Since formation of the coronary vasculature is strikingly similar in avian and mammalian species (Tomanek,2005), data obtained from quail and chicken reflect developmental patterns and mechanisms in man and other mammals. Experiments on avian hearts provide two distinct advantages over those using mammalian species, namely, accessibility and cost. The current study provides the first temporal-spatial documentation of coronary vascularization throughout the entire embryonic/fetal period.
Hirakow (1983) noted that the first definitive blood vessels in the developing human heart resembled blood islands, with “primitive erythroblasts.” Perez-Pomares et al. (1998) found that 20% of the cells within the expanding subepicardium in the embryonic hamster heart were VEGFR-2-positive, suggesting that they were endothelial or erythrocyte progenitors. Subsequently, CD45-positive hematopoietic precursors were shown to precede blood vessel formation in the quail heart (Kattan et al.,2004). Most recently, using retroviral tagging, we documented that erythrocytes in blood islands are derivatives of the proepicardium (Tomanek et al.,2006). Thus, the proepicardium is a source of not only all vascular components (Mikawa et al.,1992,1996), but also erythrocytes that associate with endothelial cells that form vascular channels. Our current study and previous work (Tomanek et al.,2002) also document a very strong expression of VEGF protein in epicardial and subepicardial cells. Thus, this ligand is available in high quantity at the site of cells that delaminate from the mesothelium of the epicardium and that express the VEGFR-2 receptor. The earliest tube formation (capillaries) occurs in the subepicardium, where there is the highest level of VEGF expression.
Vrancken Peeters et al. (1997) noted that the myocardial tubular network has nonluminized strands that connect to the sinusoids that are continuous with the ventricular lumen. They suggested that flow through the capillary bed prior to connections to the aorta is possible. Subsequently, Manner (2000) grafted quail proepicardium to chick hearts at E3 and incubated the chimeras until E10. He found small patches of quail-derived endocardium in communication with the coronary vasculature, which he described as “ventriculo-coronary communications.” Although the question of flow between the ventricular lumen and the coronary system is not entirely resolved, our electron micrographs of ink-injected beating hearts show that capillary profiles lack carbon particles prior to the establishment of the coronary arteries. Even when erythrocytes were present in the vascular tubes of E7 hearts, carbon particles were lacking. Thus, blood islands lacked connections with the ventricular lumen. As the compact region of the ventricle expands, tubulogenesis keeps pace, a finding consistent with our previous work on embryonic chicken hearts (Tomanek et al.,1999). In that study, we showed that experimentally accelerated myocardial growth is matched by enhanced tubulogenesis. The magnitude of growth of the compact component of the left ventricle during the embryonic period, as noted by wall thickness, is very similar in quail (the current study) and chicken (Sedmera et al.,2000).
Coronary Arterial System: Arteriogenesis
Since the first evidence that the coronary arteries are formed by ingrowth, rather than outgrowth, into the aorta (Bogers et al.,1989), this concept has been corroborated by a variety of approaches in several species (Waldo et al., 1990; Poelmann et al.,1993; Tomanek et al., 1996). Ando et al.(2004) showed that endothelial strands in the quail penetrated the aorta at E6 and E7 and fused by E8 to form the coronary stems. Our own findings agree with this temporal sequence, but we also provide evidence that the tubular network at the coronary ostia expands into a venous-capillary plexus, surrounds the coronary artery stems, and persists through venous-capillary plexus, and the coronary arteries and the parasympathetic ganglia remain in close proximity throughout this period (Fig. 4).
It is not surprising, as documented in this study, that parasympathetic ganglia are closely associated with the roots of the two coronary arteries and also with their immediate branches. The presence of these ganglia has been documented to be necessary for normal development of coronary artery vascular smooth muscle (Hood and Rosenquist,1992). Their study used neural crest ablation to demonstrate a spatial disordering of smooth muscle, the presence of only one coronary ostium and anomalous coronaries from the subclavian artery. Moreover, cardiac ganglion cells in the chick have been shown to originate from the cardiac neural crest (Verberne et al.,1998). The latter are not coronary artery smooth muscle precursors, but rather are positioned at the base of coronary arteries (Waldo et al.,1994), a finding that suggests their signaling may contribute to formation of the coronary ostia. The locations of the intracardiac ganglia in human neonate (Smith,1971) and adult (Pauza et al.,2000) have been described in detail.
