Acetylation and phosphorylation of SRSF2 control cell fate decision in response to cisplatin

Abstract

SRSF2 is a serine/arginine-rich protein belonging to the family of SR proteins that are crucial regulators of constitutive and alternative pre-mRNA splicing. Although it is well known that phosphorylation inside RS domain controls activity of SR proteins, other post-translational modifications regulating SRSF2 functions have not been described to date. In this study, we provide the first evidence that the acetyltransferase Tip60 acetylates SRSF2 on its lysine 52 residue inside the RNA recognition motif, and promotes its proteasomal degradation. We also demonstrate that the deacetylase HDAC6 counters this acetylation and acts as a positive regulator of SRSF2 protein level. In addition, we show that Tip60 downregulates SRSF2 phosphorylation by inhibiting the nuclear translocation of both SRPK1 and SRPK2 kinases. Finally, we demonstrate that this acetylation/phosphorylation signalling network controls SRSF2 accumulation as well as caspase-8 pre-mRNA splicing in response to cisplatin and determines whether cells undergo apoptosis or G(2)/M cell cycle arrest. Taken together, these results unravel lysine acetylation as a crucial post-translational modification regulating SRSF2 protein level and activity in response to genotoxic stress.

Conflict of interest statement

The authors declare that they have no conflict of interest.

Figures

Figure 1

SRSF2 is acetylated on its lysine 52 (K52) residue by Tip60. (A) SAOS2 cells were transfected for 48 h with a HA-tagged SRSF2 expression vector. Whole-cellular extracts were subjected to immunoprecipitation (IP) with either anti-HA antibody or irrelevant immunoglobulin (IgG) as a negative control, followed by immunoblotting with anti-acetyl-lysine (Ac-K) or anti-HA antibody. (B) Endogenous acetylated SRSF2 protein was detected from H69 nuclear-enriched extract after immunoprecipitation of acetylated proteins with an anti-Ac-K (left panel) or anti-SRSF2 (right panel) antibody and immunodetection using anti-SRSF2 or anti-Ac-K antibody, respectively. (C) Representation of recombinant truncated GST–SRSF2 fusion proteins and recombinant His-tagged Tip60212–513 and GCN5 acetyltransferases. (D) Purified GST–SRSF2 fusion proteins were incubated with recombinant His-tagged Tip60212–513 or hGCN5 in the presence of [14C]acetyl-CoA. Acetylation was revealed after autoradiography (upper panel). Equivalent amounts of various recombinant proteins were assessed by Coomassie staining (lower panel). (E) H1299 cells were transfected for 48 h with either mismatch or Tip60 siRNAs. Total cellular extracts were subjected to immunoprecipitation with an anti-Ac-K (upper panel) or an anti-SRSF2 (middle panel) antibody followed by immunoblotting with anti-SRSF2, anti-SRSF1 or anti-Ac-K antibody as indicated. Neutralization of Tip60 was controlled by quantitative RT–PCR (lower panel). (F) GST–SRSF2 (1–60) wild-type (WT) or point mutants (K17R; K52R) recombinant proteins were incubated with [14C]acetyl-CoA in the presence (+) or absence (−) of recombinant His-tagged Tip60212–513. Analysis of SRSF2 acetylation was performed as in D. K17R and K52R: substitution of the indicated lysine by arginine (lower panel). A full-colour version of this figure is available at The EMBO Journal Online.

Figure 2

Acetylation by Tip60 controls SRSF2 turnover. (A) H358 and H1299 cell lines were transfected for 72 h with mismatch or Tip60(2) siRNAs. SRSF2, SRSF1 and Tip60 protein levels were analysed by western blotting. Actin was used as a loading control (upper panel). SRSF2 mRNA levels were quantified by quantitative RT–PCR. G3pdh was used as an internal control (lower panel). (B) H1299 cells were co-transfected for 72 h with mismatch or Tip60(2) siRNAs in the presence or absence of either wild-type (WT) HA-Tip60 or mutant (G380) HA-Tip60G380 with impaired HAT domain. SRSF2 protein expression was assessed by western blotting (upper panel). Neutralization or overexpression of Tip60 was verified by quantitative RT–PCR (lower panel). (C) HA-tagged SRSF2 protein was co-expressed for 36 h in H1299 cells with either wild-type (WT) HA–Tip60 or mutant (G380) HA–Tip60G380, and treated (+) or not (−) for an additional 18 h with MG132. Western blot analysis using anti-HA antibody is presented (upper panel). SRSF2 and actin signal intensities were quantified using the ImageJ software and the relative densitometric areas for SRSF2–HA were determined according to actin signal in each condition (lower panel). (D) H1299 cells were transfected for 36 h with plasmid encoding either wild-type (WT) HA–Tip60 or mutant (G380) HA–Tip60G380 protein and treated (+) or not (−) for an additional 18 h with MG132. SRSF2 protein level was analysed by western blot. (upper panel) The relative densitometric areas for SRSF2 were determined as in C (lower panel). (E) H1299 cells were transfected for 36 h with plasmid encoding lysine 52 mutant (K52R) HA-tagged SRSF2 protein in the presence or absence (−) of either wild-type (WT) HA–Tip60 or mutant (G380) HA–Tip60G380 and treated (+) or not (−) for an additional 18 h with MG132 as indicated. SRSF2(K52R) and Tip60 proteins were detected by western blotting using anti-HA antibody (upper panel) and the relative densitometric areas for SRSF2(K52R) were determined (lower panel). (F) H1299 cells were transfected for 48 h with HA–SRSF2 WT or K52R expression vector then treated with cycloheximide (CHX) for the indicated times. Whole-cellular extracts were subjected to western blotting using anti-HA antibody (upper panel). HA–SRSF2 densitometric signals were normalized to actin. A value of 1 was arbitrarily assigned to the signal obtained at 0 time of CHX treatment (lower panel).

