Histone acetylation homeodynamics navigates cell survival and apoptosis

Abstract

The balance between inhibitor of apoptosis proteins (IAPs) and pro-apoptotic proteins (PAPs) tightly and precisely regulates cellular homeostasis. However, the epigenetic mechanism by which this balance is maintained in vivo remains largely unknown. Here we show that in various Drosophila tissues, the homeodynamics of H3K14ac/H3K27ac/H4K8ac on the promoters/enhancers of E93 and PAPs (rpr/hid), modulated by P300-CtBP/HDAC3, directs the decision between cell survival and the activation of hormone-induced developmental apoptosis. Concurrently, the homeodynamics of H3K14ac/H3K27ac/H4K8ac in IAPs (Diap1) promoters, modulated by Tip60/P300-CtBP/HDAC3, sustains cellular homeostasis by antagonizing the activities of PAPs. Notably, the epigenetic mechanism revealed in Drosophila is partially conserved in mammals. Moreover, disrupting the histone acetylation homeodynamics attenuates tumorigenesis through altering the balances between IAPs and PAPs in Drosophila and mice. In conclusion, histone acetylation homeodynamics navigates cell survival and apoptosis, suggesting potential epigenetic targets for the treatment of diseases or tumors caused by the imbalance between IAPs and PAPs.

Fig. 1. Blocking CtBP induces apoptosis by inducing the expression of E93.

a–e Fat body-specific CtBP-i with ppl-Gal4, ppl-Gal4>wt was used as control. Developmental time statistics (a), lethality statistics (b), phenotype observation at pupal stage (c), TEM image of the nucleus (N, nucleus) at the early wandering stage (d), and caspase 3 activity measurement (e), the fat body from 5 lavar as one group. In b and e, mean ± SD; n = 3 independent samples. In e, two-tailed paired t test: ***p < 0.001. f Fat body-specific CtBP-i results in gene expression changes at three developmental stages detected by RNA-seq. Heatmap to show FPKM of genes encoding E93, PAPs, MPAPs, IAPs, and caspases. E93, PAPs, and Diap1 is marked with yellow rectangles. EW, early wandering; LW, late wandering; WPP, white prepupae. g 4’,6-diamidino-2-phenylindole (DAPI) staining was used to detect the induction of apoptosis in the fat body (FB) for there is no developmental apoptosis occurring according to Cyt-c and active caspase 3 (A-cas 3) staining shown in Supplementary Fig. 1i. h, i’ After CtBP-i, UAS-E93, CtBP-i & E93-i, and CtBP-i & UAS-Diap1 at the EW, DAPI staining was performed to label nuclei in the fat body using ppl-Gal4 (h) and Flp-out lines (i). In i GFP clone cells represent cells subjected to gene RNA-i or overexpression. The size of the cell nucleus statistics of clone cells compared to wt cells in i (i’). In i’, mean ± SD; n = 6 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). j Immunofluorescence (IF) staining of Cyt-c and A-cas 3 were used to detect both the induction and blockage of apoptosis in the salivary glands (SG) for there is developmental apoptosis occurring shown in Supplementary Fig. 1j. k and k’ After CtBP-i, UAS-E93, CtBP-i & E93-i, and CtBP-i & UAS-Diap1 at the EW, IF staining of A-cas 3 and cytoplasmic Cyt-c in the salivary glands using Flp-out line (k). The fluorescence intensity statistics of clone cells compared to wt cells (k’). In k’, mean ± SD; n = 4 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). l IF staining of A-cas 3 and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining were used to detect the induction of apoptosis in the wing discs (WD) for there is little developmental apoptosis occurring shown in Supplementary Fig. 1k. m and m’ After CtBP-i, UAS-E93, CtBP-i & E93-i, and CtBP-i & UAS-Diap1 at the EW, IF staining of A-cas 3 in the wing discs using en-Gal4 or Gal80^ts, en-Gal4 line (m). GFP region indicates the posterior (P) boundaries where gene RNA-i or overexpression. The fluorescence intensity statistics of the P boundaries compared to anterior (A) boundaries (m’). In m’, mean ± SD; n = 4 independent wing disc. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). n Diagram showing that CtBP suppresses E93-mediated apoptosis. Source data are provided as a Source Data file. The genotypes are provided in Supplementary Table 3.

