A genome-wide screen had previously shown that knocking down miR-98 and let-7g, two miRNAs of the let-7 family, leads to a dramatic increase in terminal myogenic differentiation. In the present paper, we report that a transcriptomic analysis of human myoblasts, where miR-98 was knocked down, revealed that approximately 240 genes were sensitive to miR-98 depletion. Among these potential targets of miR-98, we identified the transcriptional repressor E2F5 and showed that it is a direct target of miR-98. Knocking down simultaneously E2F5 and miR-98 almost fully restored normal differentiation, indicating that E2F5 is involved in the regulation of skeletal muscle differentiation. We subsequently show that E2F5 can bind to the promoters of two inhibitors of terminal muscle differentiation, ID1 (inhibitor of DNA binding 1) and HMOX1 (heme oxygenase 1), which decreases their expression in skeletal myoblasts. We conclude that miR-98 regulates muscle differentiation by altering the expression of the transcription factor E2F5 and, in turn, of multiple E2F5 targets.

INTRODUCTION

Skeletal muscle differentiation is a complex process regulated at multiple levels. Although the transcriptional control of gene expression during determination and differentiation has been very well studied, the knowledge of the affect of post-transcriptional regulation in these processes is less advanced. However, multiple studies have identified miRNAs as important factors in translational regulation of gene expression in skeletal muscle, characterizing a large number of myogenic miRNAs (myomiRs), as well as some of their mRNA targets (reviewed in [1]). In our previous work, we have shown that as many as 63 miRNAs have an effect on terminal differentiation of human skeletal myoblasts [2]. Whereas the majority of these miRNAs are positive regulators of differentiation, a small group was shown to suppress this process. The most striking example of such differentiation-suppressor miRNA was the let-7 family, more specifically, miR-98 and let-7g.

The let-7 family is one of the most widely and abundantly expressed, best-conserved and functionally important miRNA families known to date (reviewed in [3,4]). In humans, this family consists of 13 highly similar miRNAs (let-7a–i, miR-98 and miR-202) that share the same seed sequence and in most cases were shown to target the same mRNAs and to have redundant functions (reviewed in [3]). In most instances, let-7 miRNAs had a positive effect on differentiation [57]. However, in skeletal muscle cells, miR-98 (and let-7g) appeared to down-regulate the differentiation. In order to address how let-7 could have this unusual behaviour, we used two different strategies to identify let-7 gene targets in these cells. First, transcriptomic analyses led to the identification of a number of putative target genes with modified expression in miR-98 loss-of-function or gain-of-function assays. Secondly, a co-suppression assay, in which both the miRNA and the putative target genes were concomitantly knocked down, demonstrated that the transcriptional repressor E2F5 is largely responsible for the hypertrophic phenotype of human and mouse myoblasts following miR-98 KD (knock down). We show that E2F5 is a direct target of miR-98. Our further analyses identified several novel direct targets of E2F5 that are known down-regulators of muscle differentiation, such as ID1 (inhibitor of DNA binding 1). In conclusion, we have deciphered an anti-differentiation network controlled by miR-98 in skeletal muscle cells.

MATERIALS AND METHODS

Cell culture

All cell lines were grown at 37°C in a humidified 5% CO2 atmosphere. C2C12 mouse myoblast and HeLa uterine cervical carcinoma cell lines were purchased from ATCC (American Type Culture Collection) and cultured as indicated on the ATCC website. LHCN–M2 immortalized human myoblasts [8] were cultured under standard conditions. Differentiation was induced by switching to DMEM (Dulbecco's modified Eagle's medium)/antibiotics supplemented with 0.01 mg/ml insulin and 0.1 mg/ml transferrin (Invitrogen).

RNA interference

Transient transfection of siRNA was performed using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's instructions. The efficiency of the siRNA-mediated KD of gene expression was evaluated by quantitative reverse transcription PCR (qRT-PCR) and/or Western blotting (WB). The siRNA referred to as E2F5_1 was purchased from Qiagen (Hs_E2F5_2 siRNA, catalog number SI00030443). The siRNA referred to as control 1 was designed in Guido Kroemer laboratory. The siRNAs referred as control 2, E2F5_2 and E2F5_3 were purchased from Dharmacon (ON-TARGETplus siRNAs D-001810-01, J-003263-10 and J-003263-12 respectively).

siRNA target sequences were: control 1, AAGCCGGTATGCCGGTTAAGT; E2F5_1, AUCAAUCUAAAGAGUCAUUCA; E2F5_2, GGUGCUGGCUGUAAUACUA; E2F5_3, GAGGUACCCAUUCCAGAAA.

Plasmid transfection

Transient transfection of plasmids was performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Efficiency of transfection was evaluated by immunofluorescence (IF) and/or WB. Expression vector for E2F5 (MGC Mouse E2F5 cDNA) was purchased from Thermo Fisher Scientific.

Immunofluorescence

IF was performed as previously described [9]. Image acquisitions were carried out on a Zeiss LSM510 Meta confocal microscope with a Plan Neofluar 100×, NA (numerical aperture)=1.3 oil immersion objective. Pinhole apertures were set to one Airy unit for each wavelength.

Quantitative real-time RT-PCR

For qRT-PCR, RNA was extracted using TriZOL reagent (Life Technologies) according to the manufacturer's instructions. RT was performed using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative real-time RT-PCR was then performed using Power SYBR-Green Master Mix (Applied Biosystems) on an Applied 7500 Real-Time PCR system (Applied Biosystems). Primers used for qRT-PCR were as follows: human/mouse cyclophilin A (cyclo A), forward, GTCAACCCCACCGTGTTCTT and reverse, CTGCTGTCTTTGGGACCTTGT; human E2F5, forward, CCAGAAATGGGTCAGAATGGA and reverse, CCATGCTGGATTTCTGTGGA; mouse E2F5, forward, CACCAATGTCTTAGAGGGAATTGA and reverse, TGCTTTGCTGTAGCCACAACTT; human HMOX1 (heme oxygenase 1), forward, TCAACATCCAGCTCTTTGAGGA and reverse, GCTGAGTGTAAGGACCCATCG; human ID1, forward, GAGGCGGCATGCGTTC and reverse, CCCAGGCTGGATGCAGTTA; human MSTN (myostatin), forward, GAAGATGGGCTGAATCCGTTT and reverse, AATCCAATCCCATCCAAAAGC; human transforming growth factor β (TGFB) 2, forward, CGAGAGGAGCGACGAAGAGT and reverse, CAACTGGGCAGACAGTTTCG; human TGFB3, forward, GAGCAGAATTCCGGGTCTTG and reverse, ATCTGGCCGAAGGATCTGG.