Figure 3

HDAC6 acts as a positive regulator of SRSF2 protein level by counteracting Tip60-mediated effects. (A) H1299 cells were co-transfected for 48 h with HA-tagged SRSF2 and Flag-tagged HDAC6 expression vectors. SRSF2 protein level was analysed by western blotting using an anti-HA antibody. (B) H358 and H810 cells were transfected for 72 h with mismatch or HDAC6 siRNAs. Expression of SRSF2 and SRSF1 proteins was studied by western blotting. (C) H1299 cells were transfected for 72 h with either mismatch or HDAC6 siRNAs. Total cellular extracts were subjected to immunoprecipitation (IP) with anti-Ac-K52 SRSF2 antibody followed by immunoblotting with anti-SRSF2 antibody (upper panel). H1299 cells were transfected for 48 h with HA-tagged SRSF2 WT or K52R expression vector, in the presence (+) or absence (−) of Flag-tagged HDAC6 and subjected to western blot analysis for detection of HA–SRSF2 expression. The relative densitometric areas for HA–SRSF2 were determined according to actin signal (middle panel). H1299 cells were co-transfected for 48 h with HA-tagged SRSF2 WT or K52R expression vector, in the presence or absence of mismatch or HDAC6 siRNA as indicated and subjected to western blot analysis for detection of HA–SRSF2 or HDAC6 expression (lower panel). (D) Whole-cellular extracts derived from 3T3 cell lines wild-type (WT), HDAC6−/− (KO) or HDAC6−/− re-expressing the wild-type HDAC6 (+WT) were used to analyse SRSF2 protein level (upper panel). One of the HDAC6-deficient clones was used to establish new lines re-expressing wild-type HDAC6 (+WT) or a catalytically dead (+HDm) or a non-ubiquitin-binding mutant of HDAC6 (Ubm; Boyault et al, 2007b). Whole-cellular extracts were used to study SRSF2 expression (lower panel). (E) Same cells were treated (+) or not (−) with proteasome inhibitor MG132 for 18 h and SRSF2 expression was analysed by western blot. (F) HA-tagged SRSF2 WT or K52R protein was co-expressed for 48 h with HA-tagged Tip60 protein in H1299 cells, in the presence of increasing amount of Flag-tagged HDAC6 expression vector. SRSF2 protein expression was detected by western blotting using an anti-HA antibody.

Figure 4

Tip60 downregulates SRSF2 phosphorylation by preventing SRPK1/2 nuclear translocation. (A) H1299 cells were co-transfected for 48 h with HA-tagged SRSF2 and either HA–Tip60 WT or HA–Tip60G380 expression vectors and treated for 18 h with MG132. SRSF2 phosphorylation was studied by western blotting using either mAb104 or phosphorylated SRSF2 protein (P-SRSF2) antibody. (B) H1299 cells were transfected for 48 h with plasmid encoding HA-tagged SRSF2 protein in the presence or absence of HA-tagged Tip60 expression vector and treated with MG132 18 h before harvesting. Cellular extracts were subjected to immunoprecipitation (IP) with an anti-HA antibody followed by immunoblotting with an anti-P-SRSF2 or anti-SRSF2 antibody. (C) H358 cells were transfected for 72 h with either mismatch or Tip60(1) siRNAs. SRSF2 phosphorylation was studied by western blotting using a specific P-SRSF2 antibody (left panel). Neutralization of Tip60 was controlled by quantitative RT–PCR (right panel). (D) HA-tagged SRSF2 protein was co-expressed in H1299 cells with HA-tagged Tip60 protein. Total cellular extracts (upper panel) or cytoplasmic (C) and nuclear (N) extracts (lower panel) were subjected to western blot. Anti-α-tubulin and anti-histone H3 antibodies were used to assess the fractionation efficiency. (E) H810 cells were transfected for 72 h with mismatch or Tip60(1) siRNAs. Total cellular extracts were immunoblotted with anti-SRPK1 or anti-SRPK2 antibody (upper panel). Cytoplasmic (C) and nuclear (N) extracts were subjected to immunoblotting using the indicated antibodies. (lower panel; F) H810 cells were transfected for 72 h with mismatch or Tip60(2) siRNA. Immunolocalization of either SRPK1 or SRPK2 protein (green) was visualized by immunofluorescence. DAPI staining (blue) was used to counterstain nuclei.

