SEARCH

SEARCH BY CITATION

Keywords:

  • coronary vessels;
  • heart development;
  • proepicardial organ;
  • vasculogenesis;
  • angiogenesis;
  • growth factors;
  • apoptosis;
  • ischemia;
  • arteriosclerosis;
  • VEGF;
  • FGF

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

Earlier views of the development of the coronary vasculature included angiogenic budding and growth of arteries from the aortic sinuses and veins from the coronary sinus. The current concept begins with the establishment of the epicardium from the proepicardial organ, an outgrowth of the dorsal wall of the pericardial cavity. Capillaries form in a subepicardial mesenchymal population, extending as a plexus toward the truncus arteriosus and the atria. Multiple vessels grow from a peritruncal ring of capillaries, preferentially invading the newly formed aorta. In a process involving apoptotic changes of both the aortic wall and the invading capillaries, orifices open at the level of the aortic sinuses. Smooth muscle cells and pericytes, recruited from the surrounding mesenchyme, contribute to the vessel walls, and the definitive coronary artery pattern is established. Similar events are occurring on the venous side of the coronary circulation, following a slightly earlier time course. Multiple factors govern this process, including VEGF and FGF-1 stimulating vasculogenesis and angiogenesis, and the angiopoietins and their tyrosine kinase receptors modulating interactions between endothelial cells and the mural components. As remodeling of the capillary plexus and the coronary orifices progresses, TGFβ released by apoptotic cells or from other sources likely modulates VEGF and FGF-1, and also contributes to further apoptotic changes. A better appreciation of the controls of the mechanisms of coronary vessel development may direct further research in the prevention of arteriosclerosis and ischemic tissue injuries. Anat Rec (New Anat) 269:198–208, 2002. © 2002 Wiley-Liss, Inc.


INTRODUCTION: EARLY HEART DEVELOPMENT

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

As an embryo grows and develops, it soon reaches a size that prohibits simple diffusion of sufficient nutrients and oxygen from the surrounding liquid environment to all tissues. In humans, this occurs by early in the third week. It is fortuitous that the development of the circulatory system is already ongoing at this critical time. The elaborate plumbing system is designed to compensate for increased metabolic demand by conveying the needed materials to the various spaces and masses of tissues that make up the growing individual.

The complex but elegant development of the heart is perhaps the most notable feature at this period, first recognized as paired endocardial tubes that soon fuse beneath the forming foregut. During this process, the endocardial tube acquires a layer of presumptive cardiac muscle (known as the myocardial mantle) from the mesoderm adjacent to the forming heart. The single heart tube then undergoes a sequence of changes apparent from the outside, looping and folding to rearrange the straight tube into the familiar shape of the heart, with the input and output pathways brought closer together and the main pumping ventricle positioned at the opposite apex. During the same period, a well-choreographed process of remodeling also progresses on the inside, resulting in the partitions and valves that are the familiar anatomical features of the adult four-chambered heart with a separate pulmonary artery trunk and aorta. The myocardium begins its contractions at approximately the third week (in a human embryo), driving the beginnings of circulation. The heart continues to pump during the complex morphogenetic changes in its shape and blood flow paths and then throughout the entire lifetime of the individual.

At the earliest stages, a coronary circulation does not yet exist, and the blood flowing through the lumen of the still rather simple heart tube must nourish the endocardium and myocardium. As the heart wall increases in thickness, it also outgrows the diffusion limits and soon an improved means of nutrient and oxygen supply must be provided to the heart itself. This is ultimately accomplished by the formation of a dedicated vascular system over the surface of the heart, consisting of coronary arteries, coronary veins, and a capillary bed penetrating and supplying the myocardium. The developmental pattern of the coronary circulation is remarkably consistent; in fact, the incidence of congenital coronary arterial abnormalities is low.

The concept of how the rather complex pattern of the coronary vascular system develops has been a story undergoing its own dynamic changes, especially during the past two or three decades.

The concept of how the rather complex pattern of the coronary vascular system develops has been a story undergoing its own dynamic changes, especially during the past two or three decades.

Many aspects of this development have features not found in other areas of embryonic tissue development or organogenesis. The clinical importance of health and integrity of the coronary circulation is obvious to the most naive of first-year medical students. However, until very recently, even basic information about how the coronary arteries develop has been inexplicably omitted from most textbooks used in medical embryology courses.

EARLY CONCEPTS, CHANGING CONCEPTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

Can the Heart Be Supplied From Blood Flow in Its Lumen?

As long as the wall of the heart remains only a few cells thick, simple diffusion allows the passage of needed nutrients into the endocardium and the closely applied myocardial cells. Continued development results in increased cell mass of the heart wall, especially in regions around the atrioventricular canal, bulbus cordis, and truncus arteriosus as the cardiac jelly (extracellular matrix) thickens between the endocardium and the myocardium to form the endocardial cushions responsible for partitioning these areas and forming valve structures. As is the case with any muscle, the constant exercise and growth of the myocardium results in changes in size and thickness as it does its work as a pump. These mass changes soon prohibit supply by simple diffusion from the blood being pumped through the heart lumen.

Are Intertrabecular Sinusoids the Origin of the Dedicated Coronary Blood Supply?

