), withstaining in cartilage and perichondrium (Figure 3A,parts ix and xii). In the mouse embryo, expression indeveloping long bones was less apparent; however, in thestage E18 embryo, whole-mount staining and sectioningof isolated joints showed clear expression in the growthplate and perichondrium (Figure 3B, parts ix and x).Expression was also seen in the interdigital region(Figure 3B, parts iii–vi) and in the sutures of thedeveloping skull (Figure 3B, parts vii and viii).
Expression of miRNAs in human articular carti-lage. The miRNA fraction was purified from humanarticular cartilage obtained during total hip replacementfor either OA or fracture to the neck of the femur.Similar samples have previously been used and validatedas controls in profiling studies (29,30). Measurement of
Figure 2. Expression of microRNA-455-3p (miR-455-3p), miR-140-5p, Col27a1, and Wwp2 in the ATDC5 cell model. Microarray data werevalidated using quantitative reverse transcription–polymerase chain reaction in the individual replicate samples (n � 6). The expression of bothmiR-455-3p (A) and miR-140-5p (B) increased over the course of chondrogenesis. Expression of the genes in which the microRNAs are encoded,Col27a1 (for miR-455) (C) and Wwp2 (for miR-140) (D), was induced earlier and returned to noninduced levels on day 42. Values are the mean �SEM.
MicroRNAs IN CHONDROGENESIS AND OSTEOARTHRITIS 1913
miR-140-5p was taken from a TaqMan low-density arrayused to profile the expression of 365 miRNAs in thesesamples (Young DA: unpublished observations), withmiR-455-3p measured using a separate TaqMan assay inthe same samples. The data were normalized using arecently described method based on the mean expres-sion value of all expressed miRNAs in a sample (31), and
the results for miR-140-5p and miR-455-3p are shown inFigure 4A. Both miRNAs were expressed at higherlevels in the OA samples compared with the fracturecontrols. We also localized expression of miR-455-3p inadult articular cartilage from OA knees using in situhybridization. As shown in Figure 4B, expression waspredominantly in the intermediate zone.
Figure 3. Expression of microRNA-455 (miR-455) during chick and mouse development. A, In situ hybridization was performed on whole-mountchick embryos (i–iii), dissected limbs (iv, v, vii, viii, x, and xi), and sections through these at different stages of development (vi, ix, and xii). Expressionwas not detected at Hamburger-Hamilton stage 21 or stage 27 (i and ii) but was seen from Hamburger-Hamilton stage 30, approximately day 6.5,onward (iii–xii). Sections through limb buds showed staining in the perichondrium (vi, ix, and xii). B, In mouse embryos (i and ii), expression wasseen in the interdigital regions (iii–vi) and in the developing joint in the growth plate and perichondrium (ix). Sectioning showed staining in andaround the cartilage (x). Staining of the newborn calvaria showed strong staining in the developing sutures (vii and viii). hl � hindlimb; fl � forelimb.
Figure 4. Expression of microRNA-455 (miR-455) in adult human articular cartilage. A, Human articular cartilage samples obtained from thefemoral heads of patients with osteoarthritis (OA; n � 10) were compared with those from patients undergoing hip replacement following fractureof the femoral neck (NOF; n � 10). RNA was purified, reverse transcribed, and assayed by quantitative reverse transcription–polymerase chainreaction for miR-455-3p and miR-140-5p. Bars represent the means. �� � P � 0.01; ��� � P � 0.001, by Mann-Whitney U test. B, In situhybridization of human articular cartilage from an OA knee was performed, showing staining predominantly in the intermediate zone. MEV � meanexpression value.
