miR-130 Suppresses Adipogenesis by Inhibiting Peroxisome ...

MOLECULAR AND CELLULAR BIOLOGY, Feb. 2010, p. 626–638 0270-7306/10/$12.00 doi:10.1128/MCB.00894-10 Copyright © 2010, American Society for Microbiology. All Rights Reserved.

Vol. 31, No. 4

miR-130 Suppresses Adipogenesis by Inhibiting Peroxisome Proliferator-Activated Receptor ␥ Expression䌤 Eun Kyung Lee,1 Mi Jeong Lee,2 Kotb Abdelmohsen,1 Wook Kim,3 Mihee M. Kim,1 Subramanya Srikantan,1 Jennifer L. Martindale,1 Emmette R. Hutchison,4 Hyeon Ho Kim,1 Bernard S. Marasa,1 Roza Selimyan,1 Josephine M. Egan,3 Steven R. Smith,5 Susan K. Fried,2 and Myriam Gorospe1* Laboratory of Molecular Biology and Immunology, NIA-IRP, NIH, Baltimore, Maryland 212241; Department of Medicine, Boston University School of Medicine, Boston, Massachusetts 021182; Laboratory of Clinical Investigation, NIA-IRP, NIH, Baltimore, Maryland 212243; Laboratory of Neurosciences, NIA-IRP, NIH, Baltimore, Maryland 212244; and Translational Research Institute, Florida Hospital-Burnham Institute, Winter Park, Florida 327895 Received 2 August 2010/Returned for modification 1 September 2010/Accepted 18 November 2010

Adipose tissue development is tightly regulated by altering gene expression. MicroRNAs are strong posttranscriptional regulators of mammalian differentiation. We hypothesized that microRNAs might influence human adipogenesis by targeting specific adipogenic factors. We identified microRNAs that showed varying abundance during the differentiation of human preadipocytes into adipocytes. Among them, miR-130 strongly affected adipocyte differentiation, as overexpressing miR-130 impaired adipogenesis and reducing miR-130 enhanced adipogenesis. A key effector of miR-130 actions was the protein peroxisome proliferator-activated receptor ␥ (PPAR␥), a major regulator of adipogenesis. Interestingly, miR-130 potently repressed PPAR␥ expression by targeting both the PPAR␥ mRNA coding and 3ⴕ untranslated regions. Adipose tissue from obese women contained significantly lower miR-130 and higher PPAR␥ mRNA levels than that from nonobese women. Our findings reveal that miR-130 reduces adipogenesis by repressing PPAR␥ biosynthesis and suggest that perturbations in this regulation is linked to human obesity. microRNAs (miRNAs), which modulate the stability and translation of mRNAs encoding adipogenic factors. For example, at the onset of adipogenesis, the RBP HuR selectively promotes expression of the target mRNAs encoding C/EBP␤ (13); C/EBP␤ in turn contributes to increasing the expression of other adipogenic factors such as PPAR␥ and C/EBP␣ (39, 40). The RBP CUGBP1 was also shown to modulate adipogenesis by binding the C/EBP␤ mRNA and enhancing the translation of a liver-enriched inhibitory protein (LIP) isoform of C/EBP␤ (17). MicroRNAs are small noncoding RNAs that associate with the RNA-induced silencing complex (RISC) and bind target mRNAs with partial complementarity. MicroRNAs typically repress expression of the target mRNA by lowering its stability and/or translation (3). Through their influence on target mRNAs, microRNAs are involved in numerous physiologic and pathological processes, such as tissue development, cell proliferation, apoptosis, energy metabolism, immune response, and tumorigenesis (2, 21, 32). Some microRNAs have been identified as being differentially expressed during adipogenesis, including several microRNAs that alter cell proliferation (e.g., miR-24, miR-31, and the miR-17-92 cluster), repress Wnt signaling (miR-8), or target PPAR␥ expression (miR-27) (16, 18–20, 25, 33, 38). The mouse 3T3-L1 preadipocyte cell line has been a useful in vitro system for unraveling the molecular events that occur along the course of adipocyte differentiation (4, 26). However, human adipose tissue also contains numerous preadipocytes in more advanced stages of commitment to adipocyte differentiation (14, 34). Therefore, it is important that the role of microRNAs be examined during human adipogenesis. Here,

In obese individuals, the increase in adiposity results from increases in the number and size of adipocytes, while the degree of hypertrophy relative to hyperplasia influences the level of body fat and the metabolic consequences of obesity (5, 11). Therefore, a thorough understanding of the mechanisms that regulate the formation of adipose tissue could have clinical relevance in light of the ongoing worldwide obesity epidemic. The conversion of progenitor mesenchymal cells into fully functional adipocytes involves dramatic changes in gene expression programs. Many such changes are elicited at the transcriptional level. Prominent among the transcriptional regulators of adipogenesis are the peroxisome proliferator-activated receptor ␥ (PPAR␥) and CCAAT/enhancer-binding protein ␣ (C/EBP␣), which function with other adipogenic transcription factors to regulate the expression of adipogenic gene products like adipsin, lipoprotein lipase (LPL), and the adipocyte fatty acid-binding protein 4 (FABP4) (reviewed in reference 10). Additionally, adipogenesis involves the transient expansion of confluent preadipocytes, which requires cell cycle regulatory proteins such as E2Fs (a family of transcription factors) and pocket proteins like pRB, p107, and p130, which regulate E2F activity and hence expression of adipogenesisrelated proteins (9). Besides the transcription factors that modulate adipogenesis, there is increasing recognition of posttranscriptional regulatory factors such as RNA-binding proteins (RBPs) and

* Corresponding author. Mailing address: LMBI, NIA-IRP, NIH, 251 Bayview Blvd., Baltimore, MD 21224. Phone: (410) 558-8443. Fax: (410) 558-8386. E-mail: [email protected]. 䌤 Published ahead of print on 6 December 2010. 626

