Lung (1990) Suppl: 35-42
Lung
©
Springer-Verlag
New York Inc. 1990
Endothelium-derived Relaxing Factors and the Human Pulmonary Circulation D. McCormack Victoria Hospital, London, Ontario, Canada
Abstract. The vasodilator effect of acetylcholine on the pulmonary circulation was first described over 30 years ago, however, the mechanism remained unknown until Furchgott described the endothelium-dependent relaxation of certain vasodilators. It was not until 1987 that endotheliumderived relaxant factor (EDRF) was demonstrated to dilate human pulmonary arteries in vitro. Despite this work, the physiologic role of EDRF in the pulmonary circulation is not known. It has been suggested that hypoxiainduced inhibition of EDRF action or release from pulmonary artery endothelial cells may have a role in hypoxic pulmonary vasoconstriction (HPV) but present evidence suggests that loss of EDRF activity is not directly involved in the phenomenon of HPV. It is more likely that EDRF is released from pulmonary artery endothelial cells during hypoxia and this released EDRF then modulates HPV. If EDRF does modulate HPV in vivo then the role of EDRF in the altered HPV found in disease merits attention. It is known that in disease states such as acute lung injury and pneumonia there is loss or attenuation of HPV which inevitably leads to increased V/Q mismatch and hypoxemia. Whether this attenuation of HPV is due to release of an endogenous vasodilator such as EDRF is presently being investigated. Additionally, there is in vitro evidence that loss of EDRF activity may be important in the genesis of pulmonary hypertension such as found in severe cystic fibrosis. During the next decade the role of EDRF in the human pulmonary circulation in both health and disease will undoubtedly be elucidated. Key words: Pulmonary artery--Endothelium-derived relevant factor--Hypoxic pulmonary vasoconstriction--Pulmonary hypertension. Offprint requests to: Dr. D. McCormack, Victoria Hospital, 375 South Street, London CDNOntario, Canada.
36
D. McCormack
Discussing the role of EDRF in the human pulmonary circulation is at once both stimulating and difficult. Stimulating because of the interesting potential this compound possesses as an endogenous vasodilator in health and disease, and difficult because so little is known. Much work has been done concerning EDRF and the pulmonary circulation in animals but very little has been done in humans. There are obvious reasons as to why this is so, first, the pulmonary circulation is difficult to study noninvasively; second, removal or damage of the endothelium experimentally usually results in acute lung injury and pulmonary edema; and thirdly, the general enthusiasm of clinicians and scientists regarding pulmonary hypertension is less than that for systemic hypertension. Additionally, human tissue is difficult to obtain to work with in the laboratory and even when obtained, as many of us can attest to, it is exceedingly more difficult to handle and experiment with than animal tissue. With these points in mind this article will be divided into two main parts: the first part will deal with the known and possible role of EDRF in control of the pulmonary circulation in health, and the second part will deal with the potential role of EDRF in some of the disease states that concern the pulmonary vasculature.
EDRF and the Pulmonary Circulation in Health
The first paper to describe the effects of acetylcholine (ACh) on the pulmonary circulation in vivo was in 1957 by a group of physicians and surgeons from New York [1]. They described a fall both in pulmonary artery pressure and pulmonary vascular resistance in patients after ACh was infused. At the time, the explanation for this phenomenon was not clear although it was suggested that ACh was somehow acting directly on the pulmonary vascular smooth muscle. Since then we have come a long way. The major breakthrough was in 1980 when Furchgott and colleagues reported that ACh caused relaxation of precontracted rabbit aortic strips only when endothelial cells were present [2] and the term "endothelium-derived relaxant factor" was coined. Not long after this the effect of ACh was being investigated in a variety of different vascular beds from many different species. Additionally, it is now known that vasodilators in both the pulmonary and systemic circulation can be divided into endothelium-dependent and endothelium-independent vasodilators. Furchgott suggested that EDRF release after ACh was the result of stimulation of a muscarinic receptor. We have demonstrated that in the pulmonary vasculature of the rat the muscarinic receptor involved is of the M3 subtype [3]. This was done by looking at the effect of the muscarinic antagonists 4-DAMP, methoctramine, and pirenzipine on the ACh-induced, endothelium-dependent vasodilation of isolated rat pulmonary artery rings. Despite this explosion of knowledge regarding EDRF in both the pulmonary and systemic circulation there has been an astonishing lack of reported information concerning EDRF and the humau pulmonary circulation. Indeed it was not until 1987 that Greenberg and co-workers [4] published data showing that it was possible to mechanically remove the endothelium from normal
Endothelium-derived RelaxingFactors
37
