If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported by grants from the National Natural Science Foundation of China (Grants 90813035, 30890044, 30570731, and 30771036), the National Basic Research Program of China (973 Program) (Grants 2006CB503908, 2006CB503909, 2004CB518603, and 2003CB715905), the “111” Project, and the Natural Science Foundation of Jiangsu Province (Grants BK2004082 and BK2006714). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The on-line version of this article (available at http://www.jbc.org) contains a supplemental table and a supplemental figure. 1 These authors contributed equally to this work.
UCP2 and UCP3, two novel uncoupling proteins, are important regulators of energy expenditure and thermogenesis in various organisms. The striking disparity between UCP2 mRNA and protein levels in muscle tissues prompted initial speculation that microRNAs are implicated in the regulatory pathway of UCP2. We found, for the first time, that the repression of UCP2 expression in cardiac and skeletal muscle resulted from its targeting by a muscle-specific microRNA, miR-133a. Moreover, our findings illustrate a novel function of UCP2 as a brake for muscle development. We also show that MyoD can remove the braking role of UCP2 via direct up-regulation of miR-133a during myogenic differentiation. Taken together, our current work delineates a novel regulatory network employing MyoD, microRNA, and uncoupling proteins to fine-tune the balance between muscle differentiation and proliferation during myogenesis.
UCP2 is a member of the uncoupling proteins (UCPs)
). One of the main puzzles encountered in the studies of UCP2 is the ambiguous regulatory mechanisms controlling its expression. Previous study has suggested that the expression of UCP2 mRNA in a specific tissue does not necessarily correlate with the expression at the protein level. For example, although UCP2 mRNA is found in mouse heart and skeletal muscle, no UCP2 protein could be detected in these tissues (
). Given the disparity between UCP2 mRNA and protein in mouse muscle tissues, it is quite likely that a post-transcriptional regulatory mechanism exists. Recently, a class of RNA regulatory genes known as microRNAs (miRNAs) has been found to introduce a whole new layer of protein regulation after transcription (
). Their widespread and important role in eukaryotes is highlighted by recent discoveries that they control a range of physiological processes including development, differentiation, proliferation, apoptosis, and metabolism (
). In the present study, we identified UCP2 as a target of the muscle-specific miR-133a and demonstrated that UCP2 plays a role in inhibiting muscle differentiation and promoting myoblast proliferation. Therefore, we suggested that down-regulation of UCP2 might be an indispensable step during myogenic differentiation. The specific mechanism through which miR-133a and UCPs regulate the myogenic pathway has also been well described in this study.
Cells and Reagents—Mouse lung fibroblast cells L929, mouse Lewis lung carcinoma cells LLC, and mouse mesenchymal stem cells C2C12 were obtained from China Cell Culture Center (Shanghai, China). L929 and LLC cells were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum. C2C12 myoblasts were maintained at subconfluent densities in growth medium (Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum). When cells reached ∼75% confluence, the medium was switched to differentiation medium (Dulbecco's modified Eagle's medium supplemented with 2% horse serum). Medium was changed every other day. After incubation for several days, cells were harvested and subjected to various experiments. Antibodies, including anti-UCP2 (C-20), anti-UCP3 (C-20), anti-MyoD (M-318), anti-cytochrome c (A-8), and anti-glyceraldehyde-3-phosphate dehydrogenase (6C5), were purchased from Santa Cruz Biotechnology. Precursor oligonucleotides (pre-miR-133a, pre-miR-1, and pre-miR-210), antisense oligonucleotides (anti-miR-133a), and scrambled negative control oligonucleotides were purchased from Ambion.
PCR Analysis—Assays to quantify mature miRNAs were conducted as described previously (
). For real-time PCR analysis, the relative amount of each gene to internal control was calculated by using the equation 2-ΔCT, in which ΔCT = CTgene-CTcontrol. For semiquantitative RT-PCR analysis, 20-30 PCR cycles were performed. All primers used are listed in supplemental Table 1.
Overexpression or Knockdown of miR-133a—L929, LLC, or C2C12 cells were seeded on 6-well plates or 60-mm dishes and were transfected the following day by using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. For each well, equal doses of precursor oligonucleotides, antisense oligonucleotides, or scrambled negative control RNA were used. Cells were harvested 48 h after transfection for semiquantitative RT-PCR, real-time PCR analysis, and Western blotting.
