MicroRNA-378 Targets the Myogenic Repressor MyoR during Myoblast Differentiation*

MicroRNAs play important roles in many cell processes, including the differentiation process in several different lineages. For example, microRNAs can promote differentiation by repressing negative regulators of transcriptional activity. These regulated transcription factors can further up-regulate levels of the microRNA in a feed-forward mechanism. Here we show that MyoD up-regulates miR-378 during myogenic differentiation in C2C12 cells. ChIP and high throughput sequencing analysis shows that MyoD binds in close proximity to the miR-378 gene and causes both transactivation and chromatin remodeling. Overexpression of miR-378 increases the transcriptional activity of MyoD, in part by repressing an antagonist, MyoR. The 3′ untranslated region of MyoR contains a direct binding site for miR-378. The presence of this binding site significantly reduces the ability of MyoR to prevent the MyoD-driven transdifferentiation of fibroblasts. MyoR and miR-378 were anticorrelated during cardiotoxin-induced adult muscle regeneration in mice. Taken together, this shows a feed-forward loop where MyoD indirectly down-regulates MyoR via miR-378.

Skeletal muscle differentiation from specification of the myotome through the formation of mature myofibers is driven by basic helix-loop-helix transcription factors referred to as the muscle regulatory factors (MRFs) 2 . MRFs heterodimerize with the ubiquitous E-proteins E12/E47 and bind to DNA motifs known as E-boxes (1). Although all MRFs have similar DNA binding motifs, their spatiotemporal regulation and role in development is unique. A well characterized MRF is MyoD, which was noted for its ability to differentiate fibroblasts into myotubes (2). MyoD has roles in both lineage specification and terminal differentiation. MyoD has a broad range of targets and has been shown to bind to thousands of locations in the genome (3). In the C2C12 myoblast cell line, MyoD is expressed but is not active as a transactivator at loci of differentiation-associated genes until myotube formation is induced by low-serum conditions. There are a number of inhibitors of MyoD that prevent its premature activation. Among these are the inhibitor of DNA binding proteins (4), Twist (5), and a competing basic helix-loop-helix called MyoR or Musculin (6). MyoR inhibits MyoD by binding to E-proteins and binding directly to MyoD target DNA sequences. MyoR is present during early embryonic stages but down-regulated during secondary embryogenesis. MyoR is expressed in C2C12 myoblasts, and down-regulated upon differentiation to myotubes.
A second level of regulation of the myogenic program involves microRNAs. They are initially transcribed as a long transcript by Pol II or Pol III, which is referred to as the pri-miRNA. The RNA then folds into a hairpin and is cut by Drosha/DGCR8 into a hairpin-shaped pre-miRNA and exported to the cytoplasm. Most miRNAs are then processed by Dicer into a 19-to 24-bp single-stranded mature miRNA. One strand of the hairpin is then preferentially loaded into the RNAinduced silencing complex. Silencing is achieved through destabilizing target mRNAs and blocking translation. Several miRNAs that are normally induced during myogenic differentiation can initiate the myogenic program even in the presence of high serum in C2C12 (7,8). The targets of these myogenic miRNAs include cell cycle machinery (9), Pax family transcription factors (10 -12), and chromatin remodelers (13). Because the activity of MyoD as a transactivator increases during differentiation, we explored whether differentiation-induced microRNAs have an impact on targets that regulate MyoD activity. In this paper, we show that miR-378, a microRNA that is up-regulated during differentiation by MyoD, plays a role in the activation of MyoD by targeting its inhibitor MyoR. This interaction illustrates a new mechanism that allows MyoD to change its transcriptional program in the transition from proliferating myoblasts to differentiating myotubes.

MATERIALS AND METHODS
Cell Culture-C2C12 myoblasts were cultured in DMEM with 20% FBS and 1% penicillin/streptomycin growth medium. 10T1/2 fibroblasts were cultured similarly but with 10% FBS. Cells were induced to differentiate by replacing 20% FBS with DMEM containing 2% horse serum (differentiation medium (DM)). Transfection with miRNA mimics or inhibitors was performed as described before (12). Retroviruses for the creation of stable cells were generated in 293T cells. All transfections were performed with Lipofectamine 2000 (Invitrogen). Cells were transfected with the viral vector as well as a vesicular stomatitis virus glycoprotein-and a gag/pol-encoding plasmid. After 48 h, the supernatant was removed, centrifuged to pellet any 293T cells, and passed through a 0.45-m filter. This viral supernatant was then added to the target cells in the presence of polybrene. Stable cell lines were selected by puromycin (1.5 mg/ml) for 48 h.