human pulmonary arteries and they demonstrated in vitro that human vessels have endothelium-dependent vasodilation in response to ACh and ATP but not to vasoactive intestinal peptide. As physicians we know that it is one thing to describe a phenomenon in an organ bath, however, quite another to relate this to the intact patient. Unfortunately there has been little work evaluating the role of the endothelium in the control of pulmonary vascular tone in the human. Despite the lack of hard experimental data and the difficulty in extrapolating results from animal studies directly to humans there is room for hypothesizing on the role of EDRF in the human pulmonary circulation. Although the physiologic role of EDRF is not yet fully known, the fact that it is known to have a very short half-life suggests a localized site of action on subjacent vascular smooth muscle, and this has been shown experimentally. Hemoglobin bound to haptoglobin in plasma rapidly inactivates EDRF providing a mechanism to limit the downstream action within the vascular lumen in vivo. This would allow for precise local regulation of the local blood flow, something that is essential in the pulmonary circulation. The pulmonary vasculature is unique in two ways: (1) it normally has a very low resting tone and as a result the resting pulmonary arterial pressure is much lower than systemic arterial pressure and (2) it contracts in response to hypoxia. This hypoxic pulmonary vasoconstriction makes teleologic sense and is felt to be the most important mechanism in the matching of perfusion to ventilation in the lung. The role of the endothelium and EDRF in the maintenance of low pulmonary vascular tone and HPV is not clear but several important points have been established. First, as regards the possible role of EDRF in HPV, in 1978 Dr. Weir proposed that the phenomenon of hypoxic pulmonary vasoconstriction could be explained ff the pulmonary vasodilation present during normoxia were actively maintained by a vasodilator substance and hypoxia then lead to a decrease in this vasodilation with resultant vasoconstriction [5]. The question was therefore whether hypoxia resulted in the release of a pulmonary vasoconstrictor substance or if hypoxia inhibited the release or action of a vasodilator substance. Dr. Weir at that time proposed that bradykinin may have a role as this vasodilator, however, interest has lately focused on a possible role for EDRF in this capacity. To support this hypothesis of HPV being due to a loss of active normoxic vasodilation by EDRF it must be demonstrated that there is a basal release of EDRF from vessels. Griffith et al. [6] have done this using an experimental setup in which a bioassay strip of coronary artery is precontracted with 5-HT and then superfused. They demonstrated that EDRF is both continuously released from aortic endothelial cells in the basal state and this release can be stimulated by ACh. However, not all systemic vessels appear to have identical basal EDRF production [7] and we have performed experiments using columns of pulmonary artery endothelial cells grown on beads and perfused with Kreb's solution to evaluate the basal production of EDRF. We found that although the endothe-
38
D. McCormack
lial cells could be stimulated to produce EDRF and relax the test strip of endothelial-denuded pulmonary artery, there was no basal production o f EDRF by the cultured pulmonary artery endothelial cells in this system. Therefore, there may be reason to suspect that systemic and pulmonary artery endothelial cells do not have the same basal production of EDRF. Thus there is reason to doubt that there is sufficient basal production of EDRF by pulmonary artery endothelial cells to account for inhibition of EDRF leading to HPV. What then is the role of EDRF in the important phenomenon of HPV? To answer this we need to examine the known modulating effects of EDRF in the pulmonary circulation. Yamaguchi and co-workers have demonstrated that EDRF has a significant role in modulating the contractile sensitivity of isolated pulmonary vessels to KC1, angiotensin II and noradrenaline [8]. EDRF is thus important not only in endothelial-dependent pulmonary vasodilation but also as a modulator of pulmonary vasoconstriction. It is likely that the modulation of vasoconstriction by the endothelium and EDRF varies with agonist and species as well as with arterial segment. What then is the role of EDRF in modulating HPV? Rodman et al. have reported that isolated rat pulmonary artery rings in vitro contract as the pO2 of the surrounding bath is lowered [9]. They looked at the effect of methylene blue, an inhibitor of guanylate cyclase, in a dose of 10-5 M which has been shown to inhibit approximately 75% of an EDRF-stimulated relaxation. They found that the contraction caused by hypoxia was augmented by methylene blue between 10% and 3% Oz. This is consistent with the hypothesis that EDRF is released during hypoxia and modulates HPV such that inhibition of EDRF leads to an augmentation of the resoonse. There are problems with using only the model of isolated pulmonary arteries as a model for HPV as usually for in vitro experiments, conduit arteries are used which are almost certainly not the arteries involved in HPV in vivo. Brashers and co-workers therefore examined the effect of EDRF inhibition using the isolated perfused and ventilated lung of the rat [10]. They examined the effect of various chemically dissimilar and structurally unrelated pharmacologic agents such as ETYA, NDGA, and hydroquinone, all of which are known to block or attenuate the actions of EDRF. They found that all three antagonists augmented the hypoxic pressor response. These studies suggest that EDRF maybe an important modulator of HPV. Two points merit emphasis from these data. First, the administration of the EDRF antagonists did not affect the baseline pressures on room air ventilation, suggesting that in this isolated lung system there is no significant basal EDRF production. This point is consistent with the conclusion that the augmentation of I-IPV is not due to a decrease in basal EDRF activity. Second, the augmentation of HPV b y ETYA, NDGA, and hydroquinone implies that EDRF activity and/or release is increased in hypoxia. In a separate but similar study, Mazmanian and co-workers found that methylene blue and hemoglobin, both inhibitors of E D R F , augmented both the hypoxic and angiotensin II pressor response [11]. These studies provide some evidence that EDRF may have an important
Endothelium-derived Relaxing Factors
39
role in the modulation of the hypoxic pressor response. It is possible that by blocking EDRF-induced increase in cGMP levels there is a potentiation of hypoxic vasoconstriction, since the relaxant factor EDRF no longer modulates the contractile response. Furthermore, from the data presented so far, it would appear that EDRF is not tonically released to actively maintain low pulmonary vascular tone.
