RA induces the degradation of SRC-3 by the proteasome. (A) COS-1 cells were cotransfected with the full-length B10-SRC-3 vector (0.2 μg) along with RARα (0.1 μg) and the DR5-tk-CAT reporter gene (1 μg), and treated with RA (1 μM). Whole-cell extracts (WCEs) were analyzed by immunoblotting with B10, RARα and β-actin antibodies. (B) MG132 reverses the degradation of SRC-3 induced by RA at 16 h in COS-1 cells transfected as in (A). (C) Extracts from COS-1 cells cotransfected with Flag-tagged SRC3 (WT or AAA), RARα and RXRα vectors were incubated with a DR5 RARE, RA-treated and immunoprecipitated with RARα antibodies. Bound RARα and SRC-3 were analyzed by immunoblotting. Lanes 1 and 2 correspond to 1% of the amount of immunoprecipitated extracts. (D) Unlike SRC-3 WT, SRC-3AAA overexpressed in COS-1 cells as in (A), and is not degraded in response to RA. (E) Same as in (A) except that RXRα was cotransfected and RA was added for 16 h. (F) COS-1 cells were cotransfected with increasing concentrations of the SRC-1, TIF2/SRC-2 or B10-SRC-3 vectors along with RARα and the DR5-tk-CAT reporter gene and RA-treated for 16 h. WCEs were immunoblotted with SRC-1, TIF2, B10 or β-actin antibodies.

Inhibition of p38MAPK enhances the interaction of SRC-3 with RARα. (A) COS-1 cells cotransfected with the B10-SRC-3, RARα, RXRα and dnp38MAPK vectors were RA-treated for 2 h. WCEs were immunoprecipitated with RARα antibodies and bound RARα and SRC-3 proteins were analyzed by immunoblotting. The expression of dnp38MAPK was tested by immunoblotting of the extracts before immunoprecipitation. Lane 1 corresponds to 5% of the amount of immunoprecipitated extracts. (B) Extracts from COS-1 cells cotransfected as in (A) were incubated with a DR5 RARE in the absence or presence of the peptides corresponding to the LXD1, LXD2 and LXD3 motifs (40 μM), RA-treated and immunoprecipitated. In lane 3, a mutated DR5 (mDR5) element was used. (C) Same as in (B) with WCEs from cells cotransfected or not with the dnp38MAPK vector and with increasing concentrations (40, 120 and 240 μM) of the LXD2 peptide. (D) COS-1 cells cotransfected with the SRC-1 vector along with RARα, RXRα and dnp38MAPK were processed as in (C), in the absence or presence of the peptides specific for each SRC-1 LXXLL domain (40 μM).

Overexpressed SRC-3 compensates for RA-induced SRC-3 degradation. (A) Increasing amounts of dnp38MAPK, p38MAPK, SRC-3, SRC-1 and SRC-2 vectors were transfected in COS-1 cells along with the DR5-tk-CAT reporter gene and RARα. CAT activity was determined 16 h after RA treatment. The results are the mean±s.d. of three experiments. (B) Extracts corresponding to lanes 1 and 8–10 in (A) were analyzed for expression and degradation of SRC-3 and RARα by immunoblotting. (C) COS-1 cells were transfected with the SRC-3 WT or SRC-3(S860A) vectors along with DR5-tk-CAT reporter gene and RARα. CAT activity was analyzed as in (A). (D) Extracts from lanes 2 and 3 in (C) were analyzed for expression and degradation of RARα and SRC-3 by immunoblotting.

p38MAPK controls the RA-induced maturation of NB4 and HL60 cells through SRC-3. NB4 and HL60 cells, seeded at an initial concentration of 150 000 cells/ml, were treated for 3 days with vehicle, RA (0.01 μM for NB4 cells and 1 μM for HL60 cells), PD 169316 (1 μM), SB203580 (5 μM), SP600125 (5 μM) either alone or in combination. In all experiments, the percentage of viable cells was >90%. (A) Immunoblotting analysis of RA-induced SRC-3 degradation. (B) NBT-R activity was measured in cell homogenates. Each value represents the mean±s.d. of three independent culture flasks from two independent experiments. (C) FACS analysis of the CD11b, CD11c and CD33 markers. The results correspond to the mean associated fluorescence (MAF) for each marker. (D) Western blot analysis of STAT1(p91), β-actin and cEBPβ (the 36–38 kDa LAP form) at 5 and 72 h. (E) Knockdown of SRC-3 in NB4 cells electroporated with siSRC-3 (80 and 160 nM) was demonstrated by immunoblotting with siGAPDH (100 nM) as a negative control. Protein loading was controlled by immunoblotting of β-actin. Knockdown of SRC-3 by siRNA decreased the effects of RA and PD 169316 on the expression of cEBPβ (F) and NBT-R activity (G).

Schematic representation of (A) RARα1 and RARγ2 proteins with the phosphorylation sites and (B) SRC-3 with the known functional domains and the p38MAPK phosphorylation sites. The truncated mutants are also shown.