VOL. 31, 2010

miR-130 REPRESSES PPAR␥ AND ADIPOGENESIS

we identify subsets of microRNAs in preadipocytes isolated from abdominal subcutaneous adipose tissue of human donors, whose levels change during differentiation in culture. Of the microRNAs downregulated with differentiation, miR-130a and miR-130b assumed prominence because we found that they interacted with the coding region (CR) and 3⬘ untranslated region (UTR) of PPAR␥ and potently repressed its production, thereby blocking the expression of PPAR␥-regulated genes. Accordingly, adipogenesis was inhibited by miR-130 overexpression and enhanced by reducing miR-130 abundance. These effects on expression of PPAR␥ and adipogenic marker genes were recapitulated in mouse 3T3-L1 preadipocytes in which miR-130 levels were modulated by lentiviral constructs. A survey of female donors with different degrees of obesity, quantified by measuring their body mass indices (BMIs; i.e., kg body weight/m2 height), revealed a strong correlation between the levels of PPAR␥ mRNA (high in the obese group, low in the lean group) and miR-130 (low in the obese, high in the lean). Our results underscore a key function for the adipogenesis-regulated miR-130 in repressing PPAR␥, thereby controlling adipocyte gene expression programs, and suggest that reduced levels of miR-130 in combination with obesity are linked to the elevated levels of PPAR␥ and adipocyte expansion.

MATERIALS AND METHODS Adipose tissue handling and preparation of human primary preadipocytes. The study protocol was approved by the Institutional Review Boards of the University of Maryland, Boston University, Pennington Biomedical Research Center, and the National Institute on Aging, National Institutes of Health. Adipose tissue samples were either immediately frozen in liquid nitrogen in the clinic or transferred to the lab in medium 199. For preparation of preadipocytes, adipose tissue was obtained during elective surgeries (abdominoplasties in women who had lost weight following bariatric surgeries), and minced adipose tissue was digested with collagenase to obtain stromal cells, as described previously (12, 14). Macrophages were removed by replating, and no endothelial cells or endothelial markers were detected; this isolation procedure yielded 90% preadipocytes. Briefly, adipose tissue was digested in collagenase (1 mg/ml for 2 h at 37°C) and filtered through a 250-␮m mesh, and stromal cells (preadipocytes) were pelleted and plated in growth medium after erythrocytes were lysed. For measurement of miRNA-130 expression in lean and obese women, abdominal subcutaneous adipose tissue samples from nondiabetic, normotensive volunteers not taking medications that could affect adipose biology (including steroids or oral contraceptives) were obtained by needle aspiration and immediately frozen in liquid nitrogen. Cell culture, transfection, infection, differentiation, small RNAs, and plasmids. Human primary preadipocytes and HeLa cells were cultured in Dulbecco’s modified essential medium–nutrient mixture F-12 (DMEM-F12) and DMEM (Invitrogen), respectively, supplemented with 10% fetal bovine serum. To induce adipocyte differentiation, cells were stimulated with insulin, hydrocortisone, IBMX (hydrocortisol, 3-isobutyl-1-methylxanthine), and rosiglitazone for 3 days and maintained in differentiation medium containing insulin, hydrocortisone, transferrin, biotin, T3, and pantothenate (14). Differentiation of adipocytes was fully established by day 10 and was characterized by the presence of lipid droplets, positive oil red O staining, elevated triglyceride (TG) levels (quantified using a triglyceride assay kit from Cayman Chemical), and reverse transcription (RT) followed by quantitative PCR (RT-qPCR) analysis of adipocyte-specific marker transcripts. Undifferentiated 3T3-L1 preadipocytes were cultured in DMEM containing 10% calf serum and infected with lentiviral vectors (SBI Biosciences) to reduce miR-130a function (miRZip) or to overexpress miR-130a. To induce differentiation, 3T3-L1 cells were incubated with 0.5 mM IBMX, 1 ␮M dexamethasone, and 167 nM insulin for 2 days and further incubated with 167 nM insulin for 2 days. Differentiated 3T3-L1 cells were maintained in 10% fetal bovine serum (FBS)-DMEM for up to 8 days. miRNAs (Ambion), control small interfering RNA (siRNA) (Qiagen), PPAR␥ siRNA (Santa Cruz Biotechnology), and enhanced green fluorescent protein (EGFP) reporter plasmids were transfected with Lipofectamine (Invitrogen). Cells undergoing differentiation

