AKT phosphorylates H3-threonine 45 to facilitate termination of gene transcription in response to DNA damage

Abstract

Post-translational modifications of core histones affect various cellular processes, primarily through transcription. However, their relationship with the termination of transcription has remained largely unknown. In this study, we show that DNA damage-activated AKT phosphorylates threonine 45 of core histone H3 (H3-T45). By genome-wide chromatin immunoprecipitation sequencing (ChIP-seq) analysis, H3-T45 phosphorylation was distributed throughout DNA damage-responsive gene loci, particularly immediately after the transcription termination site. H3-T45 phosphorylation pattern showed close-resemblance to that of RNA polymerase II C-terminal domain (CTD) serine 2 phosphorylation, which establishes the transcription termination signal. AKT1 was more effective than AKT2 in phosphorylating H3-T45. Blocking H3-T45 phosphorylation by inhibiting AKT or through amino acid substitution limited RNA decay downstream of mRNA cleavage sites and decreased RNA polymerase II release from chromatin. Our findings suggest that AKT-mediated phosphorylation of H3-T45 regulates the processing of the 3' end of DNA damage-activated genes to facilitate transcriptional termination.

Figure 1.

AKT phosphorylates H3–45 in response to DNA damage. (A) AKT substrate sequences conserved in various proteins, including histones H2B and H3. (B) Immunofluorescence staining of phosphorylated AKT-serine 473 (p-AKT-S473) and phosphorylated H3-threonine 45 (p-H3-T45) in MCF10A cells. Cells were treated with DMSO, 100 μM etoposide (ETPS), 0.4 μg/ml adriamycin (ADR) or 50 J/m² UV irradiation (UV) for 18 h. DNA counterstained with DAPI; scale bar, 20 μm. (C) Western blot of samples in (B). (D) MCF10A cells were treated with DMSO, 0.4 μg/ml ADR, 0.2 μM AKT inhibitor IV (IV) or 0.4 μg/ml ADR and 0.2 μM AKT inhibitor IV for 18 h. Total cell extracts were probed by western blot. Data shown are the representative of three independent experiments.

Figure 2.

AKT1 phosphorylates H3-T45 in vitro and in vivo. (A) In vitro kinase assay of recombinant His₆-tagged core histones purified from Escherichia coli. (B) AKT1 in vitro kinase assay of His₆-H3. F/L indicates the full-length of H3. 1–35 and 1–51 represent the corresponding amino acid sequences of purified H3. (C) AKT1 in vitro kinase assay of full-length H3, wild-type (WT) and T45 mutated to alanine (T45A). Autophosphorylated AKT1 (auto-p-AKT1) is shown as a kinase loading control. (D) In vitro AKT1 kinase assay with/without non-radiolabeled ATP or MgCl₂. Assay samples were probed with anti-phosphorylated H3-T45. (E) IP kinase assay of Myc-tagged blank vector and dominant-negative (DN) and constitutively-active myristoylated (Myr) AKT1. AKT1 constructs were transfected into HEK293T cells and cell lysates were immunoprecipitated with anti-Myc and subjected to in vitro kinase assay using His₆-H3 as a substrate. (F) Immunofluorescence staining of Myc-tagged DN/Myr-AKT1 overexpressed in MCF10A cells. DNA counterstained with DAPI. Scale bar, 20 μm. All data above are the representative of three independent experiments.

Figure 3.

H3-T45 phosphorylation signal is most abundant near the TTS. MCF10A cells were treated with 0.4 μg/ml ADR for 18 h and analyzed by ChIP-seq. (A) Functional annotation of H3-T45 phosphorylation peak. (B) Pie chart showing the proportion of transcript coordinates for H3-T45 phosphorylation peaks (1041 regions). Proportion of transcript coordinates for H3-T45 phosphorylation peaks were compared with RefSeq transcripts. (C) Average profiles for phosphorylated Pol II-S2, S5, H3K36me3 and phosphorylated H3-T45 were plotted around ADR-induced H3-T45 phosphorylation-enriched genes (610 genes). (D) Phosphorylated RNA Pol II-S2 and phosphorylated H3-T45 ChIP peak distribution. (E and F) ChIP binding profiles of indicated genes. Scale data ranges are indicated on the right side of the individual track. Red boxes indicate the highest peak of phosphorylated H3-T45 signal. (G) Real-time qPCR analysis of CDKN1A mRNA in DMSO/ADR-treated MCF10A cells. (H) ChIP assay covering the CDKN1A locus above with the indicated antibodies. (I) ChIP-qPCR of the indicated gene locus with anti-phosphorylated H3-T45 and anti-phosphorylated RNA Pol II-S2. (J) ChIP-qPCR of promoter and TTS of CDKN1A, using anti-pan AKT. (K) ChIP-qPCR using anti-phosphorylated AKT-S473. qPCR was performed with complementary primers to the TTS of the indicated genes. ChIP-pPCR values were normalized with 1% input DNA. Real-time qPCR and ChIP assay data shown are the average values of at least three independent experiments. Standard deviations are indicated as error bars. * P < 0.05, **P < 0.001.

