Convergent proliferative response and divergent morphogenic pathways induced by epicardial and endocardial signaling in fetal heart development

Abstract

Heart development requires cardiomyocyte proliferation, coupled with morphogenic differentiation of the inner trabecular myocardium and the outer compact zone myocardium. In mouse embryos lacking the retinoic acid receptor RXRalpha, proliferation and morphogenesis of the compact zone fails. We demonstrated previously that epicardial cells, in response to retinoic acid, secrete an activity that promotes cell proliferation. In this study, we have investigated downstream signaling pathways that are elicited in response to this factor. We find that cells treated in culture activate PI3 kinase and Erk pathways, and that these are required for a proliferative response. In vivo, phosphorylation of Akt and GSK3beta (PI3K pathway) and of Erk1/2 and p90rsk (Erk pathway) is substantially reduced in RXRalpha-deficient heart tissue. Neuregulin, a mitogen secreted from the endocardium which promotes cardiomyocyte proliferation and trabecular differentiation, also activates proliferation via PI3K and Erk pathways. However, the epicardial factor is not neuregulin, and does not function via the neuregulin receptor. Gene markers known to be selectively expressed in trabecular or compact myocardium in vivo are differentially activated in cell culture by treatment with neuregulin or epicardial factor, and are misexpressed in RXRalpha(-/-) heart tissue. We therefore conclude that epicardial and endocardial signals converge on common proliferative components, but diverge in downstream pathways that lead to compact vs. trabecular morphogenic differentiation.

PDGF‐A as an epicardial mitogen during heart development

“…As assayed in primary cardiomyocytes and in NIH3T3 cells, rCM induced the phosphorylation of the signaling intermediates Akt and Erk, and activation of the PI3 kinase and MAP kinase pathways was required for a mitogenic response. We also noted the rapid tyrosine phosphorylation of an approximately 180-kDa membrane-associated protein in NIH3T3 cells in response to rCM (Kang and Sucov, 2005), which is consistent with this protein being a member of the receptor tyrosine kinase (RTK) family; such receptors tyrosine autophosphorylate themselves in response to cognate ligands as the first step in signal transduction. We took two parallel approaches to identify this putative receptor.…”

“…The same response was obtained in NIH3T3 cells and other mammalian tissue cell lines that are more amenable to experimental analysis (Chen et al, 2002;Kang and Sucov, 2005). As assayed in primary cardiomyocytes and in NIH3T3 cells, rCM induced the phosphorylation of the signaling intermediates Akt and Erk, and activation of the PI3 kinase and MAP kinase pathways was required for a mitogenic response.…”

“…The Wnt-responsive TOP-Flash luciferase reporter was not activated in transfected cells in response to rCM (data not shown), suggesting that secreted Wnts acting through the canonical ␤-catenin pathway are not present in rCM. EMC cells also do not express neuregulin (Kang and Sucov, 2005). The unknown factor is unlikely to be a ligand for a receptor tyrosine kinase, as our survey of the RTK family in Hep3B cells did not reveal any receptor other than PDGFR␣ as being phosphorylated in response to rCM (although, we did not verify that all tested receptors are expressed in Hep3B cells).…”

“…This may be sufficient to support an in vivo function, although in the in vitro mitogenic assay that we use, we suspect that this level of expression may not be adequate to yield a significant proliferative response. Thus, the mitogenic activity we have previously observed in primary embryonic cardiomyocytes in response to EMC cell factors (Chen et al, 2002;Kang and Sucov, 2005) may primarily be a response to other factors made by EMC cells, and not to PDGF-A.…”

“…For proliferative responses, serum-starved cells were treated for 24 hr and assayed by thymidine incorporation as previously described (Chen et al, 2002). For analysis of signaling components, after 10 min, cells were washed, lysed, and extracts processed for Western blotting as previously described (Kang and Sucov, 2005). For immunoprecipitation, cell lysates (500 g protein) were incubated with 2 g of antibody overnight at 4°C, then incubated with protein G agarose (20 l of a 50% slurry; Upstate) for 2 hr; after washing 3ϫ with lysis buffer, the suspensions were spun down for 2 min and the precipitates used for Western blot analysis.…”