627

were imaged at a total magnification of ⫻200 with an Axio Observer microscope (Zeiss) with AxioVision 4.7 Zeiss image processing software. RNA analysis. Total RNA was prepared directly from the cells or tissues using Trizol (Invitrogen) or Qiagen columns, according to the manufacturers’ protocols. After RT using random hexamers and SSII reverse transcriptase (Invitrogen), the abundance of transcripts was assayed by RT-qPCR using SYBR green PCR master mix (Applied Biosystems) and gene-specific primer sets (below). RT-qPCRs were performed on Applied Biosystems model 7300 and 7900 instruments. The forward and reverse primers used, respectively, were as follows (sequence, 5⬘33⬘): adiponectin, GGCATGACCAGGAAACCAC and TTCACCGATGTC TCCCTTAGG; mouse adiponectin, TGTTCCTCTTAATCCTGCCCA and CC AACCTGCACAAGTTCCCTT; leptin, GAACCCTGTGCGGATTCTTGT and TCCATCTTGGATAAGGTCAGGAT; GAPDH, TGCACCACCAACTGCT TAGC and GGCATGGACTGTGGTCATGAG; mouse GAPDH, AGGTCGG TGTGAACGGATTTG and TGTAGACCATGTAGTTGAGGTCA; PPAR␥, GCTGTGCAGGAGATCACAGA and GGGCTCCATAAAGTCACCAA; mouse PPAR␥, TCGCTGATGCACTGCCTATG and GAGAGGTCCACAGA GCTGATT; FABP4, AACCTTAGATGGGGGTGTCCTG and TCGTGGAA GTGACGCCTTTC; LPL, CTGGACGGTAACAGGAATGTATGAG and CA TCAGGAGAAAGACGACTCGG; adipsin, CAAGCAACAAAGTCCCGAGC and CCTGCGTTCAAGTCATCCTC; mouse adipsin, CATGCTCGGCCCTAC ATGG and CACAGAGTCGTCATCCGTCAC; GLUT4, TCAACAATGTCC TGGCGGTG and TTCTGGATGATGTAGAGGTAGCGG; PPAR␥, ATCGC TCGAGGAAAGCCTTTTGG and CGCCGGATCCGAATAATGACAGC; PPAR␥-UTR, ATTCCTCGAGACTAGCAGAGAGTCCTG and GAGCGGA TCCCATATTCTAAAACCTTT; mouse aP2, GGGGCCAGGCTTCTATTCC and GGAGCTGGGTTAGGTATGGG. microRNA qPCR array (miRNome analysis) and microRNA detection. MicroRNA qPCR array analysis was performed using the hsa-miRNome microRNA profiling kit (System Biosciences) to examine miRNA expression in preadipocytes and adipocytes from two different donors. Two micrograms of RNA was reverse transcribed using the QuantiMir RT kit (System Biosciences), and miRNA profiling was performed by qPCR using microRNA-specific primers and SYBR green PCR master mix (Applied Biosystems). Expression levels in samples were normalized to U6 expression. The fold change indicates the expression level of microRNAs from adipocytes relative to that of preadipocytes. Individual microRNAs were further quantified using a TaqMan microRNA detection assay (Applied Biosystems) or the QuantiMir cDNA kit (System Biosciences). For the TaqMan microRNA detection assay, miRNA-specific primer sets supplied by the manufacturer were used in RT-qPCRs. For QuantiMir cDNA analysis, forward primers were designed to be exact sequences of the miRNAs in the miRBase database and a universal reverse primer was used from the kit. EGFP reporters containing PPAR␥ mRNA sequences. EGFP reporters were cloned by insertion of specific fragments from the PPAR␥ 3⬘-UTR and CR into the 3⬘-UTR and CR of pEGFP-C1 (BD Bioscience), respectively, as described previously (1). The seed regions of miR-130 binding sites on PPAR␥ mRNA were mutated by site-directed mutagenesis (Stratagene). Western blot analysis. Whole-cell lysates were prepared using radioimmunoprecipitation assay (RIPA) buffer, separated by electrophoresis in SDS-containing polyacrylamide gels, and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore). Incubations with primary antibodies to detect PPAR␥ or GFP (Santa Cruz Biotech), or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or ␤-actin (Abcam), were followed by incubations with the appropriate secondary antibodies conjugated with horseradish peroxidase (HRP) and by detection using enhanced luminescence (GE Healthcare).

RESULTS Differential expression of microRNAs during differentiation of human adipocytes. To investigate changes in expressed microRNAs during adipocyte differentiation, human primary adipocytes were obtained during elective surgeries. Stromal cells isolated by collagenase digestion were differentiated with a cocktail of insulin, hydrocortisone, IBMX, and rosiglitazone using standard protocols (14) (see Materials and Methods). RNA extracted from preadipocytes and from fully differentiated adipocytes was used to identify the collections of differentially expressed microRNAs by miRNome analysis (Fig. 1A),

628

LEE ET AL.

MOL. CELL. BIOL.

FIG. 1. Differentially expressed miRNAs during human adipogenesis. (A) Schematic flow of miRNome analysis (see Materials and Methods) to compare microRNA expression profiles in preadipocytes and in fully differentiated adipocytes from two individual subjects. (B) MicroRNAs downregulated in differentiated adipocytes (a complete list of microRNAs can be obtained from the authors). Boldface microRNAs were studied further. (C) miR-130 expression levels were assayed by RT-qPCR analysis of 5 individual donors using miR-130-specific primer sets; *, P ⬍ 0.05.

VOL. 31, 2010

a platform that uses RT and real-time qPCR amplification of individual microRNAs (Materials and Methods). MicroRNAs showing ⬎3-fold lower abundance in differentiated adipocytes than in preadipocytes are shown in Fig. 1B (a complete list of microRNAs can be obtained from the authors). The specific reductions in microRNA abundance with adipocyte differentiation were verified by single-miRNA RT-qPCR analysis using preadipocytes from five different donors (Fig. 1C). The microRNAs miR-27a, previously reported to lower PPAR␥ (18, 20, 25), and miR-301, computationally predicted to target the PPAR␥ mRNA (unpublished data), revealed no significant change in abundance (data not shown). Thus, our subsequent efforts focused on miR-130a and miR-130b, two microRNAs predicted to target PPAR␥ mRNA (below). Human adipogenesis model. To investigate whether miR130a and miR-130b influenced human adipocyte differentiation, we studied the process of adipogenesis using preadipocytes obtained from individual human donors. The preadipocytes (⬎90% of the cell preparation, as described in Materials and Methods) were subjected to a standard differentiation protocol that resulted in their differentiation into mature adipocytes by ⬃day 10 (Fig. 2A). During this process, they gradually accumulated lipid droplets that were visualized using oil red O (Fig. 2B), produced increasing amounts of triglycerides (Fig. 2C), and expressed increasing concentrations of FABP4 and GLUT4 mRNAs, two classic markers of mature adipocytes (Fig. 2D). In this system, we measured the time-dependent decline in miR-130a and miR-130b levels (Fig. 3A). Among the mRNAs showing increased expression during human adipogenesis (Fig. 3B), PPAR␥ mRNA was predicted to be a target of miR-130a and miR-130b. In a well-established model of murine adipogenesis (3T3-L1 preadipocytes differentiating into adipocytes), there was a similar decline in the conserved murine miR-130a and miR-130b and a concomitant increase in mRNAs encoding adipocyte-specific proteins (Fig. 3C and D). The parallel changes in human and murine mRNAs and miR-130 during adipogenesis further support the suitability of the human adipogenesis model. Altering miR-130 levels modulates adipocyte differentiation. To ascertain whether the reduction of miR-130a and miR-130b levels influences adipocyte differentiation, miR-130a and miR130b mimics were transfected separately into preadipocytes that were then subjected to the differentiation protocol (Fig. 2A). This intervention resulted in a marked impairment of differentiation, as determined by morphological changes and oil red O staining (Fig. 4A and data not shown) and reduced TG production (Fig. 4B). Conversely, decreasing miR-130a or miR-130b by transfection of antisense (AS) oligonucleotides complementary to miR-130a and miR-130b (“antagomirs”) enhanced the differentiation phenotype (Fig. 4A) and increased TG content (Fig. 4B). The abundance of miR-130 following transfection is shown in Fig. 4C. By 24 h (day 1) after transfection, miR-130a overexpression was ⬃150-fold higher and miR-130b was ⬃400-fold higher; by 24 h following transfection of the antagomirs, miR-130a levels were ⬃60% lower and miR-130b levels were 50% lower (Fig. 4C). It must be noted that antagomirs can neutralize microRNAs without actively lowering their steady-state abundance, so the concentration of