EDRF and the Pulmonary Circulation in Disease Having suggested some possible roles of EDRF in the pulmonary circulation in health what can be said about EDRF in the pulmonary circulation in disease? First, in pulmonary hypertensive disease there may be a failure of release or action of EDRF resulting in an increased resistance in the pulmonary circulation. This suggestion comes from experiments which have looked at isolated arterial rings from spontaneously hypertensive animals. In these studies it has been found that endothelium-independent vasodilators relax thoracic aortic rings from normal rats to the same extent as those from hypertensive animals [12]. By contrast, endothelium-dependent vasodilators do not effect the same relaxation in arteries from hypertensive animals compared arteries from normals. In addition to these in vitro findings demonstrating decreased action of EDRF in vessels from hypertensive animals, similar findings have been demonstrated in vivo using the cerebral circulation and looking at the vessels using a video image device [13]. Again, these investigators found that relaxation caused by the endothelium-independent vasodilator adenosine was similar in hypertensive and normal animals, however, the relaxation caused b y a n endothelial-dependent vasodilator such as ACh was markedly diminished in the cerebral circulation from spontaneously hypertensive animals. Thus, there has been speculation that this decrease in EDRF activity in hypertensive states may be important in the pathogenesis of pulmonary hypertension. The only published study so far addressing this in the human pulmonary circulation was by the group from Cambridge who used isolated pulmonary arteries from patients with cystic fibrosis who had pulmonary hypertension and were undergoing heart-lung transplantation. They demonstrated that vessels from these patients had diminished relaxation to acetylcholine and ADP [14]. It is not clear as to whether this diminished relaxation is the cause of or the result of the pulmonary hypertension however the authors supported the latter hypothesis. Obviously, further studies need to be done in vivo and in vitro in patients with pulmonary hypertension to answer this question and these studies have already started. Removal of endothelium from intrapulmonary arteries can transform the relaxations of ACh and bradykinin into pulmonary vasoconstrictors. It is possible that damage in vivo to arterial and arteriolar endothelial cells or destruction of endothelial cells by a variety of mechanisms such as release of proteolytic enzymes, leukocyte-endothelial and platelet-endothelial interactions leading to
40
D. M c C o r m a c k
the release of lysosomal enzymes, etc. may lead to pulmonary hypertension found in ADRS. So what about the role of the endothelium and EDRF in acute lung injury? It has become clear in the last few years that in animal models of acute lung injury there is abnormal pulmonary vascular reactivity. Rounds et al. [15] treated rats with intraperitoneal o~-naphthylthiourea (ANTU) resulting in a model of endothelial cell damage and an acute permeability pulmonary edema much like ARDS. Then, using an isolated, perfused lung setup, they evaluated the vascular reactivity of the pulmonary circulation to bolus infusions of angiotensin II. They found that in the lungs from rats treated with A N T U the pressor response to angiotensin II was greater compared to the pressor response to the same agent in the normal lungs treated with the ANTU vehicle Tween. This suggests that abnormal endothelial cell function, and perhaps abnormal EDRF release or activity is associated with enhanced pulmonary vascular reactivity, similar to the hypothesis regarding EDRF and hypoxic pulmonary vasoconstriction discussed earlier. However, it is important to note that not all models of acute lung injury are similar in so far as pulmonary vasoreactivity is concerned. As noted above, in the ANTU-treated model of acute lung injury there is increased vasoreactivity to angiotensin II. However, others have shown that there is decreased pulmonary vasoreactivity in different models of lung injury and this may have important clinical implications. For example, when rats are treated with intratracheal bleomycin there is an acute permeability pulmonary edema maximal about 48 h after the insult. At this time lungs from these animals can be removed, isolated, and ventilated and perfused. By ventilating the isolated lungs with hypoxic gas mixtures the pressor response to hypoxia may be evaluated. We have performed such experiments and have demonstrated that the injured lungs have virtual abolishment of HPV compared to lungs from saline-treated animals. Similar findings to ours have been noted in lungs from animals exposed to 100% 02 for 52 h to induce hyperoxic lung injury, another model of endothelial cell damage. Thus, in a least two models of acute lung injury there is a diminished pressor response in the pulmonary circulation to hypoxia. This pulmonary vascular hyporeactivity is not found only in acute lung injury. Graham et al. used a rat model of chronic Pseudomonas pneumonia and in vivo measured the hypoxic pressor response of these animals. They demonstrated that in control and infected animals there was a similar resting P A P but that the pressor response to hypoxia was diminished in the infected animals compared to controls [16]. What is the mechanism of this decreased pulmonary pressor response to hypoxia in these models of acute lung injury, lung inflammation and endothelial damage? The answer is not known however abnormal control of the release of EDRF may play a role. It is possible, for example, that when endothelial cells are damaged there may be loss of the normal regulation of EDRF stimulation and release and as a result, increased basal production and release of this compound. The vascular hyporeactivity in sepsis may be secondary to en-