Seemingly in response to this situation, the heart wall acquires a series of distinctive thickenings on its luminal side, with the formation of lacunar spaces into the thickened myocardium. The bar-shaped masses of heart tissue would appear to act as buttressing supports of the heart walls or perhaps provide extra muscular components, but why these trabeculae form in the walls of the heart is not well understood. Many of the trabeculae persist in the heart and are especially notable in the adult auricles and enlarged in the ventricles. This trabeculated rearrangement of the tissues results in the partitioning of endocardial-lined sinusoids, folding into the heart wall. As the heart contracts, blood is exchanged between these spaces and the lumen of the ventricle. The intertrabecular spaces seem to have the effect of reducing the diffusion distance from the heart lumen into the muscle mass, and the heart wall could conceivably be supplied in this manner by the blood being pumped through the heart, if the wall thickness remains relatively small.

Earlier studies, in fact, suggested that the vessels penetrating the myocardium and communicating with the coronary arteries (Thebesian vessels) take their origin from the intertrabecular spaces (Grant, 1926). Although they seem to provide partial blood supply to the myocardium in some fish and amphibians, the sinusoidal system in mammals and birds never quite forms into an effective and defined internal coronary circulatory supply. Thus, the responsibility of supplying blood to the myocardium is left to the coronary vessels that are found on the outside surface of the heart (Rýchter and Oštádal, 1971; Waldo et al., 1990).

What Was Previously Thought To Be the Source of the Coronary Vessels?

Earlier interpretations of histological studies in both bird and mammal embryos seemed to present two variations on a theme of the developmental origins of coronary arteries. Some of the earliest studies suggested that coronary arteries might begin as single solid cords of endothelial cells growing from the developing aortic wall at the sinus. These would form tubular channels by a process of canalization, opening from the aorta down onto the heart in the direction of the ventricles (Grant and Regnier, 1926; Bennett, 1936; Goldsmith and Butler, 1937; Dbalý et al., 1968; Voboril and Schiebler, 1969; Rýchter and Oštádal, 1971; Aikawa and Kawano, 1982; Hirakow, 1983). Alternatively, tubular branches from the aorta at the sinuses were thought to grow by a process of angiogenesis (for a review of angiogenesis, see Tomanek and Schatteman, 2000). These branches would begin as aortic evaginations of the endothelium that would extend and take on smooth muscle cell outer layers to become the proximal roots of the coronary arteries.

Many studies seemed to support the latter model, found generally in common in several species, including shark and rabbit (Lewis, 1904; Grant, 1926), pig (Bennett, 1936, Goldsmith and Butler, 1937), rat (Dbalý et al., 1968; Voboril and Schiebler, 1969), mouse (Virágh and Challice, 1981), chick (Rýchter and Oštádal, 1971), and even human (Vernall, 1962; Hutchins et al., 1988). Whether formed by canalization of cell cords or by evagination, these vascular outgrowths would eventually establish communication between the aorta and a developing plexus of capillaries in the extracellular matrix space beneath the epicardium to become the well-recognized pattern of coronary arteries. Either alternative seemed to be an acceptable concept, because the coronary arteries have always been thought of and continue to be taught as being the first direct branches of the aorta (Moore and Dalley, 1999). Although this is certainly correct from an anatomical point of view, it does not quite conform to the current concepts of how the coronary arteries develop.

What Is Now Thought To Be the Source of the Coronary Vessels?

More recent studies began to cast some new light on the process of formation of coronary arteries and their connections to the aorta. Key to the new line of thinking were the observations that the approximation of small vessels to the developing aorta always preceded the formation of coronary artery orifices in the aortic sinuses (not the reverse) and that evaginations of the endocardium at the aortic sinuses of the sort expected in an angiogenic process were not found (Gittenberger-de-Groot et al., 1987; Bogers et al., 1988). One significant study that attempted to define the onset of the aortic endothelial evagination failed to observe that process and concluded that the arteries instead arise from a ring of capillaries that can be found at and around the bulbus cordis portion of the outflow region of the heart. These vessels grow toward and eventually attach to the aorta at the aortic sinuses in what was described as a “controlled invasion of the aorta” (Waldo et al., 1990). In addition, as Waldo et al. (1990) point out, the current concepts of angiogenesis indicate that new vessels can originate only from vessels the size of small venules and capillaries, making it unlikely that branches can form at the walls of the aorta. So now it is more generally accepted that the proximal coronary arteries do not develop as a budding and branching process arising from the aorta, but rather begin as part of a complex capillary network that develops from epicardially derived mesenchyme found in the subepicardial space. The capillaries grow around and then penetrate into the developing aorta, connecting with and being supplied by the blood already flowing out of the developing heart. How this happens is one of the unique features of the embryonic development story, and it begins with the development of the epicardium and a subepicardial capillary bed.

SUBEPICARDIAL CAPILLARY BED

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

What Is the Origin of the Epicardium?

After the folding of the early heart tube is essentially completed, it consists of only two layers, the endocardial lining and the outer myocardium (Manasek, 1969; Virágh and Challice, 1981). Over subsequent hours, a layer of epicardial cells covers the bare myocardium, spreading from the proepicardial organ that initially contacts the heart in the region of the atrioventricular canal. The proepicardial organ forms as an outgrowth of the dorsal wall of the intraembryonic coelom adjacent to the developing liver (septum transversum) and near the attachment of the sinus venosus to the cardinal veins and other venous returns to the developing heart.