629

FIG. 2. Human adipocyte differentiation model. (A) Schematic of preadipocyte transfection and differentiation into mature adipocytes. Preadipocytes were stimulated to differentiate on day 1 and completed differentiation by ⬃day 10. (B) Formation of lipid droplets was observed by light microscopy and by staining with oil red O. (C) Triglyceride (TG) concentration was measured as described in Materials and Methods. (D) The abundance of FABP4 and GLUT4 mRNAs was normalized to the levels of GAPDH mRNA. In panels C and D, the data represent the means ⫹ SD for three individual donors, each assayed in triplicate.

functional miR-130 after (AS)miR-130a/b transfection is likely lower than that measured by RT-qPCR. The impairment in differentiation following miR-130 overexpression lowered the levels of differentiation markers such as leptin, adiponectin, and PPAR␥ mRNAs (Fig. 5A, left); other microRNAs (miR-30a, let-7g) which did not affect differentiation based on morphological parameters (not shown) did not significantly affect the levels of differentiation markers (Fig. 5A, right). The transfection of antagomirs to lower miR-130 levels elevated the expression of differentiation markers (Fig. 5B). To study whether modulating miR-130 levels had the

630

LEE ET AL.

MOL. CELL. BIOL.

FIG. 3. Comparison of human and murine adipogenesis models. (A and B) The levels of miR-130a and miR-130b (A), as well as PPAR␥ mRNA and several adipogenic marker transcripts (B), were measured in human preadipocytes subjected to a differentiation protocol as described in the legend to Fig. 2. (C and D) The levels of miR-130a and miR-130b (C), as well as PPAR␥ mRNA and several adipogenic marker transcripts (D), were measured during differentiation of 3T3-L1 mouse preadipocytes. The data represent the means ⫾ SD of three independent experiments.

same influence on mouse 3T3-L1 preadipocytes, we obtained lentiviruses to express miR-130 (Lenti-miR-130a) or neutralize miR-130a function [Lenti-(AS)miR-130a, a miRZip lentiviral vector; see Materials and Methods]. 3T3-L1 cells were infected with each of the three lentiviruses (Fig. 5C) and cultured under adipogenesis differentiation conditions; 7 days later, the levels of adipogenic markers were measured by RT-qPCR analysis (Fig. 5D). As shown, overexpression of miR-130a in 3T3-L1 cells (Lenti-miR-130a) significantly lowered the abundance of PPAR␥ mRNA and differentiation markers adipsin and adiponectin mRNAs; reducing miR-130a function [Lenti-(AS)miR-130a] significantly upregulated adipsin and PPAR␥ mRNA levels. These results indicate that modulating miR-130 levels can also influence adipogenesis in a mouse model. The changes in miR-130 abundance seen by 24 h (Fig. 4C) rapidly modulated the levels of PPAR␥ mRNA, a key regulator of adipogenesis. As shown, the PPAR␥ mRNA concentration was lowered to ⬃50% of the original abundance after transfection of miR-130, while it was increased by ⬃50% after transfection of (AS)miR-130 (Fig. 5E). After joint transfection of miR-130a and miR-130b, PPAR␥ mRNA levels (Fig. 5F) and PPAR␥ protein levels (Fig. 5G) were also potently reduced. Together, these results indicate that miR-130a and miR-130b decrease adipocyte differentiation, while lowering miR-130a and miR-130b increases adipocyte differentiation. miR-130 specifically inhibits PPAR␥ expression via both coding region and 3ⴕ-UTR determinants. Since PPAR␥ mRNA (but not leptin or adiponectin mRNAs) was found to

be a putative target of miR-130a and miR-130b, we investigated whether miR-130 directly inhibited PPAR␥ expression. The PPAR␥ mRNA has a long (1.4-kb) CR and relatively short 5⬘- and 3⬘-UTRs (each ⬃0.2-kb long) and bears two predicted sites of miR-130 interaction, one in its CR and one in its 3⬘-UTR (Fig. 6A and B). In order to test whether these sequences specifically contribute to the miR-130-mediated regulation of PPAR␥ mRNA expression, we generated EGFP reporter constructs containing segments (⬃130 nucleotides long each) that spanned the miR-130 sites of the PPAR␥ CR and 3⬘-UTR (Fig. 6A, plasmids pEGFP-PPAR␥CR and pEGFP-PPAR␥UTR). We generated additional reporter plasmids (Fig. 6A, pEGFP-PPAR␥CRmut and pEGFPPPAR␥UTRmut) in which the “seed” regions (the regions of extensive complementarity between miRNAs and target mRNAs that are critical for microRNA function) were mutated by changing 4 nucleotides to disrupt the putative interaction of PPAR␥ mRNA with miR-130 (Fig. 6B). Expression of these reporter constructs was assessed in human cervical carcinoma HeLa cells, as human preadipocytes are notoriously difficult to transfect with plasmid DNA. As shown in Fig. 6C, strong EGFP expression was observed in cells transfected with the control vector, pEGFP, regardless of miR-130a or miR130b abundance. In contrast, EGFP expressed from pEGFPPPAR␥CR and pEGFP-PPAR␥UTR was potently repressed by overexpression of miR-130a or miR-130b (Fig. 6C). The reduction in EGFP level was due at least in part to a reduction in the chimeric mRNAs encoding EGFP, since the abundance

