This experimental study showed that AF is accompanied by spatial remodeling of gap junctions that might induce changes in the biophysical properties of the tissue.

Antiarrhythmic peptides enhance gap junction current in pairs of cardiomyocytes and coupling in cardiac tissue. To elucidate the underlying mechanisms, we investigated the effects of the antiarrhythmic peptide AAP10 (GAG-4Hyp-PY-CONH2) on pairs of adult guinea pig ventricular cardiomyocytes and pairs of HeLa cells transfected with rat cardiac connexin 43 (Cx43). By using a double-cell voltage-clamp technique in pairs of cardiomyocytes, we found that under control conditions the gap junction conductance (gj) steadily decreased with time (by -0.292 +/- 0.130 nS/min). Use of 50 nmol/L AAP10 reversed this rundown and increased gj (by +0.290 +/- 0.231 nS/min, Pa). In HeLa-Cx43 cells, AAP10 exerted the same electrophysiological effect. In these cells, AAP10 activated PKC (determined by using ELISA) in CGP54345-sensitive manner and significantly enhanced incorporation of 32P into Cx43 with dependence on PKC. If G-protein coupling was inhibited with 1 mM GDP-BS, we found the effects of AAP10 on 32P incorporation were also completely abolished. Next, we performed a radioligand binding study with 14C-AAP10 as radioligand and AAPnat as competitor. We found saturable binding of 14C-AAP10 to cardiac membrane preparations, which could be displaced with AAPnat. The Kd of AAP10 was 0.88 nmol/L. We conclude that 1) AAP10 increases gj both in adult cardiomyocytes and in transfected HeLa-Cx43 cells, 2) AAP10 exerts its effect via enhanced PKC-dependent phosphorylation of Cx43, 3) AAP10 activates PKCa, and 4) a membrane receptor exists for antiarrhythmic peptides in cardiomyocytes.

Gap junction channels provide the basis for the electrical syncytial properties of the heart as a communicating electrical network. Cardiac gap junction channels are predominantly composed of connexin 40 or connexin 43. The conductance of these channels (g(j)) can be regulated pharmacologically: substances which activate protein kinase C, protein kinase A or protein kinase G may alter Cx43 gap junction conductance. However, for PKC, this seems to be subtype specific. Thus, antiarrhythmic peptides can enhance g(j) via activation of PKCepsilon, while FGF-2 reduces g(j) via PKCepsilon. Lipophilic drugs can uncouple the channels. Besides an acute regulation of g(j), the expression of the cardiac connexins can also be regulated. A decrease in Cx43 with a concomitant increase in Cx40 has been found in end-stage failing hearts, while in renovascular hypertension, an increase in Cx43 has been described. Mediators like endothelin-1, angiotensin-II, TGF-beta, VEGF, and cAMP have been shown to increase Cx43. Interestingly, endothelin-1 and angiotensin-II increased Cx43 but did not affect Cx40 expression. In contrast, in humans suffering from atrial fibrillation (AF), the content in Cx40 can be enhanced while Cx43 was unaltered, although in several other studies, other changes of the cardiac connexins were found, which might be related to the type of AF. Regarding the role of calcium, the content in both Cx40 and Cx43 was decreased in cultured neonatal rat cardiomyocytes after 24 h administration of 100 nM verapamil. Thus, gap junctional channels can be affected pharmacologically either acutely by modulating gap junction conductance or chronically by altering gap junction protein expression. Interestingly, it appears that the expression of Cx43 and Cx40 can be differentially regulated.

Gap junctions were assayed during re-differentiation of adult rat cardiomyocytes in long-term culture to gain insight into the processes of remodeling. Double immunostaining allowed the localization of connexins Cx40, Cx43, and Cx45 between myocytes and demonstrated co-expression and co-localization in individual cells and gap junction plaques, respectively. Immunoblots showed differential time-dependent changes in connexin expression and phosphorylation. The total amount of connexins and the ratio of phosphorylated/non-phosphorylated isoforms gradually increased during the re-establishment of intercellular communication. Dual voltage-clamp studies showed the involvement of several types of gap junction channels. Multichannel currents yielded diverse spectra of g(j,inst)=f( V(j)) and g(j,ss)=f( V(j)) relationships ( g(j,inst): instantaneous gap junction conductance; g(j,ss): conductance at steady state; V(j): transjunctional voltage), indicative of homotypic and heterotypic channels. Single-channel currents revealed two prominent conductances reflecting gamma(j,main) and gamma(j,residual). The histograms of gamma(j,main) showed four discrete peaks (41-44, 59-61, 70-76, and 100-107 pS) attributable to a combination of Cx45-Cx45, Cx40-Cx45 and Cx43-Cx45 channels (1st peak), Cx43-Cx43 and Cx40-Cx43 channels (2nd peak), Cx43-Cx43 channels (3rd peak) and Cx40-Cx40 and Cx40-Cx43 channels (4th peak). However, the presence of heteromeric channels cannot be excluded. The data are consistent with an up-regulation of Cx45 and Cx43 during re-differentiation.