1914 SWINGLER ET AL
Figure 5. MicroRNA-455 (miR-455) regulates and is regulated by Smad2/3 signaling. A and B, Human SW-1353 chondrosarcoma cells wereserum-starved for 24 hours before the addition of transforming growth factor �1 (TGF�1) or TGF�3 (5 ng/ml) (A) or activin A (20 ng/ml) (B), andmiR-455-3p was measured by quantitative reverse transcription–polymerase chain reaction (n � 3). C and D, Cells were transfected with theSmad2/3-responsive reporter (CAGA)12-Luc in the presence of miR-140 mimic, miR-455 mimic, or scrambled control at 50 nM before the additionof TGF�1 (5 ng/ml) (C), or activin A (20 ng/ml) (D) for 6 hours (n � 3). Relative light units were normalized to �-galactosidase activity from acotransfected expression construct. Bars show the mean � SEM. � � P � 0.05; �� � P � 0.01; ��� � P � 0.001 versus time 0 (A and B) and versusscrambled (C and D), by t-test.
Figure 6. MiicroRNA-455-3p targets components of the TGF� signaling pathway. A–C, Cells (3T3) were transfected with the 3�-untranslated regionof SMAD2 (n � 18) (A), ACVR2B (n � 12) (B), or CHRDL1 (n � 18) (C) cloned into pMIR-Report with or without control small interfering RNAor miR-455 mimic and incubated for 24 hours. Relative light units were normalized to �-galactosidase activity from a cotransfected expressionconstruct. Values are the mean � SEM. D, Overview of miR-455 and miR-140 impact on TGF� signaling in cartilage. scr � scrambled (see Figure5 for other definitions).
MicroRNAs IN CHONDROGENESIS AND OSTEOARTHRITIS 1915
Regulation and function of miR-455-3p. We pre-viously showed that miR-140 regulates Smad3 expres-sion and could regulate TGF�-induced signaling (10).The expression of miR-455-3p was induced by TGF�1,TGF�3, and activin A in human SW-1353 chondrosar-coma cells (Figures 5A and B) and in murineC3H10T1/2 cells (results not shown). MicroRNA-455mimic diminished Smad-dependent signaling (inducedby treatment with either TGF�1 or activin A) to a(CAGA)12-Luc construct in a manner similar to that ofmiR-140 (Figures 5C and D), although the effect ofmiR-455 was more significant.
Using the miRNA body map (http://www.mirnabodymap.org), we predicted SMAD2, ACVR2B,CHRDL1 as targets for miR-455-3p with a potentialimpact on TGF� signaling. We validated these as directtargets of miR-455-3p by cloning the 3�-UTR down-stream of the luciferase gene in the pMIR-Report vectorand showed that a miR-455 mimic reduced luciferaseactivity, while mutation of the seed sequence for miR-455 in the 3�-UTR abolished these effects (Figures6A–C). For CHRDL1 and ACVR2B, the scrambled smallinterfering RNA showed some nonspecific effects on thewild-type construct compared with the mutant.
DISCUSSION
MicroRNA-140 is selectively expressed in carti-lage in the developing skeleton (7,8), during chondro-cyte differentiation, and in human articular cartilage(14). Here, we attempted to identify additional micro-RNAs with functions in cartilage development and OA.
MicroRNA microarrays from time points acrossATDC5 cell differentiation identified 7 clusters of co-expressed miRNAs. Groups 1 and 2 showed decreasedexpression of miRNAs in the induced cultures comparedwith control, with group 1b showing this only on days5–26. Groups 3, 4, and 5 showed increased expression ofmiRNAs in the induced cultures compared with control,with group 3a showing this increase from day 15 onward,group 3b from day 10 onward, and group 4 from day 5 today 26. Group 5 showed an alternating pattern ofexpression across the time course.
A comparison of miRNAs expressed in humanmesenchymal stem cells (MSCs) in the setting of chon-drogenic differentiation (20) or mouse MSCs differenti-ated via culture on polyhydroxyalkanoates (21) showedincomplete overlap with our data. This is difficult tointerpret, because no time course of differentiation waspresented, and each model likely measured differentfacets of chondrocyte differentiation.