VOL. 31, 2010


631

FIG. 4. miR-130 regulates adipocyte differentiation. Preadipocyte transfection and differentiation into mature adipocytes were performed as described in the legend to Fig. 2. After transfection of protected miR-130a or miR130b mimics, miR-130 antagomirs [(AS)miR-130a and (AS)miR-130b] or control (Ctrl) siRNA, preadipocytes were stimulated to differentiate. After ⬃10 days, cells were harvested for analysis. (A) Formation of lipid droplets was observed by light microscopy and by staining with oil red O in cultures transfected with control siRNA, with miR-130a or miR-130b microRNAs (left), and with antagomirs (AS)miR-130a or (AS)miR-130b (right). (B) The degree of differentiation was also determined by measuring TG concentration, represented as the means ⫹ SD, of three individual donors, each assayed in triplicate (*, P ⬍ 0.05). (C) Following transfection of preadipocytes on day 0 with the small RNAs shown, followed by differentiation (as shown in Fig. 2A), the levels of miR-130a (left) and miR-130b (right) were quantified by RT-qPCR analysis on days 1, 4, and 10. Data are representative of two independent experiments yielding similar results.

632

LEE ET AL.

MOL. CELL. BIOL.

FIG. 5. miR-130 regulates adipogenic marker genes. Preadipocyte transfection and differentiation into mature adipocytes were performed as explained in the legend to Fig. 2. (A and B) By 10 days following transfection of the small RNAs shown, cells were harvested and the degree of differentiation was determined by RT-qPCR analysis of the adipocyte-specific marker transcripts shown, normalized to the levels of GAPDH

VOL. 31, 2010

of the ectopic EGFP-PPAR␥CR and EGFP-PPAR␥UTR mRNAs was significantly reduced (Fig. 6D). Taken together, these findings indicate that miR-130 interacts with the PPAR␥ CR and 3⬘-UTR, thereby repressing PPAR␥ mRNA and protein levels. Enhanced adipogenesis by (AS)miR-130 is rescued by silencing PPAR␥. The lowering of PPAR␥ expression by miR130 led to significant reductions in the abundance of PPAR␥regulated transcripts, including adipsin, LPL, and FABP4 mRNAs, none of which is a predicted miR-130 target (Fig. 7A). Conversely, the enhancement in PPAR␥ production after reducing miR-130 in adipocytes elevated the expression of the same mRNAs (Fig. 7B), indicating that miR-130 represses PPAR␥ production, which in turn lowers the mRNAs transcriptionally regulated by PPAR␥. To investigate whether the promotion of adipogenesis by (AS)miR-130 was due to the increased abundance of PPAR␥, a rescue experiment was performed. Preadipocytes were transfected with (AS)miR-130a or (AS)miR-130b and PPAR␥ abundance was either left unchanged (control siRNA group) or was silenced by using small interfering RNA directed to PPAR␥ mRNA (PPAR␥ siRNA). Transfection of (AS)miR130 promoted adipocyte differentiation, as assessed by the morphology and increased oil red O staining of the culture (Fig. 7C, left), the enhanced TG production (Fig. 7D), and the elevated production of differentiation marker genes (Fig. 7E). Importantly, however, (AS)miR-130 transfection did not increase adipogenesis when PPAR␥ was silenced (Fig. 7F): the cells showed reduced oil red O staining (Fig. 7C), lower TG production (Fig. 7D), and decreased adipogenic gene expression (Fig. 7E). Collectively, these data indicate that lowering miR-130 promotes adipogenesis by allowing higher PPAR␥ expression, since silencing PPAR␥ reverses the adipogenic phenotype. Differential miR-130 and PPAR␥ mRNA expression in lean and obese subjects. To investigate the physiologic relevance of the miR-130-elicited repression of PPAR␥, we hypothesized that miR-130 levels might be selectively reduced within obese states, since PPAR␥ expression increased and miR-130 levels declined with advancing adipogenesis (Fig. 1C) (37) and because forcing a reduction in miR-130 abundance enhanced adipogenesis (Fig. 4, 5, and 7). We also anticipated that PPAR␥ abundance would follow a pattern opposite to that of miR-130 levels. To test these possibilities, RNA isolated from abdominal depots of 12 healthy women, 5 with BMIs (kg body weight/m2 height) of ⬍25 (lean) and 12 with BMIs of ⬎25 (obese), all premenopausal (Fig. 8A), were used for the measurement of PPAR␥ mRNA, miR-130a, and miR-130b. Consistent with previous observations (37), PPAR␥ mRNA levels were significantly lower in lean women (Fig. 8B, left); recipro-


633

cally, miR-130a and miR-130b levels were selectively elevated in the lean group (Fig. 8B, right). The obese female subjects showed higher PPAR␥ mRNA levels and lower concentrations of miR-130a and miR-130b (Fig. 8B). The inverse correlation between PPAR␥ mRNA levels and miR-130 levels is shown in Fig. 8C. Importantly, the levels of miR-130 correlated inversely with BMIs, indicating a link between the degree of obesity and miR-130a and miR-130b levels (Fig. 8D). In sum, miR-130 levels decline with advancing adiposity in women, in association with elevated PPAR␥ abundance.

