It has been classically assumed that the ventricle of the primitive vertebrate heart is
composed of spongy myocardium, supplied exclusively by oxygen-poor, luminal blood.
This idea is on two facts: (1) extant agnathans have a spongy ventricular myocardium,
and (2) in avian and mammalian embryos, the formation of trabeculated myocardium
precedes the appearance of compact myocardium. Recently, it has been proposed that,
like elasmobranchs, the early gnathostomes possess a fully vascularised ventricle
composed of mixed myocardium. We tested this idea by studying the structure and
vascularisation of the ventricular myocardium in four holocephalan species of the
families Chimaeridae and Rhinochimaeridae. Chimaera monstrosa, Hidrolagus affinis
and Harriotta raleighana have a spongy myocardium covered by a thin layer of cardiac
muscle. In H. raleighana, the compacta is reduced to an extremely fine rim. In all three
species there is a well-developed coronary artery system consisting of subepicardial
vessels which give off branches that penetrate the myocardial trabeculae.
Rhinochimaera atlantica has no compacta and its ventricular coronary artery system is
reduced to subepicardial vessels that do not enter the spongy layer. This report is the
first to show that in wild living vertebrates, a coronary artery system supplying the whole
myocardium exists in the absence of a well-developed compacta, which supports
experimental work that shows that myocardial cell proliferation and coronary vascular
growth rely on genetically separated programs. We conclude that the mixed ventricular
myocardium is primitive for chondrichthyans, and that the lack of compacta in some
holocephalans is a derived character. Moreover our results support the hypotheses that
the mixed myocardium is the primitive condition in gnathostomes, and that the absence
of a compacta in different actinopterygian taxa is the result of its repeated loss during
evolution.