A number of miRNAs regulated across theATDC5 model have been described as having a role inosteoblast differentiation (20,24,25). Although no stud-ies have compared miRNA expression during osteo-genic, adipogenic, and chondrogenic differentiation inthe same starting population of precursor cells, onestudy examined both osteogenic and adipogenic differ-entiation (32). In that study, approximately half of themiRNAs regulated during differentiation were commonto both adipogenesis and osteogenesis and may have arole in the process of differentiation per se rather thandifferentiation to a specific lineage.
Group 1a contains miR-146a, miR-155, and miR-125b, all of which are regulated by inflammation medi-ators (e.g., IL-1, tumor necrosis factor �, and lipopoly-saccharide [LPS]) and have a role in regulatinginflammation/innate immunity (33–35). MicroRNA-125b inhibits osteoblastic differentiation from mouseST2 cells (36) and is negatively regulated by BMP-2treatment in C2C12 cells (24). Group 1b contains miR-29, which promotes osteogenesis, regulating collagengenes and inhibitors of osteoblast differentiation andchondrogenesis (25). Group 2 contains miR-199a, aBMP-2–responsive miRNA that regulates chondrogen-esis via suppression of Smad1 (24).
In groups 3a and 3b, miR-466, miR-467, miR-669, and miR-297 are part of a genomic cluster and arepotentially coregulated. Mmu-miR-99a and mmu-let-7c-1 also form a cluster. Group 5 contains miR-675,which is processed from a longer noncoding RNA calledH19 and regulated by SOX9 during chondrocyte differ-entiation in vitro (23).
The expression of miRNAs in group 3b increasedwith differentiation and hypertrophy; these miRNAsincluded miR-140 (miR-140-5p), and miR-140* (miR-140-3p), the passenger strand. Although the passengerstrand is generally considered to be nonfunctional, miR-140* was reportedly induced by nicotine and targetedthe 3�-UTR of the dynamin 1 gene in the nervous system(37). MicroRNA-455-3p and miR-455-5p were also ex-pressed in this group. MicroRNA-140 and miR-455 areboth located within introns of protein-coding genes(Wwp2 and Col27a1, respectively), and these were reg-ulated across the ATDC5 cell model, with kineticssimilar to those of the miRNAs. Type XXVII collagen,the product of Col27a1, is a cartilage collagen (38).Although there are no predicted miR-455 target sites inCol27a, it is possible that miR-455 could indirectlyregulate Col27a1 expression, e.g., via effects on TGF�signaling (see below).
Whole-mount in situ hybridization showed miR-
1916 SWINGLER ET AL
455 expression in the developing long bones of chicks.With time, expression became more restricted to devel-oping joints, with expression observed in the cartilageand perichondrium. There was evidence of expression inmuscle, in line with a report that miR-455 was expressedin myotubes treated with the proinflammatory cytokineTWEAK (39). In situ hybridization in mouse embryosconfirmed expression in long bones and joints. We alsoobserved expression in the sutures of the developingskull and in the interdigital region of the developingmouse paw. These processes involve apoptosis, andmiR-455 may regulate apoptosis in these tissues.Col27a1 is expressed in the cartilage anlagens of thedeveloping skeleton, most prominently in the prolifera-tive zone of the growth plate, and in adult mice stainingis seen in articular cartilage (38,40,41). Expression ofmiR-455 has been reported during the differentiation ofbrown adipocytes (42), and miR-455 may also have arole in innate immunity, because heat-killed Candidaalbicans and LPS induced its expression in macrophages(43) in an NF-�B–dependent manner. We could notdetect induction of miR-455 following stimulation ofhuman articular chondrocytes with a variety of Toll-likereceptor ligands (data not shown).
We have previously shown that miR-140 targetsSmad3 expression and regulates Smad-dependent TGF�signaling (10). Here, we demonstrated that miR-455abrogates Smad-dependent signaling and validated 3direct targets of miR-455-3p: Smad2, activin receptor2B, and chordin-like 1. The expression of miR-455 isinduced by TGF�1, TGF�3, and activin. The Smadsignaling pathway has been shown to regulate the mat-uration of some miRNAs by the Drosha complex (44),but pri-miR-455 is also induced by TGF�1 (data notshown), and both Col27a1 and Wwp2 are induced byTGF�1 (data not shown), suggesting that transcriptionalinduction is the likely mechanism.