Phosphorylation by p38MAPK signals SRC-3 degradation. (A) COS-1 cells cotransfected as in Figure 2A were RA-treated and labelled with [³²P] orthophosphate. SRC-3 phosphorylation was analyzed by autoradiography and immunoblotting after immunoprecipitation of the extracts with B10 antibodies. (B) Same as in (A) with and without dnp38MAPK (0.2 μg). (C) Kinetics of p38MAPK activation as assessed by in vitro phosphorylatiion of equal amounts of recombinant ATF2 by phospho-p38MAPK immunoprecipitated from extracts of RA-treated COS-1 cells. Phosphorylated ATF-2, immunoprecipitated P-p38 MAPK and p38MAPK in the extracts before immunoprecipitation were revealed by immunoblotting. (D) Same as in (C) in the absence and presence of dnp38MAPK. (E) In vitro phosphorylation of immobilized GST or GST-SRC-3 (aa 481–1417) proteins with P-p38MAPK immunoprecipitated from COS-1 cells treated or not with RA for 16 h was analyzed by autoradiography and immunoblotting. (F) Same as in (E), with recombinant p38MAPK or p42/p44MAPK. (G) Effect of dnp38MAPK overexpression and PD98059 (5 μM) on the degradation of SRC-3 in COS-1 cells treated with RA for 16 h. Protein loading and dnp38MAPK expression were checked by reprobing the membrane with β-actin and p38MAPK antibodies. (H) COS-1 cells were transfected with the vectors expressing FLAG-SRC-3 WT or the phosphorylation-deficient Flag-SRC-3 mutants (S505A, S547A, S860A and S867A) and RA-treated for 16 h. SRC-3 degradation was analyzed by immunoblotting with Flag antibodies.

p38MAPK controls RARα-mediated transcription through SRC-3 phosphorylation. (A) COS-1 cells transfected with the DR5-tk-CAT reporter gene along with RARα were RA-treated for the indicated times and analyzed for CAT activity. The results, which were normalized to the CAT activity in the absence of RA, are the mean±s.d. of at least four experiments. (B) Effect of dnp38MAPK, and p38MAPK on RARα-mediated activation of the DR5-tk-CAT reporter gene. (C) RARαS77A is less efficient than RARα WT in inducing CAT activity but its activity is similarly enhanced at 16 h by dnp38MAPK. The results are the average of two experiments that agreed within 15%. (D) Overexpression of dnp38MAPK does not affect the phosphorylation of RARα as determined by immunoblotting with antibodies recognizing specifically RARα, RARα phosphorylated at S77 and p38MAPK. (E) Overexpression of dnp38MAPK in COS-1 cells does not affect the degradation of RARα induced by RA.

SiRNA-induced knockdown of SRC-3 abrogates the enhancing effect of dnp38MAPK on RARα-mediated transcription. (A) COS-1 cells transfected with control, SRC-1 or SRC-3 siRNAs (50 nM) were subsequently transfected with the SRC-3 and RARα vectors and the DR5-tk-CAT reporter gene and RA-treated for 16 h. Knockdown of overexpressed SRC-3 was analyzed by immunoblotting as well as RARα expression and degradation. Protein loading was controlled by immunoblotting of β-actin. (B) Same as in (A) with overexpressed SRC-1. (C) SiRNA against SRC-3 and not siRNA against SRC-1 abrogate the enhancing effect of overexpressed SRC-3 and dnp38MAPK on RARα-mediated activation of a DR5-tk-CAT reporter gene. (D) In HeLa cells, siRNA against SRC-3 also reduce the levels of endogenous and overexpressed SRC-3 and decrease the activation of a DR5-tk-CAT reporter gene. (E) In HeLa cells, siSRC-3 and not siSRC-1 abrogate the enhancing effect of dnp38MAPK on CAT activity. The results are an average of two experiments that agreed within 15%.

When recruited to RARγ, SRC-3 is not phosphorylated nor degraded in response to RA. (A) COS-1 cells were cotransfected with the B10-SRC-3, RXRα, RARγ and dnp38MAPK vectors and WCEs were incubated with a DR5 RARE, in the absence or presence of the peptides specific for each SRC-3 LXXLL domain (40 μM) and RA-treated. After RARγ immunoprecipitation, bound proteins were analyzed by immunoblotting. Lane 1 corresponds to 5% of the amount of immunoprecipitated extracts. (B) SRC-3 phosphorylation was analyzed in transfected COS-1 cells treated with RA either alone or in combination with PD169316 (1 μM) and labeled with [³²P] orthophosphate as in Figure 3A. (C) SRC-3 and RARγ degradation was analyzed by immunoblotting. (D) RARγ phosphorylation was analyzed by immunoblotting with antibodies recognizing RARγ and its phosphorylated form. (E) CAT activity in COS-1 cells cotransfected with the RARγ vector (WT or mutated) and a DR5-tk-CAT reporter, in the absence or presence of WT or dnp38MAPK and treated with RA alone or combined to PD98059 (5 μM) or PD169316 (1 μM). The results are the mean±s.d. of at least three experiments. (F–G) HeLa cells were cotransfected with the SRC-3 and p65 vectors along with an NF-κB-luciferase reporter gene and dnp38MAPK, and treated with TNFα (100 nM) for 16 h. Extracts were immunoprecipitated with p65 antibodies and analyzed for bound SRC-3 and p65 proteins by immunoblotting (F) or for luciferase activity (G). The results are the mean±s.d. of three experiments.