The structure of the proepicardial organ has been described variously either as a villous outgrowth or as a vesicular structure resembling a cluster of grapes or a cauliflower (Ho and Shimada, 1978; Virágh and Challice, 1981; Komiyama et al., 1987; Hiruma and Hirakow, 1989; Männer, 1999). Each vesicle in the cluster is composed of multiple mesothelial cells, surrounding an apparently fluid-filled lumen. The structure arises from the mesothelial lining of the coelomic cavity that becomes the dorsal portion of the parietal pericardium (Figure 1).

thumbnail image

Figure 1. The precursor of the epicardium, the proepicardial organ, appears as an outgrowth of clustered mesothelial cell vesicles on the dorsal body wall adjacent to the atrioventricular canal of the looped heart (see inset). A: As viewed by scanning electron microscopy, the proepicardial organ (PEO) is seen near the heart surface from the dorsal wall of the pericardial cavity. A, atrium; AV, atrioventricular region; V, ventricle. B: At higher magnification, the individual vesicles clearly can be seen to be composed of multiple cells. Scale bars = 100 μ in A, 10 μ in B.

Download figure to PowerPoint

As the heart continues its growth and rhythmic contractions, repeated contact with the proepicardial organ eventually results in transfer of the mesothelial cells across onto the myocardial surface. The descriptions of how this is accomplished vary with the species studied, but it seems to happen in a manner rather unique for the transfer of cells from one location to another in an embryo. In birds, villous projections form tissue bridges to contact and attach directly to the dorsal surface of the atrioventicular region (Hiruma and Hirakow, 1989; Virágh et al., 1993; Männer, 1999). In studies of the mouse and the tree shrew, this transport has been seen as being accomplished mainly by translocation of detached cellular vesicles (Komiyama et al., 1987; Kuhn and Liebherr, 1989). In either case, the vesicular clusters of cells open on the dorsal atrioventricular canal surface. The epicardial epithelium spreads out from a circular patch around either side to eventually meet on the ventral surface, forming a ring around the atrioventricular canal. The epicardium continues to spread over the surface of the myocardium toward the atria and sinus venosus in one direction and the proximal outflow tract (bulbus cordis) in the other direction (Vrancken Peeters et al., 1995) (Figure 2).

thumbnail image

Figure 2. A: As the growing proepicardial organ (PEO) contacts the heart, vesicles open onto the bare myocardium and the cells spread across the surface of the heart, from the atrioventricular (AV) region toward the atria (A) and ventricle (V). B: At higher magnification, the sheet of epicardial cells (E) can be distinguished from the myocardium (M). C: At still higher magnification, cells at the leading edge of the spreading epicardium (E) can be seen to have projections and ruffled edges but maintain well-established cell junctions, typical of actively migrating sheets of epithelial cells. Scale bars = 100 μ in A,B, 10 μ in C.

Download figure to PowerPoint

What Happens Next to the Epicardium?

As the epicardium grows and progressively covers the heart, a layer of extracellular matrix is produced between the surface epicardial cells and the underlying myocardium. The source of this subepicardial matrix remains unclear, although it is generally thought that both the epicardium and the myocardium contribute materials (Hurle et al., 1994; Bouchey et al., 1996). The subepicardial space soon becomes populated with mesenchymal cells, apparently derived from the epicardial layer by epithelial-mesenchymal transformation similar to that seen elsewhere in the developing embryo and the developing heart (Hiruma and Hirakow, 1989; Mikawa and Fischman, 1992; Dettman et al., 1998; Markwald et al., 1996).

The epicardium seems to be the source of a majority of the mesenchymal cells in the subepicardial space (Pérez-Pomares et al., 1997). However, fate mapping studies have shown that epicardially derived mesenchymal cells become smooth muscle cells and adventitial fibroblasts of coronary arteries but not the endothelial cells (Dettman et al., 1998). Specific histochemical identification of endothelial markers and specific cell labeling suggests that a separate subpopulation of endothelial precursor cells may migrate into the established subendothelial space from the mesothelium of the proepicardial organ, the dorsal body wall of the intraembryonic coelom, or the dorsal mesocardium, just behind and beneath the leading migration of the epithelial layer of the epicardium (Olson et al., 1989; Bolender et al., 1990; Mikawa and Fischman, 1992; Poelmann et al., 1993).

Studies using a small piece of eggshell membrane to block contact of the proepicardial organ with the myocardial surface in chick hearts and the added explants of quail proepicardial organ demonstrated that its cells are the source of both the endothelial and smooth muscle cells of coronary arteries (Männer, 1999). Other studies using this model demonstrated a compensatory growth of mesothelial cells from the pharyngeal arches progressively investing the outflow tract and conotruncal regions (Gittenberger-de-Groot et al., 2000). Although this observation was made under experimentally modified conditions, it suggests an interesting alternative source of the epicardium, especially in the outflow tract, the site of the future aorta and pulmonary trunk.

How Do Capillaries Form Beneath the Epicardium?