The ATDC5 cell model relies on insulin toinduce chondrogenesis; however, TGF� expression isregulated by and regulates differentiation (45). In thegrowth plate, TGF� signaling through Smad2/3 blockschondrocyte terminal differentiation; conversely, BMPsignaling through Smad1/5/8 promotes this process. Thecommon mediator Smad4 is required for both pathways,and where this is limiting, signaling can be regulatedthrough competition for Smad4 (46). TGF� can signalthrough Smad1/5/8 in chondrocytes via the activinreceptor–like kinase 1 (ALK-1) receptor rather thanALK-5. This has led to the elegant hypothesis that achange in the ratio of ALK-5 to ALK-1 with age shiftssignaling toward the ALK-1–mediated Smad1/5/8 path-
way, with differentiation to a catabolic phenotype con-tributing to cartilage destruction.
We therefore suggest that terminal differentia-tion is regulated by miRNAs, with miR-140 and miR-455decreasing Smad2/3 and consequently decreasing TGF�signaling and promoting Smad1/5/8-dependent BMPsignaling via increasing availability of Smad4.MicroRNA-199a* regulates early chondrogenesis by tar-geting Smad1 (24), which fits with our data showing themiR-199a* expression decreased during late chondro-genesis in the ATDC5 cell model. Similarly, miR-21 hasbeen shown to target BMP receptor type II (47), and thismiRNA is repressed during late chondrogenesis. Wehave shown that activin receptor type IIB (ACVR2B) isa direct target of miR-455-3p, and recently, miR-210 wasshown to target ACVR1B to promote osteoblastic differ-entiation (25). Activin signals via the Smad2/3 pathway,so down-regulation of these targets would decreaseSmad2/3 signaling and potentially enhance Smad1/5/8signaling. Chordin-like 1 is a BMP antagonist that, whilenot implicated in OA, has been shown to have an impacton MSC proliferation.
The expression of miR-140 and miR-455 wasincreased in OA cartilage compared with control carti-lage. This finding contrasts with published data formiR-140–null mice and human OA cartilage (11,49) andis likely explained by differences in the human cartilagesamples used in each study. In our comparison ofcartilage obtained from the hips of patients with OA andcontrol cartilage obtained from patients with a fractureof the femoral head, ADAMTS5 expression (a demon-strated target of miR-140) was always decreased in OA(29). The average expression of DNPEP (another miR-140 target identified in null mouse studies) was alsodecreased in our OA samples compared with control,although the difference was not significant (50). Thekinetics of miRNA expression across disease initiationand progression will be important.
Pursuing our hypothesis, a change in miR-140/miR-455 expression would lead to altered TGF�/activinA signaling through the Smad2/3 pathway (Figure 6D).TGF� signaling is important for the maintenance ofarticular cartilage, and decreased TGF� signalingthrough the Smad2/3 pathway leads to OA-like changesin the joint (46). This provides a potential mechanisticlink between miRNAs that regulate the Smad pathwayand the pathology of OA.
In conclusion, miRNA-455 is expressed duringchondrogenesis in adult articular cartilage and is differ-entially expressed in OA. It has the potential to alter
MicroRNAs IN CHONDROGENESIS AND OSTEOARTHRITIS 1917
We would like to thank Eran Hornstein (WeizmannInstitute of Science, Israel) for his open discussion in the earlypart of this study.
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising itcritically for important intellectual content, and all authors approvedthe final version to be published. Dr. Clark had full access to all of thedata in the study and takes responsibility for the integrity of the dataand the accuracy of the data analysis.Study conception and design. Swingler, Boot-Handford, Hajihosseini,Dalmay, Young, Clark.Acquisition of data. Swingler, Wheeler, Carmont, Elliott, Barter,Abu-Elmagd, Donell, Munsterberg, Clark.Analysis and interpretation of data. Swingler, Wheeler, Elliott, Abu-Elmagd, Donell, Munsterberg, Young, Clark.