Chromatin immunoprecipitation

C2C12 cells, either in proliferation or differentiation, were washed twice with PBS, treated with formaldehyde (Sigma) diluted in PBS to a final concentration of 1% during 10 min at 37°C then stopped by adding glycine at a final concentration of 0.125 M. Fixed cells were washed twice with cold PBS and harvested. Chromatin was prepared first by a step of cell lysis in cell lysis buffer (5 mM Pipes, pH 8, 85 mM KCl and 0.5% NP-40) for 5 min on ice. Lysed cells were then centrifuged for 10 min at 600 g and resuspended in nucleus lysis buffer (50 mM Tris/HCl, pH 8, 10 mM EDTA and 1% SDS). Chromatin was then sheared by sonication (10 cycles of 1 min on, 1 min off) in a Bioruptor (Diagenode) to obtain DNA fragments of around 500 bp. Cell debris was then cleared by centrifugation at 16000 g for 10 min.

For immunoprecipitation, Protein A/G ultralink beads (Thermo Fisher Scientific) were blocked overnight at 4°C in IP dilution buffer (16.7 mM Tris/HCl, pH 8, 1.2 mM EDTA, 167 mM NaCl, 0.01% SDS and 1.1% Triton X-100) supplemented with 1 mg/ml BSA (New England Biolabs) and 1 mg/ml salmon sperm DNA (Life Technologies). Beads were then washed for 4 h at 4°C in IP dilution buffer.

Chromatin was diluted in 1 ml of IP dilution buffer and pre-cleared for 4 h at 4°C by the addition of pre-blocked beads. Beads were removed and 10 μg of chromatin was immunoprecipitated with 5 μg of the specific antibody or rabbit IgG as a negative control, overnight at 4°C. Immunocomplexes were recovered by the addition of pre-blocked beads, incubated for 1 h at 4°C. Beads were washed twice with IP dilution buffer, twice with TSE buffer (20 mM Tris/HCl, pH 8, 2 mM EDTA, 500 mM NaCl and 1% Triton X-100), twice with LiCl buffer (100 mM Tris pH 8, 500 mM LiCl, 1% desoxycholic acid, 1% NP-40), and twice with TE buffer (10 mM Tris/HCl, pH 8, and 1 mM EDTA). Immunocomplexes were then twice eluted from the beads with 250 μl of elution buffer (50 mM NaHCO3 and 1% SDS).

Eluates were supplemented with 200 mM NaCl and cross-linking was reversed overnight at 65°C. Eluates were then treated with RNAse followed by proteinase K, extracted with phenol:chloroform:isoamyl alcohol (125:24:1, pH 4.5), precipitated with 100% ethanol supplemented with 90 mM NaAc and finally resuspended in DNAse-free water. qPCRs were performed with Power SYBR-Green Master Mix (Applied Biosystems) in an Applied 7500 real-time PCR system (Applied Biosystems). Chromatin immunoprecipitation (ChIP)–qPCR results are represented as a percentage of the negative control gene neuronatin (NNAT).

The following primers were used for qPCR: HMOX1, forward, CACGTGACCCGCGTACTTAA and reverse, ACTCACTGGTTGTATGCGGAAA; ID1, forward, CAAAGCCACGCCTCTAGCA and reverse, GACCAAGGGCGGGTTCA; NNAT, forward, CCCTACCCAACCCATCCTATC and reverse, CCACCGCGGCACTTTG.

Antibodies

The following antibodies were used: for WB and IF, anti-E2F5 (sc-999 for WB, sc-374268 for IF), anti-MCK (muscle creatine kinase; sc-15161), anti-myogenin (sc-576) were obtained from Santa Cruz, monoclonal anti-myosin heavy chain (M4276, clone MY-32) and anti-actin (A5441, clone AC-15) were purchased from Sigma Aldrich; for ChIP, anti-E2F4 (sc-866×) and anti-E2F5 (sc-1083×) were obtained from Santa Cruz; goat IgG (I 9140) and mouse IgG (I 8765) were purchased from Sigma Aldrich; rabbit IgG (sc-2027) for ChIP were purchased from Santa Cruz.

Luciferase reporter assays

The 3′ UTR of human E2F5 mRNA (harbouring the miR-98 binding site) was cloned into a psiCHECK2-reporter vector (Promega) downstream of the reporter gene (Renilla luciferase). HeLa cells were seeded at 20000 cells per well in a 96-well plate. After 24 h, 10 ng of psiCHECK2–E2F5–3′ UTR was co-transfected with 50 nM of either miRNA mimic or miRNA inhibitor. Co-transfection was performed with Lipofectamine 2000 (Life Technologies). At 48 h after transfection, the relative levels of Renilla compared with firefly luciferase activity (control of transfection efficiency) were measured with a Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions. The luminescence signal was quantified on a Mithras LB 940 Multilabel Reader and analysed with Mikrowin software (Berthold Technologies).

miRNA inhibitors and mimics

For miR-98 inhibition assays, the LNA (locked nucleic acid) to miR-98 and the control LNA (targeting an irrelevant sequence) were obtained from Exiqon. For let-7 overexpression, human let-7g miScript miRNA mimic was purchased from Qiagen. AllStars negative-control siRNA (Qiagen) was used as a negative control.

