- Email: [email protected]
A model of xenograft hyperacute rejection attenuates endothelial nitric oxide production: a mechanism for graft vasospasm? 1
ORIGINAL ARTICLES BASIC SCIENCE AND IMMUNOBIOLOGY
A Model of Xenograft Hyperacute Rejection Attenuates Endothelial Nitric Oxide Production: A Mechanism for Graft Vasospasm? David G. Cable, MD, Kunikazu Hisamochi, MD, and Hartzell V. Schaff, MD Background: The deposition of complement components within grafts, complement consumption, and prolongation of graft function by complement inactivation imply a pivotal role for complement in xenograft hyperacute rejection. The current investigations examined the endothelial production of vasoactive substances in pulmonary arteries during simulated hyperacute rejection. Methods and Results: Canine pulmonary arteries were suspended in organ chambers and exposed to either autologous canine serum for 90 minutes or heterologous porcine serum for 30, 60, or 90 minutes. Following serum exposure, the vessels were allowed a one-hour equilibration in buffered crystalloid solution. Dose-response curves were obtained with acetylcholine, sodium nitroprusside, and calcium ionophore A23187 following contraction with phenylephrine (1026M) in the presence of indomethacin (1025M). Receptor-dependent, endothelial-dependent relaxations to acetylcholine (1029–1024M) were impaired with 30-, 60-, or 90-minute porcine serum exposure when compared to vessels exposed to autologous canine serum (n 5 10, 7, 9, respectively; p , .05; 2-way ANOVA). Receptor-independent, endothelial-dependent relaxations to calcium inophore (1029–1026M) were significantly impaired at 60- and 90-minute porcine exposures only (n 5 7, 8; p , .05). Endothelial-independent relaxations to sodium nitroprusside (1029–1024M) were not impaired with either canine or porcine serum exposure. Oxyhemoglobin (1026M) abolished acetylcholine-mediated relaxations, indicating that nitric oxide was the predominant mediator. Conclusions: Simulated hyperacute xenograft rejection impairs endothelium-dependent relaxation of canine pulmonary arteries. Both basal and stimulated production of nitric oxide is impaired by heterologous serum exposure and, subsequently, complement activation. Reduced production of nitric oxide may explain, in part, the vasospasm and thrombosis of xenografts during hyperacute rejection. J Heart Lung Transplant 1999;18: 177–184.
From the Cardiac Surgical Research Center and the Section of Cardiovascular Surgery, Mayo Clinic and Mayo Foundation, Rochester, Minn., USA Submitted April 10, 1998; accepted July 10, 1998. Supported in part by The Mayo Foundation. Performed during DGC tenure as a Clinical Investigator Research Fellow. Presented at the International Society for Heart and Lung Transplantation Eighteenth Annual Meeting and Scientific
Sessions, Chicago, Illinois, April 18, 1998. Reprint requests: Hartzell V. Schaff, MD, Section of Cardiovascular Surgery, Mayo Clinic and Mayo Foundation, 200 First Street SW, Rochester, MN 55905. Copyright © 1998 by the Mayo Foundation. Copyright © 1999 by the International Society for Heart and Lung Transplantation. 1053-2498/99/$–see front matter PII S1053-2498(98)00023-0
177
178
C
Cable et al.
