Vascular elastic fiber heterogeneity in health and disease

Abstract

Purpose of review

Elastin has historically been described as an amorphous protein that functions to provide recoil to tissues that stretch. However, evidence is growing that elastin’s role may not be limited to biomechanics. In this minireview, we will summarize current knowledge regarding vascular elastic fibers, focusing on structural differences along the arterial tree and how those differences may influence the behavior of affiliated cells.

Recent findings

Regional heterogeneity, including differences in elastic lamellar number, density and cell developmental origin, plays an important role in vessel health and function. These differences impact cell-cell communication, proliferation, and movement. Perturbations of normal cell-matrix interactions are correlated with human diseases including aneurysm, atherosclerosis, and hypertension.

Summary

While classically described as a structural protein, recent data suggest that differences in elastin deposition along the arterial tree have important effects on heterotypic cell interactions and human disease.

Introduction

Mature elastic fibers, composed mainly of elastin, provide the elastic recoil necessary for normal physiologic function of organs such as arteries, lungs and skin. Elastin’s primary function in the arterial system, particularly in elastic arteries, is to store energy during systole and release it during diastole, dampening the pulse pressure and allowing laminar blood flow to the microcirculation where oxygen and nutrient exchange occurs. Indeed, the amount of elastin present in arteries along the arterial tree reflects this function; in the ascending aorta for example, elastin accounts for 40–60% of its dry weight depending on the species(1). Hence, proximal conduit arteries, where pulse pressure is high, have numerous elastic lamellae separated by smooth muscle cells (SMCs), while muscular and resistance arteries only have an internal elastic lamina (IEL) that separates endothelial cells (ECs) in the tunica intima from SMCs of the media and a discontinuous external elastic lamina that separates the media from the adventitia (Figure 1 A and B).

Figure 1. Arterial organizational structure.

The ascending aorta receives the entire output from the heart, leading to it to experience high pulse pressures (A). Those pressures are dampened along the course of the arterial tree. Correspondingly, the number of elastic lamellae is highest in the proximal vessels and is reduced to only 1–2 in the smaller arteries (B). The ascending aortic IEL elastin is produced by smooth muscle cells with a small contribution from endothelial cells. In the periphery, however, endothelial cells produce the majority of the elastin (C). The IEL has small evenly spaced fenestrations in the proximal aorta. Further out, the fenestrations become larger and more sporadic (D).

As experimental methodologies evolved, so has our understanding of vascular elastic fibers. No longer are they viewed as inert physical structures or scaffolds; rather elastic fibers are now known to be active participants in the dynamic signaling that occurs between cells and their environment. In this minireview, we will summarize recent findings of regional heterogeneity of elastin along the arterial tree and will highlight the impact of these structural differences on health and disease.

Changes to the arterial wall in elastin insufficiency

That elastin is indispensable for vascular development has been recognized for over 20 years. The global deletion of the mouse elastin gene (Eln^−/−) was shown to result in lethality of the mice within two days of birth(2). Without normal elastic lamellae, Eln^−/− mice suffered luminal occlusion due to migration of medial SMCs into the subendothelial space(2). Sections from the Eln^−/− aortas revealed ~2.5X more proliferating cell nuclear antigen (PNCA) positive cells and abnormal orientation of subendothelial SMCs, without obvious disruption of the endothelial layer. In mouse lines with some remaining elastin (the Eln^+/− with 50% elastin(3) and the Eln^−/−; ELN^+/+ that has 30% of the wild-type [WT] amount of elastin due to the presence of a human ELN bacterial artificial chromosome transgene(4)), the lumen was smaller, but not fully obliterated, and the wall showed increased medial SMC layers. Parallel findings have been reported in humans with elastin insufficiency(3)—in this case additional elastic lamellae are noted in arteries throughout the body. The vessels have reduced caliber(5) and display increased stiffness(6). In humans, in addition to the global arterial vasculopathy described above, some individuals also present with focal areas of stenosis, most commonly in the supravalvular aorta and in the pulmonary arteries. Histology reveals the formation of a neointima consisting of irregularly shaped SMCs interspersed between sparse and irregular elastic fibers(5, 7, 8), in addition to the medial thickening. Studies in the Eln^−/−; ELN^+/+ mice showed that decreased circumferential growth contributes greatly to the luminal obstruction phenotype(9), but none of the global insufficiency models have produced focal arterial stenosis to the extent seen in human elastin insufficiency.