Figure 3.

H3-T45 phosphorylation signal is most abundant near the TTS. MCF10A cells were treated with 0.4 μg/ml ADR for 18 h and analyzed by ChIP-seq. (A) Functional annotation of H3-T45 phosphorylation peak. (B) Pie chart showing the proportion of transcript coordinates for H3-T45 phosphorylation peaks (1041 regions). Proportion of transcript coordinates for H3-T45 phosphorylation peaks were compared with RefSeq transcripts. (C) Average profiles for phosphorylated Pol II-S2, S5, H3K36me3 and phosphorylated H3-T45 were plotted around ADR-induced H3-T45 phosphorylation-enriched genes (610 genes). (D) Phosphorylated RNA Pol II-S2 and phosphorylated H3-T45 ChIP peak distribution. (E and F) ChIP binding profiles of indicated genes. Scale data ranges are indicated on the right side of the individual track. Red boxes indicate the highest peak of phosphorylated H3-T45 signal. (G) Real-time qPCR analysis of CDKN1A mRNA in DMSO/ADR-treated MCF10A cells. (H) ChIP assay covering the CDKN1A locus above with the indicated antibodies. (I) ChIP-qPCR of the indicated gene locus with anti-phosphorylated H3-T45 and anti-phosphorylated RNA Pol II-S2. (J) ChIP-qPCR of promoter and TTS of CDKN1A, using anti-pan AKT. (K) ChIP-qPCR using anti-phosphorylated AKT-S473. qPCR was performed with complementary primers to the TTS of the indicated genes. ChIP-pPCR values were normalized with 1% input DNA. Real-time qPCR and ChIP assay data shown are the average values of at least three independent experiments. Standard deviations are indicated as error bars. * P < 0.05, **P < 0.001.

Figure 4.

AKT1 phosphorylates H3-T45 phosphorylation more efficiently than AKT2. AKT knockdown MCF10A cells were treated with 0.4 μg/ml ADR for 18 h. (A) Real-time PCR analysis of AKT mRNA. (B) Western blot analysis of total cell extracts with the indicated antibodies. (C) Phosphorylated H3-T45 ChIP assay of the CDKN1A locus. (D) Real-time PCR analysis of CDKN1A mRNA. (E) Real-time qPCR analysis of the indicated genes in ADR-treated MCF10A cells. Real-time qPCR and ChIP assay data shown are the average values of at least three independent experiments. Standard deviations are indicated as error bars. *P < 0.05, **P < 0.001.

Figure 5.

H3-T45 phosphorylation is critical for transcriptional termination. (A) ChIP assay of phosphorylated H3-T45 from H3 WT or T45A mutant overexpressing MCF10A cells. (B) Real-time qPCR of CDKN1A mRNA in cells from (A) prior to crosslinking. (C) Real-time qPCR analysis of nascent RNA of H3 WT or T45A mutant-overexpressing MCF10A cells, treated with ADR. Values represent the fold induction, relative to WT. (D) MCF10A cells expressing WT H3 and T45A mutant were treated with ADR for 18 h. ChIP assay was performed with anti-RNA Pol II. Real-time qPCR was performed with primers corresponding to the TTS of the indicated genes. Real-time qPCR and ChIP assay data shown are the average values of at least three independent experiments. Standard deviations are indicated as error bars. *P < 0.05 **P < 0. 01 ***P < 0.001.

Figure 6.

Schematic of AKT1-mediated H3-T45 phosphorylation in transcriptional termination on DNA damage. DNA damage triggers the transcription of stress-response genes by RNA Pol II. As transcription progresses through the 3′ region of the locus, RNA Pol II-S2 undergoes hyperphosphorylation and DNA damage-activated AKT1 phosphorylates H3-T45. When RNA Pol II passes through the 3′ area of the poly(A) site, pre-mRNA is cleaved and is processed into mature mRNA (12,14). Phosphorylated H3-T45 accelerates the degradation of 5′ unprotected RNA downstream of the poly(A) site to induce the dissociation of RNA Pol II from chromatin.