DISCUSSION We report that miR-130 prevents the unscheduled differentiation of adipocytes, an effect that is mediated, at least in part, by its potent repression of PPAR␥, a master regulator of adipogenesis. Previously, miR-27 was shown to repress PPAR␥ in human multipotent adipose-derived stem cells (18) as well as in mouse preadipocyte model systems, where it was linked to a blockage of adipogenesis (20, 25). In the primary human preadipocyte model system studied here, miR-27b was decreased (albeit less than miR-130a or miR-130b [unpublished data]), but miR-27a was not significantly changed. These differences likely reflect distinct patterns of miR-27 production in each model system, but the mechanisms responsible for the production of these microRNAs remain to be elucidated. Instead, we discovered that miR-130 was markedly reduced in differentiated human adipocytes and it potently repressed PPAR␥ production. Unlike the single miR-27 site, miR-130 had two functional sites of interaction with the PPAR␥ mRNA, one in the CR and one in the 3⬘-UTR (Fig. 6). Whether miR-27b may function in conjunction with miR-130 to influence adipogenesis awaits investigation. An interesting aspect of this study is the discovery that miR130 functions by targeting both the coding region and the 3⬘-UTR of PPAR␥ mRNA. The PPAR␥ mRNA has relatively short 5⬘- and 3⬘-UTRs (173 and 211 nucleotides long, respectively), and therefore it is not surprising that so few posttranscriptional regulators have been identified for this mRNA to date. Indeed, most sites of microRNA-mediated regulation are thought to reside in the 3⬘-UTR of the mRNA (7), although a few examples of microRNAs regulating gene expression through the coding region of target mRNAs are beginning to emerge (miR-24, miR-519, and miR-148) (6, 7, 23, 24). Additional regulation of PPAR␥ expression is conferred by transcription factors (e.g., C/EBP␤, C/EBP␦, KLF5, KLF15, and SREBP-1c) which regulate PPAR␥ expression transcriptionally (8, 22, 28, 29, 40). This multilevel regulation (both transcriptional and posttranscriptional) is characteristic of factors

mRNA. (A) Left, transfection of miR-130a or miR-130b; right, transfection of miR-30a or let-7g. (B) Transfection of miR-130 antagomirs. (C and D) Mouse 3T3-L1 preadipocytes were infected with the lentiviruses shown, and 24 h later they were induced to differentiate. Seven days after that, the levels of miR-130a (C) and adipogenic markers (D) were measured by RT-qPCR analysis; data represent the means ⫹ SD from three measurements in a representative experiment. *, P ⬍ 0.05. (E) The levels of PPAR␥ mRNA by 24 h after transfection of human preadipocytes with the RNAs indicated were assessed by RT-qPCR analysis and normalized to GAPDH mRNA levels. (F and G) The levels of PPAR␥ mRNA after joint transfection of human preadipocytes with miR-130a and miR-130b were measured by RT-qPCR analysis 10 days later (F), and PPAR␥ protein abundance at the times shown was assessed by Western blot analysis and quantified by densitometry (G). In panels A, B, E, and F, the data represent the means ⫹ SD of results for three individual donors, each assayed in triplicate; *, P ⬍ 0.05; **, P ⬍ 0.01.

634

LEE ET AL.

MOL. CELL. BIOL.

FIG. 6. miR-130 directly regulates PPAR␥ expression through CR and 3⬘-UTR sites. (A) Schematic of PPAR␥ mRNA and the EGFP reporter plasmids containing the miR-130 binding site of PPAR␥ mRNA. The CR site (nucleotide positions 1243 to 1350; light gray) and the 3⬘-UTR site (nucleotide positions 1603 to 1730; dark gray) were inserted in the CR or 3⬘-UTR of the heterologous pEGFP reporter, as indicated. *, point mutations introduced to disrupt miR-130 binding. (B) miR-130 sequence and predicted binding between miR-130 and PPAR␥ mRNA. Sequences of miR-130a and miR-130b (www.mirbase.org) are shown. PPAR␥ mRNA has two different putative binding sites for miR-130, one on the coding region (left) and one on the 3⬘-UTR (right). Four nucleotides of PPAR␥ 3⬘-UTR (underlined) were replaced with TTGT by site-directed mutagenesis in order to disrupt the binding with miRNA seed regions (indicated by an asterisk in panel A). (C and D) HeLa cells were transfected with each of the five reporter constructs, together with either miR-130a, miR-130b, or the corresponding antagomirs; 48 h later, EGFP expression was determined by Western blot analysis (C) and by RT-qPCR analysis of the expressed EGFP mRNAs (D). The data represent the means ⫹ SD of three independent experiments; *, P ⬍ 0.05.

VOL. 31, 2010


635

FIG. 7. miR-130 modulates adipogenesis via PPAR␥. (A and B) Preadipocytes transfected with miR-130 (A) or (AS)miR-130 (B) were subjected to differentiation; on day 10, the expression levels of three different transcriptional targets of PPAR␥ (adipsin, LPL, and FABP4 mRNAs) were analyzed by RT-qPCR, normalized to GAPDH mRNA abundance, and represented relative to the levels in the control siRNA population. (C to E) Preadipocytes were transfected with control siRNA, (AS)miR-130a, or (AS)miR-130b and expressed normal levels of PPAR␥ (Ctrl siRNA group) or expressed reduced levels of PPAR␥ (PPAR␥ siRNA group). On day 10, the morphology and oil red O staining were assessed by light microscopy (C), the TG content was measured (D), and the abundance of adipogenic marker mRNAs was quantified by RT-qPCR analysis (E). (F) The degree of PPAR␥ silencing in panels C to E was assessed by RT-qPCR analysis. The data in panels A, B, D, E, and F represent the means ⫹ SD for three individual donors, each assayed in triplicate; *, P ⬍ 0.05.

636

LEE ET AL.

MOL. CELL. BIOL.

FIG. 8. Differential expression of miR-130 in lean and obese women. (A) Table giving the information about the individual female donors from which abdominal fat was used to prepare mRNA. Donors were either lean (BMI ⬍ 25) or obese (BMI ⬎ 25). (B) The levels of PPAR␥ mRNA (left, normalized to GAPDH mRNA), miR-130a, and miR-130b (right, normalized to U6) were measured by RT-qPCR analysis; differences between lean and obese females were significant (*, P ⬍ 0.05; **, P ⬍ 0.01). (C) Inverse correlation between PPAR␥ mRNA and miR-130a and miR-130b in the cohort studied. (D) Inverse correlation between miR-130 levels and BMI.

