On the basis of strictly morphological characteristics, arterial anatomy is divided into three components. The tunica intima includes a single layer of endothelial cells lining the vessel lumen and the internal elastic lamina membrane. The tunica media comprises the muscular portion of the blood vessel, whereas the tunica adventitia includes the external elastic lamina, terminal nerve fibers, and surrounding connective tissue, which contains fibroblasts and tissue macrophages. Although this histological segregation of vascular components was of historical origin, important functional properties of the vascular wall have been uncovered that conform to this structural framework.
Before 1980, studies addressing control of vasomotor tone focused primarily on the tunica media. Pharmacological manipulation of vascular function involved examining the effects of compounds on vascular smooth muscle reactivity. The endothelium was largely considered a hemostatic barrier and the adventitia a support structure for the vessel. In the late 1970s it became evident that the endothelium had the potential to play an important role in vasomotor responses by virtue of endothelial localization of the production of prostacyclin, an arachidonic acid-derived vasodilator. However, prostacyclin was also found in denuded vessels, suggesting that vascular smooth muscle could also synthesize this compound (1). It was not until 1980 that Furchgott and Zawadzki (24) first described the necessary role of endothelium for vasodilation to acetylcholine in the rabbit aorta. In the subsequent 19 years, thousands of studies have been published that broaden our understanding of the role of the endothelium in a variety of physiological and pathophysiological vasomotor responses. During this time, the tunica adventitia continued to receive only minor attention as a potential contributor to vascular control, but recent evidence suggests that adventitial regulation of vascular function may be important in several circumstances. Figure1 summarizes potential mechanisms by which the adventitia influences vascular function. Those situations with the greatest potential for clinical application include cerebral and coronary vasomotor function, coronary collateral development, prevention and treatment of atherosclerosis and restenosis, and neural control of the circulation.
VASOMOTOR FUNCTION AND GENE TRANSFER
In the June 1999 issue of the American Journal of Physiology: Heart and Circulatory Physiology, Tsutsui and colleagues (75) provided new insights into the role of the adventitia in regulation of cerebral blood vessels. They demonstrated that genetically altered adventitial fibroblasts may serve as a surrogate for endothelial cells in the regulation of basilar artery tone. This study is the latest in a series of investigations by the same group designed to exploit the nitric oxide (NO)-producing capabilities of cells in the vascular adventitia to restore or enhance dilation to agonists that traditionally elicit endothelium-dependent relaxations (37, 47, 51).
Gene transfer has been used successfully to insert DNA encoding a variety of proteins into vascular tissue. Although much work has focused on transfection of endothelial cells, adventitial cells, particularly adventitial fibroblasts, are also susceptible to adenoviral infection (10, 11, 55). This may be particularly helpful in situations where there is direct access to the tunica adventitia, for example, in the cerebral and epicardial coronary circulations. Not only is access for vector delivery available (intracerebroventricular or intrapericardial approaches, respectively), but a critical problem with intraluminal viral delivery is obviated by this approach, namely, the need for sustained contact time between virus and tissue, which generally requires prolonged interruptions in tissue perfusion. In addition, the vascular inflammatory response to adenoviral vectors is reduced with adventitial transfection (66). Clinical applications of gene therapy in this manner might include treatment of cerebral vasospasm following subarachnoid hemorrhage or coronary vasospasm.
The effectiveness of this approach to gene transfer in cerebral vessels was first demonstrated by Ooboshi et al. (55), who showed that intracerebroventricular injection of an adenovirus vector containing the gene for β-galactosidase effectively transfected adventitial cells of large arteries in the subarachnoid space. These findings were extended in subsequent studies that demonstrated the efficacy of transfecting adventitial fibroblasts with recombinant endothelial NO synthase (eNOS) (52, 54, 75). Adventitial cells are structurally and biochemically suited for gene transfer of eNOS. Adventitial fibroblasts contain invaginations (10) similar to caveolae, important sarcolemmal sites in endothelial cells for depalmitoylation and activation of eNOS (60). Adventitial cells also contain caveolin-1, the structural protein that sequesters eNOS in the endothelium (65).