Plasmid Construction and Luciferase Assay—The entire mouse UCP2 3′-untranslated region (UTR) segment was amplified by PCR using mouse genomic DNA as a template. The PCR products were inserted into the p-MIR-report plasmid (Ambion). Efficient insertion was confirmed by sequencing. For luciferase reporter assays, 1 μg of firefly luciferase reporter plasmid, 0.5 μg of β-galactosidase expression vector (Ambion), and equal amounts (200 pmol) of pre-miR-133a, anti-miR-133a, or scrambled negative control RNA were transfected into cells in 6-well plates. The β-galactosidase vector was used as a transfection control. At 24 h after transfection, cells were assayed using luciferase assay kits (Promega).
Cloning of miR-133a—The genomic fragment of pre-miR-133a-1 from mouse chromosomes 18 was amplified by PCR using mouse genomic DNA as a template. The PCR products were cloned into the pcDNA(-)3.1 vector (Invitrogen), and effective overexpression was qualified by semiquantitative RT-PCR analysis after transfecting the expression vector into L929 cells.
Isolation of Mitochondria and Western Blotting—Mitochondria of UCP2 knock-out mice (gift from Bradford B. Lowell, Harvard University), wild type mice and culture cells were isolated as described previously (
). UCP2 or UCP3 protein levels were quantified by Western blot analysis of 50 μg of mitochondrial proteins with antibodies against UCP2 or UCP3. Normalization was performed by blotting the same samples with an antibody to cytochrome c. The MyoD protein level was quantified by Western blot analysis of 100 μg of whole cell extracts with antibodies against MyoD, and the normalization was performed by blotting the same sample with an anti-glyceraldehyde-3-phosphate dehydrogenase antibody.
ChIP Experiments—Chromatin immunoprecipitation experiments were performed as described previously (
). Briefly, C2C12 cells were maintained in differentiation medium (DM) for 3 days and then cross-linked with formaldehyde. Cell extracts were sonicated to generate 200-1000-bp DNA fragments. Rabbit polyclonal anti-MyoD antibody and rabbit IgG were used to immunoprecipitate chromatin fragments. The DNA-protein cross-links were reversed by heating at 65 °C for 4 h, and the recovered DNA was PCR-amplified. In the PCR analysis, 26, 20, and 48 PCR amplicons roughly 200 bp in length were designed to assay for binding sites in the 5-kb sequence upstream of pre-miR-133a-1, the 5-kb sequence upstream of pre-miR-133a-2, and the 10-kb sequence upstream of UCP3. Control amplification was carried out on input chromatin before immunoprecipitation and on mock-immunoprecipitated chromatin.
Statistical Analysis—All photo images of Western blotting, semiquantitative RT-PCR, and chromatin immunoprecipitation (ChIP) experiments were representatives of at least three independent experiments. Real-time PCR and luciferase reporter assay were performed in triplicate, and the entire experiment was repeated several times. Data shown were presented as means ± S.E. of three or more independent experiments, and the differences were considered statistically significant at p < 0.05 by using the Student's t test.
Regulation of UCP2 Expression by miR-133a—We first investigated the distribution of UCP2 mRNA and protein in mouse tissues. Western blot analysis revealed the presence of UCP2 protein in spleen and lung (Fig. 1A). This detection was specific because the 33-kDa band detected by the antibody disappeared in ucp2-/- mice (Fig. 1A). Surprisingly, we were unable to detect UCP2 protein in skeletal muscle and heart (Fig. 1A) despite a clear identification of UCP2 mRNA in these tissues (Fig. 1B). Moreover, this disparity between protein and mRNA in mouse muscle tissues seemed specific for UCP2; both the protein and the mRNA of UCP3, the most homologous of UCP2 in UCPs, were normally expressed in mouse muscle tissues (Fig. 1, B and C).
Because UCP2 protein is absent from mouse muscle tissues but is highly expressed in other tissues, it is intriguing to speculate that muscle-specific miRNAs modulate the expression of UCP2. Overall, four muscle-specific miRNAs have been identified: miR-1, miR-133a, miR-206, and miR-208 (
). Given the absence of miR-206 from heart and miR-208 from skeletal muscle, we then focused on miR-1 and miR-133a. As expected, our PCR analysis revealed that mature miR-1 and miR-133a were strictly limited to cardiac and skeletal muscle (Fig. 1, B and D).