Plasmids-Mir-378-encoding oligonucleotides were cloned into a plasmid based on the miR-30 conformation as described previously (14). This was done to eliminate any effect from the miR-378* strand. The miR-378 inhibitor design was based on previous work (15) and ligated into tough decoy plasmids that were a kind gift from Dr. Hideo Iba. The short hairpin for MyoD knockdown was cloned into pLKO.1 plasmid. The target sequence on MyoD was described before (16). MyoR expression plasmids were cloned by PCR from C2C12 cDNA and ligated into pBabe-puro vectors. Mutagenesis was performed by PCR amplification and DpnI digestion to remove parental DNA. The MyoR 3Ј UTR sequence was PCR-amplified from C2C12 genomic DNA and cloned into the pRL-CMV vector. Enhancer activity was tested by cloning an approximately 500base pair region around the microRNA binding site and ligating it into a pGL3 promoter plasmid. MCK luciferase was a kind gift from Dr. Stephen Tapscott. The 4RE plasmid was ordered from Addgene (plasmid 16057).
RT-PCR-Cells were lysed, and total RNA was extracted using TriZOL reagent (Invitrogen) following the manufacturer's instructions. Ncode miRNA first-strand cDNA synthesis and quantitative RT-PCR kit (Invitrogen) were used to perform quantitative RT-PCR for microRNA detection. For mRNA detection, cDNA synthesis was carried out using the Superscript III first-strand synthesis system for RT-PCR (Invitrogen). Then, quantitative PCR was carried out using SYBR Green PCR FIGURE 1. miR-378 is up-regulated during C2C12 differentiation. A, miR-378 levels increase during C2C12 differentiation, predominantly in the first 2 days after serum withdrawal. The miR-378 level was normalized to that of snU6, and the ratio on day 0 was set to 1. Data are mean Ϯ S.D. of three measurements. B, MyoD ChIP-seq shows MyoD binding relative to the miR-378 locus in C2C12 myoblasts (top panel) and myotubes (bottom panel). The relevant portion of the Ppargc1b gene containing the miR-378 locus is shown. The y axis represents the number of independent sequence reads that overlap within a given 10-bp window in the MyoD ChIP. The primary data is from Ref. 3. C, the 3Ј binding peak of MyoD that was detected only in myotubes (denoted by the asterisk in B) contained three E-box sequences consistent with the MyoD consensus sequence. All three showed significant mammalian conservation. master mix in an ABI cycler. ABI 7300 software was used for quantification (Applied Biosystems).
ChIP and High Throughput Sequencing Mapping-We used Novoalign (version 2.05.04) to align reads from the Sequence Read Archive (SRP001761) to the mouse reference genome (NCBI v37, mm9). MACS (version 1.3.6.1) was used to identify areas enriched over the background without a control sample. From MACS, we created wig files with peaks of at least 10-fold enrichment and a window size of ten base pairs. For the H4ac experiment, the MyoD and pCLBABE samples were aligned with Novoalign using default parameters. For each sample, we collected unique reads with an alignment score of at least 50 and generated wig files of the enrichment using MACS without a control sample. The log 2 of MyoD to pCLBABE enrichment for H4Ac-bound DNA was calculated for the chromosome 18 1400000 -61600000 region. To ensure that the ratio was always defined, we set the enrichment at each position to be at least 1.
Animal Experiments-The use of animals was approved by the Animal Care and Use Committee of University of Virginia. Eight-week-old male C57BL/6 mice were anesthetized with isoflurane and sacrificed by cervical dislocation to harvest muscle. For the regeneration assay, the injury was performed on tibialis anterior muscles of mice by injecting 100 l of 10 M cardiotoxin (CTX). In groups of n ϭ 5, mice were sacrificed at 1, 3, 5, 7, and 14 days post-injection to collect the tibialis anterior muscles.

