Dissociation of cardiogenic and postnatal myocardial activities of GATA4

Abstract

Transcription factor GATA4 is a critical regulator of the embryonic and postnatal heart, but the mechanisms and cofactors required for its diverse functions are not fully understood. Here, we show that whereas the N-terminal domain of GATA4 is required for inducing cardiogenesis and for promoting postnatal cardiomyocyte survival, distinct residues and domains therein are necessary to mediate these effects. Cardiogenic activity of GATA4 requires a 24-amino-acid (aa) region (aa 129 to 152) which is needed for transcriptional synergy and physical interaction with BAF60c. The same region is not essential for induction of endoderm or blood cell markers by GATA4, suggesting that it acts as a cell-type-specific transcriptional activation domain. On the other hand, a serine residue at position 105, which is a known target for mitogen-activated protein kinase (MAPK) phosphorylation, is necessary for GATA4-dependent cardiac myocyte survival and hypertrophy but is entirely dispensable for GATA4-induced cardiogenesis. We find that S105 is differentially required for transcriptional synergy between GATA4 and serum response factor (SRF) but not other cardiac cofactors such as TBX5 and NKX2.5. The findings provide new insight into GATA4 mechanisms of action and suggest that distinct regulatory pathways regulate activities of GATA4 in embryonic development and postnatal hearts.

Fig 1

Rat GATA4 induces cardiac tissue in Xenopus animal caps independently of endogenous xGATA4. (A) Animal cap explants injected with 500 pg of indicated mRNAs have been analyzed for expression of cardiac (MLC2 and alpha-myosin heavy chain [MHCα]), endothelial (MSR), blood (globin), and liver (For1) markers at stage 34. Hex marks both liver and endothelia (34). rGATA4 and xGATA4 induce all markers tested, whereas xGATA1 induces only blood and endothelia. (B) rGATA4 induces endogenous GATA4 expression at stage 10, approximately 3 h after activation of zygotic gene expression. (C) Morpholino oligonucleotides against xGATA4 (4MO) do not prevent induction of cardiac markers by rGATA4. ODC, ornithine decarboxylase, loading control; AC, control animal caps; embryo, cDNA from sibling control embryos; −RT, no reverse transcriptase control for the embryo sample.

Fig 2

The rGATA4 domain of aa 129 to 153 is required for cardiogenesis. (A) Schematic representation of GATA4 protein showing the position of mutations/deletions in green. NLS, nuclear localization signal. (B and D) Explants injected with indicated rGATA4 variants were tested for expression of alpha-myosin heavy chain (MHCα) at stage 34 (top) and by Western blotting for the levels of rGATA4 protein at stage 10 (below). Panel B (bottom) shows nuclear localization as determined by whole-mount immunohistochemistry with anti-HA–HRP at stages 6 to 8. (C) Noncardiogenic rGATA4(201–440) protein is stable throughout the cultivation period. Samples collected at indicated stages were analyzed for rGATA4 and ERK by Western blotting. rGATA4 protein was detected by anti-HA (B and C) or anti-human GATA4 antibodies (D). (E) Stage 34 embryos injected with indicated mRNAs in anterior ectodermal precursor blastomeres were analyzed for MLC2 expression, highlighting the heart (h) and ectopic MLC2 expression (arrow). More than 50 embryos were analyzed for each construct, with 15 to 20% showing ectopic expression in the case of cardiogenic constructs.

Fig 3

Noncardiogenic mutants GATA4(153–440) and GATA4(201–440) retain transcriptional activity. (A) Transcriptional activity of various GATA4 (G4) N-terminal deletion mutants. Cotransfections in NIH 3T3 cells were carried out on 1.5 μg of the (GATA)₃-Luc reporter, and 25, 50, and 100 ng of each GATA4 construct. (B) Western blot showing protein levels of expression of GATA4 constructs (top) and gel shift assay showing the ability of GATA4 mutants to bind DNA (bottom). The probe used is the GATA element at position −120 on the Nppa (ANF gene) promoter. (C) Removal of the first 153 or 201 amino acids from rGATA4 attenuates, but does not abolish, the ability to activate a GATA site-driven reporter. Embryos were injected with a firefly luciferase reporter under the control of two GATA sites, Renilla luciferase plasmid driven by the thymidine kinase (TK) promoter and indicated mRNAs. Control, DNA alone (no RNA). Animal caps were collected 3 h after excision. GATA activity is firefly luciferase activity normalized for Renilla luciferase activity, expressed as fold activation over the control sample. A representative of three repeated experiments is shown. (D) Noncardiogenic mutants consisting of GATA4 aa 153 to 440 and 201 to 440 can induce the endogenous gata4 gene both early (stage 10) and late (stage 34), as revealed by RT-PCR analysis. (E and F) rGATA4 mutants consisting of GATA4 aa 153 to 440 and 201 to 440 cannot induce cardiomyocyte-specific markers MLC2, MHCα, and cardiac troponin I (cTnI) but retain the capacity to induce globin, a marker of blood (albeit at a lower level) as well as endodermal markers Sox17 at stage 10 and endodermin at stage 34. RT-PCR analysis was performed on stage 34 animal caps. AC, control animal caps.