Transient transfection of LNA and mimic were performed using Lipofectamin RNAiMAX (Invitrogen) according to the manufacturer's instructions.

Image acquisition and analysis of differentiation

Terminal differentiation was monitored using an Operetta (Perkin-Elmer) High Content Screening system as described [2]. The fusion index (the number of nuclei in myotubes containing at least three nuclei above the total number of nuclei) was measured and used as a parameter to monitor terminal differentiation.

E2F site prediction

Putative E2F binding-sites were identified with the software TRANSFAC from Biobase Gmbh (http://www.biobase-international.com/).

Transcriptomic analysis

Variations of the transcriptome were analysed on Affymetrix Human Gene 1.1 ST arrays. Values of a fold change for all genes in each experimental set (proliferation, early and late differentiation) were ranked from the highest to the smallest ones and, next, the inflection-point of the corresponding curve was considered as a threshold for significantly high fold change for a particular experimental set. According to this analysis, the threshold for Affymetrix data for the late differentiation set was set at 2-fold change and for the proliferation and early differentiation sets at 1.5-fold change.

RESULTS AND DISCUSSION

miR-98 KD increases differentiation in human and mouse myoblasts by massively altering gene expression

A genome-wide screen previously identified miR-98 and let-7g as negative regulators of human skeletal muscle cell differentiation: these miRNAs suppress hypertrophic and/or precocious differentiation of myotubes [2]. To further explore the mechanisms underlying this phenomenon, we performed miR-98 loss-of-function assays in human and mouse skeletal myoblasts. As we have previously shown, an LNA antisense miR-98 inhibitor is very efficient against the let-7 family in myoblasts, with a strong effect on let-7 targets [10]. LNA antisense inhibitors act by binding their target miRNAs with high affinity and efficiency [11]. Transfection of miR-98 LNA inhibitor leads to a strong and long-term (at least 5 days in human myoblasts) KD of miR-98 levels [2,10] (Figures 1A and 1B).

miR-98 KD increases differentiation in LHCN (human) and C2C12 (mouse) myoblasts by massively altering gene expression

Figure 1
miR-98 KD increases differentiation in LHCN (human) and C2C12 (mouse) myoblasts by massively altering gene expression

(A and B) Inhibition of miR-98 following transfection of LNA inhibitor in LHCN human (A) or C2C12 mouse (B) myoblasts. (C) IF staining of LHCN human myoblasts transfected with control or miR-98 inhibitor, 5 d differentiation; MHC, myosin heavy chain. Shown are typical fields out of at least 10 independent experiments performed in triplicate. (D) Quantification of miR-98 KD experiments shown in (C). (E) IF staining of C2C12 mouse myoblasts transfected with control or miR-98 inhibitor, 2 d differentiation. (F) Quantification of miR-98 KD experiments shown in (E). (G) Time course of miR-98 loss-of-function assays in LHCN myoblasts. A, proliferation; B, early differentiation; and C, late differentiation. (H) Number of significantly up-regulated genes in loss-of-function miR-98 assays performed at the indicated time points. Three independent experiments, ***P<0.001 (Student's t test). ctrl, control; d, days.

Figure 1
miR-98 KD increases differentiation in LHCN (human) and C2C12 (mouse) myoblasts by massively altering gene expression

(A and B) Inhibition of miR-98 following transfection of LNA inhibitor in LHCN human (A) or C2C12 mouse (B) myoblasts. (C) IF staining of LHCN human myoblasts transfected with control or miR-98 inhibitor, 5 d differentiation; MHC, myosin heavy chain. Shown are typical fields out of at least 10 independent experiments performed in triplicate. (D) Quantification of miR-98 KD experiments shown in (C). (E) IF staining of C2C12 mouse myoblasts transfected with control or miR-98 inhibitor, 2 d differentiation. (F) Quantification of miR-98 KD experiments shown in (E). (G) Time course of miR-98 loss-of-function assays in LHCN myoblasts. A, proliferation; B, early differentiation; and C, late differentiation. (H) Number of significantly up-regulated genes in loss-of-function miR-98 assays performed at the indicated time points. Three independent experiments, ***P<0.001 (Student's t test). ctrl, control; d, days.

The effect of miR-98 and let-7g mimics on muscle differentiation was less pronounced than the effect of the miR-98 inhibitor. The let-7 mimics were notably less efficient in modifying the expression of well-known let-7 targets than let-7 inhibitors. These results are probably due to the high expression level of endogenous let-7 miRNAs in myoblasts ([2,10], and our (C. Degerny, A. Polesskaya) unpublished work). Therefore, in the present study we preferentially used miR-98 inhibitor and only included the mimic in some experiments performed with high-sensitivity techniques.

Transfection of miR-98 inhibitor led to a 4- to 6-fold increase in the fusion index of human LHCN–M2 myoblast cells or of mouse C2C12 myoblast cells. (Figures 1C–1F), confirming that miR-98 is a negative regulator of differentiation in both cell models.

E2F5 is a novel direct target of miR-98 in differentiating myoblasts

In order to identify potential targets of miR-98 in muscle cells, we used two sequential assays. First, we analysed the transcriptome of human myoblasts transfected with miR-98 inhibitor (KD) or mimic. The miR-98 loss-of-function (KD) assays were performed at different stages: proliferation, early differentiation or late differentiation (Figure 1G). Given the phenotype of miR-98 KD myoblasts, we were most interested in the early differentiation time-point, whereas the other two were only analysed to determine whether miR-98 targets changed during the differentiation process. Thus, the gain-of-function assays were performed only in early differentiating myoblasts. Statistical analysis of the data allowed us to identify 92 significantly up-regulated genes in the early differentiation loss-of-function assay, 147 genes in the proliferation assay and 48 genes in the late differentiation assay. Interestingly, only 13 genes were found to be significantly up-regulated at all three time points in miR-98 loss-of-function assays, suggesting that the targets of let-7 in myoblasts can vary during differentiation (Figure 1H and Supplementary Table S1).