Identifying miR-378 as a Myo miRNA and Identifying That Its
Level Is Regulated by MyoD-To fully catalogue the miRNA species that are induced in myogenic differentiation, we hybridized short RNAs from proliferating and differentiating C2C12 cells to a locked nucleic acid array of probes for miRNA (12). miR-378 was one of the species that was up-regulated more than 4-fold during differentiation. We confirmed this change of mature miRNA levels by quantitative PCR. As shown in Fig. 1A, there was a slight increase on day 1, a larger increase on day 2, and no further increase up to day 4. The miR-378 coding sequence lies in the first intron of the Ppargc1b gene. There is a fair amount of conserved DNA sequence in the first intron adjacent to the coding sequence of the miRNA (Fig. 1B). We hypothesized that some of the conserved sequence is important for regulating miR-378 expression.
Of the myogenic transcription factors to consider as possible inducers of miR-378, the first candidate was MyoD because the increase of miR-378 was observed early in differentiation concurrent with MyoD transactivation activity. This would not be consistent with myogenin or other factors that are induced later during differentiation. We analyzed the sequencing data pub-lished from a MyoD ChIP-seq experiment in C2C12 (3). A conclusion of the authors was that areas with a higher read density in myotubes than myoblasts were correlated with increased expression of associated genes. Less than 1 kb downstream of miR-378, we mapped a ChIP-seq peak that was below our threshold for calling a peak in myoblasts (10 reads per 10-bp window) but had a maximum read density of 38 per 10-bp window in myotubes (Fig. 1b). MyoD has been known to bind to a DNA sequence called E-box with the canonical sequence CANNTG. The ChIP-seq study showed that CAGCTG and CAGGTG are the sequences that are specifically enriched within MyoD-bound regions. The binding peak downstream of miR-378 contained three such E-boxes, all of which had at least partial conservation among mammals (Fig. 1C).
Because not all of the MyoD-bound regions are correlated with increased gene expression during differentiation, we first tested whether MyoD binding at this peak regulates miR-378 transcription. We knocked down MyoD by an shRNA expressed from a lentiviral vector in C2C12 myoblasts ( Fig. 2A). The primary miR-378 transcript was then measured because mature sequences are very stable, and the pri-miRNA level would more accurately reflect recent changes in transcriptional activity. There was significantly less pri-miR-378 in myoblasts expressing shMyoD versus a control hairpin vector (Fig. 2B). From this we conclude that MyoD is involved in regulating miR-378 levels in myoblasts.
MyoD Regulates miR-378 by Both Chromatin Remodeling and Transcriptional Enhancer Activity-One of the ways MyoD potentiates myogenic transdifferentiation is by recruiting chromatin remodeling proteins. Specifically, MyoD is known to bind to histone acetyltransferases p300 and PCAF (17). This leads to the local enrichment of acetylated histone H4, which is a marker of transcriptional activity. The same group that performed the MyoD ChIP-seq in C2C12 also performed a ChIP-seq with a pan-acetyl-H4 antibody in fibroblasts transfected with MyoD or a control vector (3). Identifying changes in histone modifications in MyoD-transfected samples will reveal which regions are epigenetically modified during myogenic transdifferentiation. As shown in Fig. 3A, mapping the reads from this experiment showed a significant enrich-ment of acetylated H4 in the cells transfected with MyoD over the control in the region of miR-378. To determine whether MRFs other than MyoD are activating differentiation-induced genes, C3H10T1/2 fibroblasts were used because they have no inherent myogenic properties but can readily differentiate into myotubes when MyoD is expressed exogenously (2). Full differ-entiation does not occur until the cells are placed in low-serum DM. Therefore, a gene directly responsive to MyoD should increase once the cells are transfected with a MyoD-expressing plasmid, with a larger increase upon serum withdrawal. A gene that is activated by another MRF or a secondary response will only be turned on after being placed in DM. miR-378 showed a FIGURE 3. The MyoD binding peak near miR-378 function for both chromatin remodeling and as a transactivator in naïve fibroblasts. A, log base 2 ratio of histone H4 acetylation between 10T1/2 fibroblasts infected with a MyoD-expressing virus relative to control fibroblasts. For reference, the MyoD binding sites in C2C12 myotubes (Fig. 1B) are shown above. B, mature miR-378 levels in 10T1/2 fibroblasts in growth medium (GM) and low-serum DM when transfected with empty vector (EV) or MyoD and E12-expressing plasmids. miR-378 levels were increased by MyoD/E12 transfection and further increased by DM. Details are as in Fig. 1A. C, the MyoD binding site from miR-378 (the asterisk in B) was cloned into an enhancer-less luciferase vector. Luminescence increased 4-fold with MyoD/E12, but this was abrogated by adding exogenous MyoR or mutating the E-boxes. RLU, relative light unit. Data are mean Ϯ S.D. of three experiments.