As the epicardial layer completes its investment of the folded heart, the population of mesenchymal cells between the epicardium and the myocardium continues to increase. Histological examinations of these areas demonstrate the presence of capillary-like structures, very often with blood-island inclusions (Hiruma and Hirakow, 1989; Pérez-Pomares et al., 1997). These new blood containing capillaries are similar to the vessels that can be found elsewhere in the embryo at early stages of the formation of the cardiovascular system and seem to form by the same or similar mechanisms: vasculogenesis followed by angiogenesis (Poole and Coffin, 1989; Coffin and Poole, 1991).

Vasculogenesis is the process of de novo formation of a blood vessel at a specific location, the ultimate site of the vessel (Box 1). The process begins when groups of mesenchymal cells are segregated from their neighboring cells. Within these clumps of mesenchyme, the precursors of the vessel endothelium begin to differentiate. These angioblasts attach to each other and form spherical aggregates of cells, which then line up as strings, or cords, and establish the pattern of the forming vessel. The cords open to form tubes, often with the formation of blood cells (hematopoiesis) occurring in the midst of the angioblasts (Poole and Coffin, 1989).

Box 1. Vasculogenesis and Angiogenesis Defined

Vasculogenesis: the de novo formation of blood vessels at a specific site from aggregates of endothelial precursor cells (angioblasts) that form from mesenchyme

Angiogenesis: the elongation of small vessels and formation of branches by proliferation of existing endothelial cells and remodeling.

How Do the Capillaries Change?

The newly formed vessels, the products of vasculogenesis, elongate and form new branches by the process of angiogenesis (Box 1). This process of remodeling involves migratory types of changes in the cells at the end or at a particular location on the existing vessel, which then extend as a coherent structure often referred to as a sprout. The sprout lengthens and the endothelial cells proliferate adjacent to the active migratory end of the sprout (Poole and Coffin, 1989; Tomanek and Schatteman, 2000). By means of combinations of vasculogenesis and subsequent angiogenesis, a complex array of capillary-sized vessels is established in the subepicardial space, at first mostly concentrated along the atrioventricular sulcus and the dorsal interventricular sulcus. These often are seen histologically to contain blood cells, as a result of the hematopoietic nature of the mesenchymal precursors, but the vessels do not yet demonstrate coordinated circulation.

The vessels in the subepicardial space continue to develop into an extensive capillary plexus, spreading along the dorsal interventricular sulcus, into the atrioventicular sulcus and the ventral interventricular sulcus, and then around the bulbus cordis region, eventually approaching the base of the truncus arteriosus. This peritruncal ring of capillaries eventually is joined on the ventral aspect, with sprouts directed toward the now divided aorta and pulmonary trunk. However, for reasons still not understood, only the aorta is invaded and receives connections from the peritruncal capillary plexus. The vessels in the interventricular sulci also continue to extend toward the apex of the heart. The subepicardial space is thicker along these areas and can provide greater capacity for growth of the developing coronary vessels. In these areas, angiogenic changes in the capillaries also seem to be responsible for the invasion of the new vessels into the ventricular myocardial layer. This change is first notable along the interventricular septum, itself undergoing growth and extension at this period. The template, thus, is established for the eventual formation of the rather constant pattern (with minor variations) of coronary vessels seen on the surface of the adult heart. All that is required is remodeling of the capillary plexus to reduce the number and sizes of direct arteriovenous connections, leaving the intramyocardial capillaries to supply the heart wall muscle (Vrancken Peeters et al., 1997).

COMPLETING THE CORONARY CIRCULATION

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

How Does the Peritruncal Capillary Plexus Attach to the Primitive Aorta?

The process by which the peritruncal vessels invade the aorta and establish connections has been well described in chicken embryos (Bogers et al., 1989; Waldo et al., 1990), although not mechanistically. In general, it has been noted that vessels from the peritruncal ring selectively extend into the aorta (presumably by means of angiogenesis) and establish multiple persistent connections to the left, right, and, rarely, posterior aortic sinuses. The invasion and persistence of multiple vascular channels might seem counter to the description of single left and right coronary arteries generally given in anatomy texts. However, studies of human coronary artery patterning have actually shown the rather frequent occurrence (≈35%) of small accessory coronary orifices (Turner and Navaratnam, 1996).

Upon connecting to the aorta, rapid remodeling of some of the vessels and their orifices follows. By a still unknown mechanism, some of the penetrating vessels acquire a layer of vascular smooth muscle cells and persist to become the definitive coronary artery stems (Hood and Rosenquist, 1992; Poelmann et al., 1993; Waldo et al., 1994; Vrancken Peeters et al., 1997). Poelmann et al. (1993) observed that developing coronary vessels that failed to form a tunica media ultimately regressed, suggesting that the acquisition of a tunica media plays a stabilizing role in this process. Other investigators have noted an association of parasympathetic ganglia with the developing coronary artery stems and suggest that the ganglia may somehow support coronary vascular development (Hood and Rosenquist, 1992; Waldo et al., 1994). Mechanical factors such as blood flow and pulsatile stretch have been shown to stimulate vascular development in vitro (Niklason et al., 1999) and in vivo (reviewed in Hudlicka and Brown, 2000), so it seems likely that such factors probably also contribute to the stabilization and maturation of the nascent coronary arteries once they are connected to the aorta.