Secondly, we identified, among these potential targets, those that are involved in suppressing early myogenic differentiation by miR-98. To this end, we used the StarS (suppressed target screen) assay [2] in which concomitant down-regulation of phenotypically relevant gene targets together with the miRNA rescues a normal phenotype. In this assay, the transcriptional repressor E2F5 was found to be among the putative targets whose inhibition significantly rescued a normal phenotype (Figure 2A and 2B), indicating that E2F5 is, at least in part, responsible for the phenotype of miR-98 KD cells. Reporter assays using the 3′ UTR of E2F5 cloned downstream of the Renilla luciferase gene showed that E2F5 is a direct target of let-7g and miR-98 (Figure 2C). Moreover, both E2F5 mRNA and protein were shown to be strongly up-regulated in myoblasts when miR-98 was inhibited (Figures 2D and 2E). Taken together, these results demonstrate that E2F5 is a novel target of let-7 that is important in skeletal myoblasts and suggest that E2F5 is a previously uncharacterized regulator of terminal myogenic differentiation. Consistent with this finding, suppression of let-7 was recently shown to directly up-regulate E2F5 and thus decrease proliferation of chondrocytes [12].

E2F5 is a novel direct target of miR-98 in differentiating myoblasts

Figure 2
E2F5 is a novel direct target of miR-98 in differentiating myoblasts

(A) IF staining of LHCN myoblasts transfected with two different control siRNAs or three different E2F5 siRNAs, 5 days differentiation. Shown are typical fields out of at least three independent experiments performed in triplicate. (B) Quantification of E2F5/miR-98 double KD experiments shown in (A). (C) Luciferase assays using the 3′ UTR of E2F5 cloned into pSiCheck.2 reporter vector, transfected together with miR-98 inhibitor or let-7g mimic in HeLa cells. The results are shown as the mean of three independent experiments, each performed in nine replicates. (D and E) Expression of endogenous E2F5 in proliferating or differentiating LHCN (D) or C2C12 (E) myoblasts transfected for 24 h with the indicated miRNA inhibitors. (D) mRNA, the mean of three independent experiments; (E) protein, a representative experiment out of three. *P<0.05; **P<0.01; ***P<0.001 (Student's t test). conf, confluent; ctrl, control.

Figure 2
E2F5 is a novel direct target of miR-98 in differentiating myoblasts

(A) IF staining of LHCN myoblasts transfected with two different control siRNAs or three different E2F5 siRNAs, 5 days differentiation. Shown are typical fields out of at least three independent experiments performed in triplicate. (B) Quantification of E2F5/miR-98 double KD experiments shown in (A). (C) Luciferase assays using the 3′ UTR of E2F5 cloned into pSiCheck.2 reporter vector, transfected together with miR-98 inhibitor or let-7g mimic in HeLa cells. The results are shown as the mean of three independent experiments, each performed in nine replicates. (D and E) Expression of endogenous E2F5 in proliferating or differentiating LHCN (D) or C2C12 (E) myoblasts transfected for 24 h with the indicated miRNA inhibitors. (D) mRNA, the mean of three independent experiments; (E) protein, a representative experiment out of three. *P<0.05; **P<0.01; ***P<0.001 (Student's t test). conf, confluent; ctrl, control.

E2F5 is a pro-differentiation factor expressed in human and mouse skeletal myoblasts

In order to decipher the role of E2F5 in muscle differentiation, we first analysed the expression of this factor in skeletal myoblasts. Whereas E2F5 mRNA levels did not strongly change during differentiation, the E2F5 protein levels appeared to decrease (Figures 3A and 3B). Most importantly, at the same time, E2F5, which was diffuse in the cytoplasm and the nucleus of proliferating and early differentiating cells, concentrated in the nucleus in the majority of analysed cells (typical images and numbers are presented in Figure 3C). This observation is consistent with the previously reported localization of E2F5 to the nucleus in differentiating keratinocytes, where this transcription factor plays an important pro-differentiation role [13]. E2F5 loss-of-function assays in differentiating human myoblasts demonstrated that E2F5 is required for differentiation, as three distinct siRNAs all induced an almost complete inhibition of differentiation (Figures 3D and 3E). Conversely, overexpressing E2F5 in mouse skeletal C2C12 myoblasts led to precocious expression of the differentiation marker MCK, as well as of the pro-differentiation transcription factor myogenin (Figure 3F). Therefore, the miR-98 target E2F5 plays a pro-differentiation role and influences the timing of differentiation in skeletal myoblasts.

E2F5 is a pro-differentiation factor largely responsible for the phenotype of miR-98 KD myoblasts

Figure 3
E2F5 is a pro-differentiation factor largely responsible for the phenotype of miR-98 KD myoblasts

(AC) Expression of endogenous E2F5 in proliferating or differentiating C2C12 myoblasts. (A) mRNA, the mean of three independent experiments; (B) protein, a representative experiment out of three; (C) IF (shown are typical cells out of 50 analysed, representing 47/50 for 0 d, 48/50 for 1 d, 38/50 for 2 d, 42/50 for 3 d time points). (D) IF staining of LHCN myoblasts transfected with two different control siRNAs or three different E2F5 siRNAs, 7 d differentiation. MHC, myosin heavy chain. Shown are typical fields out of three independent experiments performed in triplicate. (E) Quantification of E2F5 KD experiments shown in (D). (F) Western blot analysis of differentiating C2C12 myoblasts transfected with an E2F5 expression vector or with a control vector. conf, confluent; d, days.