Endothelium-derived Relaxing Factors
41
hanced endothelial cell EDRF release caused by bacterial endotoxins or by mediators generated during sepsis that either interact with processes coupled to EDRF release or cause endothelial cell damage. This would result in abnormal and inappropriate vasodilation in the area of endothelial and lung injury. One could hypothesize that in an area of pneumonia if there was abnormal EDRF release and consequent increased blood flow to the area this would lead to increased shunt and hypoxemia. Whether this does indeed happen is not known and is presently being investigated. Does the unique physiologic action of EDRF have any implications for the treatment of pulmonary hypertensive states? The major problem associated with pulmonary vasodilator therapy is the consequent systemic hypotension that seems to invariably accompany therapy. The extremely short half-life of EDRF would suggest that it may be possible to selectively vasodilate the pulmonary circulation without effect on the systemic circulation. However, until the exact nature and pharmacologic properties of EDRF are known this possibility cannot be fully explored.
References 1. Fritts Jr. HW, Harris P, Clauss RH, Odell JE, Cournand A (1958) The effect of acetylcholine on the human pulmonary circulation under normal and hypoxic conditions. J Clin Invest 37(1):99-1 t0 2. Furchgott RF, Zawadzki JV (1980) The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 288:373-376 3. McCormack DG, MAk JC, Minnette P, Barnes PJ (1989) Muscarinic receptor subtypes mediating vasodilation in the pulmonary artery. Eur J Pharmacol 158:293-297 4. Greenberg B, Rhoden K, Barnes PJ (1987) Endothelium-dependent relaxation of human pulmonary arteries. Am J Physiol 252 (Heart Circ Physiol 21):H434-438 5. Weir EK (1978) Does normoxic pulmonary vasodilatation rather than hypoxic vasoconstriction account for the pulmonary pressor response to hypoxia? Lancet 8062:476-477 6. Griffith TM, Edwards DH, Collins P, Lewis MJ, Henderson AH (1985) Endothelium derived relaxant factor. J R Coil Phys Lond 19(2):74-79 7. Collins P, Chappell SP, Griffith TM, Lewis MJ, Henderson AH (1986) Differences in basal endothelium-derived relaxing factor activity in different artery types. J Cardiovasc Pharmacol 8:1158-1162 8. Yamaguchi T, Rodman D, O'Brien R, McMurtry I (1989) Modulation of pulmonary artery contraction by endotheliurn-derived relaxing factor. Eur J Pharmacol 161:259-262 9. Rodman D, Yamaguchi T, O'Brien R, McMurtry I (1988) Methylene blue enhances hypoxic contraction in isolated rat pulmonary arteries. Chest 93(3):93S-94S 10. Brashers VL, Peach MJ, Rose Jr. CE (1988) Augmentation of hypoxic pulmonary vasoconstriction in the isolated perfused rat lung by in vitro antagonists of endothelium-dependent relaxation. J Clin Invest 82:1495-t502 11. Mazmardan GM, Baudet B, Brink C, Cerrina J, Kirldacharian S, Weiss M (1989) Methylene blue potentiates vascular reactivity in isolated rat lungs. J Appl Physiol 66(3): 1040-1045 12. Van de Voorde J, Leusen I (1986) Endothelium-dependent and independent relaxation of aortic rings from hypertensive rats. Am J Physiol 250 (Heart Circ Physiol 19):H711-H717 13. Mayhan WG, Faraci FM, Heistad DD (1987) Impairment of endothelium-dependent responses of cerebral arterioles in chronic hypertension. Am J Physiol 253 (Heart Circ Physiol 22):H1435-H1440
42
D. McCormack
14. Dinh Xuan AT, Higenbottam TW, Pepke-Zaba J, Clelland C, Wallwork J (1989) Reduced endothelium-dependent relaxation of cystic fibrosis pulmonary arteries. Eur J Phamacol 163:401-403 15. Rounds S, Farber HW, Hill NS, O'Brien RF (1985) Effects of endothelial cell injury on pulmonary vascular reactivity. Chest 88(4):213S-216S 16. Graham LM, Vasil ML, Voelkel NF, Stenmark KR (1989) Chronic pseudomonas pneumonia results in reduced pulmonary vasoreactivity and elevated pulmonary artery pressure. Am Rev Respir Dis A71
Lung
©
Springer-Verlag
New York Inc. 1990
Endothelium-derived Relaxing Factors and the Human Pulmonary Circulation D. McCormack Victoria Hospital, London, Ontario, Canada
Abstract. The vasodilator effect of acetylcholine on the pulmonary circulation was first described over 30 years ago, however, the mechanism remained unknown until Furchgott described the endothelium-dependent relaxation of certain vasodilators. It was not until 1987 that endotheliumderived relaxant factor (EDRF) was demonstrated to dilate human pulmonary arteries in vitro. Despite this work, the physiologic role of EDRF in the pulmonary circulation is not known. It has been suggested that hypoxiainduced inhibition of EDRF action or release from pulmonary artery endothelial cells may have a role in hypoxic pulmonary vasoconstriction (HPV) but present evidence suggests that loss of EDRF activity is not directly involved in the phenomenon of HPV. It is more likely that EDRF is released from pulmonary artery endothelial cells during hypoxia and this released EDRF then modulates HPV. If EDRF does modulate HPV in vivo then the role of EDRF in the altered HPV found in disease merits attention. It is known that in disease states such as acute lung injury and pneumonia there is loss or attenuation of HPV which inevitably leads to increased V/Q mismatch and hypoxemia. Whether this attenuation of HPV is due to release of an endogenous vasodilator such as EDRF is presently being investigated. Additionally, there is in vitro evidence that loss of EDRF activity may be important in the genesis of pulmonary hypertension such as found in severe cystic fibrosis. During the next decade the role of EDRF in the human pulmonary circulation in both health and disease will undoubtedly be elucidated. Key words: Pulmonary artery--Endothelium-derived relevant factor--Hypoxic pulmonary vasoconstriction--Pulmonary hypertension. Offprint requests to: Dr. D. McCormack, Victoria Hospital, 375 South Street, London CDNOntario, Canada.