omplement activation, upon the recognition of recipient natural antibodies to graft endothelial aGal epitopes, induces hyperacute rejection of discordant xenografts. The rapid reduction in complement component levels after revascularization of a xenograft suggests a central role of complement in hyperacute rejection.1 Xenografts implanted in recipients with a congenital deficiency in complement components function longer than grafts implanted in normal recipients.2,3 Moreover, recipient complement is deposited within the xenograft as evidenced by immunohistology.4 While the pivotal role of complement activation in hyperacute rejection appears firmly established, the precise mechanism by which complement generates the sequel of hyperacute rejection is not clearly explained. The endothelium is injured during complement activation via several mechanisms. Membrane attack complexes generated form transmembrane channels.5,6 An upregulation of membrane P-selectin densities, an adherence protein for leukocytes, is concomitantly demonstrable.7,8 Endothelial release of heparan sulfate has been shown to occur during complement activation,9 impairing endothelial adhesion to the extracellular matrix.10 Less-well established is the role of endotheliumderived relaxing factors in the pathophysiology of hyperacute rejection. Endothelium-dependent relaxations can result from cyclo-oxygenase metabolites of arachidonic acid, nitric oxide synthase stimulation, and production of the endothelium-derived hyperpolarizing factor (EDHF).11 Levels of 6-keto-prostaglandin F1a, the stable metabolite of prostaglandin I2, are not elevated during hyperacute rejection.12 Therefore, the prostanoids produced by the endothelium, as a group, do not appear to be involved in hyperacute rejection. Nitric oxide is a potent endogenous vasodilator produced by the vascular endothelium, which inhibits platelet aggregation and promotes platelet disaggregation, attenuates leukocyte adhesion, and serves as a superoxide radical scavenger.13–18 We have previously demonstrated reduced endothelial nitric oxide production in canine coronary arteries during simulated hyperacute xenograft rejection is mediated by complement activation.19 The objective of the current investigation was to determine whether endothelial production of nitric oxide was impaired during simulated hyperacute rejection in pulmonary arteries, and compare this to our previously published results in coronary arteries. Further, the present study evaluated whether im-
The Journal of Heart and Lung Transplantation March 1999
pairment of endothelial production of nitric oxide occurred through receptor-dependent or independent pathway.
MATERIAL AND METHODS Humane care was provided in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH Publication No. 85-23, revised 1985). Purpose-bred, heart worm-free,20 adult mongrel dogs weighing 20 –30 kg were anesthetized with intravenous sodium pentobarbital (30 mg/kg; Fort Dodge Laboratories, Fort Dodge, IA) and rapidly exsanguinated. The lungs were excised and placed in cool, oxygenated modified Krebs–Ringer (crystalloid solution) solution of the following composition (mmol/L): NaCl, 118.3; KCl, 4.7; MgSO4, 1.2; KH2PO4, 1.2; CaCl2, 2.5; NaHCO3, 25; glucose, 11.1; pH 5 7.4.
In Vitro Studies Third-generation pulmonary arteries were delicately dissected from the lung and the overlying connective tissue and removed without injury to the intima. The arteries were divided into segments (5– 6 mm in length), and the endothelium was mechanically denuded in half the rings as previously described.21 Arterial segments, with and without endothelium, were studied in parallel organ chambers containing crystalloid solution (pH 7.4, 37° C) aerated with 95% O2–5% CO2.11,22 The blood vessel segments were suspended between two stainless steel clips connected to a strain gauge for the measurement of isometric force (Statham UC2; Gould, Cleveland, OH). The rings were placed at the optimal point of their length-tension relation by progressive stretching until contraction to potassium chloride (20 mM) was maximal.23 The presence or absence of endothelium was confirmed by response to acetylcholine (1026M) during contraction with phenylephrine (1026M). After confirmation of the presence or absence of endothelium, the segments were allowed to equilibrate for 30 minutes in the presence of indomethacin (1025M) to block endogenous cyclo-oxygenase activity, as the study was designed to specifically evaluate the nitric oxide synthase pathway.
The Journal of Heart and Lung Transplantation Volume 18, Number 3
Groups Serum was prepared as previously described,19 resulting in an acellular solution. Serum preparation significantly reduced the content of all formed blood elements. Vessels were exposed to either canine serum for 90 minutes or porcine serum for 30, 60, or 90 minutes at 37° C, aerated with 95% O2–5% CO2. Following serum exposure, the organ bath was rinsed with crystalloid solution and the segments equilibrated for 60 minutes. Dose-response curves were then performed after contraction with phenylephrine (1026M) in the presence of indomethacin (1025M). A maximum of three dose-response curves were performed on any one vessel, and we allowed a 30-minute equilibration period between dose-response curves; no curves were obtained after administration of calcium ionophore. In separate experiments, aliquots of canine and porcine sera were treated by heat-inactivation or by immunoadsorption of complement. Heat-inactivation was accomplished by raising the temperature of the serum to 56° C for 30 minutes and the subsequent cooling to 37° C prior to incubation of the vessels.24 Immunoadsorption was accomplished with 1:16 dilution of goat anti-C3 fractionated antiserum (Sigma, St. Louis, MO) at 37° C.19 Vessels were exposed to inactivated serum for 60 minutes because the prior timecourse studies determined that both acetylcholine and calcium ionophore (receptor-independent, endothelial-dependent) relaxations were impaired by this length of exposure.