Fig. 2. CtBP-HDAC3 suppresses the expression of E93.

a, b’ After HDAC3-i, HDAC3-i & E93-i, and HDAC3-i & UAS-Diap1 at the EW, DAPI staining in the fat body using ppl-Gal4 (a) and Flp-out (b) line. The size of the cell nucleus statistics of clone cells compared to wt cells in b (b’), mean ± SD; n = 6 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). (c) Relative transcript levels of E93 and PAPs/IAPs in the fat body after HDAC3-i using ppl-Gal4 at the EW, ppl-Gal4>wt was used as control. Mean ± SD; n = 3 independent samples, two-tailed paired t test: *p < 0.05, **p < 0.01. d and d’ After HDAC3-i, HDAC3-i & E93-i, and HDAC3-i & UAS-Diap1 at the EW, IF staining of Cyt-c and A-cas 3 in the salivary glands using Flp-out line (d). GFP clone cells represent cells subjected to gene RNA-i or overexpression. The fluorescence intensity statistics of clone cells compared to wt cells (d’), mean ± SD; n = 4 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). e After UAS-CtBP, UAS-HDAC3, and UAS-CtBP & UAS-HDAC3 at 8-12 h APF, IF staining of Cyt-c and A-cas 3 in the salivary glands using Flp-out line. GFP clone cells represent cells subjected to gene RNA-i or overexpression. f Co-IP results showing the interaction between CtBP-V5 and HDAC3-Flag in Kc cells. g Diagram showing that CtBP-HDAC3 suppresses E93-mediated apoptosis. Data in e and f are representative of three independent experiments with similar results. Source data are provided as a Source Data file. The genotypes are provided in Supplementary Table 3.

Fig. 3. P300-CtBP/HDAC3 modulates H3K27ac/H4K8ac homeodynamics in the E93 enhancer and promoter.

a and a’ Immunofluorescence (IF) staining of Cyt-c and A-cas 3 after P300-i, and P300-i & UAS-E93 in the salivary glands using Flp-out line at 8-12 h APF. The fluorescence intensity statistics of clone cells compared to wt cells in a (a’), mean ± SD; n = 4 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). b, c’ After UAS-P300, UAS-P300 & E93-i, and UAS-P300 & UAS-Diap1 at the EW, DAPI staining in the fat body using ppl-Gal4 (b) and Flp-out (c) line. The size of the cell nucleus statistics of clone cells compared to wt cells in c (c’), mean ± SD; n = 6 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). d and d’ After UAS-P300, UAS-P300 & E93-i, and UAS-P300 & UAS-Diap1 at the EW, IF staining of A-cas 3 and Cyt-c in the salivary glands using Flp-out line (d). The fluorescence intensity statistics of clone cells compared to wt cells (d’), mean ± SD; n = 4 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). e Diagram showing potential modifications of CtBP/HDAC3-P300-modulated histone acetylation targeting on E93. Red font indicates CtBP-regulated histone acetylation modifications screened in Supplementary Fig. 6b. f The Integrative Genomics Viewer (IGV) tracks showing global H3K27ac and H4K8ac at the E93 gene locus at 96 h AEL and 6 h APF, respectively. R1-R5, region 1-region 5 of E93B nucleotide (−9059– + 7875 bp); green fonts and boxes, 5’ UTR of E93A/B; light brown fonts and boxes, exon of E93A/B. g Dual luciferase activity driven by the 20E-induced enhancer and basic promoter of E93B. R1, 20E-induced enhancer; R3, basic promoter. Mean ± SD; n = 3 independent samples. Two-tailed paired t test: ***p < 0.001. ns, not significant. h Enrichment of EcR and Pol Ⅱ on the 20E-induced enhancer (R1) and basic promoter (R5) detected by ChIP (Chromatin Immunoprecipitation)-qPCR, after global P300-i at 6 h APF using tub-Gal4, tub-Gal4>wt was used as control. Primers for ChIP-qPCR were located in peak-R1 and peak-R3. Mean ± SD; n = 3 independent samples. Two-tailed paired t test: *p < 0.05, **p < 0.01. i and j Detection of the GFP signal indicating the activity of 20E-induced E93 enhancer in CtBP-i, HDAC3-i, UAS-P300, and UAS-EcR-B1 (i) and UAS-CtBP, UAS-HDAC3, P300-i, and UAS-EcR-B1^DN (j) clone cells of the salivary glands using Flp-out line at the EW and 8-12 h APF, respectively. k Diagram showing that P300-CtBP/HDAC3 modulates H3K27ac/H4K8ac dynamics, which regulate 20E-induced E93 expression. EcRE, ecdysone response element. Source data are provided as a Source Data file. The genotypes are provided in Supplementary Table 3.