36
D. McCormack
Discussing the role of EDRF in the human pulmonary circulation is at once both stimulating and difficult. Stimulating because of the interesting potential this compound possesses as an endogenous vasodilator in health and disease, and difficult because so little is known. Much work has been done concerning EDRF and the pulmonary circulation in animals but very little has been done in humans. There are obvious reasons as to why this is so, first, the pulmonary circulation is difficult to study noninvasively; second, removal or damage of the endothelium experimentally usually results in acute lung injury and pulmonary edema; and thirdly, the general enthusiasm of clinicians and scientists regarding pulmonary hypertension is less than that for systemic hypertension. Additionally, human tissue is difficult to obtain to work with in the laboratory and even when obtained, as many of us can attest to, it is exceedingly more difficult to handle and experiment with than animal tissue. With these points in mind this article will be divided into two main parts: the first part will deal with the known and possible role of EDRF in control of the pulmonary circulation in health, and the second part will deal with the potential role of EDRF in some of the disease states that concern the pulmonary vasculature.
EDRF and the Pulmonary Circulation in Health
The first paper to describe the effects of acetylcholine (ACh) on the pulmonary circulation in vivo was in 1957 by a group of physicians and surgeons from New York [1]. They described a fall both in pulmonary artery pressure and pulmonary vascular resistance in patients after ACh was infused. At the time, the explanation for this phenomenon was not clear although it was suggested that ACh was somehow acting directly on the pulmonary vascular smooth muscle. Since then we have come a long way. The major breakthrough was in 1980 when Furchgott and colleagues reported that ACh caused relaxation of precontracted rabbit aortic strips only when endothelial cells were present [2] and the term "endothelium-derived relaxant factor" was coined. Not long after this the effect of ACh was being investigated in a variety of different vascular beds from many different species. Additionally, it is now known that vasodilators in both the pulmonary and systemic circulation can be divided into endothelium-dependent and endothelium-independent vasodilators. Furchgott suggested that EDRF release after ACh was the result of stimulation of a muscarinic receptor. We have demonstrated that in the pulmonary vasculature of the rat the muscarinic receptor involved is of the M3 subtype [3]. This was done by looking at the effect of the muscarinic antagonists 4-DAMP, methoctramine, and pirenzipine on the ACh-induced, endothelium-dependent vasodilation of isolated rat pulmonary artery rings. Despite this explosion of knowledge regarding EDRF in both the pulmonary and systemic circulation there has been an astonishing lack of reported information concerning EDRF and the humau pulmonary circulation. Indeed it was not until 1987 that Greenberg and co-workers [4] published data showing that it was possible to mechanically remove the endothelium from normal
Endothelium-derived RelaxingFactors
37
human pulmonary arteries and they demonstrated in vitro that human vessels have endothelium-dependent vasodilation in response to ACh and ATP but not to vasoactive intestinal peptide. As physicians we know that it is one thing to describe a phenomenon in an organ bath, however, quite another to relate this to the intact patient. Unfortunately there has been little work evaluating the role of the endothelium in the control of pulmonary vascular tone in the human. Despite the lack of hard experimental data and the difficulty in extrapolating results from animal studies directly to humans there is room for hypothesizing on the role of EDRF in the human pulmonary circulation. Although the physiologic role of EDRF is not yet fully known, the fact that it is known to have a very short half-life suggests a localized site of action on subjacent vascular smooth muscle, and this has been shown experimentally. Hemoglobin bound to haptoglobin in plasma rapidly inactivates EDRF providing a mechanism to limit the downstream action within the vascular lumen in vivo. This would allow for precise local regulation of the local blood flow, something that is essential in the pulmonary circulation. The pulmonary vasculature is unique in two ways: (1) it normally has a very low resting tone and as a result the resting pulmonary arterial pressure is much lower than systemic arterial pressure and (2) it contracts in response to hypoxia. This hypoxic pulmonary