Fig 4

S105 phosphorylation is not required for cardiogenesis by rGATA4. (A) Transcriptional activity of GATA4 S105A and S105E mutants. Cotransfections in NIH 3T3 cells were carried out on 1.5 μg of the (GATA)₃-Luc reporter, and 25, 50, and 100 ng of each of the GATA4 constructs. (B) Western blot showing nuclear GATA4 protein levels in transfected cells (top) and gel shift assay showing the ability of GATA4 mutants to bind DNA (bottom). The probe used is the GATA element at position −120 on the Nppa promoter. (C) S105A and S105E rGATA4 variants induce at state 34 markers of cardiomyocyte differentiation MLC2 and MHCα. S105E induces a greater amount of cardiac markers than the WT or S105A rGATA (top). A Western blot shows levels of rGATA4 variants at stage 10, detected with anti-HA antibody (middle). Nuclear localization as determined by whole-mount immunohistochemistry with anti-HA–HRP is shown at stages 6 to 8 (bottom). (D) Indicated rGATA4 variants induce ectopic MLC2 expression in stage 34 embryos. More than 50 embryos were examined for each construct, and 15 to 20% showed ectopic expression. AC, control animal caps.

Fig 5

GATA4 regions required for functional cooperativity with cardiac cofactors. (A) Cotransfections in NIH 3T3 cells using 1.5 μg of the Nppa promoter, 15 ng of each of the GATA4 constructs, and 50 ng each of NKX2.5, TBX5, and MEF2A. The data are the mean ± SEM of two to three experiments carried out in duplicate. (B) Histograms showing the fold synergy for each GATA4 mutant. Synergy was calculated from the data in panel A. (C) Cotransfections in NIH 3T3 using 1.5 μg of the Nppa promoter, 15 ng of each of the GATA4 constructs, and 250 ng of BAF60c. The data are the mean ± SEM of three experiments carried out in duplicate. (D) Immunoprecipitation (IP) of endogenous BAF60c with rabbit anti-BAF60c (kind gift from Puri) followed by detection of transfected HA-tagged GATA4 mutants. IB, immunoblotting.

Fig 6

GATA4 N-terminal domain is essential for myofibrillar reorganization and myocyte survival. (A) Immunofluorescence in cardiomyocyte infected with the indicated adenoviral constructs. The sarcomeric alpha-actinin was labeled green, and the exogenous GATA4 proteins detected with the anti-HA antibody were labeled in red. Notice the increase in size of myocytes expressing the WT GATA4 (G4) but not the GATA4(201–440) mutant with an N-terminal deletion or the S105A GATA4 mutant. (B) Western blot confirming similar nuclear overexpression of indicated HA-tagged GATA4 proteins in cardiomyocytes infected with the corresponding adenoviral vectors. (C) Relative cell surface area was measured across 10 fields (magnification, ×40) in three separate experiments. *, P < 0.05 versus LacZ, GATA4-S105A, and GATA4(201–440). (D) Quantification of TUNEL assays in cardiomyocytes infected with the indicated adenoviruses in the presence or absence of 300 nM Dox. Notice the apoptotic effect of both GATA4-S105A and GATA4(201–440) mutants. *, P < 0.05 versus LacZ; **, P < 0.01 versus LacZ; #, P < 0.05 versus LacZ+Dox. (E) The left panel shows a Western blot of Bcl-X proteins in nuclear extracts of cardiomyocytes expressing the indicated constructs. At right, 1.5 μg of Bcl-X–luciferase reporter was cotransfected with 25 and 50 ng of each GATA4 construct. The data are the mean ± SEM of two to three experiments carried out in duplicate.

Fig 7

GATA4-S105 is required for functional cooperativity with SRF. (A) Cotransfections in NIH 3T3 using 1.5 μg of the Nppa promoter, 15 ng of each of the GATA4 constructs, and 50 ng each of NKX2.5, TBX5, MEF2A, and SRF. The data are the mean ± SEM of three experiments carried out in duplicate. (B) Histograms showing the fold synergy for each GATA4 mutant. Synergy was calculated from the data in panel A.