TEF-1 and C/EBPbeta are major p38alpha MAPK-regulated transcription factors in proliferating cardiomyocytes

Abstract

p38 MAPKs (mitogen-activated protein kinases) play important roles in the regulation of cellular responses to environmental stress. Recently, this signalling pathway has also been implicated in the regulation of processes unrelated to stress, for example, in T lymphocytes and cardiomyocytes. In order to identify molecular targets responsible for the housekeeping functions of p38 MAPKs, we have analysed the differences in the transcriptomes of normally proliferating wild-type and p38alpha knockout immortalized embryonic cardiomyocytes. Interestingly, many potential components of the myocardium extracellular matrix were found to be upregulated in the absence of p38alpha. Further analysis of the microarray data identified TEF-1 (transcriptional enhancer factor-1), a known regulator of heart-specific gene expression, and C/EBPbeta (CCAAT/enhancer-binding protein beta), as the two transcription factors the binding sites of which were most enriched in the promoters of p38alpha-regulated genes. We have focused on the study of the extracellular matrix component COL1A1 (alpha1 chain of type I collagen) and found evidence for the involvement of both TEF-1 and C/EBPbeta in the p38alpha-dependent inhibition of COL1A1 transcription. Our data therefore show that p38 MAPKs regulate TEF-1 and C/EBPbeta transcriptional activity in the absence of environmental stress and suggests a role for p38alpha in the expression of extracellular matrix components that maintain organ architecture.

Figure 1. Summary of the genes for which expression levels significantly changed in proliferating p38α^−/− versus wt cardiomyocytes

Three independent p38α^−/− and wt cardiomyocyte cultures were analysed and only those genes that showed a greater than ±1.5-fold changes in all three experiments were selected. Gene functions were initially categorized based on the NIA mouse 15K cDNA gene ID list and modified when necessary. Note that many genes that are differentially regulated in the p38α^−/− cardiomyocytes encode matrix and structural proteins.

Figure 2. Up-regulation of COL1A1 and COL3A1 in p38α^−/− cardiomyocytes

(A) and (C) Northern blotting analysis of RNAs prepared from wt and p38α^−/− cardiomyocytes. The membranes were hybridized with COL1A1, COL3A1 and GAPDH probes as indicated. (B) PhosphorImager quantification of COL1A1 and COL3A1 transcript levels (normalized to GAPDH values) in the indicated cell lines. The results were compiled from six or two independent experiments for COL1A1 and COL3A1 respectively.

Figure 3. COL1A1 up-regulation in p38α^−/− cardiomyocytes is reversed by p38α expression

(A) Cardiomyocytes were infected with a retrovirus expressing wt p38α or the kinase-dead mutant p38α-D/A. RNA was prepared from the infected cells and the COL1A1 and GAPDH transcript levels were determined by Northern blotting. PhosphorImager quantification (normalized to GAPDH) from four independent experiments is shown. (B) COL1 protein expression was analysed in the same cell lines by immunoblotting. The filter was re-probed with antibodies to MKK6, the expression of which is known to be negatively regulated by p38α activity, and tubulin as a loading control. The results are representative of three independent experiments.

Figure 4. Increased basal ERK activity in the p38α^−/− cardiomyocytes and super-induction of COL1A1 upon TNFα treatment

(A) Cell lysates were prepared from the indicated cardiomyocyte lines and analysed by immunoblotting. Antibodies that specifically recognize the phosphorylated and active forms of ERK1/2 were used. The total amount of p38α or ERK1/2 was also determined and the filter was re-probed with a tubulin antibody as the loading control. (B) Cardiomyocytes were starved overnight (0.5% foetal calf serum) and then treated with TNFα (20 ng/ml) for 6 h. Total RNA was prepared and analysed by Northern blotting. The Phosphorimager quantification data (normalized to GAPDH values) are shown at the bottom of the blots. The data are compiled from two independent experiments.

Figure 5. COL1A1 gene expression is regulated by p38α at the transcriptional level

(A) Proliferating cardiomyocytes were used for the run-on analysis with 5 μg of the indicated plasmids. The signals were visualized by autoradiography (upper panel) and quantified by PhosphorImager (arbitrary units). The data were normalized to GAPDH values and plotted at the bottom of the blots. (B) Cardiomyocytes were treated with actinomycin D (Act; 2.5 μg/ml) and RNA was prepared, at the indicated times, and analysed by Northern blotting (upper panel). The amount of COL1A1 mRNA remaining after the treatment (normalized to GAPDH values) is shown at the bottom of the blots.

Figure 6. Regulation of C/EBPβ and TEF-1 DNA binding activity by p38α

(A) C/EBP DNA binding activity was determined by band-shift analysis using nuclear extracts of the indicated cardiomyocyte lines. The specificity of the binding was determined by competition with an excess (100×) of the unlabelled oligonucleotide (C/EBP wt) or a mutated version (C/EBP μ). The C/EBPs and the non-specific (NS) complexes are indicated. Identification of C/EBP proteins present in the complexes was performed by super-shift analysis using the same nuclear extracts. The C/EBP proteins in the complex were identified by the addition of specific antibodies (1 μg) against C/EBPα (lanes 5, 11, 17 and 23) or C/EBPβ (lanes 6, 12, 18 and 24). The anti-C/EBPβ antibody recognizes both p35 and p20 proteins. An equal amount of purified IgG was used as a control (lanes 4, 10, 16, and 22). The super-shifted complexes are indicated by an asterisk. The results are representative of four independent experiments. (B) The nuclear levels of the p35 and p20 C/EBPβ proteins in the same nuclear extracts as used for EMSA were determined by Western blotting. (C) TEF-1 DNA binding activity was determined by EMSA and the specificity of the binding was established by competition with an unlabelled oligonucleotide (lanes 6–9). The participation of the TEF-1 protein in the M-CAT DNA binding complexes was verified by super-shift assay using an anti-TEF-1 specific antibody (lanes 14–17). An equal amount of purified IgG was used as a control (lanes 10–13). The TEF-1 EMSA assay was performed twice. (D) The same nuclear extracts used for EMSA were also analysed by Western blotting with an anti-TEF-1 antibody (left panel). The histogram represents the normalization of TEF-1 DNA binding activity, as determined by EMSA and super-shift assays, to the amount of nuclear TEF-1 detected by Western blotting. Quantifications were performed using the Image J program. AU, arbitrary units.

Figure 7. Differential loading of C/EBPβ and TEF-1 on to the COL1A1 promoter in wt and p38α^−/− cardiomyocytes

(A) Schematic representation of the COL1A1 promoter indicating the positions of transcription factor binding sites and the PCR primers (a and b). CBF, CCAAT-binding factor. (B) ChIP assays were performed using whole lysates prepared from 40×10⁶ cells: wt (lanes 1, 2, 8 and 11), p38α^−/− infected with MSCV-p38α (lanes 3, 9 and 12) or p38α^−/− infected with the MSCV vector alone (lanes 4, 10 and 13). The protein extracts were incubated with the anti-C/EBPβ antibody (lanes 1, 3 and 4), anti-TEF-1 antibody (lanes 8–10) or with the same amount of IgG, as a control (lanes 2, 11–13). DNA purified from one-tenth of total lysates was also used as a control (Input; lanes 5–7 and 14–16). The ChIP assay was repeated twice.