Drugs The following drugs (Sigma, St. Louis, MO) were used: acetylcholine chloride, sodium nitroprusside, calcium ionophore A23187, indomethacin, phenylephrine, and goat anti-human C3 antiserum. All drugs were prepared daily with distilled water except for indomethacin (dissolved in 1025M sodium carbonate) and calcium ionophore A23187 (dissolved in 0.5% DMSO). These doses of sodium carbonate and DMSO have been previously documented to evoke no vascular effects.25 Oxyhemoglobin was prepared as previously described.15 Concentrations of all drugs are expressed as final molar concentration in the organ chamber.
Data Analysis Results are expressed as means 6 standard error of the mean (SEM). In all experiments, n refers to the number of animals from which blood vessels were utilized. Basal release of nitric oxide was calculated
Cable et al.
179
as described by Luscher.15 Responses are expressed as percent change from the baseline contracted state. Statistical analysis of the data utilized 2-way ANOVA evaluation with post-hoc Newman–Keuls test and critical ranges where appropriate. Values of p , .05 were considered statistically significant.
RESULTS Scanning electron microscopy performed on canine pulmonary arteries following a one-hour exposure to either canine or porcine serum showed similar structure with no areas of endothelial denudement. Thus, reduced production of endothelium-derived relaxing factors could not be accounted for by reduction in endothelial cell volume, or mass. Mild cellular swelling was noted in both groups, possibly representing artifact in preparation.
Impairment of EndotheliumDependent Relaxations Receptor-dependent relaxations Canine pulmonary arteries were precontracted with phenylephrine (1026M) in the presence of indomethacin (1025M). Addition of acetylcholine (1029– 1024M), a receptor-dependent agonist, produced progressive relaxations in pulmonary arteries with an intact endothelium (Figure 1A). In pulmonary arteries in which the endothelium had been mechanically denuded, no significant relaxations were noted with increasing concentrations of acetylcholine, confirming no direct action of this agent on vascular smooth muscle. Exposure of pulmonary arteries to heterologous porcine serum significantly inhibited the endothelium-dependent relaxations to acetylcholine. This was first noted after only 30 minutes of xenoserum exposure (p 5 .001 vs autologous serum). Similarly, xenoserum exposure of 60 and 90 minutes significantly inhibited relaxations to acetylcholine (p 5 .0005 and p 5 .0008 vs autologous serum, respectively). There was no statistically significant difference between the response of pulmonary arteries exposed to heterologous serum for any period of time. The reduced relaxations of pulmonary arteries to acetylcholine was associated with a reduced sensitivity to the drug. The calculated dose, which produced a 50% maximum effect (EC50) in the autologous group was 27.24 6 0.06 logM. After only 30 minutes of heterologous serum exposure, the EC50 was significantly altered, 26.80 6 0.07 logM (p 5 .003 vs autologous). A similar shift in the doseresponse curve to acetylcholine was noted with
180
Cable et al.
The Journal of Heart and Lung Transplantation March 1999
tylcholine upon heterologous serum exposure suggests an impairment in receptor/G-protein function. Oxyhemoglobin (1026M) abolished relaxation to acetylcholine in pulmonary arteries exposed to either autologous or heterologous serum (Figure 1B).
FIGURE 1 (A) Following precontraction with
phenylephrine (1026M), increasing concentrations of acetylcholine produced no significant relaxations in pulmonary arteries denuded of endothelium. In arteries with an intact endothelium, increasing concentrations of acetylcholine was associated with progressive relaxations. Relaxations to acetylcholine were significantly reduced in canine pulmonary arteries with endothelium after 30, 60, or 90 minutes of exposure to porcine serum compared to vessels exposed to canine serum for 90 minutes. (B) In endothelium-intact coronary arteries exposed to either canine or porcine serum for 60 minutes, oxyhemoglobin (1026M) abolished the relaxations to increasing concentrations of acetylcholine after precontraction with phenylephrine (1026M) in the presence of indomethacin (1025M).