After gene transfer of eNOS to the adventitia of cerebral vessels, increased expression as well as enhanced relaxation of both normal (10,11, 52, 53) and atherosclerotic cerebral arteries (54) to NO-mediated, endothelium-dependent agonists are observed. This is associated with an approximately twofold increase in the level of cGMP in eNOS- but notlacZ-transfected vessels (10). Even after endothelial denudation, dilation to A-23187 (52) or to bradykinin (75) is observed in transfected vessel segments. Thus transfection of eNOS to adventitial cells can compensate for the loss of endothelium-derived NO during pharmacological stimuli. The therapeutic implications are important for conditions such as hypercholesterolemia, hypertension, diabetes, and subarachnoid hemorrhage, where endothelial NOS-dependent vasorelaxation is impaired (34). The feasibility of gene transfer for improving vascular function was confirmed in an animal model of subarachnoid hemorrhage, where intracisternal injections of recombinant adenovirus transfected cerebral vessels (48). In a similar model, adenoviral transfection with eNOS improved the reduced relaxation to bradykinin (51). The rate and duration of transgene expression can be varied depending on the vector used (12).
Vasospasm is a serious clinical problem from 2 to 14 days following neurosurgery. Instillation of vectors containing genes for NOS at the time of surgery may reduce complications of focal vasoconstriction because the time course of protein expression is similar to that of vasospasm.
The vast majority of the morbidity from coronary artery disease (CAD) stems from problems in the epicardial conduit arteries. Access to conduit coronary vessels has been limited to endovascular approaches, which suffer from short durations of treatment exposure and the potential for adverse systemic effects from administration of large doses of vasoactive agents. Delivery of agents into or onto the vascular wall from an endovascular approach [e.g., infiltrating catheter devices (72, 77) or impregnated stents (25)] obviates these concerns, but the potential therapeutic benefit is limited to a short vascular segment.
The epicardial coronary circulation is in continuous contact with the overlying pericardial fluid. This has prompted a pericardial approach to pharmacological or gene therapy that would allow sustained application of drug or vectors to large segments of the epicardial circulation. Similar to the case with cerebrospinal application of viral vectors, transfection of epicardial coronary arteries is of great potential clinical benefit. If transfection with eNOS can restore the loss of NO production associated with coronary artery disease or its risk factors (8, 13, 32, 59), regression or prevention of smooth muscle proliferation may be possible. Adventitial transfection with superoxide dismutase may improve vasodilator function by eliminating the quenching of endothelial or neuronally released NO by the superoxide anion (78). An exciting potential clinical application involves the instillation of angiogenic factors into the pericardium of patients with ischemic coronary disease to stimulate the development of epicardial collaterals, thereby reducing ischemic stress. This approach will have to await better differentiation of the growth factors promoting collateral angiogenesis from those involved in vascular proliferation associated with atherosclerosis.
The pericardial approach for gene therapy has been tested by Lamping et al. (40) in dogs using an adenoviral vector containing the gene for β-galactosidase. Only modest gene expression was identified in vascular tissue, but significant expression was observed in the overlying visceral and parietal pericardium. It may be possible to take advantage of the high transfection rate in pericardial cells for clinical purposes. Production of recombinant protein by pericardium may be sufficient to influence vascular reactivity or to reduce smooth muscle proliferation and narrowing of epicardial vessels. It is known that prostaglandins secreted by the pericardium exert an influence on epicardial structures (46). New advances in gene therapy may increase transfection efficiency or identify other methods to reduce physical barriers to transfection. In the study by Lamping et al. (40), ventricular transgene activity was augmented by treatment with the inflammatory stimulus doxycycline.
A limitation to clinical application of pericardial gene therapy is that introduction of catheters into the pericardial space in patients without significant effusion is associated with an unacceptable complication rate (43). Recent advances in image-guided needle insertions and newer catheter systems show great promise in reducing this risk, opening the way for therapeutic pericardial interventions (43, 68). Clinical trials in the United States utilizing the transpericardial approach are scheduled within the next year (Dr. Keith March, Indiana University; personal communication).
ATHEROSCLEROSIS AND ADVENTITIA
Focusing gene therapy on adventitial cells may have broader impact than simply replacing substances lost as a result of endothelial dysfunction. Substantial data suggest that adventitial processes are involved in the genesis of atherosclerosis. Removal of the adventitia from large arteries results in intimal and medial hyperplasia characteristic of early atherosclerotic lesions (4). This proliferative process is reversible with a time course that correlates with resolution of adventitial inflammation. The causal relationship between adventitial inflammation and neointimal proliferation has been examined in pigs and rabbits (4, 67, 69). The hypothesis has been tested that the cellular elements infiltrating the tunica media and intima in atherosclerosis are derived from adventitial myofibroblasts (67, 69). 5-Bromo-2-deoxyuridine labeling of proliferating cells in the vascular wall revealed that staining was greatest in the arterial adventitia 3 days after balloon injury (67). However, 4 days later, proliferating cells were observed mostly in the neointima. By adjusting the time of application of 5-bromo-2-deoxyuridine, it was determined that the adventitia was the source of proliferating cells found in the intima atday 7 (67, 80).
