Myc and max genome-wide binding sites analysis links the Myc regulatory network with the polycomb and the core pluripotency networks in mouse embryonic stem cells

Abstract

Myc is a master transcription factor that has been demonstrated to be required for embryonic stem cell (ESC) pluripotency, self-renewal, and inhibition of differentiation. Although recent works have identified several Myc-targets in ESCs, the list of Myc binding sites is largely incomplete due to the low sensitivity and specificity of the antibodies available. To systematically identify Myc binding sites in mouse ESCs, we used a stringent streptavidin-based genome-wide chromatin immunoprecipitation (ChIP-Seq) approach with biotin-tagged Myc (Bio-Myc) as well as a ChIP-Seq of the Myc binding partner Max. This analysis identified 4325 Myc binding sites, of which 2885 were newly identified. The identified sites overlap with more than 85% of the Max binding sites and are enriched for H3K4me3-positive promoters and active enhancers. Remarkably, this analysis unveils that Myc/Max regulates chromatin modifiers and transcriptional regulators involved in stem cell self-renewal linking the Myc-centered network with the Polycomb and the Core networks. These results provide insights into the contribution of Myc and Max in maintaining stem cell self-renewal and keeping these cells in an undifferentiated state.

Figure 1. Characterization and functional validation of stable clones expressing tagged Myc.

(A) Western blot analysis of the exogenous tagged Myc expression in the indicated stable clones. Actin (Actb) was used as a loading control. (B) Immunoprecipitation analysis of endogenous Max protein in nuclear extracts obtained from the indicated clones demonstrated that Max is able to co-immunoprecipitate with exogenous Myc in each clone analyzed. Purified rabbit IgG was used as a negative control. (C) RT-PCR analysis of ChIP assays of the indicated endogenous and tagged proteins. Suz12 and Ezh2 were used as positive controls, and H19 and Kdr were used as negative controls. The results are shown as the fold difference.

Figure 2. Comparison of Biotag-Myc versus anti-Myc antibody Chip-Seq.

(A) Heatmap representation of ChIP-Seq binding for BioMyc, Myc, and Max around transcriptional start sites (TSSs) ±5 kb for all genes rank ordered from the lowest to the most highly expressed according to thee mRNA expression levels obtained from RNA-Seq analysis. (B) Venn diagram showing the exact number of promoters bound in the indicated ChIP-Seq analyses. Bio-Myc ChIP-Seq found many more genes bound on TSSs in comparison with the endogenous Myc ChIP-Seq (4325 TSS vs. 1682 TSSs, respectively), and Bio-Myc ChIP-Seq shows more overlapping genes with Max ChIP-Seq (2306 vs. 1418, respectively). (C) Genomic occupancy profiles of the indicated Chip-Seq show a well-defined peak of enrichment on the E-Box at the Suz12 promoter (−570 bp from TSS) and the Ezh2 TSS, while there is no binding at the promoters of the negative control genes (H19 and Kdr).

Figure 3. Biotag-Myc ChIP-Seq is more sensitive without loss of specificity.

(A) Motif discovery analysis showed a significant enrichment for the E-Box sequence (CACGTG) in all the ChIP-Seq indicated. (B–C) The ROC curve highlights that Bio-Myc ChIP-Seq in comparison with Myc ChIP-Seq is more sensitive without a loss of specificity at Max or H3K4me3 target genes. The major area under the curve (AUC) for Bio-Myc indicates its better performance compared with Myc (AUC = 0.80 vs. 0.75 for Max target genes and AUC = 0.74 vs. 0.69 for H3K4me3 genes). (D–E) Bio-Myc (left panels) or endogenous Myc (right panels) ChIP-Seq reads cumulative frequency distributions are plotted for the background and foreground regions calculated for Max or H3K4me3 binding. A greater distance between the curves in Bio-Myc (d = 0.86 and d = 0.78) with respect to the endogenous Myc (d = 0.51 and d = 0.45) indicates a greater difference of signal intensity occurring between the background and bound regions.

Figure 4. Biotag-Myc identifies a large number of Myc binding sites in ESCs.

(A) Venn diagram showing the number of gene promoters bound in the indicated ChIP-Seq analyses and their overlap. (B) Bio-Myc ChIP-Seq was able to identify many more bound regions in comparison with Myc ChIP-Seq as shown by the number of peaks found at each p-value indicated. (C) The genomic colocalization of Bio-Myc peaks with Max, H3K4me3-only promoters, and active enhancers at different p-values. Bio-Myc ChIP-Seq always reveals a major number of peaks overlapping with the indicated genomic features and shows a similar distribution to the NMyc dataset. (D) About 50% of the newly identified Myc-bound promoters show overlap with NMyc targets and an enrichment in the motif sequence corresponding to the perfect Myc E-Box (p-value 1E-142). (E–F) Gene ontology analysis reveals that the newly identified Myc target genes are enriched for genes associated with the cell cycle and cell metabolism as well as genes with molecular functions involved in the metabolism of nucleic acid and chromatin structure.

Figure 5. The newly identified genes by Biotag-Myc ChIP-Seq are bound by endogenous Myc in ES cells.

(A–B) Representative genomic occupancy profiles of genes identified by Bio-Myc ChIP-Seq but not by endogenous Myc ChIP-Seq. The genes were divided in two groups. The first group is composed of the genes that are also bound by NMyc (A) while the second group is composed of the genes that are bound only by Bio-Myc (B). (C) RT-qPCR analysis of endogenous Myc ChIP samples at the promoters of the indicated genes. The Suz12 and Ezh2 genes were used as positive controls, and the H19 and Kdr genes were used as negative controls. The results are shown as the percentage (1/100) of the input. (D) RT-qPCR analysis of endogenous NMyc and Max ChIP reveals their binding to the promoters of the second group of genes. The Suz12 and Ezh2 genes were used as positive controls, the H19 and Kdr genes were used as negative controls. The results are shown as the percentage (1/100) of the input.

Figure 6. The genes identified by Biotag-Myc are regulated by Myc in ES cells.

(A) RT-PCR analysis of Myc and NMyc expression in double knockdown ES cells. The results are shown as a fold difference. (B) Western blot analysis of Myc protein level in double-knockdown ES cells. β-actin was used as a loading control. (C) RT-PCR analysis of the indicated transcripts upon a double knockdown (dKD) of Myc proteins. The Suz12 and Ezh2 genes were used as positive controls, and the H19 and Kdr genes were used as negative controls. The results are shown as the fold difference.

Figure 7. Myc/Max transcriptional targets in ESCs.

Schematic representation of Myc/Max bound genes in ES cells (the genes identified in this work are indicated in red) and their molecular and biological functions.