Figure 3
E2F5 is a pro-differentiation factor largely responsible for the phenotype of miR-98 KD myoblasts

(AC) Expression of endogenous E2F5 in proliferating or differentiating C2C12 myoblasts. (A) mRNA, the mean of three independent experiments; (B) protein, a representative experiment out of three; (C) IF (shown are typical cells out of 50 analysed, representing 47/50 for 0 d, 48/50 for 1 d, 38/50 for 2 d, 42/50 for 3 d time points). (D) IF staining of LHCN myoblasts transfected with two different control siRNAs or three different E2F5 siRNAs, 7 d differentiation. MHC, myosin heavy chain. Shown are typical fields out of three independent experiments performed in triplicate. (E) Quantification of E2F5 KD experiments shown in (D). (F) Western blot analysis of differentiating C2C12 myoblasts transfected with an E2F5 expression vector or with a control vector. conf, confluent; d, days.

E2F5 directly down-regulates the differentiation inhibitors HMOX1 and ID1 and affects the expression of myostatin and other factors of the TGFB pathway

In order to explore how E2F5 affects skeletal muscle differentiation, we sought to characterize the targets of this transcriptional repressor. To this end, we identified the genes that were predicted to be targets of E2F5 in the Transfac® database (http://www.biobase-international.com/product/transcription-factor-binding-sites) among the genes that were down-regulated in miR-98 loss-of-function assays. Some of these genes (Table 1) were known regulators of skeletal muscle differentiation, including HMOX1, a myomiR inhibitor [14], ID1, a member of a family of inhibitors of key myogenic factors [15], as well as five genes belonging to the TGFB signalling pathway including MSTN, a cytokine of the TGFB superfamily and an extremely potent inhibitor of myogenesis [16]. Not all of these potential E2F5 targets were regulated in the same manner in proliferation, early differentiation or late differentiation in miR-98 KD human myoblasts. For in-depth analysis, we focused on five genes: HMOX1, ID1, MSTN, TGFB2 and TGFB3. In order to demonstrate that these genes are indeed regulated by E2F5, we used E2F5 loss-of-function assays. HMOX1, ID1, MSTN and TGFB3, but not TGFB2, were significantly up-regulated upon E2F5 KD by two different siRNAs (Figure 4A).

Table 1
Muscle differentiation inhibitors that are both predicted E2F5 targets and down-regulated in miR-98 KD myoblasts

miR-98 loss-of-function (KD) assays were performed in LHCN cells in proliferation, in early or in late differentiation (diff), shown as the mean of three independent experiments (see Figure 1G for the time-course of the assay).

   Fold change, miR-98 loss-of function 
Gene symbol Name References Proliferation Early diff Late diff 
HMOX1 Heme oxygenase 1 [14,27−1.96 −1.63 −1.32 
DKK1 Dickkopf WNT signalling pathway inhibitor 1 [28−1.32 −1.45 −1.85 
ID1 Inhibitor of DNA binding 1 [29,30−1.38 −1.23 −1.13 
TGFB2 Transforming growth factor, β 2 [311.02 −1.16 −1.56 
TGFB3 Transforming growth factor, β 3 [321.09 1.14 −1.96 
MSTN Myostatin [33−1.53 −1.09 1.03 
CTGF Connective tissue growth factor, TGFB induced [34−1.03 1.10 −1.70 
SNAI1 Snail homologue 1, TGFB induced [35,361.02 −1.05 −1.96 
SCN4A Sodium channel, voltage-gated, type IV, α subunit [37−1.09 1.04 −2.25 
MSC Musculin (MyoR) [381.02 1.09 −1.78 
   Fold change, miR-98 loss-of function 
Gene symbol Name References Proliferation Early diff Late diff 
HMOX1 Heme oxygenase 1 [14,27−1.96 −1.63 −1.32 
DKK1 Dickkopf WNT signalling pathway inhibitor 1 [28−1.32 −1.45 −1.85 
ID1 Inhibitor of DNA binding 1 [29,30−1.38 −1.23 −1.13 
TGFB2 Transforming growth factor, β 2 [311.02 −1.16 −1.56 
TGFB3 Transforming growth factor, β 3 [321.09 1.14 −1.96 
MSTN Myostatin [33−1.53 −1.09 1.03 
CTGF Connective tissue growth factor, TGFB induced [34−1.03 1.10 −1.70 
SNAI1 Snail homologue 1, TGFB induced [35,361.02 −1.05 −1.96 
SCN4A Sodium channel, voltage-gated, type IV, α subunit [37−1.09 1.04 −2.25 
MSC Musculin (MyoR) [381.02 1.09 −1.78 

E2F5 directly down-regulates the differentiation inhibitors HMOX1 and ID1 and affects the expression of MSTN and other factors of the TGFB pathway

Figure 4
E2F5 directly down-regulates the differentiation inhibitors HMOX1 and ID1 and affects the expression of MSTN and other factors of the TGFB pathway

(A) qRT-PCR analysis of altered expression of putative E2F5 targets in E2F5 KD cells. The results are shown as the mean of six independent experiments. (B) ChIP assays using anti-E2F5 or -E2F4 antibodies in C2C12 myoblasts (the mean of five independent experiments). *P<0.05; **P<0.01; ***P<0.001 (Student's t test). (C). Model for miR-98-dependent regulation of muscle differentiation. d, days.

Figure 4
E2F5 directly down-regulates the differentiation inhibitors HMOX1 and ID1 and affects the expression of MSTN and other factors of the TGFB pathway

(A) qRT-PCR analysis of altered expression of putative E2F5 targets in E2F5 KD cells. The results are shown as the mean of six independent experiments. (B) ChIP assays using anti-E2F5 or -E2F4 antibodies in C2C12 myoblasts (the mean of five independent experiments). *P<0.05; **P<0.01; ***P<0.001 (Student's t test). (C). Model for miR-98-dependent regulation of muscle differentiation. d, days.

