Temporal changes in expression of connexin 43 after load-induced hypertrophy in vitro

Abstract

Upon remodeling of the ventricle after a provoking stimulus, such as hypertension, connections between adjacent myocytes may need to be "reformatted" to preserve a synchronization of excitation of the remodeling heart. In the mammalian heart, the protein connexin forms the gap junctions that allow electrical and chemical signaling communication between neighboring cells. We aim to elucidate whether mechanical load, in isolation, potentially changes the expression of connexin 43 (Cx43), the major isoform of the connexin family in the ventricle, and its phosphorylation. Cx43 expression levels and contractile function of multicellular rabbit cardiac preparations were assessed in a newly developed in vitro system that allows for the study of the transition of healthy multicellular rabbit myocardium to hypertrophied myocardium. We found that in mechanically loaded cardiac trabeculae, Cx43 levels remained stable for about 12 h and then rapidly declined. Phosphorylation at Ser368 declined much faster, being almost absent after 2 h of high-load conditions. No-load conditions did not affect Cx43 levels, nor did phosphorylation at Ser368. The downregulation of Cx43 under mechanical load did not correspond with the contractile changes that were observed. Furthermore, blocking paracrine activity of the muscle could only partially prevent the downregulation of Cx43. Additionally, no effect of mechanical loading on the expression of N-cadherin and zonula occludens-1 was observed, indicating a specificity of the connexin response. High mechanical load induced a rapid loss of Cx43 phosphorylation, followed by a decrease in Cx43 protein levels. Paracrine factors are partly responsible for the underlying mechanism of action, whereas no direct correlation to contractile ability was observed.

Fig. 1.

A: active developed force of cultured rabbit cardiac trabeculae (n = 8), contracting at 1 Hz, continuously for 48 h at high load. B: time to peak force. C: 90% of relaxation time.

Fig. 2.

A and B: immunoblot analysis demonstrates the expression of protein connexin 43 (Cx43) and relative amount of cardiac troponin-I (cTnI) under high-loading condition at 2, 6, 12, 24, and 48 h. Cont, control (a fresh muscle sample); No-load, a culture cardiac trabeculae contracting at 1 Hz for 48 h without pre-and afterload (i.e., slack). C and D: relative arbitrary unit of Cx43 vs. cTnI. C: each control and culture muscles were from the same heart (n = 10). D: control and culture samples were from the same and different hearts (n = 5). *P < 0.05 and **P < 0.01 vs. controls by Bonferroni's multiple comparisons after ANOVA.

Fig. 3.

Comparison the phosphorylation level of Cx43 (pCx43) at Ser368 using immunoblot analysis of high-loading samples at 2, 6, and 12 h (A) vs. no-load samples at the same point (B). cTnI was used to indicate the amount of protein loading because no change in the ratio of Cx43 to cTnI was detected in 12 h of culture shown in Fig. 2. Each set of controls and cultured muscles were from the same hearts.

Fig. 4.

Immunoblot analysis demonstrates the expression of N-cadherin (N-Cad; A) and zonula occludens-1 (ZO-1; B) vs. TnI. C: relative arbitrary unit of N-Cad vs. TnI (n = 8–10) and ZO-1 vs. TnI (n = 6 to 7). No significant difference among groups by ANOVA.

Fig. 5.

Active developed force over 24 h of high-loaded culture cardiac trabeculae in the absence (•, n = 9) and presence (○, n = 6–8) of selective pharmacological antagonists. PD-145065 (A) is a nonselective endothelin receptor antagonist, BQ-123 (B) is a specific endothelin A receptor antagonist, losartan (C) and telmisartan (D) are angiotensin II receptor antagonists, and SB-431542 (E) is a selective inhibitor of activin receptor-like kinase receptors.

Fig. 6.

A: expression of Cx43 vs. cTnI of high-loaded muscle at 24 h in the absence and presence of selective pharmacological antagonists by immunoblot analysis. B: relative arbitrary unit of the ratio of Cx43 to TnI in each group was compared to Cont. PD, PD-145065; BQ, BQ-123; SB, SB-431542; Los, losartan; Tel, telmisartan. *P < 0.05 vs. controls by Bonferroni's multiple comparisons after ANOVA.