Functional interdependence at the chromatin level between the MKK6/p38 and IGF1/PI3K/AKT pathways during muscle differentiation

Abstract

During muscle regeneration, the mechanism integrating environmental cues at the chromatin of muscle progenitors is unknown. We show that inflammation-activated MKK6-p38 and insulin growth factor 1 (IGF1)-induced PI3K/AKT pathways converge on the chromatin of muscle genes to target distinct components of the muscle transcriptosome. p38 alpha/beta kinases recruit the SWI/SNF chromatin-remodeling complex; AKT1 and 2 promote the association of MyoD with p300 and PCAF acetyltransferases, via direct phosphorylation of p300. Pharmacological or genetic interference with either pathway led to partial assembly of discrete chromatin-bound complexes, which reflected two reversible and distinct cellular phenotypes. Remarkably, PI3K/AKT blockade was permissive for chromatin recruitment of MEF2-SWI/SNF complex, whose remodeling activity was compromised in the absence of MyoD and acetyltransferases. The functional interdependence between p38 and IGF1/PI3K/AKT pathways was further established by the evidence that blockade of AKT chromatin targets was sufficient to prevent the activation of the myogenic program triggered by deliberate activation of p38 signaling.

Fig. 1. Pharmacological blockade of p38 and Pi3K in satellite cells leads to distinct undifferentiated phenotypes

Myofiber were isolated from 3 month-old C57 mice and placed in culture (20% FBS, 10% horse serum, 1% chick embryo extract-contaning medium) in the presence or in the absence of SB (10μM) or LY (20 μM). The inhibitors were replaced by fresh compounds every 24 hours. During this time, in absence of the inhibitors, satellite cells delaminate and formed myotubes with high efficiency A) Effect of p38 and Pi3K blockade on satellite cell differentiation and morphology was evaluated by phase contrast microscopy B) Expression of MCK and Cyclin A2 transcripts was measured by real time PCR in satellite cells C) ChIP with anti-AcH3 antibodies was performed on chromatin of satellite cells. The reported data represent real-time RT-PCR values normalized to input DNA and to the values obtained with normal rabbit IgG D) Global H3 acetylation level was monitored by western blot from whole cell extracts of myofiber-derived satellite cells. The values shown in A and C represent the average of standard error of three independent experiments.

Fig. 2. Inhibition of Pi3K prevents the recruitment of acetyltransferases and MyoD on chromatin of muscle-regulatory genes

A) ChIP analysis of Ebox-containing regions within the myogenin promoter and MCK enhancer was performed on chromatin of C2C12 cultured in growth medium (GM – 20% FBS) or in differentiation medium (DM – Serum free/insulin transferring containing medium), using the indicated antibodies B) Endonuclease assay of myogenin promoter, MCK enhancer and p21 promoter in the same culture conditions as A. Restriction endonuclease accessibility was performed in nuclei by digestion with BanI, followed by PCR to amplify fragments relative to the transcription start site (TSS) of myogenin promoter (top panel, left) and MCK enhancer (top panel, right). As a control, a p21 promoter fragment, which is devoid of Ban I sites was chosen. Absence or reduction of PCR signal indicates Ban I accessibility and reflects chromatin remodeling. The intensity of the amplified fragment was visualized (bottom panel, left) and quantified by densitometric analysis (bottom panel, right) C) MEF2-associated deacetylase activity was determined by deacetylation assay. Nuclear extracts from C2C12 cells undifferentiated (GM) or induced to differentiate (DM), with or without SB (5μM) and LY (20μM) were incubated with agarose-conjugated anti-MEF2A/D antibodies. Immunoprecipitates were incubated with tritium-labelled acetylated histones H4 and assayed for deacetylase activity, in presence or absence of the deacetylase inhibitor TSA (100nM). The values shown in B and C represent the average of standard error of three independent experiments.

Fig. 3. Pi3K/AKT-dependent interactions between MyoD and the p300 C-terminus during myogenic differentiation

A) Immunoprecipitation of endogenous MyoD was performed from nuclear extracts of C2C12 myoblasts either undifferentiated (GM) or incubated in differentiation medium (DM) for 24 hours) in the absence or the presence of SB or LY. Levels of precipitated MyoD and co-precipitated p300 and PCAF were detected by western blot. Nuclear extracts (N.E.) from untreated DM cells was used as input B) Simultaneous downregulation of AKT 1 and 2 or p38 α and β was achieved in C2C12 cells by RNAi and was evaluated by western blot using antibodies detecting endogenous p38 α/β and total AKT. Specific reduction of expression (about 80%) of the targeted genes is shown, as compared to control cells (transfected with scrambled oligos) C) Expression of myogenin (top) and MCK transcripts (bottom) measured by real time PCR in C2C12 depleted of AKT 1 and 2 or p38 α and β, as shown in B D) Immunoprecipitation of endogenous MyoD was performed from nuclear extracts of C2C12 myoblasts depleted of AKT 1 and 2 or p38 α and β and incubated in differentiation medium (DM) for 24 hours. Levels of precipitated MyoD and co-precipitated p300 were detected by western blot E) The effect of Pi3K blockade on the interaction between p300 and MyoD in C2C12 myoblasts induced to differentiate in DM for 24 hours was evaluated by a mammalian two-hybrid assay. Activation of the reporter reflects interactions between Gal4MyoD and p300-VP16 F) Schematic representation of AKT phosphorylation sites in the c-terminal region of p300. The values shown in C and E represent the average of standard error of three independent experiments.