The correlation between miRNAs and UCP2 was further examined by evaluating the expression of UCP2 after overexpression or knockdown of miRNAs. In these experiments, overexpression of miRNAs was achieved by transfecting cells with pre-miRNAs (synthetic RNA oligonucleotides mimicking miRNA precursors), whereas knockdown was achieved by transfecting cells with anti-miRNAs (chemically modified antisense oligonucleotides designed to specifically target against mature miRNAs). As shown in Fig. 2A, three exogenous pre-miRNAs, pre-miR-1, pre-miR-133a, and an unrelated miRNA precursor pre-miR-210, were introduced into the mouse lung fibroblast cell line L929. UCP2 levels were assayed with mitochondrial proteins isolated on day 2 after transfection. Control experiments were performed in the same manner except using cells transfected with scrambled oligonucleotides (negative control) or UCP2 siRNA (positive control). L929 cells were selected for this experiment because they expressed low levels of endogenous miR-1 and miR-133a but a high level of UCP2 protein (supplemental Fig. 1). The expression of UCP2 was significantly abolished by the introduction of pre-miR-133a or UCP2 siRNA, whereas cells treated with pre-miR-1, pre-miR-210, or scrambled oligonucleotides maintained a considerable protein level of UCP2 (Fig. 2A). Moreover, anti-miR-133a significantly delayed the down-regulation of UCP2 by pre-miR-133a. As shown in Fig. 2B, although pre-miR-133a alone was sufficient to reduce UCP2 protein level, cells treated with pre-miR-133a plus anti-miR-133a expressed constant UCP2 protein when compared with control cells treated with scrambled oligonucleotides. To determine at which level miR-133a influenced UCP2 expression, UCP2 transcript levels were evaluated after pre-miR-133a or anti-miR-133a treatment. Although miR-133a intracellular level was altered significantly after pre-miR-133a or anti-miR-133a treatment (Fig. 2C), overexpression or knockdown of miR-133a did not affect UCP2 mRNA stability (Fig. 2, C and D). Taken together, these results confirm that miR-133a blocks UCP2 translation without affecting its mRNA abundance, a hallmark of miRNA-mediated post-transcriptional regulation. Subsequently, the entire UCP2 3′-UTR was fused into a luciferase reporter plasmid, and the resulting plasmid was introduced into L929 cells combined with a transfection control plasmid (β-galactosidase) and pre-miR-133a, pre-miR-1, pre-miR-210, or scrambled oligonucleotides. As expected, overexpression of miR-133a resulted in a ∼60% reduction of firefly luciferase reporter activity (normalized against β-galactosidase activity) when compared with the scrambled oligonucleotides transfected cells (Fig. 2E). In contrast, pre-miR-1 and pre-miR-210 did not significantly affect the luciferase reporter activity (Fig. 2E). The observed alteration of luciferase activity was specific because co-transfection of L929 cells with the parental luciferase plasmid (without the UCP2 3′-UTR) and pre-miR-133a did not affect luciferase activity (Fig. 2E). To narrow down the miR-133a-responsive element, our in silico analysis further identified an ∼0.5-kb region that is most likely responsible for the down-modulation of UCP2 in muscle tissues. As shown in Fig. 2F, three hybrids in this region were observed. The predicted free energy values of these hybrids (-23.7, -21.9, and -21.6 kcal/mol) were all well within the range of true miRNA-target pairs. Subsequently, we constructed an equivalent luciferase reporter with these three putative miR-133a binding sites mutated (the sequence that interacts with the 2-4 bases of miR-133a were mutated). Mutant luciferase reporter was unaffected by overexpression of miR-133a (Fig. 2E). Thus, this region is likely the core sequence through which miR-133a recognizes UCP2 transcripts. Besides L929, the mouse lung cancer cell line LLC, which expressed neither endogenous UCP2 nor endogenous miR-133a (supplemental Fig. 1), was also used in the present study. As anticipated, the effect of overexpression of miR-133a in LLC cells was concordant with that in L929 cells. Overexpression of miR-133a significantly affected luciferase reporter activity, whereas anti-miR-133a could relieve the suppression of pre-miR-133a on UCP2 (Fig. 2G). These results unequivocally demonstrate that miR-133a directly recognizes the 3′-UTR of UCP2 transcripts. In addition to a chemically synthetic miRNA mimic, an expression plasmid of miR-133a was also constructed and transfected into L929 and LLC cells along with luciferase reporters. As shown in Fig. 2H, luciferase activity was markedly repressed by overexpression of miR-133a. In conclusion, these experiments unequivocally demonstrate that miR-133a directly represses UCP2 translation through a post-transcriptional mechanism.