60-minute (EC50 26.74 6 0.08 logM, p 5 .0001 vs autologous) and 90-minute (EC50 26.83 6 0.13 logM, p 5 .007 vs autologous) exposure to heterologous serum. There was no significant difference between the EC50 for pulmonary arteries exposed to heterologous serum for 30, 60, or 90 minutes. The change in sensitivity of pulmonary arteries to ace-
Receptor-independent relaxations Addition of increasing concentrations of calcium ionophore A23187 (1029–1026M) to precontracted pulmonary arteries with an intact endothelium caused progressive relaxations (Figure 2A). As noted with vessels exposed to acetylcholine, pulmonary arteries in which the endothelium had been mechanically denuded exhibited no significant relaxations to increasing concentrations of calcium ionophore, confirming the lack of a direct action on vascular smooth muscle. Endothelium-dependent relaxations to calcium ionophore were significantly inhibited upon exposure of pulmonary arteries to heterologous porcine serum. In contrast to acetylcholine dose-response curves, though, this was not significant with 30 minutes of xenoserum exposure (p 5 0.12 vs autologous serum). Only with exposures of 60- and 90-minute durations was calcium ionophore-mediated relaxations significantly impaired (p 5 .0008 and p 5 .001 vs autologous serum, respectively). Also, in contrast to the results noted with acetylcholine, there were significant differences between the response of pulmonary arteries exposed to heterologous serum for 30 minutes compared to those exposed for 60 minutes (p 5 .01 vs 30 min xenoserum) and 90 (p 5 .04 vs 30 min xenoserum). This suggests that although receptor-mediated release of nitric oxide is impaired, receptor-independent release is preserved early in the course of simulated hyperacute rejection. A reduced sensitivity to calcium ionophore was also noted, similar to the results obtained with acetylcholine. The EC50 in the autologous group was 27.19 6 0.10 logM. After 30 minutes of heterologous serum exposure, the EC50 was not significantly changed, 26.96 6 0.08 logM (p 5 0.10 vs autologous). A shift in the dose-response curve to calcium ionophore was noted, however, with 60-minute (EC50 26.74 6 0.06 logM, p 5 .003 vs autologous) and 90-minute (EC50 26.74 6 .07 logM, p 5 .002 vs autologous) exposure to heterologous serum. Oxyhemoglobin (1026M) abolished relaxation to calcium ionophore in pulmonary arteries exposed to either autologous or heterologous serum (Figure 2B). This suggested that the nitric oxide synthase pathway mediated the entirety of these relaxations
The Journal of Heart and Lung Transplantation Volume 18, Number 3
Cable et al.
181
FIGURE 3 Increasing concentrations of sodium
nitroprusside evoked relaxations after precontraction with phenylephrine (1026M) in the presence of indomethacin (1025M) in both endothelium-intact and endothelium-denuded vessels. No significant difference in relaxation to increasing concentrations of sodium nitroprusside was noted.
FIGURE 2 (A) Following precontraction with
phenylephrine (1026M), increasing concentrations of calcium ionophore produced no significant relaxations in pulmonary arteries denuded of endothelium. In endothelium-intact vessels, increasing concentrations of calcium ionophore A23187 evoked progressive relaxations. Relaxations to calcium ionophore A23187 were significantly reduced in canine pulmonary arteries with endothelium after 60 or 90 minutes of exposure to porcine serum compared to vessels exposed to canine serum for 90 minutes. (B) In endothelium-intact coronary arteries exposed to either canine or porcine serum for 60 minutes, oxyhemoglobin (1026M) abolished the relaxations to low concentrations of calcium ionophore A23187 (1029–1026M) after precontraction with phenylephrine (1026M) in the presence of indomethacin (1025M).
to calcium ionophore, as it had with acetylcholinemediated relaxations.
