KAT-Independent Gene Regulation by Tip60 Promotes ESC Self-Renewal but Not Pluripotency

Abstract

Although histone-modifying enzymes are generally assumed to function in a manner dependent on their enzymatic activities, this assumption remains untested for many factors. Here, we show that the Tip60 (Kat5) lysine acetyltransferase (KAT), which is essential for embryonic stem cell (ESC) self-renewal and pre-implantation development, performs these functions independently of its KAT activity. Unlike ESCs depleted of Tip60, KAT-deficient ESCs exhibited minimal alterations in gene expression, chromatin accessibility at Tip60 binding sites, and self-renewal, thus demonstrating a critical KAT-independent role of Tip60 in ESC maintenance. In contrast, KAT-deficient ESCs exhibited impaired differentiation into mesoderm and endoderm, demonstrating a KAT-dependent function in differentiation. Consistent with this phenotype, KAT-deficient mouse embryos exhibited post-implantation developmental defects. These findings establish separable KAT-dependent and KAT-independent functions of Tip60 in ESCs and during differentiation, revealing a complex repertoire of regulatory functions for this essential chromatin remodeling complex.

Keywords: Ep400; Kat5; Tip60; acetyltransferase; chromatin; development; differentiation; embryonic stem cells; self-renewal.

Figure 1. Tip60 KAT and p400 ATPase activities are dispensable for ESC self-renewal and gene regulation

(A) Alkaline phosphatase staining (AP) of Tip60^ci/ci and Ep400^ci/ci mutants and controls (Tip60^fl/+, Tip60 KD, Ep400 KD, and Ep400^hypo). Scale bars equal 200 μm. (B) Growth curve, measuring the proliferation rates of the indicated mutant and control ESCs. (C, D) Heatmaps of differentially expressed genes in Tip60^ci/ci and Tip60 KD ESCs relative to Tip60^fl/+control cells (C), or Ep400^ci/ci and Ep400^hypo ESCs relative to wild type (E14) control ESCs (D). Genes in the heatmaps are sorted from the most upregulated to the most down regulated genes in the Tip60 KD and Ep400^hypo controls, respectively. (E, F) Venn diagrams showing number of genes commonly misregulated in Tip60^ci/ci and Tip60 KD ESCs (E), or Ep400^ci/ci and Ep400^hypo ESCs (F). Genes were considered significantly misregulated in each KD or mutant if their |log₂ (fold change)| > 1 and their multiple testing-adjusted p value < 0.05.

Figure 2. KAT-independent regulation of chromatin accessibility at Tip60 target loci

(A) Example Tip60 target gene showing increased promoter-proximal chromatin accessibility in Tip60 KD but not Tip60^ci/ci relative to Tip60^fl/+ control cells. Shown are normalized ATAC-seq reads ≤ 100bp for each biological replicate, and Tip60 ChIP-seq data from (Ravens et al., 2015). (B) Aggregation plot showing average ATAC-seq signal for two biological replicates of each mutant or KD aggregated over high-quality Tip60 binding sites. A Kolmogorov–Smirnov test of differences in ATAC profiles was used to calculate p values. (C) K-means clustering (K=3) for ATAC-seq data over Tip60 binding sites. Promoter-proximal peaks are marked with a black bar to the right, promoter-distal peaks with a white bar. (D) Aggregation plot of ATAC-seq data (as in B) over Tip60–bound promoter regions aligned such that all gene bodies are to the right. Promoter-proximal regions (pro) and transcription start sites (TSS) are indicated. Tip60 ChIP-seq data (Ravens et al., 2015) are shown for reference. (E) Aggregation plot over Tip60–bound gene-distal regions.

Figure 3. The Tip60 catalytic activity is required for differentiation and post-implantation development

(A) Genotypes of embryos from Tip60^ci/+ intercrosses at different developmental stages. (B) Images of E10.5 embryos of the indicated genotypes. Scale bar equals 1 mm. (C) Embryoid body (EB) formation assay comparing EB morphology in Tip60^ci/ci mutant ESCs to Tip60^fl/+ and Tip60 KD controls. Scale bars equal 400 μm. (D) Quantification of EB size in indicated mutants and controls (n = 49 per genotype). Boxes range from the 25^th to the 75^th percentile, the dark lines indicate the median, and the whiskers indicate the lesser of either the extreme (max or min) value or 1.5 times the interquartile range (***p < 0.001, calculated using a two-sided t-test). (E) RT-qPCR analysis of indicated germ layer markers during a time course of EB differentiation. (F, G) Whole mount in situ hybridization in E6.5 and E7.5 mouse embryos staining for T transcript. Scale bars equal 100 μm (F) or 250 μm (G).

Figure 4. Delayed/impaired expression of developmental regulators in differentiating Tip60^ci/ci ESCs

(A) Heatmap indicating induction kinetics of each germ layer markers during differentiation of Tip60^fl/+ controls or Tip60^ci/ci mutant ESCs. (B) K-means clustering (K = 9) of differentially expressed genes [|log2 (fold change)| > 0.7; multiple testing-adjusted p value < 0.05] in Tip60^fl/+ controls or Tip60^ci/ci mutant ESCs during the differentiation time course. Large up-regulated clusters are noted. Key regulatory proteins with impaired induction in Tip60^ci/ci mutant ESCs are highlighted. (C) Western blots (one of two independent experiments with similar results) of phosphorylated and total Smad2/3 and Erk1/2 during differentiation in Tip60^fl/+ or Tip60^ci/ci ESCs. (D) Model indicating the KAT-independent role of Tip60 in ESC self-renewal and gene regulation, as well as pre-implantation development, and the KAT-dependent role of Tip60 in differentiation and post-implantation development. See text for additional details.