It should be noted that the development of the proximal coronary arteries and their orifices requires rather drastic and rapid reorganization of many tissues.

It should be noted that the development of the proximal coronary arteries and their orifices requires rather drastic and rapid reorganization of many tissues.

Upon contacting the aortic endothelium, a patent connection between the penetrating vessels and the aorta must be made. With further development of the coronary arteries, the wall of the aorta must accommodate the expansion of the coronary arteries and their orifices as they increase in diameter. Recently, apoptotic cells have been observed in close spatial and temporal association with all of these aspects of coronary artery development (Velkey and Bernanke, 2001) (Figure 3). The correlation of apoptosis with coronary artery orifice formation sheds new light on the possible regulatory mechanisms involved. Finally, the peritruncal ring and its numerous vascular channels will be reduced to the general coronary artery pattern observed in the adult, a process that must also involve vascular regression (Figure 4).

thumbnail image

Figure 3. A: Coronary artery attachment to the aorta involves invasion of blind-ended extensions of vessels from the peritruncal capillary plexus (PTV) into the tunica media (TM) of the developing aorta superior to the leaflets (VL) of the forming aortic tricuspid valve. Apoptotic cells (Apo) can be found in association with the invading vessels but not within the aortic endothelium (En). B: As the blind end of the invading vessel contacts the aortic endothelium, a patent coronary orifice forms by means of apoptosis. C: Even after connection to the aorta, remodeling continues, involving both apoptosis to accommodate expansion of the artery and enlargement of the orifice and recruitment of a vascular smooth muscle coat (VSMC) from the surrounding mesenchyme and aortic wall.

Download figure to PowerPoint

thumbnail image

Figure 4. A: At the final stages of coronary artery formation, multiple capillaries (Cap) form from mesenchyme by vasculogenesis within the matrix beneath the epicardium (Epic). B: The capillaries elaborate a peritruncal ring (PTR) around the aorta and pulmonary trunk, growing by angiogenesis, and establishing dominant channels of vessels with larger sizes. The multiple larger vessels approach and attach to the aorta in preference to the pulmonary trunk, in a process involving apoptosis to form orifices at the level of the coronary sinuses (see Figure 3). C: The peritruncal capillary plexus is pared down by further apoptotic events to the predominant right (RCA) and left (LCA) coronary arteries attached to the aorta at the corresponding coronary sinuses. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Download figure to PowerPoint

How Does the Development of Coronary Veins Compare With That of the Arteries?

The story of the formation of coronary veins has features similar to that of coronary artery development. Studies in a variety of research animal species have described sprouting and growth of venous precursor vessels from the sinus venosus (Lewis, 1904; Grant, 1926; Bennett, 1936; Goldsmith and Butler, 1937; Voboril and Schiebler, 1969; Rýchter and Oštádal, 1971). More recent studies, using chick-quail chimeras and specific staining for endothelial markers, did not support these observations (Poelmann et al., 1993). Instead, the precursors of the coronary veins have been seen to develop similar to the coronary arteries, as a capillary plexus, which then grows toward and into the wall of the sinus venosus. Some vessels in this portion of the capillary plexus remain as the future coronary veins, maintaining continuity with the right atrium by means of the coronary sinus, the venous return side of the heart (Vrancken Peeters et al., 1997). The venous side of the capillary plexus precedes the arterial side in making connection with the luminal circulation, but coordinated circulation through the vessels still awaits connection of the artery precursor vessels at the aortic sinuses, which follows approximately a day later in most species studied thus far.

CONTROL OF CORONARY VESSEL DEVELOPMENT

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

What Factors Influence Vasculogenesis?

Box 2 lists several growth factors and their receptors that are important in coronary vessel development. Of these, FGF-2 has been well established as the factor responsible for angioblast induction from mesenchyme in embryonic vasculogenesis (reviewed in Poole et al., 2001) and, thus, is likely to be involved in coronary endothelial induction from the perihepatic mesenchyme related to the proepicardial organ. FGF signaling has also been shown to regulate epithelial-to-mesenchymal transformation as occurs during gastrulation at Hensen's node (Bikfalvi et al., 1997), in endocardial cushions (Markwald et al., 1996), and as demonstrated in the epicardium (Dettman et al., 1998). FGF-2 is present in the myocardium and was shown by Dettman et al. (1998) to specifically induce epithelial-to-mesenchymal transformation in epicardial cells. However, despite the presence of FGF-2 in the subepicardial matrix, none of these epicardially derived mesenchymal cells were demonstrated to become angioblasts in that study.