Preservation of EndotheliumIndependent Relaxations Impaired relaxations to acetylcholine and calcium ionophore following heterologous serum exposure
could be caused by reduced production of nitric oxide from the endothelium or impairment in the ability of the vascular smooth muscle to respond. To exclude the latter, pulmonary arteries precontracted with phenylephrine (1026M), were stimulated with sodium nitroprusside, a spontaneous donor of nitric oxide. Arterial segments exposed to either autologous or heterologous serum relaxed similarly with increasing concentrations of sodium nitroprusside (1029–1024M), suggesting that, in fact, the ability of the smooth muscle to respond to nitric oxide was intact following heterologous serum exposure (Figure 3). Stated differently, the guanylate cyclase/ cGMP pathway was intact in all arterial segments.
Effect of Complement Inactivation Pulmonary artery segments exposed to heat-inactivated heterologous serum for one hour exhibited normal relaxations to acetylcholine (Figure 4A). Similarly, heterologous serum in which the complement cascade was blocked by immunoadsorption with anti-C3 antiserum had no effect on pulmonary artery relaxations to acetylcholine (Figure 4B).
Basal Production of Nitric Oxide The previous experiments demonstrated a reduction in the endothelial production of nitric oxide following heterologous serum exposure, an observation that could not be explained by structural degradation of the endothelium. This finding correlates with
182
Cable et al.
The Journal of Heart and Lung Transplantation March 1999
DISCUSSION
FIGURE 4 (A) Heat-inactivated porcine serum had no
effect on relaxations of endothelium intact vessels to increasing concentrations of acetylcholine. (B) Similarly, canine coronary artery rings responded normally to increasing concentrations of acetylcholine after 60 minutes of exposure to C3-inactivated porcine serum.
stimulated production of nitric oxide. To determine if basal production was also affected, we utilized an indirect determination of basal nitric oxide production as previously described.26,27 The addition of oxyhemoglobin (1026M) to the organ bath following a one-hour serum exposure resulted in a relative contraction of the pulmonary arteries. Subsequent contraction of these arteries with phenylephrine (1026M) resulted in a tension of 5.2 6 0.1 gm in the autologous serum group and 4.8 6 0.2 gm in the heterologous serum group (n 5 5, p 5 .17). The calculated basal release of nitric oxide was 12.5 6 1.9% in the autologous serum group and 7.9 6 1.7% in the heterologous serum group (n 5 5, p 5 .005); this represented a 38 6 8% reduction of basal nitric oxide production with one hour of xenoserum exposure.
The purpose of this investigation was to determine if nitric oxide production of pulmonary arteries is impaired during simulated hyperacute xenograft rejection and if this impairment is attributable to complement activation. The major findings of the present study were: (1) exposure of canine pulmonary arterial segments to porcine serum for 60 and 90 minutes significantly attenuated endotheliumdependent vasodilatation; (2) exposure of canine pulmonary arteries to porcine serum did not alter endothelium-independent vasodilatation; and (3) inhibition of complement with either heat-inactivation or immunoadsorption prevented impairment of endothelium-dependent vasodilatation. The current study utilized an animal combination that has previously been shown to cause complement activation during hyperacute rejection.28 –32 Mazzoni et al demonstrated capillary thrombosis within 20 minutes in canine-to-porcine renal xenografts.33 Furthermore, we have previously described reproducible impairment of endothelial nitric oxide production attributable to complement activation in canine coronary arteries treated with porcine serum.19 We recognize that data from this animal combination may not be transferable to humans, but the model is useful in laboratory studies of hyperacute rejection. The present investigations demonstrated the reduction of endothelium-dependent relaxations to acetylcholine following exposure to heterologous porcine serum. Relaxations to sodium nitroprusside, a nitric oxide donor, were unchanged, suggesting that the function of the vascular smooth muscle was intact; ie, the ability of the vascular smooth muscle to relax to nitric oxide was not impaired by exposure to heterologous serum. This fact argues against nonspecific activation of membrane complexes and, subsequent, ubiquitous vascular injury. Rather, the failure of calcium ionophore to fully correct this impairment in endothelium-dependent relaxation signals a defect in the nitric oxide synthase pathway directly during simulated hyperacute rejection. Several pharmacologic tools are available to illustrate action of nitric oxide in vascular relaxation. Many are familiar with the L-arginine analogues, but two factors precluded their use in the present study. First, many have nonspecific actions, such as inhibition of L-arginine transport, which are difficult to predict and compound the experimental variables. Secondly, and more importantly for this study, they require a finite period of time to inhibit nitric oxide
