Comparison of the T-tubule system in adult rat ventricular and atrial myocytes, and its role in excitation–contraction coupling and inotropic stimulation

Abstract

Narrow, tubular, inward projections of the sarcolemma (‘T-tubules’) are an established feature of adult mammalian ventricular myocytes that enables them to generate the whole-cell Ca 2+ transients and produce coordinated contraction. Loss of T-tubules can occur during ageing and under pathological conditions, leading to altered cardiac excitation–contraction coupling. In contrast to adult ventricular cells, atrial myocytes do not generally express an extensive T-tubule system at any stage of development, and therefore rely on Ca 2+ channels around their periphery for the induction of Ca 2+ signalling and excitation–contraction coupling. Consequently, the characteristics of systolic Ca 2+ signals in adult ventricular and atrial myocytes are temporally and spatially distinct. However, although atrial myocytes do not have the same regularly spaced convoluted T-tubule structures as adult ventricular cells, it has been suggested that a proportion of adult atrial cells have a more rudimentary tubule system. We examined the structure and function of these atrial tubules, and explored their impact on the initiation and recovery of Ca 2+ signalling in electrically paced myocytes. The atrial responses were compared to those in adult ventricular cells that had intact T-tubules, or that had been chemically detubulated. We found that tubular structures were present in a significant minority of adult atrial myocytes, and were unlike the T-tubules in adult ventricular cells. In those cells where they were present, the atrial tubules significantly altered the on-set, amplitude, homogeneity and recovery of Ca 2+ transients. The properties of adult atrial myocyte Ca 2+ signals were different from those in adult ventricular cells, whether intact or detubulated. Excitation–contraction coupling in detubulated adult ventricular myocytes, therefore, does not approximate to atrial signalling, even though Ca 2+ signals are initiated in the periphery of the cells in both of these situations. Furthermore, inotropic responses to endothelin-1 were entirely dependent on T-tubules in adult ventricular myocytes, but not in atrial cells. Our data reveal that that the T-tubules in atrial cells impart significant functional properties, but loss of these tubular membranes does not affect Ca 2+ signalling as dramatically as detubulation in ventricular myocytes.