Not all studies conclude that the adventitia plays an important role in the development of atherosclerosis. Examining the response of the rabbit iliac artery to balloon injury, Le Feuvre et al. (41) found relatively few proliferating cells in the adventitia. Most were observed in the media and intima. Macrophages were not a major component of histological changes postangioplasty, at least up to 28 days. Atherosclerosis following cardiac transplantation may also be independent of adventitial disease. Immunohistochemical staining suggests that the tunica media is the source of new vascular smooth muscle cells that migrate to both the adventitia and intima (2).
A devastating consequence of coronary atherosclerosis in the coronary circulation involves acute plaque rupture and myocardial infarction. Mast cells in the adventitia may play an important role in this process. Adventitial mast cells are present in higher concentrations in infarct-related coronary arteries compared with control vessels (38). The proportion of these mast cells that are degranulated is greatest in segments with plaque rupture. Released products include histamine, leukotrienes, and prostaglandins, which may contribute to focal vasospasm and the ensuing ischemic damage.
Gene therapy applied to the adventitia, as described in the paper by Tsutsui et al. (75), may be helpful in treating early vascular changes associated with atherosclerosis. Support for this idea stems from the observation that adventitial delivery of antisense to c-myb oligonucleotide suppresses intimal proliferation in rat carotid arteries (70). In separate studies, transfection of atherosclerotic carotid arteries from Watanabe heritable hyperlipidemic rabbits with an adenovirus containing the transgene for eNOS improved vasodilator responses to acetylcholine (54). Histochemical staining revealed transgene expression in both the adventitia and endothelium, although it was not determined which transfection site was responsible for improvement in function.
Direct adventitial application of heparin or a congener that has antiproliferative but not anticoagulant properties inhibits rat carotid artery smooth muscle neoplasia following endothelial injury (20, 50). This perivascular drug delivery avoids bleeding problems associated with systemic administration of heparin. This approach has also been used for local vascular treatment with high doses of calcium antagonists to reduce neointimal proliferation following carotid balloon injury in rats (30). Perivascular preventive approaches may be important following regression of atherosclerosis. In animal models of regression, inflammatory cellular elements disappear from intimal and medial layers of the vessel but remain in the adventitia (57). Vasomotor responses to activation of these adventitial leukocytes persist following regression of atherosclerosis.
Important from a clinical standpoint, intimal proliferation is attenuated by endovascular irradiation. Histological studies indicate that the radiation acts to reduce adventitial cell proliferation and correspondingly prevents intimal lesion development (81). The relatively frequent and rapid remodeling and luminal narrowing that is common after angioplasty could result from adventitial disruption and may be reduced by radiation.
It has been suggested that the mechanism of adventitial induction of neointimal proliferation involves obstruction of the vasa vasorum with subsequent vascular wall hypoxia (3, 76). External stenting of vein grafts is associated with interruption of adventitial perfusion and resultant neointimal formation. This can be prevented by using porous external stents, which allow microangiogenesis, thereby minimizing graft hypoxia (31). After the adventitia of rabbit carotid arteries is stripped, local inflammation and development of intimal lesions are observed. This is associated with heightened expression of platelet-derived growth factor, occurring first in the adventitial layer (67). Intimal lesions regress within 28 days, concurrent with revascularization of the adventitia. Interventions used to restore the adventitial vasa vasorum have been successful in inhibiting the development of intimal hyperplasia (4). These data suggest that disruption of the vasa vasorum is closely linked to the development of intimal plaque through stimulation and migration of adventitial fibroblasts.
Other mechanisms of adventitia-mediated intimal proliferation may also be operant. For example, in rats, chronic systemic dilator treatment (4 days) with prazosin leads to an eightfold increase in the proliferation of arteriolar adventitial fibroblasts without trauma or disruption of the vessel wall (62). Adventitial fibroblasts are also stimulated by increases in arterial pressure. DNA synthesis is increased 600% in adventitial fibroblasts proximal to an aortic obstruction in a rat model of hypertension (9).
To the extent that these observations apply to humans, prophylactic epicardial therapy targeted to reduce adventitial hypoxia may be useful in patients following coronary bypass grafting or during cardiac transplantation, where the surgical procedure affords ready exposure to the surface of the heart. If the risk of coronary puncture is reduced, pericardial instillation of antiproliferative agents may become a useful adjunct to angioplasty or stent placement to prevent early restenosis resulting from adventitial stretch or damage.