2. Mackie EJ, Tatarczuch L, Mirams M. The skeleton: a multi-functional complex organ. The growth plate chondrocyte andendochondral ossification. J Endocrinol 2011;211:109–21.
3. Aigner T, Fundel K, Saas J, Gebhard PM, Haag J, Weiss T, et al.Large-scale gene expression profiling reveals major pathogeneticpathways of cartilage degeneration in osteoarthritis. ArthritisRheum 2006;54:3533–44.
4. Liu X, Fortin K, Mourelatos Z. MicroRNAs: biogenesis andmolecular functions. Brain Pathol 2008;18:113–21.
5. Chang TC, Mendell JT. MicroRNAs in vertebrate physiology andhuman disease. Annu Rev Genomics Hum Genet 2007;8:215–39.
6. Kobayashi T, Lu J, Cobb BS, Rodda SJ, McMahon AP, SchipaniE, et al. Dicer-dependent pathways regulate chondrocyte prolifer-ation and differentiation. Proc Natl Acad Sci U S A 2008;105:1949–54.
7. Wienholds E, Kloosterman WP, Miska E, Alvarez-Saavedra E,Berezikov E, de Bruijn E, et al. MicroRNA expression in zebrafishembryonic development. Science 2005;309:310–1.
8. Tuddenham L, Wheeler G, Ntounia-Fousara S, Waters J, Haji-hosseini MK, Clark I, et al. The cartilage specific microRNA-140targets histone deacetylase 4 in mouse cells. FEBS Lett 2006;580:4214–7.
9. Nicolas FE, Pais H, Schwach F, Lindow M, Kauppinen S, MoultonV, et al. Experimental identification of microRNA-140 targets bysilencing and overexpressing miR-140. RNA 2008;14:2513–20.
10. Pais H, Nicolas FE, Soond SM, Swingler TE, Clark IM, Chantry A,et al. Analyzing mRNA expression identifies Smad3 as amicroRNA-140 target regulated only at protein level. RNA 2010;16:489–94.
11. Miyaki S, Sato T, Inoue A, Otsuki S, Ito Y, Yokoyama S, et al.MicroRNA-140 plays dual roles in both cartilage development andhomeostasis. Genes Dev 2010;24:1173–85.
12. Nakamura Y, Inloes JB, Katagiri T, Kobayashi T. Chondrocyte-specific microRNA-140 regulates endochondral bone develop-ment and targets Dnpep to modulate bone morphogenetic proteinsignaling. Mol Cell Biol 2011;31:3019–28.
13. Jones SW, Watkins G, Le Good N, Roberts S, Murphy CL,
Brockbank SM, et al. The identification of differentially expressedmicroRNA in osteoarthritic tissue that modulate the production ofTNF-� and MMP13. Osteoarthritis Cartilage 2009;17:464–72.
14. Iliopoulos D, Malizos KN, Oikonomou P, Tsezou A. IntegrativemicroRNA and proteomic approaches identify novel osteoarthritisgenes and their collaborative metabolic and inflammatory net-works. PLoS One 2008;3:e3740.
15. Tardif G, Hum D, Pelletier JP, Duval N, Martel-Pelletier J.Regulation of the IGFBP-5 and MMP-13 genes by the microRNAsmiR-140 and miR-27a in human osteoarthritic chondrocytes. BMCMusculoskelet Disord 2009;10:148.
16. Akhtar N, Rasheed Z, Ramamurthy S, Anbazhagan AN, Voss FR,Haqqi TM. MicroRNA-27b regulates the expression of matrixmetalloproteinase 13 in human osteoarthritis chondrocytes. Ar-thritis Rheum 2010;62:1361–71.