Box 2. Growth Factors and Receptors Significant to Coronary Artery Development

Growth Factors

Ang1: angiopoietin 1

Ang2: angiopoietin 2

bFGF: basic fibroblast growth factor (now known as FGF-2)

FGF-2: fibroblast growth factor–2 (formerly known as basic FGF)

PDGF: platelet derived growth factor

TGF-β: transforming growth factor-beta

VEGF: vascular endothelial growth factor

Receptors

TIE1: tyrosine kinase with immunoglobulin and EGF* homology domains 1

TIE2: tyrosine kinase with immunoglobulin and EGF* homology domains 2

TEK: endothelial tyrosine kinase (also known as TIE2)

* EGF = epidermal growth factor

The process of angioblast migration and coalescence into tubes has also been well studied (although not extensively in the heart). The results of such studies indicate that VEGF is most likely the primary factor involved as both a diffusible chemoattractant for migrating endothelial cells and as a substrate-bound vessel stabilization factor once tubes are formed (Tomanek et al., 1999; reviewed in Poole et al., 2001). It has been shown that differential splicing of the VEGF gene produces protein isoforms of varying size (e.g., 120, 164, and 188 amino acid residues in the mouse) and with varying degrees of heparin binding affinity. The shorter forms do not bind heparin well and are thus more diffusible, whereas the larger forms are less diffusible due to their higher heparin affinities (reviewed in Ferrara, 2000). Studies support the hypothesis that diffusible VEGF is involved in endothelial chemotaxis (Cleaver and Krieg, 1998), whereas VEGF isoforms with higher heparin binding affinities affect endothelial coalescence and vessel patterning (Poole et al., 2001) as well as vessel stability (Cheng et al., 1997).

What Factors Influence Angiogenesis?

The evidence is strong that FGF and VEGF, along with many other factors, also play crucial roles in the angiogenic expansion of the early network of coronary vessels.

The evidence is strong that FGF and VEGF, along with many other factors, also play crucial roles in the angiogenic expansion of the early network of coronary vessels.

Angiogenesis begins with the disintegration of the endothelial cell basement membrane followed by the proliferation and migration of endothelial cells away from the parent vessel. VEGF signaling is known to induce expression of serine proteases that mediate dissolution of basement membranes while also stimulating endothelial proliferation and migration (reviewed in Ferrara, 2000). The same appears to be true for FGF-2 (Basilico and Moscatelli, 1992; Tomanek et al., 1996). Direct evidence that FGF-2 and VEGF affect coronary vasculogenesis and angiogenesis comes from a study by Tomanek et al. (1998) in which they systemically administered these two compounds to chicken embryos well in advance of coronary vessel formation. In this study, both FGF-2 and VEGF enhanced tube formation during the vasculogenic phase of vessel formation while also increasing vessel numerical density at later stages indicating proangiogenic effects as well.

How Is Mural Recruitment Involved and Controlled?

In addition to the endothelium, all vessels, even capillaries, have additional cells associated with their vessel walls known as pericytes, and larger vessels also add a complement of smooth muscle cells. The cells of the vessel wall are acquired or “recruited” from the surrounding mesenchyme. In coronary vessels, the process of mural recruitment occurs quite rapidly upon connection of the vascular plexus to the aorta, as these formerly low-pressure, capillary-sized vessels must now accommodate much higher systolic blood pressures (see Figure 3C). Ultrastructural examination of this process by Hood and Rosenquist (1992) showed that mural recruitment in coronary arteries proceeds in an orderly manner, beginning proximally at the site of aortic attachment and proceeding distally. Experimental results by Mikawa and Gourdie (1996) and Dettman et al. (1998) indicated that the pericytes and vascular smooth muscle cells of coronary arteries are recruited from epicardially derived mesenchyme by an as yet unknown mechanism. The phenomenon of mural recruitment has been known to involve factors produced by endothelial cells such as PDGF, FGF-2, and EGF as well as cell-to-cell contacts (reviewed in Hirschi and D'Amore, 1996; Lu et al., 2001). The evidence thus far suggests that PDGF produced by endothelial cells induces the proliferation and migration of mesenchymal cells toward developing vessels (Hirschi et al., 1999; Lu et al., 2001). Upon contacting the endothelium, these mesenchymal cells are then induced to differentiate into mural components (as indicated by expression of smooth muscle actin) in a TGF-β–dependent manner. Additionally, cell-to-cell contact between mural cells and endothelium suppresses proliferation (even in the presence of PDGF) in both cell types, thus, further stabilizing the vessel.

The role of TGF-β in this process, however, appears to be quite complicated. In general, TGF-β is considered to be a growth factor largely involved with the differentiation of cells during development. This finding certainly seems to be the case in vascular development as well. A recent study of the effect of various vascular growth factors on vessel formation from heart explants has shown that TGF-β repressed vascular tube formation (Tomanek et al., 2001). It has also been shown that TGFβ can suppress proliferation and even induce apoptosis in endothelial cells (Pollman et al., 1999), while conversely promoting smooth muscle differentiation (Shah et al., 1996). Once differentiated, these smooth muscle cells can respond to TGF-β by increasing production of PDGF (Gadson et al., 1997), thus even further enhancing mural recruitment. However, TGF-β can also modulate secretion of endothelial mitogens such as VEGF from cardiac myocytes (Li et al., 1997) and macrophages or other inflammatory cells (Nicosia and Villaschi, 1999). Thus, although TGF-β seems to be capable of directly promoting vessel stability and maturation by means of the repression of new vascular growth and the recruitment and differentiation of mural components in existing vessels, it would also seem that this growth factor can indirectly promote vascular instability and endothelial proliferation by means of VEGF signaling.

What Are the Regulators of Vessel Remodeling?