vasoconstriction makes teleologic sense and is felt to be the most important mechanism in the matching of perfusion to ventilation in the lung. The role of the endothelium and EDRF in the maintenance of low pulmonary vascular tone and HPV is not clear but several important points have been established. First, as regards the possible role of EDRF in HPV, in 1978 Dr. Weir proposed that the phenomenon of hypoxic pulmonary vasoconstriction could be explained ff the pulmonary vasodilation present during normoxia were actively maintained by a vasodilator substance and hypoxia then lead to a decrease in this vasodilation with resultant vasoconstriction [5]. The question was therefore whether hypoxia resulted in the release of a pulmonary vasoconstrictor substance or if hypoxia inhibited the release or action of a vasodilator substance. Dr. Weir at that time proposed that bradykinin may have a role as this vasodilator, however, interest has lately focused on a possible role for EDRF in this capacity. To support this hypothesis of HPV being due to a loss of active normoxic vasodilation by EDRF it must be demonstrated that there is a basal release of EDRF from vessels. Griffith et al. [6] have done this using an experimental setup in which a bioassay strip of coronary artery is precontracted with 5-HT and then superfused. They demonstrated that EDRF is both continuously released from aortic endothelial cells in the basal state and this release can be stimulated by ACh. However, not all systemic vessels appear to have identical basal EDRF production [7] and we have performed experiments using columns of pulmonary artery endothelial cells grown on beads and perfused with Kreb's solution to evaluate the basal production of EDRF. We found that although the endothe-
38
D. McCormack
lial cells could be stimulated to produce EDRF and relax the test strip of endothelial-denuded pulmonary artery, there was no basal production o f EDRF by the cultured pulmonary artery endothelial cells in this system. Therefore, there may be reason to suspect that systemic and pulmonary artery endothelial cells do not have the same basal production of EDRF. Thus there is reason to doubt that there is sufficient basal production of EDRF by pulmonary artery endothelial cells to account for inhibition of EDRF leading to HPV. What then is the role of EDRF in the important phenomenon of HPV? To answer this we need to examine the known modulating effects of EDRF in the pulmonary circulation. Yamaguchi and co-workers have demonstrated that EDRF has a significant role in modulating the contractile sensitivity of isolated pulmonary vessels to KC1, angiotensin II and noradrenaline [8]. EDRF is thus important not only in endothelial-dependent pulmonary vasodilation but also as a modulator of pulmonary vasoconstriction. It is likely that the modulation of vasoconstriction by the endothelium and EDRF varies with agonist and species as well as with arterial segment. What then is the role of EDRF in modulating HPV? Rodman et al. have reported that isolated rat pulmonary artery rings in vitro contract as the pO2 of the surrounding bath is lowered [9]. They looked at the effect of methylene blue, an inhibitor of guanylate cyclase, in a dose of 10-5 M which has been shown to inhibit approximately 75% of an EDRF-stimulated relaxation. They found that the contraction caused by hypoxia was augmented by methylene blue between 10% and 3% Oz. This is consistent with the hypothesis that EDRF is released during hypoxia and modulates HPV such that inhibition of EDRF leads to an augmentation of the resoonse. There are problems with using only the model of isolated pulmonary arteries as a model for HPV as usually for in vitro experiments, conduit arteries are used which are almost certainly not the arteries involved in HPV in vivo. Brashers and co-workers therefore examined the effect of EDRF inhibition using the isolated perfused and ventilated lung of the rat [10]. They examined the effect of various chemically dissimilar and structurally unrelated pharmacologic agents such as ETYA, NDGA, and hydroquinone, all of which are known to block or attenuate the actions of EDRF. They found that all three antagonists augmented the hypoxic pressor response. These studies suggest that EDRF maybe an important modulator of HPV. Two points merit emphasis from these data. First, the administration of the EDRF antagonists did not affect the baseline pressures on room air ventilation, suggesting that in this isolated lung system there is no significant basal EDRF production. This point is consistent with the conclusion that the augmentation of I-IPV is not due to a decrease in basal EDRF activity. Second, the augmentation of HPV b y ETYA, NDGA, and hydroquinone implies that EDRF activity and/or release is increased in hypoxia. In a separate but similar study, Mazmanian and co-workers found that methylene blue and hemoglobin, both inhibitors of E D R F , augmented both the hypoxic and angiotensin II pressor response [11]. These studies provide some evidence that EDRF may have an important
Endothelium-derived Relaxing Factors
39
role in the modulation of the hypoxic pressor response. It is possible that by blocking EDRF-induced increase in cGMP levels there is a potentiation of hypoxic vasoconstriction, since the relaxant factor EDRF no longer modulates the contractile response. Furthermore, from the data presented so far, it would appear that EDRF is not tonically released to actively maintain low pulmonary vascular tone.