17. Abouheif MM, Nakasa T, Shibuya H, Niimoto T, Kongcharoen-sombat W, Ochi M. Silencing microRNA-34a inhibits chondrocyteapoptosis in a rat osteoarthritis model in vitro. Rheumatology(Oxford) 2010;49:2054–60.
18. Karlsen TA, Shahdadfar A, Brinchmann JE. Human primaryarticular chondrocytes, chondroblasts-like cells, and dedifferenti-ated chondrocytes: differences in gene, microRNA, and proteinexpression and phenotype. Tissue Eng Part C Methods 2011;17:219–27.
19. Lin L, Shen Q, Zhang C, Chen L, Yu C. Assessment of theprofiling microRNA expression of differentiated and dedifferen-tiated human adult articular chondrocytes. J Orthop Res 2011;29:1578–84.
20. Sorrentino A, Ferracin M, Castelli G, Biffoni M, Tomaselli G,Baiocchi M, et al. Isolation and characterization of CD146�multipotent mesenchymal stromal cells. Exp Hematol 2008;36:1035–46.
21. Yan C, Wang Y, Shen XY, Yang G, Jian J, Wang HS, et al.MicroRNA regulation associated chondrogenesis of mouse MSCsgrown on polyhydroxyalkanoates. Biomaterials 2011;32:6435–44.
22. Sumiyoshi K, Kubota S, Ohgawara T, Kawata K, Nishida T, ShimoT, et al. Identification of miR-1 as a micro RNA that supportslate-stage differentiation of growth cartilage cells. Biochem Bio-phys Res Commun 2010;402:286–90.
23. Dudek KA, Lafont JE, Martinez-Sanchez A, Murphy CL. Type IIcollagen expression is regulated by tissue-specific miR-675 inhuman articular chondrocytes. J Biol Chem 2010;285:24381–7.
24. Lin EA, Kong L, Bai XH, Luan Y, Liu CJ. miR-199a, a bonemorphogenic protein 2-responsive microRNA, regulates chondro-genesis via direct targeting to Smad1. J Biol Chem 2009;284:11326–35.
25. Kapinas K, Delany AM. MicroRNA biogenesis and regulation ofbone remodeling. Arthritis Res Ther 2011;13:220.
27. Sweetman D, Rathjen T, Jefferson M, Wheeler G, Smith TG,Wheeler GN, et al. FGF-4 signaling is involved in mir-206expression in developing somites of chicken embryos. Dev Dyn2006;235:2185–91.
28. Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysisand display of genome-wide expression patterns. Proc Natl AcadSci U S A 1998;95:14863–8.
29. Davidson RK, Waters JG, Kevorkian L, Darrah C, Cooper A,Donell ST, et al. Expression profiling of metalloproteinases andtheir inhibitors in synovium and cartilage. Arthritis Res Ther2006;8:R124.
30. Kevorkian L, Young DA, Darrah C, Donell ST, Shepstone L,Porter S, et al. Expression profiling of metalloproteinases andtheir inhibitors in cartilage. Arthritis Rheum 2004;50:131–41.
31. Mestdagh P, Van Vlierberghe P, De Weer A, Muth D, Wester-mann F, Speleman F, et al. A novel and universal method for
microRNA RT-qPCR data normalization. Genome Biol 2009;10:R64.
32. Oskowitz AZ, Lu J, Penfornis P, Ylostalo J, McBride J, Fleming-ton EK, et al. Human multipotent stromal cells from bone marrowand microRNA: regulation of differentiation and leukemia inhib-itory factor expression. Proc Natl Acad Sci U S A 2008;105:18372–7.
33. Stanczyk J, Pedrioli DM, Brentano F, Sanchez-Pernaute O,Kolling C, Gay RE, et al. Altered expression of microRNA insynovial fibroblasts and synovial tissue in rheumatoid arthritis.Arthritis Rheum 2008;58:1001–9.