Contributions of endothelial cells and smooth muscle cells to arterial elastic fibers

It is generally accepted that SMCs are the major elastin-producing cell in the vessel wall(10–12). While evidence for elastin production by ECs in culture has been reported(13–15), whether ECs contribute to the IEL in vivo has been controversial. Recently, Lin et al.(16) utilized the Cre/LoxP system to generate mice with specific deletion of Eln in either ECs or SMCs and assessed the effect of cell-specific elastin loss on arterial structure. Their findings point to complex differences in elastin deposition along the arterial tree. Unlike mice with global elastin deletion, mice lacking SMC-produced elastin lived until ~10–15 days old and manifested coarctation of the aorta, but did not completely occlude their aortas. The ascending aorta of these mice had a highly fragmented IEL, while the IEL of the descending thoracic aorta and all other smaller arteries examined appeared intact. This suggests that EC-derived elastin is sufficient for the production of an IEL along most of the vascular tree, while in the ascending aorta, SMC-derived elastin is necessary for IEL integrity. Whether this requirement is related to physical (higher pulse pressure in the ascending aorta) or biological (neural crest vs. mesenchymal origin of SMCs(17, 18)) differences along the arterial tree remains to be determined. Interestingly, when elastin was deleted from SMCs in this system, a neointima of SMC origin was observed in the ascending aorta(16). The neointima was reminiscent of what is seen in humans with elastin insufficiency and tended to occur in areas where the IEL was highly disrupted, indicating that the presence of an IEL provides a barrier for SMC proliferation and migration.

When elastin was deleted from ECs, elastic arteries (ascending aorta, descending thoracic aorta, common carotid artery, subclavian artery, iliac artery and internal thoracic artery) had an intact IEL, while smaller muscular (mesenteric artery, renal artery, inferior epigastric artery and femoral artery) possessed a thin and fragmented IEL(16). Taken together, these data suggest a gradient in which SMCs are primarily required for functional IEL production in proximal elastic arteries while ECs are predominantly responsible for the IEL in muscular/resistance arteries (Figure 1C). However, it is likely that both ECs and SMCs contribute to the IEL in both locations.

Heterogeneity in elastic fiber organization/structure along the arterial tree

In addition to variations in cellular contributions to elastic fibers along the arterial tree, evidence suggests that significant differences exist in elastic fiber structure within an arterial segment and along the arterial tree. In early scanning electron microscopy studies of the elastic fiber, Roach(19) noted that elastin in the IEL and luminal medial elastic lamellae of sheep thoracic aorta was organized as concentric fenestrated sheets, while it formed a fibrous network on the adventitial side. Similar observations have been made more recently in rodents using 3D confocal microscopy(20). When comparing canine IEL along different segments of the aorta, Song and Roach(21) noted that the IEL of the proximal descending thoracic aorta had a rough “felt-like” surface with uniformly distributed fenestrae or holes. The IEL became increasingly smooth with larger more randomly distributed fenestrae as it transitioned from the thoracic to the abdominal aorta and further to the iliac arteries (Figure 1D); the same was true for sheep, rabbits, and rodents(22–25).

Postnatal developmental changes in IEL fenestration size and density have been examined in conduit and muscular arteries. When comparing the carotid artery IEL of 3 week-old vs. 23 week-old rabbits, Wong and Langille(25) found that the fenestrae surface area increased from 11.3±0.7 μm² at 3 weeks to 61.2±5.5 μm² at 23 weeks. Similarly, the number of fenestrae increased with age. The density of fenestrae (i.e. fenestrae/mm² vessel) decreased by 26% with age however, suggesting that the increase in fenestrae number did not keep up with vessel growth. Similar findings were observed in the iliac and renal arteries, although there was a greater decrease in fenestrae density (up to 70% in iliac arteries). Similar fenestrae increases in size and number with age have also been shown in rodent arteries(24, 26).