Human atherosclerosis is characterized by bony mineralization of vascular lesions. Current radiographic screening for CAD relies on the detection of small amounts of calcification in coronary plaques (79). This calcification process produces noncompliant conduit arteries and may increase the complications of interventions designed to treat focal stenoses (e.g., balloon angioplasty and stent placement). Calcified and stiff coronary arteries are more prevalent in patients with diabetes and contribute to the increased mortality (42).
Recent evidence suggests that the process of calcification is highly organized, involving the production of lamellar bone by osteoblasts (6). The process may be initiated by induction of transcription factors, which stimulate osteoblast formation, as has been observed in low-density lipoprotein receptor-deficient mice fed a high-fat diet (74). Pertinent to this editorial, the spatial pattern of gene expression related to vascular calcification suggests that osteoprogenitor cells originate in the adventitia, possibly from adiposites, with little evidence for involvement of the tunica media and intima (74). A case can be made for the tunica adventitia playing a key role in the development of atherosclerosis from early intimal hyperplasia to calcification of chronic vascular lesions. Thus therapies that focus on the adventitial contribution to the development of atherosclerosis and restenosis may have strong clinical promise.
NEURAL REGULATION OF VASCULAR FUNCTION
The vascular innervation is an obvious example of how vasomotor function may be influenced by adventitial elements. In the coronary circulation, sympathetic and parasympathetic nerve terminals are located in the adventitial-medial border of epicardial conduit vessels (15, 17, 19), and their role in regulating tissue perfusion and conduit vessel diameter is well recognized (22, 29, 44). Neurotransmitters released from sympathetic nerve endings such as norepinephrine and neuropeptide Y diffuse into the tunica media and act on receptors on the vascular smooth muscle to cause constriction. Other neurally released substances including substance P, acetylcholine, and calcitonin gene-related peptide produce vasodilation (14, 28,64). NO may be released directly from sympathetic nerve terminals or from endothelial cells activated by substances released from nerve endings in the adventitia (27, 33).
An example of this adventitial-endothelial-medial paracrine pathway for vasodilation is the parasympathetic innervation of the coronary vasculature (7, 23). Acetylcholine, released from parasympathetic nerves in the adventitia, diffuses through the medial layer of the vascular wall and activates muscarinic receptors on the endothelium to release NO in an N G-nitro-l-arginine methyl ester-inhibitable manner (7). NO then relaxes the underlying vascular smooth muscle as the final process in cholinergic vasodilation. This physiological response indicates that endothelium-dependent dilator agents may be able to traverse the vascular media and elicit responses even when applied to the adventitial surface. This concept is confirmed by observing endothelium-dependent dilation after application of acetylcholine to the adventitial surface of a conduit dog artery (73). The physiological importance of cholinergic innervation derives from several reflex pathways (e.g., chemoreflex, baroreflex, Bezold-Jarisch reflex), which modulate coronary and peripheral vasomotor tone through parasympathetic activation. In disease states such as pacing-induced heart failure, Bezold-Jarisch and chemoreflex coronary vasodilation are blunted due to endothelial dysfunction (84). Because even severe epicardial stenoses usually have a compliant segment that can be modulated by vasomotor influences, impaired parasympathetic vasodilation could increase the propensity for vasospasm and ischemia in patients with CAD (56). Efforts to improve endothelial function [e.g., through exercise (56)] may restore normal reflex parasympathetic responses. Moreover, gene therapy with eNOS (to increase NO production) or superoxide dismutase (to reduce superoxide anion) via the pericardial route may restore dilator responses to parasympathetic activation in diseased arteries, reducing the risk of epicardial coronary obstruction and thrombus formation.
The direct coronary vascular effects of sympathetic activation include α1- and α2-adrenergic vasoconstriction and α2-adrenergic activation of the endothelium, with release of the vasodilator NO. In addition, constitutively produced NO from the neuronal isoform of NOS in some tissues may be released from sympathetic nerve terminals to elicit vasodilation (16). Disease states including diabetes and atherosclerosis, where increased levels of superoxide anion quench NO (5, 61), or application of NOS inhibitors (27) augment reflex sympathetic coronary vasoconstriction (49, 82), with potentially detrimental effects on coronary flow in the presence of obstructive atherosclerotic disease. Therapeutic approaches directed at the adventitia have the capacity to abate the adverse autonomic effects on coronary vasomotion in patients with CAD.