Figure 5

SRSF2 accumulates in a hypoacetylated/phosphorylated form in response to cisplatin treatment and is required for apoptosis. (A, B) H358 cells were transfected for 48 h with either mismatch or SRSF2 siRNAs and treated (+) or not (−) for an additional 24 h with cisplatin (100 μM). (A) Western blot analysis was performed using anti-SRSF2 or anti-active caspase-3 antibody (left panel). Apoptosis was quantified by flow cytometry analysis of active caspase-3 in untreated (grey bars) or treated (black bars) cells (right panel). (B) RT–PCR analysis using specific primers of caspase-8 splice variants. Gapdh was used as an internal control (left panel). Densitometric signals were quantified using the Image J software and the relative ratio caspase-8L/caspase-8a was calculated in either untreated (grey bars) or treated (black bars) cells (right panel). (C) H810 cells were treated or not for 24 h with 100 μM cisplatin. Nuclear-enriched extracts were prepared and immunoprecipitation (IP) of acetylated proteins was performed using an anti-acetyl-lysine (Ac-K) antibody followed by immunoblotting using either an anti-SRSF2 or anti-E2F1 antibody. (D) H810 cells were treated or not for 24 h with 100 μM cisplatin and subjected to western blot analysis (left panel). Tip60 transcript level was quantified by quantitative RT–PCR (right panel). (E) H358 cells were transfected for 48 h with mismatch or HDAC6 siRNAs and treated (+) or not (−) for an additional 24 h with 100 μM cisplatin. Western blot analysis was performed for the detection of SRSF2 or HDAC6 protein (upper panel). Wild-type or HDAC6 knockout 3T3 cells were treated (+) or not (−) with 100 μM cisplatin for 24 h and SRSF2 protein level was assessed by immunoblotting (lower panel). (F) Total cellular extracts were obtained from H358 and H1299 cells treated or not for 24 h with 100 μM cisplatin. Phosphorylated SRSF2 protein (P-SRSF2) and SRSF2 protein levels were analysed by western blotting. (G) Cytoplasmic and nuclear extracts from H810 cells treated or not for 24 h with 100 μM cisplatin were subjected to western blotting.

Figure 6

SRPK2 but not SRPK1 is required for SRSF2 phosphorylation and apoptosis in response to cisplatin. (A, B) H358 cells were transfected for 48 h with either mismatch, SRPK1 or SRPK2 siRNAs and treated (+) or not (−) for an additional 24 h with 100 μM cisplatin. (A) Western blot analyses of the indicated proteins (left panel). Apoptosis was quantified in untreated (grey bars) or treated (black bars) cells by analysis of caspase-3 activation by flow cytometry (right panel). (B) RT–PCR analyses of caspase-8 splice variants. Gapdh was used as an internal control (left panel). The relative ratio caspase-8L/caspase-8a was calculated in either untreated (grey bars) or treated (black bars) cells after densitometric quantification of the signals (right panel). (C) H358 cells were transfected for 48 h with mismatch or SRPK2 siRNAs and treated or not for an additional 24 h with 100 μM cisplatin. Cell cycle distribution was analysed by flow cytometry. Percentages of cells in the different phases are indicated.

Figure 7

A model for the roles of Tip60, HDAC6, SRPK1 and SRPK2 proteins in the control of SRSF2 protein in both unstressed (A) and genotoxic stressed (B) conditions. (A) In unstressed cells, Tip60 acetylates SRSF2 on its lysine 52 residue and prevents SRPK1 and SRPK2 nuclear localization. SRSF2 accumulates in a hyperacetylated/hypophosphorylated form that is subjected to proteasomal degradation. This effect is counterbalanced by the HDAC6 deacetylase, which positively controls SRSF2 protein level by deacetylating SRSF2 and preventing its proteasomal degradation. (B) On cisplatin treatment, Tip60 protein level strongly decreases leading to the nuclear accumulation of both SRPK1 and SRPK2 kinases, to the stabilization of SRSF2 in a hypoacetylated/phosphorylated form by a mechanism involving HDAC6. SRSF2 and SRPK2 cooperate to induce apoptosis through regulation of the splicing switch of caspase-8 pre-mRNA. In contrast, SRPK1 appears to act as an anti-apoptotic factor. In the absence of SRPK2, apoptosis is blocked and treated cells accumulate in G2/M phase indicating that SRPK-mediated SRSF2 phosphorylation controls cell fate decision.