The delicate balance between vascular stability and instability is, of course, critical to both vasculogenesis and angiogenesis. On the one hand, the development of vessels requires vascular precursor cells to disassociate from each other to proliferate and migrate. Yet, these same cells must then re-associate and establish stable interactions with each other if they are to form a structurally sound vessel. We have seen how VEGF, FGF-2, PDGF, and TGF-β can modulate vessel stability by means of their varied effects on endothelial and mural cells. However, it has become clear that the actual interactions between endothelial and mural cells are most directly mediated by a relatively novel class of growth factors, the angiopoietins (Ang1 and Ang2), and the tyrosine kinase receptors TIE1 and TIE2 (or TEK) (recently reviewed in Jones et al., 2001).

Disruption of TIE1 or TIE2 in mouse embryos has little early effect on vasculogenesis as blood vessels are observed to form readily (Sato et al., 1995). However, by embryonic day 9.5 (for TIE2 mutants) and embryonic day 13.5 (TIE1 mutants), massive hemorrhaging can be observed, indicating a defect in vessel structural integrity. Disruption of Ang1 in mouse embryos (Suri et al., 1996) also produced similar defects along with abnormalities in pericardial capillary branching, insufficient vessel remodeling in yolk sacs, and insufficient heart development. Ultrastructural analyses have shown that endothelial cells fail to associate properly with pericytes in embryos deficient in either Ang1 or TIE2 (Suri et al., 1996). It has also been shown that cells expressing TIE2 are stimulated upon addition of Ang1 to bind extracellular matrix molecules such as fibronectin and collagen in an integrin-dependent manner (Sato et al., 1998). Conversely, addition of Ang2 to this system seems to block the ability of Ang1 to activate TIE2, presumably because Ang2 binds Ang1 and inactivates its signaling potential. This process leads to the dissociation of endothelial cells from their extracellular matrix (Maisonpierre et al., 1997). In this particular state, the endothelial cells are then responsive to exogenous Ang1 and will, in fact, migrate toward an Ang1 source. Takakura et al. (2000), recently demonstrated this chemotactic effect of Ang1 in experiments showing that Ang1 derived from hematopoietic stem cells could induce vessel sprouting into otherwise avascular areas. In the absence of any outside signaling (such as VEGF or Ang1), the dissociated endothelial cells will undergo apoptosis and the vessels will regress.

In the case of coronary vascular development, there have been descriptions of such vascular remodeling (Vrancken Peeters et al., 1997). The primitive vascular plexus established by vasculogenesis must somehow undergo significant expansion and remodeling to form the common pattern of coronary vessels observed in the adult heart. The previous studies that described defective pericardial capillary branching in Ang1-deficient embryos (Suri et al., 1996; Takakura et al., 2000) strongly suggest that the roles ascribed to TIE2, Ang1, and Ang2 probably apply to coronary vascular development. However, no studies to date have specifically addressed this issue.

Why Are Coronary Arteries Excluded From the Primitive Pulmonary Trunk?

As the common outflow tract (truncus arteriosus) is being divided by the aorticopulmonary septum to form the aorta and pulmonary artery, multiple small vessels from the surrounding peritruncal capillary plexus bypass the pulmonary artery and selectively penetrate the aorta. The controls on the specificity of this selective connection remain to be fully elucidated. A subpopulation of the cranial neural crest, called the cardiac neural crest, has been determined to play an important organizational (although not necessarily inductive) role in this process. Cardiac neural crest cells have been shown to contribute to the tunica media of the aorta and other pharyngeal arch arteries (Bergwerff et al., 1998), the aorticopulmonary septum (Poelmann et al., 1998), and parasympathetic ganglia of the cardiac plexus (Waldo et al., 1994). Additionally, cardiac neural crest cells seem to be necessary for the normal retraction of a sheath of myocardial tissue that surrounds the outflow tract at the location of the future coronary orifices (Rosenquist et al., 1990; Poelmann et al., 1998). However, cardiac neural crest cells do not contribute any part of the coronary vessels (Waldo et al., 1994; Bergwerff et al., 1998). Nonetheless, when cardiac neural crest cells are surgically ablated in chicken embryos, coronary artery anomalies are observed. Such anomalies include ectopic sites of origin (such as the subclavian artery), aberrant vessel patterning, and asymmetric development of the tunica media (Hood and Rosenquist, 1992).

It has also been noted by several researchers that both ectopic (Hood and Rosenquist, 1992; Poelmann et al., 1993) and normal (Waldo et al., 1994) coronary arteries are always found in close association with parasympathetic ganglia. However, not all of the vascular channels from the peritruncal plexus that initially penetrate the aorta have ganglia associated with them (Waldo et al., 1994). Vessels not associated with any ganglia ultimately regress, whereas those that are associated with ganglia persist to become the definitive coronary arteries. This finding has led to the conclusion that the ganglia of the cardiac plexus do not induce vascular invasion of the aorta per se, but the ganglia probably perform a stabilizing function by promoting vascular smooth cell development and regulating blood flow in the nascent coronary arteries (Waldo et al., 1994).