EDRF and the Pulmonary Circulation in Disease Having suggested some possible roles of EDRF in the pulmonary circulation in health what can be said about EDRF in the pulmonary circulation in disease? First, in pulmonary hypertensive disease there may be a failure of release or action of EDRF resulting in an increased resistance in the pulmonary circulation. This suggestion comes from experiments which have looked at isolated arterial rings from spontaneously hypertensive animals. In these studies it has been found that endothelium-independent vasodilators relax thoracic aortic rings from normal rats to the same extent as those from hypertensive animals [12]. By contrast, endothelium-dependent vasodilators do not effect the same relaxation in arteries from hypertensive animals compared arteries from normals. In addition to these in vitro findings demonstrating decreased action of EDRF in vessels from hypertensive animals, similar findings have been demonstrated in vivo using the cerebral circulation and looking at the vessels using a video image device [13]. Again, these investigators found that relaxation caused by the endothelium-independent vasodilator adenosine was similar in hypertensive and normal animals, however, the relaxation caused b y a n endothelial-dependent vasodilator such as ACh was markedly diminished in the cerebral circulation from spontaneously hypertensive animals. Thus, there has been speculation that this decrease in EDRF activity in hypertensive states may be important in the pathogenesis of pulmonary hypertension. The only published study so far addressing this in the human pulmonary circulation was by the group from Cambridge who used isolated pulmonary arteries from patients with cystic fibrosis who had pulmonary hypertension and were undergoing heart-lung transplantation. They demonstrated that vessels from these patients had diminished relaxation to acetylcholine and ADP [14]. It is not clear as to whether this diminished relaxation is the cause of or the result of the pulmonary hypertension however the authors supported the latter hypothesis. Obviously, further studies need to be done in vivo and in vitro in patients with pulmonary hypertension to answer this question and these studies have already started. Removal of endothelium from intrapulmonary arteries can transform the relaxations of ACh and bradykinin into pulmonary vasoconstrictors. It is possible that damage in vivo to arterial and arteriolar endothelial cells or destruction of endothelial cells by a variety of mechanisms such as release of proteolytic enzymes, leukocyte-endothelial and platelet-endothelial interactions leading to
40
D. M c C o r m a c k
the release of lysosomal enzymes, etc. may lead to pulmonary hypertension found in ADRS. So what about the role of the endothelium and EDRF in acute lung injury? It has become clear in the last few years that in animal models of acute lung injury there is abnormal pulmonary vascular reactivity. Rounds et al. [15] treated rats with intraperitoneal o~-naphthylthiourea (ANTU) resulting in a model of endothelial cell damage and an acute permeability pulmonary edema much like ARDS. Then, using an isolated, perfused lung setup, they evaluated the vascular reactivity of the pulmonary circulation to bolus infusions of angiotensin II. They found that in the lungs from rats treated with A N T U the pressor response to angiotensin II was greater compared to the pressor response to the same agent in the normal lungs treated with the ANTU vehicle Tween. This suggests that abnormal endothelial cell function, and perhaps abnormal EDRF release or activity is associated with enhanced pulmonary vascular reactivity, similar to the hypothesis regarding EDRF and hypoxic pulmonary vasoconstriction discussed earlier. However, it is important to note that not all models of acute lung injury are similar in so far as pulmonary vasoreactivity is concerned. As noted above, in the ANTU-treated model of acute lung injury there is increased vasoreactivity to angiotensin II. However, others have shown that there is decreased pulmonary vasoreactivity in different models of lung injury and this may have important clinical implications. For example, when rats are treated with intratracheal bleomycin there is an acute permeability pulmonary edema maximal about 48 h after the insult. At this time lungs from these animals can be removed, isolated, and ventilated and perfused. By ventilating the isolated lungs with hypoxic gas mixtures the pressor response to hypoxia may be evaluated. We have performed such experiments and have demonstrated that the injured lungs have virtual abolishment of HPV compared to lungs from saline-treated animals. Similar findings to ours have been noted in lungs from animals exposed to 100% 02 for 52 h to induce hyperoxic lung injury, another model of endothelial cell damage. Thus, in a least two models of acute lung injury