Global MEF2 target gene analysis in cardiac and skeletal muscle reveals novel regulation of DUSP6 by p38MAPK-MEF2 signaling

Abstract

MEF2 plays a profound role in the regulation of transcription in cardiac and skeletal muscle lineages. To define the overlapping and unique MEF2A genomic targets, we utilized ChIP-exo analysis of cardiomyocytes and skeletal myoblasts. Of the 2783 and 1648 MEF2A binding peaks in skeletal myoblasts and cardiomyocytes, respectively, 294 common binding sites were identified. Genomic targets were compared to differentially expressed genes in RNA-seq analysis of MEF2A depleted myogenic cells, revealing two prominent genetic networks. Genes largely associated with muscle development were down-regulated by loss of MEF2A while up-regulated genes reveal a previously unrecognized function of MEF2A in suppressing growth/proliferative genes. Several up-regulated (Tprg, Mctp2, Kitl, Prrx1, Dusp6) and down-regulated (Atp1a2, Hspb7, Tmem182, Sorbs2, Lmod3) MEF2A target genes were chosen for further investigation. Interestingly, siRNA targeting of the MEF2A/D heterodimer revealed a somewhat divergent role in the regulation of Dusp6, a MAPK phosphatase, in cardiac and skeletal myogenic lineages. Furthermore, MEF2D functions as a p38MAPK-dependent repressor of Dusp6 in myoblasts. These data illustrate that MEF2 orchestrates both common and non-overlapping programs of signal-dependent gene expression in skeletal and cardiac muscle lineages.

Figure 1.

Identification of MEF2A target genes in MBs and CMs using ChIP-exo. (A) Workflow of ChIP-exo analysis. C2C12 (48 h DM; MB) and primary CMs were collected to identify MEF2A target genes using ChIP-exo. A non-specific IgG antibody was used as a control. (B) The number of common MEF2A enriched peaks in MB and CM identified in ChIP-exo are indicated in a Venn diagram. (C) The percentage of peaks within the proximal promoter (±5 kb), upstream (−5 to −50 kb), downstream (+5 to +50 kb) or intergenic region (>50 kb from any annotated gene) identified in ChIP-exo using GREAT. Location is relative to the TSS. (D) The five most dominant transcription factor binding motifs found within MEF2A-enriched peaks as determined by CENTDIST (P-value < 0.05). (E) Biological Processes and Cellular Component GO terms of MEF2A enriched peaks from MB and CM.

Figure 2.

RNA-seq analysis of MEF2A depleted skeletal MBs. (A) RNA-seq analysis workflow. MB were transfected with 30 nM of siMEF2A-2 or a scrambled siRNA control. Samples were prepared for RNA-seq analysis in duplicate. Differentially expressed genes were assessed using edgeR P-value < 0.05. (B) Two different siRNA targeting Mef2a were transfected into MBs at 30 nM and allowed to differentiate for 48 h in DM. Cells were harvested and protein was extracted to assess changes in MEF2A using immunoblotting. (C) Distinguished roles for MEF2A in skeletal myogenesis were revealed when up- and down-regulated genes were grouped separately prior to GO (Biological Processes) term analysis. (D) The differentially expressed genes that were also identified as MEF2A target genes in MB were determined (ChIP (+)). Differentially expressed genes that were not identified as MEF2A targets are labeled ChIP (−). (E) Binding profiles of MEF2A recruitment to associated genes in MB based on their differential expression in RNA-seq analysis. The raw gene count is indicated within each section.

Figure 3.

Functional analysis of MEF2A target genes. (A) The percentage of differentially expressed genes in MB that were also identified as MEF2A target genes (ChIP (+)) in MB alone or MB and CM. (B) Comparative analysis of 10 putative MEF2A target genes. Selected genes were differentially expressed in RNA-seq analysis in MB and shared overlapping MEF2A enrichment peaks in MB and CM. The locations of common MEF2A recruitment peaks relative to the TSS and nearby MEF2 consensus sequences are indicated. (C) MEF2A recruitment was assessed in C2C12 (48 h DM) using ChIP-qPCR. Acta2 was used as a negative control locus. Error bars represent ±SD, n = 3. (D) Screenshot from IGV depicting C2C12 ChIP-exo read density and MACS peak calls. Read densities are in purple, MACS peak calls are in black and RefSeq genes are in blue. (E) Two different siRNA targeting Mef2a were transfected into C2C12 at 30 nM and allowed to differentiate for 48 h in DM. Cells were harvested and RNA was extracted to assess changes gene expression using qRT-PCR. Samples were normalized to β-actin. Error bars represent ±SD, n = 3. (F) Knockdown of individual target genes in MB. C2C12 were transfected with 50 nM siRNA and harvested 24 h later. mRNA was assessed similar to that in (E).

Figure 4.

siRNA mediated gene silencing of MEF2 in CMs or MBs induces DUSP6 expression. (A) MEF2A is recruited to the Dusp6 promoter in primary CMs. Gapdh was used as a negative control locus. Error bars represent ±SD, n = 3. (B) Knockdown of MEF2A or MEF2D (C) in primary CMs up-regulates Dusp6 expression. siRNA was added at a final concentration of 200 nM. Protein was harvested and analyzed by immunoblotting with the indicated antibodies. (D) Knockdown of MEF2D up-regulates DUSP6 expression in MB. Mef2a or Mef2d were targeted using 30–70 nM siRNA. C2C12 were harvested at 24 h DM for immunoblot analysis.

Figure 5.

MEF2D inhibits DUSP6 in a p38MAPK-dependent manner in MBs. (A) C2C12 were allowed to differentiate, harvested at the times specified and analyzed by immunoblotting with the indicated antibodies. (B) Myogenesis is inhibited when C2C12 are treated with p38MAPK inhibitor SB 203580 (5 μM). Media was changed to DM for the indicated time and protein was assessed by immunoblotting. Control cells were treated with an inactive analogue, SB 202474. (C) C2C12 were transfected with 30 nM siMEF2A or 70 nM siMEF2D and treated with SB 203580 (5 μM) for 24 h in DM. Cells were harvested as described above. (D) COS7 were transfected with Dusp6-luc and the indicated plasmids. One day after transfection cells were treated with 5 μM SB 203580 or inactive analogue for 24 h. Luciferase values were normalized to Renilla. Error bars represent ±SEM, n = 3. Corresponding immunoblots are shown. (E)Dusp6 is negatively regulated by MEF2D in a p38MAPK-dependent manner.