The Journal of Heart and Lung Transplantation Volume 18, Number 3
production, averaging 45 minutes to 1 hour. As the current investigations demonstrated a time-dependent phenomena, it required a pharmacologic agent which would have immediate action. Oxyhemaglobin served this purpose. It allowed rapid inhibition of nitric oxide following serum exposure. It might be argued that reduced production of nitric oxide during hyperacute rejection might represent nonspecific cellular injury. Complement activation generates membrane attack complexes; these create pores, cellular swelling ensues, and nitric oxide production would be expected to decrease as the cell undergoes death. Thus, impaired nitric oxide generation could be a marker of cellular dysfunction, but this hypothesis is not supported by our data. Electron microscopic examination of rings after exposure to heterologous serum demonstrated only mild endothelial cell swelling without significant lysis. Basal production of nitric oxide was only partially reduced with heterologous serum. Furthermore, vascular smooth muscle function, as assessed by sodium nitroprusside response, was unchanged. Another explanation is that impaired nitric oxide production in pulmonary arteries following heterologous serum exposure was the result of selective damage to the nitric oxide pathway. It should be emphasized that in the current study we specifically excluded the effects of prostanoids by the use of indomethacin; Hammer et al previously demonstrated that stable metabolite of prostaglandin I2 is not elevated during hyperacute rejection.12 More importantly, though, we eliminated in our preparation the interaction of platelets to the endothelium. It is known that stimulation of endothelial cells during hyperacute rejection increases generation of platelet-activating factor, von Willebrand Factor, and glycoprotein GPIb,34 –36 as well as thrombin release, and this may promote platelet aggregation.37 It was not our intention in this study to evaluate all aspects of hyperacute rejection, but rather we sought to simplify our preparation so that we could specifically evaluate the endothelial production of nitric oxide. Differences were noted in the present findings compared to our earlier work with canine coronary arteries.19 First, the pulmonary endothelium appears more susceptible to the injury induced by heterologous serum exposure because a shorter duration of exposure produced significant impairment of endothelium-dependent relaxations in pulmonary compared to coronary arteries. Pulmonary arteries were impaired after only 30 minutes of exposure, but coronary arteries required 60 minutes
Cable et al.
183
of exposure. Secondly, the present study demonstrated a reduction in both receptor-dependent and receptor-independent relaxations following heterologous serum exposure; canine coronary arteries demonstrated only a significant impairment of receptor-dependent relaxations. It is not clear whether this represents an initial attack on the receptor/Gprotein complex by complement activation or as suggested in our previous communication, a significant contribution of endothelium-derived hyperpolarizing factor (EDHF) to relaxations in coronary arteries. If it is true that EDHF relaxations are unaffected by complement activation, as previous work has demonstrated,19 then the nitric oxide synthase pathway would be the only endotheliumderived relaxing factor impaired during hyperacute rejection. In conclusion, basal release of nitric oxide in pulmonary arteries is reduced following xenoserum exposure, and stimulated endothelium-dependent relaxations are attenuated. By electron microscopy, there were minimal structural alterations of endothelial cells. Attenuation of nitric oxide release would further enhance the prothrombotic state of hyperacute rejection because nitric oxide promotes platelet disaggregation and inhibits platelet aggregation. Moreover, a reduction in nitric oxide during hyperacute rejection would be expected to lead to vasoconstriction, leukocyte adhesion, and augmentation of reperfusion injury. The authors wish to thank Marilyn Oeltjen for devoted technical support and Donna Stucky for the diligent preparation of this manuscript.