Recent preliminary observations suggest that adventitial neural elements can compensate for the loss of endothelial production of NO. Lamping et al. (39) showed that in murine coronary arteries, endothelial denudation abolishes dilation to acetylcholine. However, in heterozygotic eNOS knockout mice, dilation to acetylcholine was present and inhibited by the arginine analog nitro-l-arginine (39). An antagonist specific for neuronal NOS blocked dilation in eNOS knockout mice. These data suggest that NO released from nerve terminals in the coronary adventitia compensates for the absence of endothelial production of NO in this model. Future studies are needed to determine whether this adventitia-mediated improvement in dilator function operates in diseases associated with endothelial dysfunction.
VASCULAR INFLAMMATION AND OXIDANT STRESS
During systemic inflammatory states, marked elevations in NO are observed, contributing to peripheral vasodilation and cardiac dysfunction (71). The source of NO in these cases is from an inducible isoform of NOS (iNOS) known to be present in both vascular smooth muscle and endothelial cells (83). Using in situ hybridization and immunohistochemistry, Zhang et al. (83) have recently demonstrated iNOS in aortic adventitial fibroblasts of rats treated with lipopolysaccharide. Using a similar model, Kleschyov et al. (35) determined that the amount of NO produced by adventitial cells in response to lipopolysaccharide is 3.5-fold greater than produced by adjacent medial smooth muscle cells. Because inflammation of the adventitia is an early component of the atherosclerotic vascular proliferation, adventitial production of NO may serve to modulate the atherosclerotic process.
Many chronic disease states are characterized by enhanced production of reactive oxygen species. Superoxide anion can inactivate NO, thereby impairing endothelium-dependent vasodilation. Superoxide anion production is enhanced throughout the vessel wall in animal models of hypercholesterolemia (45). Recent evidence suggests that even in nondiseased rats, superoxide anion production occurs endogenously in aortic tissue and is derived primarily from NADPH oxidase in the adventitia (58, 78). Superoxide anion from this source inactivates NO as measured by relaxation responses in vitro, because dilation to NO but not nitroprusside is reduced in vessels where the adventitia faces outward versus inward. Superoxide dismutase blocks this difference in dilation due to vessel orientation (78).
The pathophysiological implications of adventitia-produced superoxide anion relate to vessel dysfunction in disease states where excess superoxide anion is formed (26, 63). In an animal model of hypertension, superoxide anion production is elevated by up to 15-fold, with the majority arising from adventitial fibroblasts (18). The excess superoxide anion is causally related to spontaneous oscillations in vessel tone, possibly contributing to the hypertension. Of potentially greater importance, quenching of NO by superoxide anion results in the production of another highly reactive and relatively long-lived radical species, peroxynitrite. In atherosclerosis, vascular inflammation leading to enhanced production of NO through activation of iNOS, together with enhanced oxidative stress, could produce high local concentrations of peroxynitrite with resultant impairment of endothelial function and vascular responsiveness (21, 36). Therapy targeted at reducing the amount of peroxynitrite formed in the arterial adventitia may prevent many of the local complications associated with atherosclerosis.
In summary, the adventitia plays more than a structural role in vascular function. Adventitial fibroblasts may contribute to the subintimal proliferation that precedes atherosclerosis. Vascular calcification, characteristic of chronic atherosclerotic lesions, originates from cells in the adventitia. Neurotransmitters released from sympathetic and vagal efferent fibers regulate vasomotor tone and can provide compensatory dilation when endothelial function is reduced. A challenge for the future will be to modulate vascular proliferation (neointimal hyperplasia as well as collateral growth) and vasomotor function using periadventitial delivery systems (pericardial catheters, intravascular delivery, intracerebroventricular application) to apply genes or pharmacological agents. Recent progress indicates that when equipped with the appropriate enzyme systems, through viral transfection, adventitial fibroblasts are capable of transducing signals from pharmacological stimuli to generate NO. Better understanding of the role of the adventitia in regulating vascular function will provide additional tools with which to modify the development and consequences of disease.
The expert advice and comments provided by Dr. Donald D. Heistad are acknowledged.
Address reprint requests to D. D. Gutterman, Medical College of Wisconsin, Cardiovascular Research Center, 8701 Watertown Plank Rd., Milwaukee, WI 53226-0509 (E-mail:).
This editorial was supported by the Cora And John H. Davis Foundation and by grants of the American Heart Association, National Institutes of Health, and the Veterans Administration Medical Center.
- Copyright © 1999 the American Physiological Society