Bradley W. Doble

The carboxy-tail of connexin-43 localizes to the nucleus and inhibits cell growth.

Gap junctions are plasma membrane intercellular communication channels that in addition to ensuring electrical coupling and coordinated mechanical activity, can act as growth suppressors. To define the role of a non-channel forming domain of connexin-43 (Cx43), the main constituent of cardiomyocyte gap junctions, on growth regulation, we expressed its C-terminal portion (CT-Cx43) in cardiomyocytes and HeLa cells. In addition to broad cytoplasmic localization, CT-Cx43 was also localized to the nucleus of both cell types, detected by immunofluorescence as well as immunoblotting of subcellular fractions. Furthermore, stable expression of CT-Cx43 in HeLa cells induced a significant decrease in proliferation. It is therefore suggested that plasma membrane localization and formation of channels are not required for growth inhibition by Cx43, and that nuclear localization of CT-Cx43 may exert effects on gene expression and growth.

Fibroblast Growth Factor-2 Decreases Metabolic Coupling and Stimulates Phosphorylation as Well as Masking of Connexin43 Epitopes in Cardiac Myocytes

Cardiac gap junction (GJ) channels, composed of connexins, allow electrical and metabolic couplings between cardiomyocytes, properties important for coordinated action of the heart as well as tissue homeostasis and control of growth and differentiation. Fibroblast growth factor-2 (FGF-2) is an endogenous growth-promoting protein, believed to participate in the short- and long-term responses of the heart to injury. We have examined short-term effects of FGF-2 on cardiac myocyte GJ-mediated metabolic coupling, using cultures of neonatal rat cardiomyocytes. FGF-2 decreased coupling between cardiomyocytes assessed by scrape dye loading as well as microinjection and dye transfer within 30 minutes of administration. Genistein blocked the effects of FGF-2. To determine the mechanism, we next assessed the effect of FGF-2 on expression, distribution, and phosphorylation of connexin43 (Cx43), which is a major cardiomyocyte connexin. FGF-2 did not affect Cx43 mRNA or protein accumulation and synthesis, and it did not change Cx43 localization at sites of intercellular contact as assessed by immunostaining with a polyclonal anti-Cx43 antibody raised against a synthetic peptide containing residues 346 to 363 of Cx43. FGF-2, however, decreased staining intensity at sites of intermyocyte contact when a monoclonal anti-Cx43 antibody was used, suggesting a localized masking of epitope(s) recognized by the monoclonal but not the polyclonal antibody. These epitopes appear to reside within residues 261 to 270 of Cx43, as indicated by full quenching of monoclonal antibody staining with synthetic peptides. In addition, FGF-2 induced a more than twofold increase in Cx43 phosphorylation. Phosphoamino acid analysis indicated increased phosphorylation of Cx43 on serine residues. Although tyrosine phosphorylation of Cx43 was not detected in either treated or control cells, a fraction of Cx43 was immunoprecipitated with anti-phosphotyrosine-specific antibodies in FGF-2-treated myocytes, suggesting interaction (and hence coprecipitation) with phosphotyrosine-containing protein(s). In conclusion, we have identified Cx43 and intercellular communication as targets of FGF-2-triggered and tyrosine phosphorylation-dependent signal transduction in cardiac myocytes. It is suggested that phosphorylation of Cx43 on serine induced by FGF-2 contributes to decreased metabolic coupling between cardiomyocytes.

Basic fibroblast growth factor stimulates connexin-43 expression and intercellular communication of cardiac fibroblasts

Gap junctions (GJ) are membrane specializations responsible for intercellular communication and for ensuring electrical and/or metabolic coupling between cells. They are composed of connexins, a family of related proteins. Connexin-43 (Cx43) is a major connexin of the rat heart, expressed by myocytes as well as non-muscle cells. In this communication we have examined expression of Cx43 by cardiac fibroblasts and regulation of its expression by an endogenous mitogen, basic fibroblast growth factor (bFGF). Recombinant human bFGF, administered to cultured cells which had been maintained in 0.5% serum for 48 h, induced dose-dependent and statistically significant increases in Cx43 mRNA as well as protein accumulation, at 6 h after addition. Intercellular communication was also increased at 6 h but not 30 min after bFGF treatment, as assessed using a scrape-loading protocol. It is concluded that the bFGF-induced stimulation of Cx43 expression caused increased coupling between cardiac fibroblasts. This would be of importance in injured myocardium, the increased bFGF content of which might stimulate electrical coupling involving fibroblasts of the scar tissue.

Regulation of basic fibroblast growth factor (bFGF) and FGF receptors in the heart.

Cardiac myocytes lose their ability for a prompt hyperplastic and hence regenerative response soon after birth, responding to increased physiological or pathological functional demand by hypertrophy. As a consequence, irreversibly damaged myocardium as occurs after myocardial infarction becomes replaced by scar tissue, the remaining myocytes adapt to the increased workload by hypertrophy, and beyond a certain potential for compensatory responses, cardiac failure ensues. It is generally accepted that interventions aimed at reducing scar size, such as stimulation of myocardial regeneration, would improve cardiac prognosis. Although very little is known about the molecular triggers of hyperplastic and/or hypertrophic growth in vivo, signaling polypeptides such as growth factors have been strongly implicated in this context. A variety of growth factors have been identified in the myocardium by us and others.‘*2 We have focused our attention on basic fibroblast growth factor (bFGF), since it is a potent stimulator of DNA synthesis not only in immature but also adult cardiomyocytes in c ~ l t u r e . ~ ~ Here we will provide a brief overview of current information about bFGF and its receptors as well as our own data.