VOL. 31, 2010

like PPAR␥ that are implicated in the control of key biological processes. Given miR-130’s potent role as an adipogenic inhibitor, it will be important to elucidate the processes that control miR130 production. The expression of miR-130a and miR-130b increases after treatment with tumor necrosis factor alpha (data not shown). Experiments are under way to test whether NF-␬B, whose activity reduces the differentiation and function of adipocytes (31), increases miR-130 expression by promoting the transcription of miR-130 primary transcripts. The posttranscriptional regulation of miR-130 biosynthesis is also being examined. Together, these analyses could lead to the identification of important molecular links between inflammation, adipogenesis, insulin resistance, and obesity. The results of this investigation underscore the usefulness of human preadipocytes as a model system to study human adipogenesis. Side-by-side comparison with the well-established model of murine adipogenesis (mouse 3T3-L1 preadipocytes differentiating into adipocytes) reveals many parallels. Human adipocytes undergo changes in morphology and oil red O staining (Fig. 2B, 4A, and 7C), hallmarks of 3T3-L1 adipogenesis. In addition, human and mouse preadipocytes undergo parallel reductions in miR-130a and miR-130b and comparable increases in the levels of PPAR␥ mRNA and other mRNAs encoding adipogenic proteins (Fig. 3). Moreover, miR-130 overexpression in both human and mouse preadipocytes caused comparable decreases in PPAR␥ expression, while lowering miR-130 function elevated PPAR␥ abundance (Fig. 5A and D). In this regard, while the human primary adipocyte constitutes a more challenging system to study adipogenesis, mouse 3T3-L1 adipocytes likely do not fully recapitulate human adipogenesis. Thus, human preadipocytes are increasingly used as a valuable (and likely a more relevant) model to investigate human adipogenesis (e.g., see references 14, 27, 30, 35, and 36). Finally, miR-130 levels were lower among obese women than among lean women (Fig. 8), which is associated with an increase in PPAR␥ mRNA levels in this relatively young, healthy population. The lower expression of miR-130 in the obese humans may reflect a relative depletion of preadipocyte numbers due to their recruitment into the adipocyte pool (15). Alternatively, the reduced miRNA-130 abundance may in part reflect its lower expression in mature adipocytes and is consistent with the higher expression of PPAR␥ in the obese group. Further studies are needed to characterize the cell types driving these associations, their links to interindividual differences in the numbers of preadipocytes, their location within the body, and their potential for spontaneous differentiation. Incidentally, in obese (ob/ob) mice (mice that have a mutated, biologically inactive form of leptin and accordingly become grossly obese), the levels of miR-27a/b, which is also a repressor of adipogenesis and PPAR␥ gene expression, are higher than in lean mice (25). This finding is in contrast to the situation in humans, where miR-130 is lower in obese than in lean subjects. These differences may reflect the high level of inflammation and adipocyte dysfunction in the ob/ob mice, in agreement with the downregulation of PPAR␥ observed in the mouse adipocytes (20). The discovery that miR-130 influences PPAR␥ expression and adipocyte differentiation highlights the