pattern of directly induced genes, with a 2-fold induction after cotransfection with MyoD and E12 and another 2-fold induction after switching to DM (Fig. 3B). The MyoD binding site near miR-378 was PCR-amplified out of genomic DNA and cloned into an enhancer-less luciferase plasmid that contained a core promoter. Cotransfection of MyoD and E12 with this luciferase reporter stimulated luciferase activity by 4-fold (Fig. 3C), exactly corresponding with the increase of miR-378 detected by quantitative PCR in Fig. 3B. To prove that this is a direct function of MyoD, increasing amounts of MyoR were added. MyoR (100 ng) reduced activation to baseline levels. In addition, mutation of the E-boxes in the miR-378 enhancer prevented activation of the luciferase reporter by MyoD. Taken together, these results suggest that the E-boxes near miR-378 constitute both a site for chromatin remodeling and a MyoD-responsive enhancer that can be inhibited by MyoR.

miR-378 Modulates MyoD Activity by Repressing MyoR-To
elucidate the biological function of miR-378 we first tested whether it had any effect on MyoD transcriptional activity. One of the simple ways to measure MyoD transcriptional activity is to use a firefly luciferase-expressing plasmid that contains a multimerized E-box upstream from a minimal promoter (4RE). When C2C12 cells were cotransfected with this reporter and a plasmid that constitutively expresses miR-378, there was an increase in luminescence (Fig. 4A). A more physiologically relevant target sequence, the promoter of MCK fused to firefly luciferase, was also activated by miR-378 expression.
We then wanted to test whether modulation of miR-378 levels can have an effect on the overall kinetics of C2C12 differentiation. After 24 h in DM, overexpression increased mRNA levels of myogenin (Fig. 4, B and C). Likewise, expression of myosin heavy chain was increased after 72 h in DM. Inhibition of miR-378 with 2Ј-O-methyl antisense was associated with a decrease  Fig. 3C. B, addition of a synthetic miR-378 mimic increases markers of myogenic differentiation in C2C12. The ratio of myogein or MHC to GADPH in cells infected with control (GL2) is set to 1. Data are mean Ϯ S.D. of three experiments. C, blocking miR-378 by addition of an inhibitor decreases the mRNA levels of myogenin in C2C12 differentiation. D, schematic of the WT binding site in the MyoR 3Ј UTR for miR-378. The mutated seed sequence was used as a negative control. E, a luciferase construct containing the MyoR 3Ј UTR is repressed by addition of miR-378, and the repression is relieved by mutating the seed sequence binding site (MyoR Mut). miR-378cs is a positive control that contains a perfect match of miR-378 in the 3Ј UTR of luciferase. Data are mean Ϯ S.D. of three experiments. JUNE 3, 2011 • VOLUME 286 • NUMBER 22 in levels of myogenin mRNA but a minor decrease in MHC that was not statistically significant. These data suggest that although miR-378 can promote differentiation, it is not the only factor that can do so. Thus induction of the microRNA promotes differentiation, but inhibition of the microRNA has a minor effect on differentiation.

MicroRNA-378 Targets the Myogenic Repressor MyoR
By using bioinformatic prediction software, we discovered a putative target site of miR-378 in the 3Ј UTR of MyoR (Fig. 4C). There is perfect complementarity to base pairs 2-8 of the miRNA. Perfect Watson-Crick pairing in this "seed sequence" is the most important property for miRNA targeting (18). We cloned the 3Ј UTR of MyoR into a reporter as the UTR of luciferase. Adding a plasmid encoding miR-378 reduced the luminescence of the MyoR 3Ј UTR containing luciferase to 60% of the control, consistent with the notion that it contains a functional miR-378 target site. When the miR-378 binding site is mutated, luminescence returned to control levels (Fig. 4D), indicating that this site was indeed the unique sequence targeted by miR-378 in the MyoR 3Ј UTR.