As one of the functions of the arterial system is to accommodate physiological demands, partly by sensing hemodynamic stress, the effect of shear stress or blood flow on IEL remodeling has been examined. In both rabbits and rodents, reducing shear stress by ligating the external carotid and hence decreasing blood flow in the ipsilateral common carotid artery resulted in smaller IEL fenestrae size without a change in number(25, 27). While increasing shear stress (carotid artery contralateral to the ligation) resulted in increased IEL fenestrae size in rabbits(25); it did not affect the IEL in rats(27). The discrepancy in IEL changes in response to high shear stress in rabbits vs. rats could be related to the duration of high shear stress exposure (5 weeks for the rabbits vs. 1 week for the rats), extent of blood flow increase with carotid artery ligation, or species differences. In addition to growth and hemodynamic factors, IEL fenestrae incidence is also, at least in part, genetically determined as arteries of different normotensive rat strains (Brown Norway, Long Evans, Sprague Dawley and Wistar) have differing IEL fenestrae numbers(28, 29).

EC-SMC-IEL interactions – lessons learned from different pathologies

The IEL forms a physical barrier between ECs of the intima and SMCs of the media, preventing diffusion of substances across the arterial wall. Fenestrae reduce this barrier. Penn and Chisolm showed that EC injury increased transmural uptake of horse radish peroxidase from the blood to the media of rat aorta in vivo(30). This permeability is important for EC-SMC interaction or myoendothelial signaling via soluble mediators such as vasoactive substances or mitogens as well as through direct contact between the cells, via gap junctions(31–36). While a cause-effect relationship has not been fully demonstrated, abnormalities in IEL fenestrae and SMC-EC communication have been noted in different disease states.

Aneurysms —

The first indication for the presence of abnormalities in elastic fibers in aneurysmal disease came from Campbell and Roach in 1981. They examined human cerebral arteries, which contain a single layer of elastin, the IEL, and showed that the diameter and density of fenestrations were larger at the apex of arterial bifurcations, where aneurysms tended to develop(37). Changes in elastic fiber fenestrations have also been quantified in mouse models of aneurysmal disease(38, 39). In general, fenestrae increased in size and number in aneurysmal vessels, however, there were regional differences where the proximal aorta and convex regions of the aorta showed the greatest increase in fenestrae density(38). While an in depth discussion of aneurysm pathophysiology is beyond the scope of this review, the identification of genes involved in heritable forms of aneurysms has perhaps shed the greatest light on how perturbations in elastic fiber structure and/or SMC contractility leads to vessel wall weakness and aneurysm development(40, 41). In particular, disruption of the physical connection between elastin projections and proteins on the neighboring SMCs, the so-called “elastin contractile unit,” due to genetic alterations in the proteins involved, may lead to an inability of the artery, particularly the thoracic aorta, to withstand the high pulse pressure, increasing aneurysm risk(41). It is important to note that disruption of elastin-contractile unit is unlikely to be the inciting event for all aneurysm development as mutations in genes involved in transforming growth factor beta (TGFβ) signaling have also been shown to lead to thoracic aortic aneurysms(42–44). Furthermore, absence of elastin (hence lack of elastin-contractile unit) in the Eln^−/− mice led to aortic occlusion rather than dilation(2). These data point to the complexity of aneurysm pathophysiology and to the importance of cell-matrix interactions in maintaining arterial integrity.

Atherosclerosis —

Lipid accumulation in the arterial wall is an initial step in atherosclerotic plaque formation that starts early in life. Initially, lipid, particularly low density lipoprotein (LDL), accumulates in the subendothelial space where it is oxidized (oxLDL). OxLDL induces an inflammatory response in which peripheral monocytes and T cells are also recruited. This process stimulates SMCs to migrate from the media to the intima, where they proliferate and secrete ECM that contributes to atheroma formation (reviewed in (45)). Changes in elastic fiber structure are integral to the development of atherosclerotic plaques as migration of SMCs from the media to the intima requires the holes to be greater than 3–4 μm wide(46). Indeed, experimental models of hypercholesterolemia showed increased IEL fenestrae size in both elastic and muscular arteries, likely a result of elastolysis(47–49). Furthermore, increased plasma low density lipoprotein (LDL) resulted in medial LDL accumulation only when the IEL was damaged(50).