To find out whether this regulation was directly dependent on E2F5 transcriptional repression, we performed ChIP using E2F5 antibody. As a positive control in these experiments, we used E2F4, a transcriptional repressor belonging to the same family, with known binding sites in the genome of C2C12 myoblasts (http://genome.ucsc.edu/ENCODE/dataMatrix/encodeDataMatrixMouse.html, GEO sample accession GSM915187). The promoters of both HMOX1 and ID1 were strongly enriched in E2F5 ChIP samples both in proliferation and in differentiation, confirming that these inhibitors of myogenesis are indeed novel direct targets of the transcriptional repressor E2F5 and indirect targets of miR-98 in muscle (Figure 4B).

The promoters of MSTN and TGFB3 were not found to be direct targets of E2F5 or E2F4 in our ChIP assays (result not shown). However, we found it extremely intriguing that these potent inhibitors of myogenesis are differentially regulated by a KD of miR-98, as well as by a KD of the miR-98 target E2F5. Moreover, a group of genes belonging to the TGFB pathway was down-regulated in miR-98-depleted human myoblasts (Table 1). The search for the most enriched pathways in the sets of genes with significantly high fold-change in miR-98 KD experiments was performed using gene ontology and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway databases. The TGFB signalling pathway was found to be the most enriched one in the list of genes significantly increased during early differentiation (14.3% of associated genes with P value corrected by the Benjamini-Hochberg algorithm=6.64×10−9), and the third most enriched one in the list of genes in the late differentiation set (37% and 1.57×10−20). Further studies are needed to find a link between miR-98, E2F5 and the TGFB regulation of myogenesis.

let-7 miRNAs, which influence the timing of Caenorhabditis elegans development, have been implicated in regulating cell proliferation and differentiation of numerous cell and tissue types, such as neuronal, cardiac, adipogenic and others ([1721], reviewed in [3]) in mammals. In a few cases, a let-7 miRNA acted as a negative regulator of differentiation, via an unknown mechanism [22]. The effect and mechanisms of let-7-dependent regulation of these processes remain contradictory, even in the context of highly similar experimental systems, such as mouse skeletal myoblasts C2C12 [23,24]. The contradictions can be partially resolved by our observation that let-7 targets can change depending on the stage of terminal myogenic differentiation, highlighting the complexity of let-7 functions.

In our previous studies, we showed that the let-7 family strongly delays skeletal muscle terminal differentiation [2]. In the present paper, we show that this effect is mediated through the inhibition of previously unidentified targets of let-7 in skeletal muscle, including the transcriptional repressor E2F5, which, in turn, acts on known inhibitors of differentiation such as ID1 and HMOX1. ID1 is known to impair the activity of myogenic regulatory factors (MRFs), including the key transcription factor MyoD, by sequestering their E-protein partners [25]. HMOX1 is also capable of repressing the expression of MyoD, as well as of myomiRs miR-1, miR-133 and miR-206, important activators of myogenesis [1,14]. MSTN and TGFB were shown in multiple studies to repress the expression and activity of various MRFs and their partners in skeletal muscle (reviewed in [26]). let-7 miRNAs thus appear to regulate a large set of key myogenic genes through their action on E2F5, which acts directly or indirectly on several regulatory pathways of differentiation. It is likely that E2F5 activates myogenesis by increasing the expression of myomiRs miR-1/133/206, and by acting at multiple levels on the expression and activity of MRFs. Consequently, let-7 could regulate a very large set of genes via its action on the mRNA of a single gene, E2F5, which seems to act here as an ‘amplifier’ of the let-7 anti-differentiation effect.

To summarize, let-7 has a specific panel of direct and indirect targets in skeletal muscle, which accounts for its anti-differentiation effect in this tissue. These data underscore the importance of the cellular environment in miRNA function: unlike transcription factors, for which target sites are constitutively present in all cells and tissues (even though not always accessible), miRNA activity strongly depends on the expression of gene target mRNAs. Thus, it is particularly important to study let-7 and other miRNA targets in the right cellular context, in order to reconstitute the regulatory networks controlled by the members of this family of miRNAs.

AUTHOR CONTRIBUTION

The present study is part of the Ph.D. thesis work of Jeremie Kropp, with the help of Cindy Degerny, under the direction of Annick Harel-Bellan and Anna Polesskaya. Jeremie Kropp designed, performed, analysed and interpreted most of the experiments. Nadezda Morozova performed the bioinformatic analysis. Julien Pontis helped with design and analysis of ChIP assays. Anna Polesskaya, Annick Harel-Bellan and Jeremie Kropp wrote the manuscript.

We thank Dr G. Pinna and Dr S. Ait-Si-Ali for help with statistical analysis and ChIP assays, Dr V. Mouly for the LHCN cells and Dr R. Gorelik and Dr C. Mann for critical reading of the manuscript.

FUNDING

This work was supported by the French National Research Agency (ANR) [grant number ANR-09-GENO-107-01]; the French Ministry of Higher Education and Research; and the Association Française contre les Myopathies.

Abbreviations

     
  • ATCC

    American Type Culture Collection

  •  
  • ChIP

    chromatin immunoprecipitation

  •  
  • HMOX1

    heme oxygenase 1

  •  
  • ID1

    inhibitor of DNA binding 1

  •  
  • IF

    immunofluorescence

  •  
  • KD

    knock down

  •  
  • LNA

    locked nucleic acid

  •  
  • MCK

    muscle creatine kinase

  •  
  • MRF

    myogenic regulatory factor

  •  
  • MSTN

    myostatin

  •  
  • myomiR

    myogenic miRNA

  •  
  • NNAT

    neuronatin

  •  
  • qRT

    quantitative reverse transcription

  •  
  • cyclo A

    cyclophilin A

  •  
  • TGFB

    transforming growth factor β

  •  
  • WB

    Western blotting

References

References
1
McCarthy
J. J.
The MyomiR network in skeletal muscle plasticity
Exerc. Sport Sci. Rev.
2011
, vol. 
39
 (pg. 
150
-
154
)
[PubMed]
2
Polesskaya
A.
Degerny
C.
Pinna
G.
Maury
Y.
Kratassiouk
G.
Mouly
V.
Morozova
N.
Kropp
J.
Frandsen
N.
Harel-Bellan
A.
Genome-wide exploration of miRNA function in mammalian muscle cell differentiation
PLoS ONE
2013
, vol. 
8
 pg. 
e71927
 