there is a diminished pressor response in the pulmonary circulation to hypoxia. This pulmonary vascular hyporeactivity is not found only in acute lung injury. Graham et al. used a rat model of chronic Pseudomonas pneumonia and in vivo measured the hypoxic pressor response of these animals. They demonstrated that in control and infected animals there was a similar resting P A P but that the pressor response to hypoxia was diminished in the infected animals compared to controls [16]. What is the mechanism of this decreased pulmonary pressor response to hypoxia in these models of acute lung injury, lung inflammation and endothelial damage? The answer is not known however abnormal control of the release of EDRF may play a role. It is possible, for example, that when endothelial cells are damaged there may be loss of the normal regulation of EDRF stimulation and release and as a result, increased basal production and release of this compound. The vascular hyporeactivity in sepsis may be secondary to en-
Endothelium-derived Relaxing Factors
41
hanced endothelial cell EDRF release caused by bacterial endotoxins or by mediators generated during sepsis that either interact with processes coupled to EDRF release or cause endothelial cell damage. This would result in abnormal and inappropriate vasodilation in the area of endothelial and lung injury. One could hypothesize that in an area of pneumonia if there was abnormal EDRF release and consequent increased blood flow to the area this would lead to increased shunt and hypoxemia. Whether this does indeed happen is not known and is presently being investigated. Does the unique physiologic action of EDRF have any implications for the treatment of pulmonary hypertensive states? The major problem associated with pulmonary vasodilator therapy is the consequent systemic hypotension that seems to invariably accompany therapy. The extremely short half-life of EDRF would suggest that it may be possible to selectively vasodilate the pulmonary circulation without effect on the systemic circulation. However, until the exact nature and pharmacologic properties of EDRF are known this possibility cannot be fully explored.
References 1. Fritts Jr. HW, Harris P, Clauss RH, Odell JE, Cournand A (1958) The effect of acetylcholine on the human pulmonary circulation under normal and hypoxic conditions. J Clin Invest 37(1):99-1 t0 2. Furchgott RF, Zawadzki JV (1980) The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 288:373-376 3. McCormack DG, MAk JC, Minnette P, Barnes PJ (1989) Muscarinic receptor subtypes mediating vasodilation in the pulmonary artery. Eur J Pharmacol 158:293-297 4. Greenberg B, Rhoden K, Barnes PJ (1987) Endothelium-dependent relaxation of human pulmonary arteries. Am J Physiol 252 (Heart Circ Physiol 21):H434-438 5. Weir EK (1978) Does normoxic pulmonary vasodilatation rather than hypoxic vasoconstriction account for the pulmonary pressor response to hypoxia? Lancet 8062:476-477 6. Griffith TM, Edwards DH, Collins P, Lewis MJ, Henderson AH (1985) Endothelium derived relaxant factor. J R Coil Phys Lond 19(2):74-79 7. Collins P, Chappell SP, Griffith TM, Lewis MJ, Henderson AH (1986) Differences in basal endothelium-derived relaxing factor activity in different artery types. J Cardiovasc Pharmacol 8:1158-1162 8. Yamaguchi T, Rodman D, O'Brien R, McMurtry I (1989) Modulation of pulmonary artery contraction by endotheliurn-derived relaxing factor. Eur J Pharmacol 161:259-262 9. Rodman D, Yamaguchi T, O'Brien R, McMurtry I (1988) Methylene blue enhances hypoxic contraction in isolated rat pulmonary arteries. Chest 93(3):93S-94S 10. Brashers VL, Peach MJ, Rose Jr. CE (1988) Augmentation of hypoxic pulmonary vasoconstriction in the isolated perfused rat lung by in vitro antagonists of endothelium-dependent relaxation. J Clin Invest 82:1495-t502 11. Mazmardan GM, Baudet B, Brink C, Cerrina J, Kirldacharian S, Weiss M (1989) Methylene blue potentiates vascular reactivity in isolated rat lungs. J Appl Physiol 66(3): 1040-1045 12. Van de Voorde J, Leusen I (1986) Endothelium-dependent and independent relaxation of aortic rings from hypertensive rats. Am J Physiol 250 (Heart Circ Physiol 19):H711-H717 13. Mayhan WG, Faraci FM, Heistad DD (1987) Impairment of endothelium-dependent responses of cerebral arterioles in chronic hypertension. Am J Physiol 253 (Heart Circ Physiol 22):H1435-H1440
42
D. McCormack
14. Dinh Xuan AT, Higenbottam TW, Pepke-Zaba J, Clelland C, Wallwork J (1989) Reduced endothelium-dependent relaxation of cystic fibrosis pulmonary arteries. Eur J Phamacol 163:401-403 15. Rounds S, Farber HW, Hill NS, O'Brien RF (1985) Effects of endothelial cell injury on pulmonary vascular reactivity. Chest 88(4):213S-216S 16. Graham LM, Vasil ML, Voelkel NF, Stenmark KR (1989) Chronic pseudomonas pneumonia results in reduced pulmonary vasoreactivity and elevated pulmonary artery pressure. Am Rev Respir Dis A71