Fig. 4. Pi3K/AKT-dependent phosphorylation of the C/H3 domain of p300 during myogenic differentiation

A) Phosphorylation of endogenous p300 was evaluated after metabolic labelling with ³²P orthophosphate C2C12 myoblasts either undifferentiated (GM) or induced to differentiate (DM) in the absence or in the presence (lower panel) of LY. Endogenous p300 was immunoprecipitated with anti-300 antibodies. Top panel: ³²P incorporation is revealed by autoradiography; bottom panel: levels of immunoprecipitated p300 revealed by western blot B) Phosphopeptide mapping of HA-C/H3 1640-1840 of p300 with constitutively active AKT1 myristoylated (myr) or AKT1 kinase defective (k.d.). 293 cells were transfected with the HA-tagged p300C/H3 fragment (aa 1649-1840) along with either AKT myr or AKT k.d. After metabolic labelling with ³²P the C/H3-p300 was immunoprecipitated with anti-HA. Immunoprecipitates were subjected to SDS PAGE and then to phospho-peptide mapping C) Phospho-aminoacid analysis of the two most prominent spots detected in B after the bands were excissed D) Phosphorylation of HA-tagged C/H3-p300 in C2C12 myoblasts either undifferentiated (GM) or induced to differentiate (DM) in the absence or presence of LY. C2C12 myoblasts were transfetced with HA-tagged p300C/H3; after metabolic labelling with ³²P orthophosphate the C/H3 p300 was immunoprecipitated with anti-HA. Top: p32 incorporation is revealed by autoradiography; bottom: levels of immunoprecipitated p300 C/H3 revealed by anti-HA E) Phosphorylation of HA-tagged C/H3-p300 wild type (wt) in C2C12 myoblasts either undifferentiated (GM) or induced to differentiate (DM) in the absence or presence of LY, and the HA-C/H3-p300 phospho-mutant (mt). After transfection and metabolic labeling with ³²P orthophosphate the C/H3 p300 was immunoprecipitated with anti-HA. Top: ³²P incorporation is revealed by autoradiography; bottom: levels of immunoprecipitated p300 CH3 revealed by anti-HA with alkaline phosphatase (AP) western blot. Quantification of two independent experiments was determined by densitometric analysis of the signal relative to the immunoprecipitated C/H3-p300 (see band indicated by arrow) and calculated as ³²P/AP signal ratio (right panel). The values represent the average of standard error of three independent experiments.

Fig. 5. Replacement of two potential AKT-target serines with non-phosphorylatable alanines or genetic ablation of one AKT isoform impairs the interaction between the p300 and MyoD during myogenic differentiation

The effect of differentiation cues (A) or the ectopic expression of constitutive active myristoylated AKT1 (myr) (B) on the interaction between Gal4-MyoD and the C/H3 domain of p300, either wt or the phospho-mutant (S1734/1834A) fused to VP16 (C/H3 VP16) was evaluated in C2C12 cells by a mammalian two hybrid assay, upon co-transfection with Gal4-luciferase reporter C) Co-immunoprecipitation of endogenous MyoD and ectopically expressed full length p300 wt and single mutants S1734A and S1834A performed in C2C12 cells after transfection of increasing concentrations (3 and 6 μg) of p300 mutants. Levels of precipitated MyoD and co-precipitated Flag-tagged p300 (top panel) were detected by western blot. Bottom panel shows total level of Flag-tagged p300 in nuclear extract of transfected cells D) Mammalian two-hybrid assay using Gal4MyoD and C/H3 p300 VP16 cotransfected along with Gal4-luciferase reporter in MEFs from wild type, AKT1 and AKT2 null mice E) ChIP analysis of Ebox-containing regions within the MCK enhancer was performed on chromatin of MEFs from wt, AKT 1 and 2 knock-out (KO) or MKK3/6 double KO mice, after adenoviral-mediated delivery of MyoDand culture in DM for 48 hours. The values shown in A, B, D and E represent the average of standard error of three independent experiments

Fig. 6. Functional interdependence at the chromatin level between the IGF1 and the p38 pathways

The p38 pathway was stimulated in C2C12 myoblasts (GM) by ectopic expression of the constitutive active form of the p38 upstream kinase, HA-MKK6EE, via adenoviral infection A) Expression level of myogenin, MCK, cyclin A, GAPDH and MKK6EE-HA transcripts was evaluated by RT-PCR. B) Activation of endogenous p38 (total and phosphorylated form), expression of myogenin, MyHC, cyclin A and tubulin were evaluated by western blot C) ChIP analysis of Ebox-containing regions within the myogenin promoter was performed using anti-acetyl H3 antibodies D) Endonuclease assay of myogenin promoter and p21 promoter (control). The intensity of the amplified fragment was visualized (botton panel, left) and quantified by densitometric analysis (bottom panel, right). E) Nuclei of C2C12 myoblasts (GM) were microinjected with a mixture of cDNA coding GFP (as an indicator of productive injection) and MKK6EE or empty vector, in the absence or presence of the p300 inhibitor LysCoA (0.5 mM) and the expression of the endogenous differentiation marker, MyHC, was analyzed by immunofluorescence after 48 hours. The percentage (%) of GFP/myogenin double positive cells, among the GFP positive cells, is reported as an index of differentiation. F) 10T1/2 fibroblasts were converted by transfection of MyoD, and the effect of co-expression of MKK6EE and increasing amounts of C/H3 p300 (either wt or the phosphorylation resistant double mutant) was analyzed after 48 hours of incubation in DM. The expression of MyHC was analyzed by immunofluorescence. The % of MyoD/myogenin double positive cells, among the MyoD positive cells, is reported as an index of differentiation. The values shown in C, D, F and E represent the average of standard error of three independent experiments

Fig. 7

Illustration of the composition of muscle transcriptosome on the chromatin of muscle genes that reflects different chromatin status and distinct cellular phenotypes generated in response to Pi3K or p38α/β inhibition, as compared to untreated cells.