[PubMed]
3
Mondol
V.
Pasquinelli
A. E.
Let's make it happen: the role of let-7 microRNA in development
Curr. Top. Dev. Biol.
2012
, vol. 
99
 (pg. 
1
-
30
)
[PubMed]
4
Wang
X.
Cao
L.
Wang
Y.
Liu
N.
You
Y.
Regulation of let-7 and its target oncogenes
Oncol. Lett.
2012
, vol. 
3
 (pg. 
955
-
960
)
[PubMed]
5
Bao
M. H.
Feng
X.
Zhang
Y. W.
Lou
X. Y.
Cheng
Y.
Zhou
H. H.
Let-7 in cardiovascular diseases, heart development and cardiovascular differentiation from stem cells
Int. J. Mol. Sci.
2013
, vol. 
14
 (pg. 
23086
-
23102
)
[PubMed]
6
Toledano
H.
The role of the heterochronic microRNA let-7 in the progression of aging
Exp. Gerontol.
2013
, vol. 
48
 (pg. 
667
-
670
)
[PubMed]
7
Boyerinas
B.
Park
S. M.
Hau
A.
Murmann
A. E.
Peter
M. E.
The role of let-7 in cell differentiation and cancer
Endocr. Relat. Cancer
2010
, vol. 
17
 (pg. 
F19
-
F36
)
[PubMed]
8
Zhu
C. H.
Mouly
V.
Cooper
R. N.
Mamchaoui
K.
Bigot
A.
Shay
J. W.
Di Santo
J. P.
Butler-Browne
G. S.
Wright
W. E.
Cellular senescence in human myoblasts is overcome by human telomerase reverse transcriptase and cyclin-dependent kinase 4: consequences in aging muscle and therapeutic strategies for muscular dystrophies
Aging Cell
2007
, vol. 
6
 (pg. 
515
-
523
)
[PubMed]
9
Boudoukha
S.
Cuvellier
S.
Polesskaya
A.
Role of the RNA-binding protein IMP-2 in muscle cell motility
Mol. Cell. Biol.
2010
, vol. 
30
 (pg. 
5710
-
5725
)
[PubMed]
10
Boudoukha
S.
Rivera Vargas
T.
Dang
I.
Kropp
J.
Cuvellier
S.
Gautreau
A.
Polesskaya
A.
MiRNA let-7g regulates skeletal myoblast motility via Pinch-2
FEBS lett.
2014
, vol. 
588
 (pg. 
1623
-
1629
)
[PubMed]
11
Petersen
M.
Wengel
J.
LNA: a versatile tool for therapeutics and genomics
Trends Biotechnol.
2003
, vol. 
21
 (pg. 
74
-
81
)
[PubMed]
12
Papaioannou
G.
Inloes
J. B.
Nakamura
Y.
Paltrinieri
E.
Kobayashi
T.
let-7 and miR-140 microRNAs coordinately regulate skeletal development
Proc. Natl. Acad. Sci. U.S.A.
2013
, vol. 
110
 (pg. 
E3291
-
E3300
)
[PubMed]
13
Apostolova
M. D.
Ivanova
I. A.
Dagnino
C.
D’Souza
S. J.
Dagnino
L.
Active nuclear import and export pathways regulate E2F-5 subcellular localization
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
34471
-
34479
)
[PubMed]
14
Kozakowska
M.
Ciesla
M.
Stefanska
A.
Skrzypek
K.
Was
H.
Jazwa
A.
Grochot-Przeczek
A.
Kotlinowski
J.
Szymula
A.
Bartelik
A.
, et al. 
Heme oxygenase-1 inhibits myoblast differentiation by targeting myomirs
Antioxid. Redox. Signal.
2012
, vol. 
16
 (pg. 
113
-
127
)
[PubMed]
15
Benezra
R.
Davis
R. L.
Lockshon
D.
Turner
D. L.
Weintraub
H.
The protein Id: a negative regulator of helix-loop-helix DNA binding proteins
Cell
1990
, vol. 
61
 (pg. 
49
-
59
)
[PubMed]
16
McPherron
A. C.
Lawler
A. M.
Lee
S. J.
Regulation of skeletal muscle mass in mice by a new TGF-beta superfamily member
Nature
1997
, vol. 
387
 (pg. 
83
-
90
)
[PubMed]
17
Meza-Sosa
K. F.
Pedraza-Alva
G.
Perez-Martinez
L.
microRNAs: key triggers of neuronal cell fate
Front. Cell. Neurosci.
2014
, vol. 
8
 pg. 
175
 
[PubMed]
18
Yang
Y.
Ago
T.
Zhai
P.
Abdellatif
M.
Sadoshima
J.
Thioredoxin 1 negatively regulates angiotensin II-induced cardiac hypertrophy through upregulation of miR-98/let-7
Circ. Res.
2011
, vol. 
108
 (pg. 
305
-
313
)
[PubMed]
19
Ventayol
M.
Vinas
J. L.
Sola
A.
Jung
M.
Brune
B.
Pi
F.
Mastora
C.
Hotter
G.
miRNA let-7e targeting MMP9 is involved in adipose-derived stem cell differentiation toward epithelia
Cell Death Dis.
2014
, vol. 
5
 pg. 
e1048
 