The formation of the coronary ostia at E8–E9 is dependent on a variety of factors (reviewed by Tomanek,2005). Foremost is a functional epicardium, including the timing of its outgrowth (Eralp et al.,2005). Our recent work has shown that VEGF family members, especially VEGF-B, play a critical role in the formation of coronary ostia (Tomanek et al.,2006). Further growth and remodeling of the coronary arteries have not received much attention. That arteriolar growth is dependent on FGF-2 has been demonstrated in neonatal rats (Tomanek et al.,2001). In that study, we showed that arteriolar growth was attenuated in response to anti-FGF-2 neutralizing antibodies. When both FGF-2 and VEGF are experimentally decreased, there occurs a shift in the arteriolar hierarchy, i.e., length density of the smallest arterioles is decreased and increased in the largest arterioles. Moreover, precocious expression of FGF-2 induces abnormal coronary artery branching (Mikawa,1995). Current studies in our laboratory on embryonic quail indicate that PDGF-B, which is expressed in coronary endothelial cells (Van Den Akker et al.,2005), in combination with FGF-2 are critical for coronary artery formation (data not shown).
As noted in the current study, coronary artery growth is rapid. Ratajska et al. (2000) noted that coronary artery diameter in the rat increases fourfold in a 3-day period prior to birth. Our findings in the quail show a progressive increase in coronary artery diameter that follows the increase in ventricular thickness. By analysis of serial sections, we have also noted a consistent pattern of branching and distribution of the coronary arterial tree, including a dominant septal artery. The substantial increases in artery diameter must involve a rapid remodeling process, which occurs with increased flow, as shown in adult models of remodeling (Gibbons and Dzau,1994; Sho et al.,2003). It has been reported that coronary artery formation predates the periarterial Purkinje fibers (Harris et al.,2002). The periarterial Purkinje fibers thus follow the course of coronary arteries. Interestingly, the large main septal artery lies between the two main bundle branches of the conduction system. A recent review of apoptosis in the chick heart notes that the greatest numbers of apoptotic figures occur in the outflow tract in HH 19–26 hearts, with a concentration noted just below the aorta at HH 31 (Martinsen,2005). This concentration, also noted by Schaefer et al. (2004), at this developmental stage corresponds to the site of coronary ostia formation. Velkey and Bernanke (2001) documented apoptosis within the aorta at the site of penetration by the peritruncal capillary plexus, as well as pruning of the capillary plexus around the aorta, a finding that is consistent with observations on quail (Ando et al.,2004). Although it is well known that apoptosis functions in vascular pruning and reduction in vessel diameter (reviewed by Fisher et al.,2000), less attention has been paid to its role in angiogenesis and arteriogenesis. In this regard, it has been shown that inhibition of apoptosis attenuates capillary formation in vivo (Segura et al.,2002). It has been noted that apoptosis occurs mainly in the “nonmyocardial compartment” (Poelmann et al.,2000), with a high rate in neural crest cells (Poelmann et al.,2005).
Apoptotic figures were noted in coronary arteries of HH 40 (E14) chick hearts (Cheng et al.,2002). Our study documents apoptosis in various-sized arterial vessels at E10 and later in both the medial and adventitial tunics. This suggests a role for apoptosis in coronary vessel remodeling. It is important to understand apoptosis in the developing heart because it plays a role in cardiac malformations (Fisher et al.,2000).
The importance of the extracellular matrix in endothelial morphogenesis and stabilization has been reviewed (Davis and Senger,2005). Evidence is provided to support the idea that fibronectin and collagen matrices stimulate tubulogenesis, whereas laminin-rich matrices affect stabilization. This supports our previous finding on rat embryonic development of the coronary vasculature (Rongish et al.,1996). In that study, we documented a role for fibronectin as scaffolding for vascular cells and suggested that laminin deposition coincides with tube formation. The current study, which focused on the arterial system, also indicates the importance of fibronectin for cell migration as evidenced by smooth muscle cell contact with fibronectin strands. We show that in addition to the presence of fibronectin and laminin in the artery's media, the latter forms a distinct ring outside of the tunica adventitia. Both of these noncollagenous glycoproteins are cell adhesion molecules that are highly expressed in the embryonic human coronary arteries (Kim et al.,1999). However, our findings should not be interpreted to mean that fibronectin and/or laminin are the only extracellular matrix components regulating vessel development. For example, fibulin-1 and fibulin-2 are expressed in tunica adventitia and basement membrane of the endothelium in human embryonic blood vessels (Miosge et al.,1996).
Myocardial vascularization includes many phenomena that are spatially and temporally regulated. The current study illustrates the progression of myocardial vascularization in the quail during the embryonic and fetal periods, i.e., E6–E18. Most of our data concern the development of the arterial hierarchy, i.e., arteriogenesis, a component of vascular development that has received less attention than the earlier stages of vascularization, i.e., vasculogenesis and angiogenesis.
Our findings, some of which confirm and complement data from other studies, provide new insights into several developmental events. Using electron microscopy, we show some features of coronary precursors, including blood islands, and demonstrate that the tubular network is almost entirely without connections to the endocardium. The hierarchy of the arterial system, based on serial sections, has been revealed and its development noted with respect to innervation, extracellular matrix, and apoptosis. Thus, the findings presented here provide an overview of a relatively broad period of vascular development. Our goal in providing these data was to chart the components of vascular development and provide a foundation for future studies on the mechanism underlying formation of the coronary vasculature.