To test if this binding site could be used to functionally regulate MyoR, we tested the ability of MyoR to prevent MyoDmediated transdifferentiation of 10T1/2 fibroblasts, where MyoR is normally transcribed at a very low level. 10T1/2 cells were infected with a retrovirus encoding different constructs of MyoR, and stable cell lines were created with puromycin selection. These cells were then transfected with MyoD, and differentiation was assessed quantitatively by cotransfection of MCK luciferase. Cells were placed in differentiation media 24 h after transfection and harvested 48 h later. The MyoR ORF without the 3Ј UTR dramatically reduced differentiation, consistent with work published previously (6). A vector with both the ORF and the 3Ј UTR was less effective in blocking MCK luciferase activity. Mutation of the seed sequence of the binding site, preventing repression by miR-378, mostly restored this block of MyoD activity (Fig. 5A). The finding was confirmed by a more physiological readout, and the endogenous MHC mRNA levels were measured by suing quantitative RT-PCR (Fig. 5B). As judged by MHC mRNA, differentiation was decreased by MyoR. The 3Ј UTR with an intact miR-378 target site mediated miR-378 inhibition of MyoR activity and so allowed better differentiation.
The inhibitory effect of microRNAs on protein level is achieved primarily through destabilization of target mRNA (19). To show that a change in MyoR expression was behind the differences observed above, we assayed MyoR mRNA in 10T1/2 cells after 2 days in growth and differentiation conditions. The ORF showed a slight reduction, and including the 3Ј UTR significantly decreased MyoR mRNA levels. Mutating the miR-378 binding site alleviated some but not all of this repression (Fig. 5C). This suggests the possibility of a separate repressor that binds the 3Ј UTR, possibly another miRNA. We expect the MyoR protein levels to follow the change in mRNA, but the available MyoR antibodies did not give specific signals on Western blot analyses. These results suggest that the repression of MyoR by miR-378 is sufficient to affect the ability of MyoR to inhibit differentiation.

MyoR and miR-378 Levels Are Anticorrelated during Muscle
Regeneration-To show the regulation of MyoR by miR-378 in skeletal muscle regeneration, we measured the RNA level in mouse skeletal muscle recovering from cardiotoxin treatment. We injected a cohort of mice in the tibialis anterior simultaneously and harvested five mice per time point up to 2 weeks after injection. By quantitative RT-PCR, we analyzed miR-378 levels along with MyoD and MyoR. miR-378 levels decreased dramatically after injury, down to 0.5% of preinjection levels, as was expected for a miRNA that is induced during myogenic differentiation. By day 14, it had returned to 50% of saline-injected mice (Fig. 6A). The regulation follows a similar pattern as canonical myomiRs, miR-1, and miR-133, but unlike miR-206, whose level saturates in the first week (20). miR-378 levels were roughly anticorrelated with MyoR, which peaked on day 5 of regeneration and then continually declined at each subsequent time point (Fig. 6B). This decrease in MyoR occurred at the same time when miR-378 levels increased, suggesting that our findings in C2C12 cells likely reflect the regulation of MyoR by miR-378 in vivo.
It is interesting to note that miR-378 levels were not correlated at all with MyoD transcript levels or even the MyoD:MyoR ratio. This makes it unlikely that MyoD is the only regulator of miR-378 levels in regenerating skeletal muscle. However, the pattern of expression of miR-378 is similar to miR-1 and miR-133, two microRNAs known to be regulated by MyoD, myogenin, MEF2 (21), and serum response factor (22).

DISCUSSION
In this study, we identify miR-378 as another miRNA that plays a role in skeletal muscle differentiation. MiR-378 forms the basis of a simple positive feedback loop whereby it is up-regulated by MyoD and then targets MyoR, a repressor of MyoD transcriptional activity (Fig. 7).