Fig. 4. P300-CtBP/HDAC3 modulates H3K14ac/H4K8ac homeodynamics in the rpr and hid promoters.

a–c In E93 mutant background, after CtBP-i, CtBP-i & UAS-Diap1, HDAC3-i, and HDAC3-i & UAS-Diap1 at the EW, DAPI staining in the fat body using ppl-Gal4 (a). In E93 mutant background, after UAS-P300 and UAS-P300 & UAS-Diap1 at the EW, DAPI staining in the fat body using Gal80^ts; ppl-Gal4 (b). The size of cell nucleus statistics is in a and b (c), mean ± SD; n = 12 independent image. Bars labeled with different lowercase letters are significantly different by one-way ANOVA (P < 0.05). d and d’ In E93 mutant background using mosaic analysis with a repressible cell marker (MARCM), after CtBP-i, CtBP-i & UAS-Diap1, HDAC3-i, HDAC3-i & UAS-Diap1, UAS-P300, and UAS-P300 & UAS-Diap1 at the EW, IF staining of A-cas 3 and Cyt-c in the salivary glands (d). GFP clone cells represent cells subjected to gene RNA-i or overexpression manipulation in E93 mutant. The fluorescence intensity statistics of clone cells compared to wt cells in d (d’), mean ± SD; n = 4 independent clone cell and wt cell. Two-tailed paired t test: ***p < 0.001. e and f Transcription of PAPs and IAPs after CtBP-i, HDAC3-i (e), and UAS-P300 (f) in E93 mutant background at the EW using ppl-Gal4 or Gal80^ts; ppl-Gal4. Mean ± SD; n = 3 independent samples. Two-tailed paired t test: *p < 0.05, **p < 0.01, ***p < 0.001. ns, not significant. g and h IGV tracks showing global H3K14ac/H4K8ac at the rpr (g) and hid (h) gene loci at 96 h AEL and 6 h APF. (i and j) Enrichment of H3K14ac and H4K8ac in the rpr (i) and hid (j) promoters detected by ChIP-qPCR, after global CtBP-i, HDAC3-i, and P300-i in E93 mutant background at 6 h APF using tub-Gal4. Primers for ChIP-qPCR were located in H3K14ac/H4K8ac promoters of rpr and hid. Mean ± SD; n = 3 independent samples. Two-tailed paired t test: *p < 0.05, **p < 0.01, ***p < 0.001. k Diagram showing that CtBP/HDAC3-P300 modulates H3K14ac/H4K8ac homeodynamics on the promoter of PAPs (rpr/hid). Source data are provided as a Source Data file. The genotypes are provided in Supplementary Table 3.

Fig. 5. Tip60/P300-CtBP/HDAC3 modulates H3K14ac/H3K27ac/H4K8ac homeodynamics in the Diap1 promoter.

a After Tip60-i and UAS-Tip60-i & UAS-Diap1 at the EW, DAPI staining in the fat body using ppl-Gal4, ppl-Gal4>wt was used as control. b and b’ After Tip60-i, UAS-Tip60-i & UAS-Diap1, and UAS-Tip60-i & UAS-p35 at the EW, DAPI staining was performed to label nuclei in the fat body using Flp-out line (b). The size of the cell nucleus statistics of clone cells compared to wt cells (b’), mean ± SD; n = 6 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). (c and c’) After Tip60-i, UAS-Tip60-i & UAS-Diap1, and UAS-Tip60-i & UAS-p35 at the EW, IF staining was performed to evaluate A-cas 3 and cytoplasmic Cyt-c in the salivary glands using Flp-out line (c). The fluorescence intensity statistics of clone cells compared to wt cells (c’), mean ± SD; n = 4 independent clone cell and wt cell. one-way ANOVA: different lowercase letters are significantly different (P < 0.05). d Transcription of IAPs (Diap1), PAPs (rpr, hid and grim) and E93 after Tip60-i in the fat body at 6 h APF using ppl-Gal4, ppl-Gal4>wt was used as control. Mean ± SD; n = 6 independent samples. Two-tailed paired t test: ***p < 0.001. ns, not significant. e Diagram showing potential CtBP/HDAC3-Tip60-modulated histone acetylation modifications targeting on Diap1. Red font indicates CtBP-regulated histone acetylation modifications sreened in Supplementary Figs. 6b and 10c. f IGV tracks showing global H3K14ac/H3K27ac/H4K8ac at the Diap1 gene locus at 96 h AEL and 6 h APF. R1, H3K14ac region; R2, H3K27/H4K8ac region; R3, H4K8ac region. Green box, 5’ UTR of Diap1. g Enrichment of H3K14ac/H3K27ac/H4K8ac in the R1, R2 and R3 regions detected by ChIP-qPCR, after Tip60-i and P300-i at 6 h APF using tub-Gal4, tub-Gal4>wt was used as control. Primers for ChIP-qPCR were located in H3K14ac-R1, H3K27/H4K8ac-R2, and H4K8ac-R3 of Diap1. Mean ± SD; n = 3 independent samples. Two-tailed paired t test: **p < 0.01, ***p < 0.001. h and i Detection of the GFP signal indicating H3K14ac/H3K27ac/H4K8 levels in the Diap1 promoter in Tip60-i and P300-i (h), UAS-Tip60 and UAS-P300 (i) clone cells of the salivary glands at 8-12 h APF using Flp-out line. RFP clone cells indicate Tip60-i, P300-i, UAS-Tip60 or UAS-P300 clone cells. j Diagram showing that Tip60/P300-CtBP/HDAC3 modulates H3K14ac/H3K27ac/H4K8ac homeodynamics at the promoters of IAPs (Diap1). Data in a is representative of three independent experiments with similar results. Source data are provided as a Source Data file. The genotypes are provided in Supplementary Table 3.

