E2f3b plays an essential role in myogenic differentiation through isoform-specific gene regulation

Abstract

Current models posit that E2F transcription factors can be divided into members that either activate or repress transcription, in part through collaboration with the retinoblastoma (pRb) tumor suppressor family. The E2f3 locus encodes E2f3a and E2f3b proteins, and available data suggest that they regulate cell cycle-dependent gene expression through opposing transcriptional activating and repressing activities in growing and quiescent cells, respectively. However, the role, if any, of E2F proteins, and in particular E2f3, in myogenic differentiation is not well understood. Here, we dissect the contributions of E2f3 isoforms and other activating and repressing E2Fs to cell cycle exit and differentiation by performing genome-wide identification of isoform-specific targets. We show that E2f3a and E2f3b target genes are involved in cell growth, lipid metabolism, and differentiation in an isoform-specific manner. Remarkably, using gene silencing, we show that E2f3b, but not E2f3a or other E2F family members, is required for myogenic differentiation, and that this requirement for E2f3b does not depend on pRb. Our functional studies indicate that E2f3b specifically attenuates expression of genes required to promote differentiation. These data suggest how diverse E2F isoforms encoded by a single locus can play opposing roles in cell cycle exit and differentiation.

Figure 1.

Expression of E2F family members during myogenesis and their role in differentiation. (A) Expression of E2F and pocket protein family members during myogenic differentiation. Each protein (indicated on the left) was detected by Western blot analysis in whole-cell lysates of myoblasts and of myotubes differentiated for 48 or 96 h (T48h and T96h/MT, respectively). Myotubes were isolated at 96 h. (B) Each E2F family member was ablated using miRNAs as indicated at the top. Green fluorescent protein (GFP) was coexpressed with each miRNA and serves as an expression marker. Loading control: β-actin and α-tubulin. (*) Nonspecific bands. (C) After ablation of the indicated E2F family members, cells were induced to differentiate for 48 and 96 h and stained with MHC and DAPI. (D) RT–qPCR analysis of the relative expression levels of the differentiation marker genes Myh7 and Mef2c in the miRNA-depletion experiment shown in C. Error bars depict the standard error of the mean (SEM) derived from three independent experiments. P-values were derived from a one-tail paired t-test between miR.Luc and miR.pan-E2F3-transduced cells. (E) Cells ablated for E2F family members were stained for BrdU incorporation at the indicated time points (as in C). The proportion of BrdU-positive nuclei is shown for each population. (F) An outline for dissecting E2F3 function during myogenic differentiation.

Figure 2.

Identification of E2f3a and E2f3b targets. (A) Strategy for genome-wide discovery E2F3 targets and the epitopes of the two E2F3 antibodies used. (B) Enrichment test of E2F3 isoforms. (Top) Nuclear extracts from myoblasts and myotubes were immunoprecipitated with the indicated antibodies and detected by Western blotting with antibodies against E2f3a/b. (C) Venn diagrams indicating the intersection between target genes in growing myoblasts and myotubes (see the text for details). Isoform-specific E2F3 target groups are labeled A–E. (D) Analysis of the percent overlap between E2F3 and E2F4 targets in myotubes expressing the percentage of bound targets as a function of E2F4 target genes bound by E2F3 in each group. Corresponding Venn diagrams are shown in Supplemental Fig S3. (E) Gene Ontology (GO) analysis of E2F3 target genes expressed as the percentage of the total number of independent GO annotations within each group. Groups are indicated on the right, and GO categories on the bottom.

Figure 3.

Verification of isoform-specific targets by conventional ChIP. qPCR analysis of a select number of genes representing the five E2F3 isoform-specific gene categories. Enrichment >0.05 is significant. Error bars depict the standard error of the mean (SEM) derived from three independent experiments.

Figure 4.

Correlation between E2F3 isoform binding and gene expression during differentiation and changes in E2F3 target gene expression when depleting E2F3. (A–E) Proportion of genes with the indicated fold change in gene expression in myotubes versus growing myoblasts (Log₂ FC MT/GM) for each group of E2F3 targets compared with genes that are bound (red trace) or not bound (blue trace). P-values were derived from the hypergeometric distribution test. (F) RT–qPCR analysis of mRNA levels of a select number of E2F3 target genes following E2F3 depletion. Error bars and P-values as in Figure 1D.

Figure 5.

E2f3b is uniquely required to promote myogenic differentiation and restores normal expression of target genes. (A) Differentiation was scored by staining for MHC and nuclei were visualized with DAPI as described in Figure 1. Combinations of virus transductions are indicated on the top. (B) RT–qPCR analysis of the genes in Figure 4 after forced expression of either E2f3a or E2f3b in E2F3-depleted cells. Error bars and P-values as in Figure 1D.

Figure 6.

ChIP analysis of promoter binding by E2f3 isoforms in the rescue experiment in Figure 5. Chromatin was prepared from each combination of transduced C2C12 cells as described and ChIP was performed using the polyclonal E2f3 antibody that recognizes both isoforms in order to ensure equal enrichment for E2f3a and E2f3b. Enriched DNA was analyzed by qPCR, and error bars depict the standard error of the mean (SEM) derived from three independent experiments.

Figure 7.

Model describing the role of E2F and pocket protein family members in myogenic differentiation. (A) Relative changes in the amount of E2f3a and E2f3b target genes during differentiation. (B) Model differential E2F3 isoform binding during myogenesis depicting the proposed existence of an E2f3b-specific targeting partner (X, purple hexagon).