The Braking Role of UCP2 in Myogenesis Differentiation—Because miR-133a is capable of regulating UCP2 expression in vitro, we next used the C2C12 model system to mimic muscle maturation and investigate the effect of endogenous miR-133a on UCP2 during myogenesis. C2C12 cells were propagated as undifferentiated, mononucleated myoblasts under high serum conditions. Upon serum depletion, mononucleated myoblasts fused to each other to form multinucleate myotubes (
). The differentiation in this model system was very efficient, as indicated by marked induction of myogenic markers including myogenin, myosin heavy chain (Fig. 3A, MHC), MyoD, MEF2, and skeletal α-actin (Fig. 3A, sk.a-actin). Accelerated myogenic differentiation was accompanied by a decrease in cell proliferation, as indicated by a significant decrease in expression of Ki-67 and proliferating cell nuclear antigen (Fig. 3A). MiR-133a was also robustly induced during the myoblast-myotube transition. As shown in Fig. 3, A and B, the undifferentiated myoblasts in proliferative condition did not express miR-133a. Upon differentiation into myotubes, a strong up-regulation of miR-133a was observed. UCP2 transcripts were detectable in undifferentiated primary myoblasts, but their steady-state levels increased many folds and then remained constant during the myoblast-myotube transition (Fig. 3, A and C). However, UCP2 proteins were present in proliferative myoblasts but nearly completely eliminated in differentiated myotubes (Fig. 3D) despite an increase in mRNA expression. In sharp contrast, both UCP3 transcripts and proteins were undetectable in myoblasts but strongly induced upon differentiation (Fig. 3, A and D). Hence, miR-133a and UCP3 levels were inversely related to UCP2 levels in a manner consistent with muscle development. To further investigate the suppression of elevated endogenous miR-133a on UCP2 during myogenic differentiation, we transfected proliferative myoblasts with wild type or mutant luciferase reporter plasmids. Because the luciferase activity generally wanes after a long time of cell culture, reporter activity continued to be reduced throughout the differentiation process (Fig. 3E). However, the accelerated reduction of wild type plasmid when compared with mutant plasmid (Fig. 3E) suggests that endogenous miR-133a may partially contribute to the drop of luciferase activity. Furthermore, anti-miR-133a oligonucleotides significantly delayed the reduction of luciferase activity during myogenesis, whereas pre-miR-133a hastened the drop of luciferase activity when compared with scrambled negative control oligonucleotide treatment (Fig. 3E). These results indicate that exogenous antisense oligonucleotides could partially rescue the suppression of endogenous miR-133a on UCP2. In summary, both the loss of UCP2 protein upon myotube formation without a similar reduction in mRNA and the constant inhibition of a luciferase reporter carrying the UCP2 3′-UTR during myogenic differentiation demonstrate a miRNA-mediated UCP2 translational repression mechanism.
The continuous suppression of UCP2 and induction of UCP3 during myogenesis prompt the speculation that the down-regulation of UCP2 and up-regulation of UCP3 might be indispensable to myotube formation. To verify this hypothesis, a UCP2 expression plasmid designed to specially express the UCP2 open reading frame (ORF) without the miR-133a-responsive 3′-UTR was constructed. We transfected this plasmid into C2C12 cells to rescue the suppressive effect of miR-133a on UCP2. Strikingly, miR-133a-resistant UCP2 strongly inhibited differentiation but enhanced proliferation of C2C12 cells, as indicated by the dramatic decrease of myogenin, myosin heavy chain, MyoD, MEF2, and skeletal α-actin and the marked increase of Ki-67 and proliferating cell nuclear antigen (Fig. 3, F-H, PCNA). UCP2 siRNA significantly attenuated this effect of UCP2 because cells co-transfected with the UCP2 ORF and UCP2 siRNA exhibited constant levels of myogenic and proliferative markers when compared with green fluorescent protein transfected cells (Fig. 3F, GFP). In contrast, overexpression of UCP3 caused only a slight reduction of myogenic markers and had almost no effect on proliferative markers (Fig. 3, F and H). Despite their structural and functional similarities, therefore, UCPs may play different roles in muscle cell differentiation. UCP2 may play a role at an earlier stage in the differentiation process, perhaps in maintaining the undifferentiated state and promoting myoblasts proliferation. On the other hand, miR-133a may serve to maintain and define cell types by suppressing unwanted UCP2 transcripts. Alternatively, the loss of basal UCP2 functions (e.g. mediating proton leak and controlling the production of reactive oxygen species) are compensated for by UCP3, which has a comparatively lower impact on muscle development. Hence, down-regulation of UCP2 and up-regulation of miR-133a may form a checkpoint of the myoblast-myotube transition. Taken together, our findings illustrate a novel function of UCP2 in fine-tuning myogenic differentiation of muscle cells.