Hypertension —

Hypertensive remodeling of both resistance and elastic arteries has been described in animal models as well as in humans(51, 52). In small resistance arteries and arterioles, which play a major role in blood pressure regulation(53), remodeling leads to increased media-to-lumen ratio and is mediated by complex mechanisms involving growth, apoptosis, inflammation, fibrosis and chronic vasoconstriction(54). In large elastic arteries on the other hand, hypertension-associated remodeling involves collagen and fibronectin deposition and elastin fragmentation, leading to increased large artery stiffness(55, 56).

While it is clear that hypertension leads to vessel wall changes, evidence from genetic studies indicates that changes in elastin also affect blood pressure control. Global hemizygous loss of elastin in mice (Eln^+/−) resulted in hypertension(3). Studies examining elastic fibers in other genetic models of hypertension, such as the spontaneously hypertensive rat (SHR) model, showed that hypertension was associated with changes in the IEL of large and small arteries (stiffer elastin and smaller fenestrae) and these changes preceded arterial narrowing(57–61). Interestingly, the differences in IEL fenestrae size and number varied with age and vessel size as Sandow and colleagues showed that, at 3 weeks of age, SHR superior mesenteric arteries have larger hole width and hole density than the normotensive Wistar Kyoto Rats (WKY), however, at 28 weeks of age, hole size was similar but hole number was lower in SHR leading to reduced hole density(24).

In resistance arteries, where elastin content is low, cellular responsiveness to vasoactive substances and EC-SMC communication are thought play a very important role in determining vascular tone. Studies comparing several hypertensive and normotensive rodent models found no correlation between IEL fenestrae size and number and direct EC-SMC contact through myoendothelial gap junctions(24). Rather, changes in IEL fenestrae size and number were thought to alter sites of diffusion of vasoactive substances between ECs and SMCs leading to altered reactivity. Consistent with this, mesenteric artery reactivity studies in Eln^+/− mice, showed impaired endothelial-dependent vasorelaxation to acetylcholine and increased contractility to the vasoconstrictive substances, angiotensin II and phenylephrine(62). These data highlight the effect of structural changes in elastin on cellular functions. It is interesting to note that, unlike the Eln^+/− mice, mice with hemizygous SMC-specific deletion of elastin (SM-Eln^F/+) were not hypertensive despite having large artery stiffness(16). This may be related to normal elastin deposition by ECs in SM-Eln^F/+ resistance arteries(16). Ultrastructural and functional comparison of resistance arteries in the two models might prove insightful.

Conclusion

While often described as structurally amorphous, it is clear that elastin deposition and its contribution to organ function are highly complex, active processes with significant regional differences. Within the arterial system, SMCs were found to be important for IEL formation in the ascending aorta while ECs were mainly responsible for its deposition in resistance or muscular arteries. This regional heterogeneity extends beyond the molecule itself to other elastic fiber associated proteins(63). In addition to its structural heterogeneity, elastin serves varying roles along the arterial tree, which include maintaining vascular integrity to prevent neointimal formation and facilitating cell-cell communication by diffusion or direct contact through its fenestrae. Despite increasing complexity, understanding regional differences in elastin assembly and function during development and disease conditions will be important to develop new therapeutic strategies.

Key points:

1. Elastin is an important component of the arterial wall, providing elastic recoil and facilitating local cell communication and movement.

2. The internal elastic lamina is generated by predominantly by smooth muscle cells in the ascending aorta and endothelial cells in the more peripheral arteries, but contributions from both cell types are probably present in all locations. Fenestrations of the elastic lamina also vary in size and number along the tree.

3. Regional characteristics of elastic lamellae contribute to a range of inherited and acquired vascular conditions.