34. Tili E, Michaille JJ, Cimino A, Costinean S, Dumitru CD, Adair B,et al. Modulation of miR-155 and miR-125b levels followinglipopolysaccharide/TNF-� stimulation and their possible roles inregulating the response to endotoxin shock. J Immunol 2007;179:5082–9.
35. Tili E, Michaille JJ, Costinean S, Croce CM. MicroRNAs, theimmune system and rheumatic disease. Nat Clin Pract Rheumatol2008;4:534–41.
36. Mizuno Y, Yagi K, Tokuzawa Y, Kanesaki-Yatsuka Y, Suda T,Katagiri T, et al. miR-125b inhibits osteoblastic differentiation bydown-regulation of cell proliferation. Biochem Biophys Res Com-mun 2008;368:267–72.
37. Huang W, Li MD. Nicotine modulates expression of miR-140*,which targets the 3�-untranslated region of dynamin 1 gene(Dnm1). Int J Neuropsychopharmacol 2009:12:537–46.
38. Boot-Handford RP, Tuckwell DS, Plumb DA, Rock CF, PoulsomR. A novel and highly conserved collagen (pro�1(XXVII)) with aunique expression pattern and unusual molecular characteristicsestablishes a new clade within the vertebrate fibrillar collagenfamily. J Biol Chem 2003;278:31067–77.
39. Panguluri SK, Bhatnagar S, Kumar A, McCarthy JJ, SrivastavaAK, Cooper NG, et al. Genomic profiling of messenger RNAs andmicroRNAs reveals potential mechanisms of TWEAK-inducedskeletal muscle wasting in mice. PLoS One 2010;5:e8760.
40. Pace JM, Corrado M, Missero C, Byers PH. Identification, char-
acterization and expression analysis of a new fibrillar collagengene, COL27A1. Matrix Biol 2003;22:3–14.
41. Plumb DA, Dhir V, Mironov A, Ferrara L, Poulsom R, Kadler KE,et al. Collagen XXVII is developmentally regulated and forms thinfibrillar structures distinct from those of classical vertebrate fibril-lar collagens. J Biol Chem 2007;282:12791–5.
43. Monk CE, Hutvagner G, Arthur JS. Regulation of miRNAtranscription in macrophages in response to Candida albicans.PLoS One 2010;5:e13669.
44. Hata A, Davis BN. Regulation of pri-miRNA processing throughSmads. Adv Exp Med Biol 2011;700:15–27.
45. Kawai J, Akiyama H, Shigeno C, Ito H, Konishi J, Nakamura T.Effects of transforming growth factor-� signaling on chondrogen-esis in mouse chondrogenic EC cells, ATDC5. Eur J Cell Biol1999;78:707–14.
46. Van der Kraan PM, Goumans MJ, Blaney Davidson E, ten DijkeP. Age-dependent alteration of TGF-� signaling in osteoarthritis.Cell Tissue Res 2012;347:257–65.
47. Qin W, Zhao B, Shi Y, Yao C, Jin L, Jin Y. BMPRII is a directtarget of miR-21. Acta Biochim Biophys Sin (Shanghai) 2009;41:618–23.
48. Fernandes H, Dechering K, van Someren E, Steeghs I, ApothekerM, Mentink A, et al. Effect of chordin-like 1 on MC3T3-E1 andhuman mesenchymal stem cells. Cells Tissues Organs 2010;191:443–52.
49. Miyaki S, Nakasa T, Otsuki S, Grogan SP, Higashiyama R, InoueA, et al. MicroRNA-140 is expressed in differentiated humanarticular chondrocytes and modulates interleukin-1 responses.Arthritis Rheum 2009;60:2723–30.
50. Swingler TE, Waters JG, Davidson RK, Pennington CJ, PuenteXS, Darrah C, et al. Degradome expression profiling in humanarticular cartilage. Arthritis Res Ther 2009;11:R96.
MicroRNAs IN CHONDROGENESIS AND OSTEOARTHRITIS 1919