The Model of UCPs and miR-133a-mediated Myogenic Differentiation Pathway—The observation that both miR-133a and UCP3 are specifically up-regulated during myogenesis leads to the hypothesis that their activation may be controlled by myogenic transcription factors. Therefore, two computer-aided algorithms, MATCH (
), were employed to search for putative myogenic transcription factor binding sites in miR-133a and UCP3 regulatory regions. A 5-kb sequence upstream from the pre-miR-133a-1 stem loop on chromosome 18, a 5-kb sequence upstream from pre-miR-133a-2 on chromosome 2, and a 10-kb sequence upstream from UCP3 start codon were screened. Within these regions, several cis elements conserved across species were noted (data not shown). These elements represented potential binding sites for essential myogenic transcription factors like MyoD, MEF2, serum response factor, and GATA-binding protein 4. Because MyoD has long been thought to represent the core regulators of myogenesis (
). As anticipated, ChIP assays using antibody against MyoD showed robust PCR product enrichment indicative of MyoD binding at -2.1 kb and -4.1 kb upstream from pre-miR-133a-1, -2.3 kb and -4.1 kb upstream from pre-miR-133a-2, and -6.7 kb and -9.7 kb upstream from UCP3 (Fig. 4, A and B). No enrichment was observed in mock-immunoprecipitated samples (without antibody) or rabbit IgG-immunoprecipitated samples (Fig. 4A). Furthermore, ChiPs experiments using pre-miR-210 promoter as a negative control for MyoD binding were performed. No binding was detected at the pre-miR-210 regulatory regions (Fig. 4A). Likewise, pre-miR-133a-1 and pre-miR-133a-2 have already been shown to be direct targets of MyoD (
). In sum, a novel MyoD-regulated signaling pathway controlling myogenesis and development could be established (Fig. 4C). As shown in this figure, differentiation cues stimulate the expression of MyoD in muscle tissues. This stimulation allows them to activate the muscle-specific gene UCP3 and simultaneously induce miR-133a to repress UCP2 expression. Down-regulation of UCP2 then removes the differentiation suppression exerted by UCP2. In other MyoD-negative tissues (e.g. spleen, lung), UCP2 is normally expressed, whereas UCP3 is silenced. UCP2 and UCP3 are generally expressed in a non-overlapping manner. Taken together, these results not only identify miRNAs as a novel link between MyoD and myogenesis but more importantly reveal a critical role for UCP2 in muscle differentiation and development. In particular, such a positive feedback loop will be expected to push the balance between muscle cell proliferation and differentiation toward the differentiation direction and maintain muscle cells in a permanently differentiated state.
The lack of reliable and specific methods for biological target validation hampers the full understanding of the mechanisms by which miRNAs execute their functions. Only a few miRNAs have thus far been assigned target mRNAs, and the conventional methodologies are still labor-intensive (
). Nevertheless, a novel approach for target identification, which employs a reverse strategy searching for miRNAs that are attached to the mRNA of a predefined target, might greatly improve the current knowledge. Using this approach, we have unambiguously addressed a key role for miR-133a in mediating UCP2 expression (Fig. 2). According to our current results, UCP2 might serve as a brake for myogenic differentiation (Figs. 3, F-H, and 4C). MyoD can remove the braking role of UCP2 by directly up-regulating miR-133a (Fig. 4, A-C), which in turn represses UCP2 expression. Thus, specific repression of UCP2 by miRNA-133a might be an indispensable step in the myoblast-myotube transition.
Although UCP2 and UCP3 share high identity with UCP1 and all three UCPs have activity in uncoupling, the primary physiological function of UCP2 and UCP3 remains unknown. A possible explanation of why UCP2 is unique in preventing muscle differentiation may be the expression pattern of UCPs. UCP1 is specifically expressed in brown fat, and UCP3 is also restricted in certain muscle tissues (
Besides mediating normal physiological processes, miR-133a targeting UCP2 may also be implicated in pathological processes. Here, we speculate that UCP2 and miR-133a may play important roles in regulating heart failure development. This postulate is supported by several pieces of evidence. Enhanced UCP2 expression has recently been reported to play an important role in the development of heart failure (
). UCP2, by virtue of its proton leak activity, decreases the generation of ATP and therefore causes myocardial energy deficiency. Interestingly, decreased miR-133a expression has been found in murine models and human disease states of cardiac hypertrophy and heart failure (
). Together, these clues point to a role for miR-133a targeting UCP2 as a heretofore hidden disease regulator of heart failure.
Taken together, our findings illustrate a novel function of UCP2 in fine-tuning MyoD-mediated myogenic differentiation. Precisely understanding the role of UCP2 and its underlying mechanisms may open up an exciting new method of treatment for diseases such as heart failure and skeletal muscle atrophy.