[PubMed]
20
Polesskaya
A.
Cuvellier
S.
Naguibneva
I.
Duquet
A.
Moss
E. G.
Harel-Bellan
A.
Lin-28 binds IGF-2 mRNA and participates in skeletal myogenesis by increasing translation efficiency
Genes Dev.
2007
, vol. 
21
 (pg. 
1125
-
1138
)
[PubMed]
21
Pelosi
A.
Careccia
S.
Lulli
V.
Romania
P.
Marziali
G.
Testa
U.
Lavorgna
S.
Lo-Coco
F.
Petti
M. C.
Calabretta
B.
, et al. 
miRNA let-7c promotes granulocytic differentiation in acute myeloid leukemia
Oncogene
2013
, vol. 
32
 (pg. 
3648
-
3654
)
[PubMed]
22
Davoodian
N.
Lotfi
A. S.
Soleimani
M.
Mola
S. J.
Arjmand
S.
Let-7f microRNA negatively regulates hepatic differentiation of human adipose tissue-derived stem cells
J. Physiol. Biochem.
2014
, vol. 
70
 (pg. 
781
-
789
)
[PubMed]
23
Kallen
A. N.
Zhou
X. B.
Xu
J.
Qiao
C.
Ma
J.
Yan
L.
Lu
L.
Liu
C.
Yi
J. S.
Zhang
H.
, et al. 
The imprinted H19 lncRNA antagonizes let-7 microRNAs
Mol. Cell
2013
, vol. 
52
 (pg. 
101
-
112
)
[PubMed]
24
Dey
B. K.
Pfeifer
K.
Dutta
A.
The H19 long noncoding RNA gives rise to microRNAs miR-675-3p and miR-675-5p to promote skeletal muscle differentiation and regeneration
Genes Dev.
2014
, vol. 
28
 (pg. 
491
-
501
)
[PubMed]
25
Jen
Y.
Weintraub
H.
Benezra
R.
Overexpression of Id protein inhibits the muscle differentiation program: in vivo association of Id with E2A proteins
Genes Dev.
1992
, vol. 
6
 (pg. 
1466
-
1479
)
[PubMed]
26
Kollias
H. D.
McDermott
J. C.
Transforming growth factor-beta and myostatin signaling in skeletal muscle
J. Appl. Physiol.
2008
, vol. 
104
 (pg. 
579
-
587
)
[PubMed]
27
Stachurska
A.
Ciesla
M.
Kozakowska
M.
Wolffram
S.
Boesch-Saadatmandi
C.
Rimbach
G.
Jozkowicz
A.
Dulak
J.
Loboda
A.
Cross-talk between microRNAs, nuclear factor E2-related factor 2, and heme oxygenase-1 in ochratoxin A-induced toxic effects in renal proximal tubular epithelial cells
Mol. Nutr. Food Res.
2013
, vol. 
57
 (pg. 
504
-
515
)
[PubMed]
28
Han
X. H.
Jin
Y. R.
Seto
M.
Yoon
J. K.
A WNT/beta-catenin signaling activator, R-spondin, plays positive regulatory roles during skeletal myogenesis
J. Biol. Chem.
2011
, vol. 
286
 (pg. 
10649
-
10659
)
[PubMed]
29
Sun
L.
Trausch-Azar
J. S.
Ciechanover
A.
Schwartz
A. L.
Ubiquitin-proteasome-mediated degradation, intracellular localization, and protein synthesis of MyoD and Id1 during muscle differentiation
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
26448
-
26456
)
[PubMed]
30
Melnikova
I. N.
Bounpheng
M.
Schatteman
G. C.
Gilliam
D.
Christy
B. A.
Differential biological activities of mammalian Id proteins in muscle cells
Exp. Cell Res.
1999
, vol. 
247
 (pg. 
94
-
104
)
[PubMed]
31
McLennan
I. S.
Koishi
K.
The transforming growth factor-betas: multifaceted regulators of the development and maintenance of skeletal muscles, motoneurons and Schwann cells
Int. J. Dev. Biol.
2002
, vol. 
46
 (pg. 
559
-
567
)
[PubMed]
32
Schabort
E. J.
van der Merwe
M.
Niesler
C. U.
TGF-β isoforms inhibit IGF-1-induced migration and regulate terminal differentiation in a cell-specific manner
J. Muscle Res. Cell Motil.
2011
, vol. 
31
 (pg. 
359
-
367
)
[PubMed]
33
Dschietzig
T. B.
Myostatin–from the mighty mouse to cardiovascular disease and cachexia
Clin. Chim. Acta
2014
, vol. 
433
 (pg. 
216
-
224
)
[PubMed]
34
Vial
C.
Zuniga
L. M.
Cabello-Verrugio
C.
Canon
P.
Fadic
R.
Brandan
E.
Skeletal muscle cells express the profibrotic cytokine connective tissue growth factor (CTGF/CCN2), which induces their dedifferentiation
J. Cell. Physiol.
2008
, vol. 
215
 (pg. 
410
-
421
)
[PubMed]
35
Zhou
Q.
Yang
M.
Lan
H.
Yu
X.
miR-30a negatively regulates TGF-β1-induced epithelial-mesenchymal transition and peritoneal fibrosis by targeting Snai1
Am. J. Pathol.
2013
, vol. 
183
 (pg. 
808
-
819
)
[PubMed]
36
Soleimani
V. D.
Yin
H.
Jahani-Asl
A.
Ming
H.
Kockx
C. E.
van Ijcken
W. F.
Grosveld
F.
Rudnicki
M. A.
Snail regulates MyoD binding-site occupancy to direct enhancer switching and differentiation-specific transcription in myogenesis
Mol. Cell
2012
, vol. 
47
 (pg. 
457
-
468
)
[PubMed]
37
Burge
J. A.
Hanna
M. G.
Novel insights into the pathomechanisms of skeletal muscle channelopathies
Curr. Neurol. Neurosci. Rep.
2012
, vol. 
12
 (pg. 
62
-
69
)
[PubMed]
38
Lu
J.
Webb
R.
Richardson
J. A.
Olson
E. N.
MyoR: a muscle-restricted basic helix-loop-helix transcription factor that antagonizes the actions of MyoD
Proc. Natl. Acad. Sci. U.S.A.
1999
, vol. 
96
 (pg. 
552
-
557
)
[PubMed]

Author notes

1

These authors contributed equally towards the article.

Supplementary data