Fig. 6. Histone acetylation homeodynamics finely regulates PAP/MPAP and IAP/MIAP expression in mammalian cells.

a–c Evaluation of PAPs/MPAPs and IAPs/MIAPs transcription after knockdown of HomoCtBP (a), HomoHDAC3 (b), and HomoKAT5 (c) in A549 cells; shScramble was used as the control. Genes in red font: directly target genes by HomoCtBP, HomoHDAC3, and HomoKAT5. Mean ± SD; n = 3 independent samples. Two-tailed paired t test: *p < 0.05, **p < 0.01, ***p < 0.001. ns, not significant. d Heatmap showing the normalized signal values (Col Scale) of H3K27ac and H4K8ac peaks in PAPs/MPAPs and IAPs/MIAPs promoters after knockdown of HomoCtBP and HomoHDAC3 in A549 cells. Signal values of H3K27ac and H4K8ac in the squares are origin values enriched by CUT&Tag, and the enrichment regions were labeled with black rectangle in Supplementary Fig. 12b. e Heatmap showing the normalized signal values of H3K14ac peaks in PAPs/MPAPs and IAPs/MIAPs promoters after knockdown of HomoKAT5 in A549 cells. Signal values of H3K14ac in the squares are origin values enriched by CUT&Tag, and the enrichment regions were labeled with black rectangle in Supplementary Fig. 12d. f–h’ After knocking down HomoCtBP, HomoHDAC3, HomoCREBBP, and HomoKAT5 in A549 cells, phosphatidylserine exposure was evaluated by flow cytometry-based Annexin V‒PE staining (f). Changes in the mitochondrial membrane potential were evaluated by TMRE staining, CCCP was used as the positive control (g), and statistical analysis of the fluorescence data (g’), n = 8 independent image. Detection of changes in the cytoplasmic Cyt-c level by western blotting (h), and gray value statistics of protein bands (h’), n = 3 independent samples. Data are presented as mean ± SD; Two-tailed paired t test: *p < 0.05, **p < 0.01, ***p < 0.001. In f gating strategy of flow cytometry was in Supplementary Fig. 14a. Source data are provided as a Source Data file.

Fig. 7. Disruption of histone acetylation homeodynamics attenuates tumorigenesis in Drosophila and mice.

a, b’ Evaluation of tumorigenesis after CtBP-i, HDAC3-i, and Tip60-i in Ras^V12 (a) or Yki^3SA (b) overexpression-induced tumors in Drosophila wing discs using Flp-out line, with statistical analysis of the tumor size in a (a’) and b (b’). Tumor cells were labelled by GFP. Data are presented as mean ± SD. Two-tailed paired t test: ***p < 0.001. In a’, n = 8, 8, 8, 8 independent wing disc. In b’, n = 9, 8, 9, 8 independent wing disc. c, d’ Evaluation of tumorigenesis after CtBP-i, HDAC3-i, and Tip60-i in scrib-i (c) or lgl2-i (d)-induced tumors between the two domains of the dorsal medial fold in Drosophila wing discs using upd-Gal4, with statistical analysis of the tumor size in c (c’) and d (d’). The two domains of the dorsal medial fold were labelled by GFP, and the tumor location was marked by red arrows. Data are presented as mean ± SD. Two-tailed paired t test: *p < 0.05, ***p < 0.001. In c’, n = 22, 29, 27, 27 independent wing disc. In d’, n = 15, 16, 8, 10 independent wing disc. e–g After subcutaneous injection of stable HomoCtBP, HomoHDAC3, HomoCREBBP, and HomoKAT5 knockdown A549 cells, Scramble knockdown A549 cells was used as the control group. The mean tumor volume in each group before sacrifice (e), mean ± SEM. n = 6 independent mice. Images (f) and wet weights (g) of the tumors from each group after sacrifice, mean ± SD; n = 6 independent mice. Two-tailed paired t test: *p < 0.05, **p < 0.01. In a’-d’ and g, box plots display the distribution of tumor size and wet weights. The top and bottom of the box represent the 75th (Q3) and 25th (Q1) percentiles, respectively. The line inside the box indicates the median. The whiskers extend to the farthest data points within 1.5× interquartile range (IQR) from the quartiles, and points beyond the whiskers are defined as outliers and plotted as individual dots. Source data are provided as a Source Data file. The genotypes are provided in Supplementary Table 3.