637

potential relevance of including microRNAs in our efforts to understand, prevent, and manage human obesity. ACKNOWLEDGMENTS We thank Stephen R. Farmer for advice and insight into this study. This research was funded by the Intramural Research Program of the National Institute on Aging, National Institutes of Health, by grants DK46200, DK072488, DK080448, and DK052398 (to S.K.F. and M.J.L.) and by grant DK072476 (to S.R.S.). We indicate no potential conflicts of interest. REFERENCES 1. Abdelmohsen, K., S. Srikantan, Y. Kuwano, and M. Gorospe. 2008. miR-519 reduces cell proliferation by lowering RNA-binding protein HuR levels. Proc. Natl. Acad. Sci. U. S. A. 105:20297–20302. 2. Baltimore, D., M. P. Boldin, R. M. O’Connell, D. S. Rao, and K. D. Taganov. 2008. MicroRNAs: new regulators of immune cell development and function. Nat. Immunol. 9:839–845. 3. Bartel, D. P. 2009. MicroRNAs: target recognition and regulatory functions. Cell 136:215–233. 4. Cowherd, R. M., R. E. Lyle, and R. E. McGehee, Jr. 1999. Molecular regulation of adipocyte differentiation. Cell Dev. Biol. 10:3–10. 5. Danforth, E., Jr. 2000. Failure of adipocyte differentiation causes type II diabetes mellitus? Nat. Genet. 26:13. 6. Duursma, A. M., M. Kedde, M. Schrier, C. le Sage, and R. Agami. 2008. miR-148 targets human DNMT3b protein coding region. RNA 14:872–877. 7. Eulalio, A., E. Huntzinger, and E. Izaurralde. 2008. Getting to the root of miRNA-mediated gene silencing. Cell 132:9–14. 8. Fajas, L., et al. 1999. Regulation of peroxisome proliferator-activated receptor gamma expression by adipocyte differentiation and determination factor 1/sterol regulatory element binding protein 1: implications for adipocyte differentiation and metabolism. Mol. Cell. Biol. 19:5495–5503. 9. Fajas, L., et al. 2002. The retinoblastoma-histone deacetylase 3 complex inhibits PPARgamma and adipocyte differentiation. Dev. Cell 3:903–910. 10. Farmer, S. R. 2006. Transcriptional control of adipocyte formation. Cell Metab. 4:263–273. 11. Flier, J. S. 2004. Obesity wars: molecular progress confronts an expanding epidemic. Cell 116:337–350. 12. Fried, S. K., C. D. Russell, N. L. Grauso, and R. E. Brolin. 1993. Lipoprotein lipase regulation by insulin and glucocorticoid in subcutaneous and omental adipose tissues of obese women and men. J. Clin. Invest. 92:2191–2198. 13. Gantt, K., J. Cherry, R. Tenney, V. Karschner, and P. H. Pekala. 2005. An early event in adipogenesis, the nuclear selection of the CCAAT enhancerbinding protein ␤ (C/EBP␤) mRNA by HuR and its translocation to the cytosol. J. Biol. Chem. 280:24768–24774. 14. Hauner, H., T. Skurk, and M. Wabitsch. 2001. Cultures of human adipose precursor cells. Methods Mol. Biol. 155:239–247. 15. Isakson, P., A. Hammarstedt, B. Gustafson, and U. Smith. 2009. Impaired preadipocyte differentiation in human abdominal obesity: role of Wnt, tumor necrosis factor-␣, and inflammation. Diabetes 58:1550–1557. 16. Kajimoto, K., H. Naraba, and N. Iwai. 2006. MicroRNA and 3T3-L1 preadipocyte differentiation. RNA 12:1626–1632. 17. Karagiannides, I., et al. 2006. Increased CUG triplet repeat-binding protein-1 predisposes to impaired adipogenesis with aging. J. Biol. Chem. 281: 23025–23033. 18. Karbiener, M., et al. 2009. microRNA miR-27b impairs human adipocyte differentiation and targets PPARgamma. Biochem. Biophys. Res. Commun. 390:247–251. 19. Kennell, J. A., I. Gerin, O. A. MacDougald, and K. M. Cadigan. 2008. The microRNA miR-8 is a conserved negative regulator of Wnt signaling. Proc. Natl. Acad. Sci. U. S. A. 105:15417–15422. 20. Kim, S. Y., et al. 2010. miR-27a is a negative regulator of adipocyte differentiation via suppressing PPAR␥ expression. Biochem. Biophys. Res. Commun. 392:323–328. 21. Kloosterman, W. P., and R. H. Plasterk. 2006. The diverse functions of microRNAs in animal development and disease. Dev. Cell 11:441–450. 22. Kudo, M., A. Sugawara, A. Uruno, K. Takeuchi, and S. Ito. 2004. Transcription suppression of peroxisome proliferator-activated receptor gamma2 gene expression by tumor necrosis factor alpha via an inhibition of CCAAT/ enhancer-binding protein delta during the early stage of adipocyte differentiation. Endocrinology 145:4948–4956. 23. Lal, A., et al. 2008. p16INK4a translation suppressed by miR-24. PLoS One 3:e1864. 24. Lee, E. K., and M. Gorospe. Coding region: the neglected post-transcriptional code. RNA Biol., in press. 25. Lin, Q., Z. Gao, R. M. Alarcon, J. Ye, and Z. Yun. 2009. A role of miR-27 in the regulation of adipogenesis. FEBS J. 276:2348–2358. 26. MacDougald, O. A., and M. D. Lane. 1995. Transcriptional regulation of gene expression during adipocyte differentiation. Annu. Rev. Biochem. 64: 345–373.

638

LEE ET AL.

27. Mikkelsen, T. S., et al. 2010. Comparative epigenomic analysis of murine and human adipogenesis. Cell 143:156–169. 28. Mori, T., et al. 2005. Role of Kru ¨ppel-like factor 15 (KLF15) in transcriptional regulation of adipogenesis. J. Biol. Chem. 280:12867–12875. 29. Oishi, Y., et al. 2005. Kru ¨ppel-like transcription factor KLF5 is a key regulator of adipocyte differentiation. Cell Metab. 1:27–39. 30. Pettersson, A. T., et al. 13 October 2010, posting date. Twist1 in human white adipose tissue and obesity. J. Clin. Endocrinol. Metab. [Epub ahead of print]. 31. Ruan, H., H. J. Pownall, and H. F. Lodish. 2003. Troglitazone antagonizes tumor necrosis factor-alpha-induced reprogramming of adipocyte gene expression by inhibiting the transcriptional regulatory functions of NF-kappaB. J. Biol. Chem. 278:28181–28192. 32. Stefani, G., and F. J. Slack. 2008. Small non-coding RNAs in animal development. Nat. Rev. Mol. Cell Biol. 9:219–230. 33. Sun, F., et al. 2009. Characterization of function and regulation of miR-24-1 and miR-31. Biochem. Biophys. Res. Commun. 380:660–665. 34. Tchkonia, T., et al. 2005. Abundance of two human preadipocyte subtypes with distinct capacities for replication, adipogenesis, and apoptosis varies among fat depots. Am. J. Physiol. Endocrinol. Metab. 288:E267–277.

MOL. CELL. BIOL. 35. Tchkonia, T., et al. 2007. Identification of depot-specific human fat cell progenitors through distinct expression profiles and developmental gene patterns. Am. J. Physiol. Endocrinol. Metab. 292:E298–307. 36. Tiller, G., et al. 2009. Effects of TWEAK (TNF superfamily member 12) on differentiation, metabolism, and secretory function of human primary preadipocytes and adipocytes. Endocrinology 150:5373–5383. 37. Vidal-Puig, A. J., et al. 1997. Peroxisome proliferator-activated receptor gene expression in human tissues. Effects of obesity, weight loss, and regulation by insulin and glucocorticoids. J. Clin. Invest. 99:2416–2422. 38. Wang, Q., et al. 2008. miR-17-92 cluster accelerates adipocyte differentiation by negatively regulating tumor-suppressor Rb2/p130. Proc. Natl. Acad. Sci. U. S. A. 105:2889–2894. 39. Wu, Z., N. L. R. Bucher, and S. R. Farmer. 1996. Induction of peroxisome proliferator activated receptor gamma during the conversion of 3T3 fibroblasts into adipocytes is mediated by C/EBP␤, C/EBP␦, and glucocorticoids. Mol. Cell. Biol. 16:4128–4136. 40. Wu, Z., Y. Xie, N. L. R. Bucher, and S. R. Farmer. 1995. Conditional ectopic expression of C/EBP beta in NIH-3T3 cells induces PPAR gamma and stimulates adipogenesis. Genes Dev. 9:2350–2363.