Protein kinase C-epsilon mediates phorbol ester-induced phosphorylation of connexin-43.

We have used adenoviral vectors to express dominant negative variants of protein kinase C epsilon (PKCϵ) or mitogen kinase kinase 1 (MKK1) to investigate their involvement in phorbol ester-induced connexin-43 (Cx43) phosphorylation in cardiomyocytes. Stimulation of cardiomyocytes with phorbol 12-myristate 13-acetate (PMA) increased the fraction of the slower migrating (≥45 kDa) and more extensively phosphorylated Cx43 species. Expression of dominant negative MKKI did not prevent the effect of PMA on Cx43 phosphorylation. Selective inhibition of PKCϵ significantly decreased baseline levels of Cx43 phosphorylation and the PMA-induced accumulation of ≥45 kDa Cx43. Thus, production of the more extensively phosphorylated species of Cx43 in cardiomyocytes by PMA requires activation of PKCϵ.

CUG‐initiated FGF‐2 induces chromatin compaction in cultured cardiac myocytes and in vitro*

Fibroblast growth factor‐2 (FGF‐2) is a mitogen found in CUG‐initiated 21–25 kDa (“hi”) or AUG‐initiated 16–18 kDa (“lo”) forms. Previously we demonstrated that “hi”—but not “lo”—FGF‐2 caused a distinct nuclear phenotype characterized by apparently condensed chromatin present as separate clumps in the nucleus of cardiac myocytes. In this manuscript we investigated whether these effects were related to apoptosis or mitosis and whether they reflected a direct effect of “hi” FGF‐2 on chromatin. Myocytes overexpressing “hi” FGF‐2 and presenting the clumped chromatin phenotype: (i) were not labeled above background with antibodies to phosphorylated histones H1 and H3 used as indicators of mitotic chromatin condensation; (ii) did not stain positive for TUNEL; (iii) their nuclear lamina, visualized by anti‐laminB immunofluorescence, appeared intact; (iv) neither caspase inhibitors, nor Bcl‐2 or “lo” FGF‐2 overexpression prevented the manifestation of the compacted nuclear phenotype. Purified recombinant “hi” FGF‐2 was more potent than “lo” FGF‐2 in promoting the condensation/aggregation of chick erythrocyte chromatin partially reconstituted with histone H1 in vitro. We conclude that the DNA phenotype induced by “hi” FGF‐2 in cardiac myocytes likely reflects a direct effect on chromatin structure that does not require the engagement of mitosis or apoptosis. By affecting chromatin compaction “hi” FGF‐2 may contribute to the regulation of gene expression.

PKC-Dependent Phosphorylation May Regulate the Ability of Connexin43 to Inhibit DNA Synthesis

Phosphorylation affects several biological functions of connexin43 (Cx43), although its role on Cx43-mediated inhibition of DNA synthesis is not known. Previous studies showed increased Cx43 phosphorylation on serine in response to growth factor stimulation of cardiomyocytes, mediated by protein kinase C-epsilon (PKCε). Here we report that activation of PKCε is also necessary for stimulation of cardiomyocyte DNA synthesis and mitosis. We have investigated the participation of specific serine residues that are putative PKC targets in producing phosphorylated Cx43 species and also in regulating DNA synthesis in cardiomyocytes. Interference with the PKC signaling system and/or the phosphorylation of specific amino-acids of Cx43 may allow regulation of the mitogenic response.

Effects of Ischemia on Cardiomyocyte Connexin-43 Distribution and Phosphorylation Studied in in vivo and in vitro Models

The gap junction protein connexin-43 (Cx43) exists mainly in the phosphory-lated state in the normal heart. We have investigated short-term effects of ischemia on cardiac Cx43 phosphorylation and distribution, in four models: global ischemia of the ex vivo perfused heart, left ventricular ischemia induced by irreversible coronary ligation in vivo, simulated ischemia on isolated adult myocyte pellets, and neonatal cardiomyocytes incubated in a hypoxia chamber. Antibody AB. 13-800 that recognizes specifically the 41kDa nonphospho-rylated form of cardiac Cx43 labeled intercalated discs (ICDs) in myocytes from perfused rat hearts subjected to 30min global ischemia; also in myocytes at the infarct border 6 hours post-infarction. Ischemia induced a sharp increase in the 41 kDa Cx43 from perfused hearts, isolated adult myocyte pellets and neonatal myocyte cultures subjected to hypoxia. The protein phosphatase type 1/2A inhibitors okadaic acid and calyculin A, tested in the in vitro models, decreased ischemia-induced Cx43 dephosphorylation. The 41 kDa Cx43 was present in both Triton-soluble as well as Triton-insoluble (enriched in ICDs) cardiac membrane fractions, assessed by western blotting. We conclude that ischemia causes dephosphorylation of car-diomyocyte Cx43 in vivo as well as in vitro, and that this phenomenon occurs irrespectively of stage (neonatal or adult) or presence of cell contact. Cx43 dephosphorylation occurs at ICDs, is mediated at least in part by PPl/2-type phosphatases, and would be expected to affect GJ function and contribute to ischemia-induced conductance and contractile changes.