Cardiac neural crest cells may also have a secondarily inductive effect in the growth of peritruncal vessels into the aorta. Based on data by Brauer and Yee (1993), Poelmann et al. (1998) recently described a subset of cardiac neural crest cells in the aorticopulmonary septum and myocardial sheath that are fated to undergo apoptosis and may release vascular regulatory factors such as TGFβ into the local aortic mesenchymal tissue, close in both space and time to the development of the proximal portions of the coronary arteries. In our own studies, we have observed apoptotic cells in close association with the developing coronary artery stems and orifices (Velkey and Bernanke, 2001). The local release of TGFβ from apoptotic cells may play a role in directing either the specific regression of some advancing vessels or the specific connection of others to the developing aorta.

TGFβ has been shown to induce apoptosis in endothelial cells (Pollman et al., 1999) and to promote smooth muscle differentiation (Shah et al., 1996). Topouzis and Majesky (1996) reported that TGFβ increased cell proliferation in neural crest derived smooth muscle cells while inhibiting growth of mesodermally derived smooth muscle. Gadson et al. (1997) observed greater increases in DNA synthesis by smooth muscle derived from neural crest than in smooth muscle from mesoderm. The neural crest–derived cells also produced PDGF, suggesting a proangiogenic effect (although indirect) for TGFβ as well. The diverse and wide-ranging effects of TGFβ and the hypothesis that apoptotic neural crest cells in the aorticopulmonary septum and other spatially related tissues might release TGFβ seems to indicate an important organizational role in proximal coronary development and in vascular invasion of the aorta. However, the possibility also exists that the apoptosis associated with proximal coronary artery development is not a cause of vascular invasion, but rather, an effect. The invading vessels might induce the surrounding tissue to undergo apoptosis to accommodate further development of the coronary arteries. Future research may determine what trophic factors might be guiding peritruncal vessels toward the aorta to form their connections as coronary arteries.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

The story of the development of coronary vessels is far from complete, with recent studies concentrating on the controls on the developmental mechanisms continuing to add to the understanding of the process. The earlier concept of angiogenic outgrowth of coronary arteries from the aorta has been replaced by our current understanding of the development of the epicardium from the proepicardial organ (see Figures 1 and 2).

The earlier concept of angiogenic outgrowth of coronary arteries from the aorta has been replaced by our current understanding of the development of the epicardium from the proepicardial organ.

Within the mesenchyme populated subepicardial matrix, capillaries develop by the process of vasculogenesis and elaborate a plexus that eventually forms a ring structure at the base of the outflow tract (see Figure 4). Capillaries from the peritruncal capillary ring approach and preferentially contact the newly formed aorta. Apoptotic changes are involved in passage of the extending vessels through the wall of the aorta, forming arterial orifices at the aortic sinuses (see Figure 3). The coronary venous system is established earlier by connections of the distal portion of the subepicardial capillary plexus to the sinus venosus, remaining as the coronary sinus at the right atrium. Lesser vascular channels regress and the more major routes are maintained, thus establishing the coronary vascular pattern seen in the adult. The studies of proximal coronary artery development have produced a rich descriptive understanding of this event and set the stage for edification at the molecular level.

It is obvious that proximal coronary artery development involves vasculogenesis, angiogenesis, and apoptosis, but very few studies have focused on how factors known to affect these processes (such as FGF-2, VEGF, PDGF, and TGF-β) might regulate development of the coronary artery stems. A deeper understanding of proximal coronary artery development may have clinical implications as well. The pathological hyperplasia of vascular smooth muscle associated with arteriosclerosis is dependent on an artery's becoming abnormally permissive of the proliferation and invasion of vasa vasorum into the developing plaque (Barger et al., 1984). It has been observed that the ascending aorta in the adult is far less prone to developing such plaques compared with other arteries perhaps due to its ability to inhibit overproliferation of vasa vasorum (Pollman et al., 1999). Yet, during development, the aorta permits the invasion and proliferation of vascular channels that will become the coronary arteries. Understanding the mechanisms underlying this process and its development-related changes may be of direct benefit to the millions of patients worldwide who suffer from diseases related to arteriosclerosis and ischemic tissue injury.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

The authors acknowledge the work of Mr. Michael P. Shenk, Director of the Department of Medical Illustration, University of Mississippi Medical Center, for Figure 4 of this article.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

Biographical Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION: EARLY HEART DEVELOPMENT
  4. EARLY CONCEPTS, CHANGING CONCEPTS
  5. SUBEPICARDIAL CAPILLARY BED
  6. COMPLETING THE CORONARY CIRCULATION
  7. CONTROL OF CORONARY VESSEL DEVELOPMENT
  8. CONCLUSIONS
  9. Acknowledgements
  10. LITERATURE CITED
  11. Biographical Information

Dr. Bernanke is an Associate Professor in the Department of Anatomy at the University of Mississippi Medical Center. A member of the American Association of Anatomists for two decades, his research interests have ranged from mesenchyme movement and extracellular matrix changes during heart development, to structural changes in cerebral arteries after stroke, to scar formation and wound healing. Current research centers on the formation and connections of coronary arteries at the developing aorta. Mr. Velkey began his graduate studies with Dr. Bernanke in Mississippi with a study of apoptosis at the attachment of coronary arteries to the developing aorta. He is now a Ph.D. candidate in the Department of Cell and Developmental Biology at the University of Michigan, investigating early neural development by using embryonic stem cells as a model system.