MyoD has often been called the "master switch" of skeletal muscle differentiation because of its ability to singularly transdifferentiate fibroblasts. This simplicity of a single reprogramming factor has not been reproduced in other systems such as cardiomyocyte differentiation (23). It has been known that miRNAs are regulated by the MRFs, specifically MyoD and myogenin (24 -27), although most of the work has focused on miR-1, miR-133, and miR-206. Additionally, miR-22, miR-100, miR-138, and miR-191 have been identified as having MyoD binding peaks in their promoters (28), but their biology remains largely unknown, and most were shown not to have altered expression levels in MyoD knockout myoblasts (29). This miR-mediated regulation was missed in the earlier studies because they were based upon ChIP-chip experiments where the probes on the array were restricted to promoter sequences. Here we have identified a MyoD binding region that is very proximate to miR-378, with properties consistent with a role in transcriptional control. There is a large increase in the amount of histone H4 acetylation at the miR-378 locus when naïve fibroblasts are transfected with MyoD. The MyoD binding site functions as an enhancer in a luciferase assay, and the function is ablated when MyoR is added or the E-boxes are mutated. Knockdown of MyoD reduces levels of the miR-378 primary FIGURE 6. miR-378 is repressed in skeletal muscle during regeneration after snake venom cardiotoxin treatment. A, mature miR-378 decreases dramatically to 0.5% of the control 3 days after venom injection and then gradually recovers (normalized to snU6 transcript level, ratio in control injected animals set to 1). Average of technical triplicate. n ϭ 5. Data are mean Ϯ S.E. B, MyoR transcript levels peak 5 days after injection and then decreases, with a general anticorrelation to miR-378. The MyoD transcript peaks on day 3. Both transcripts are normalized to GAPDH. Other details are as in Fig. 6A.

MicroRNA-378 Targets the Myogenic Repressor MyoR
transcript in myoblasts. miR-378 levels during cardiotoxin-induced injury, and regeneration mimics the established MyoDresponsive miRNA genes miR-1 and miR-133. We conclude that the MyoD-binding site near miR-378 is functional and results in increased transcription early in myoblast differentiation. The fact that miR-378 levels lag behind MyoD in cardiotoxin-induced muscle regeneration should not be interpreted to mean that this regulation is not functional in vivo; rather, that understanding the regulation of miR-378 is incomplete. Myogenic cells have many cell signaling pathways that cross-talk with MyoD, including Notch, FGF, and TGF-␤ (30), and these cross-talks could possibly delay the up-regulation of MyoD targets. As miR-378 levels so closely mimic other myomiRs during cardiotoxin-mediated muscle regeneration, it will be interesting to see if transcription factors such as MEF2 and SRF also affect miR-378 levels.
The down-regulation of MyoR during C2C12 differentiation was established soon after it was identified as an inhibitor of myogenesis (6), but the mechanism was not established. Further studies have suggested the importance of Hes6 (31), but the evidence was correlative without demonstrating direct binding of Hes6 to the MyoR gene. Notch has been postulated as a positive regulator of MyoR (32), but no binding site for the canonical Notch DNA-binding protein Rbpj has been located in the promoter. Here, we have demonstrated the first direct interaction of a differentiation-induced repressor of MyoR within the MyoR gene or transcript. Interestingly, Hes6 is also thought to be up-regulated by MyoD (33) and therefore could be acting redundantly with miR-378 in a feed-forward loop from MyoD to repress MyoR. This could also explain why the repression of MyoR and the induction of miR-378 after cardiotoxin injury were not perfectly correlated. Notch signaling inhibits MyoD (34); therefore, it is possible that the up-regulation of MyoR that was reported after stimulation of the Notch pathway is just reflective of the loss of miR-378 and Hes6, two repressors of MyoR that are normally induced by MyoD.
Myogenic transcription factors and the repressors have long been accepted as forming regulatory loops that allow the graduated induction of myogenesis interspersed with metastable progenitors. Two recent discoveries now begin to interpolate microRNAs in these regulatory loops. miR-206 was recently shown to be induced by MyoD to inhibit Pax7 and thus decrease Id2, a repressor of MyoD (12). In this report, we demonstrate that miR-378 plays a similar role in that it is induced by MyoD to directly repress MyoR. It will be interesting to learn of other such examples in myogenic differentiation. Also, similar regulatory loops involving the interaction of transcription factors, microRNA, and inhibitors of differentiation may occur in other types of tissue.