REFERENCES 1. Gewurz H, Clark DS, Finstad J, et al. Role of the complement system in graft rejections in experimental animals and man. Ann NY Acad Sci 1966;129:673. 2. Zhow XJ, Niesen N, Pawlowski I, et al. Prolongation of survival of discordant kidney xenografts by C6 deficiency. Transplantation 1990;50:896. 3. Brauer RB, Baldwin III WM, Daha MR, et al. Use of C6-deficient rats to evaluate the mechanism of hyperacute rejection of discordant cardiac xenografts. J Immunol 1993; 151:7240. 4. Platt JL, Fischel RJ, Matas AJ, et al. Immunopathology of hyperacute xenograft rejection in a swine-to-primate model. Transplantation 1991;52:214. 5. Weisman HF, Bartow T, Leppo MK, et al. Soluble human complement receptor type 1: in vivo inhibitor of complement suppressing post-ischemic myocardial inflammation and necrosis. Science 1990;249:146 –51. 6. Rus HG, Niculescu F, Vlaicu R. Presence of C5b-9 complement complex and S-protein in human myocardial areas with necrosis and sclerosis. Immunol Lett 1987;16:15–20. 7. Mulligan MS, Poley MJ, Bayer RJ, et al. Neutrophil-depen-
184
8.
9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
20.
21.
22.
Cable et al.
dent acute lung injury: requirement for P-selectin (GMP140). J Clin Invest 1992;90:1600 –7. Ma X, Weyrich AS, Lefer DJ, et al. Diminished basal nitric oxide release after myocardial ischemia and reperfusion promotes neutrophil adherence to coronary endothelium. Circ Res 1993;72:403–12. Platt JL, Vercellotti GM, Lindman BJ, et al. Release of heparan sulfate from endothelial cells: implications for pathogenesis of hyperacute rejection. J Exp Med 1990;171: 1363– 8. Platt JF, Dalmasso AP, Lindman BJ, et al. The role of C5a and antibody in the release of heparan sulfate from endothelial cells. Eur J Immunol 1991;16:15–20. Vanhoutte PM, Boulanger CM, Mombouli JV. Endotheliumderived relaxing factors and converting enzyme inhibition. Am J Cardiol 1995;76:3E–12E. Hammer C, Schutz A, Pratschke J, et al. Bridging to transplant: allogeneic heart transplantation after xenografting. J Heart Lung Transplant 1992;11:S182– 8. Ignarro LF, Buga GM, Wood KS, et al. Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide. Proc Natl Acad Sci USA 1987;84:9265. Radomski MW, Palmer RMJ, Moncada S. Comparative pharmacology of endothelium-derived relaxing factor, nitric oxide and prostacyclin in platelets. Br J Pharmacol 1987;92:181. Rodomski MW, Palmer RMJ, Moncada S. The anti-aggregating properties of vascular endothelium: interactions between prostacyclin and nitric oxide. Br J Pharmacol 1987;92:639. Schafer AI, Alexander RW, Handin RI. Inhibition of platelet function by organic nitrate vasodilators. Blood 1980;55:649. Gryglewski RJ, Plamer RM, Moncada S. Superoxide anion is involved in the breakdown of endothelium-derived relaxing factor. Nature 1986;320:454 – 6. Rubanyi GM, Vanhoutte PM. Superoxide anions and hyperoxia inactivate endothelium-derived relaxing factors. Am J Physiol 1986;250:H822–7. Cable DG, Hisamochi K, Schaff HV. Complement mediates attenuation of endothelium-dependent relaxations in canine coronary arteries following porcine serum exposure: a mechanism for vascular thrombosis in xenograft hyperacute rejection. Circulation 1997;96:II58 – 64. Kaiser L, Tithof PK, Williams JF. Depression of endothelium-dependent relaxation by filarial parasite products. Am J Physiol 1990;259:H648 –52. Pearson PJ, Schaff HV, Vanhoutte PM. Long-term impairment of endothelium-dependent relaxations to aggregating platelets after reperfusion injury in canine coronary arteries. Circulation 1990;81:1921–7. Pearson PJ, Schaff HV, Vahoutte PM. Acute impairment of endothelium-dependent relaxations to aggregating platelets following reperfusion injury in canine coronary arteries. Circ Res 1990;67:385–93.
The Journal of Heart and Lung Transplantation March 1999 23. Cohen RA, Shepherd JT, Vanhoutte PM. Inhibitory role of the endothelium in the response of isolated coronary arteries to platelets. Science 1983;221:273– 4. 24. Jonas W, Stankiewicz M. Sheep serum complement sensitisation of sheep erythrocyte-rabbit antibody complexes for hemolysis by guinea-pig complement plus EDTA or MgEGTA. Arch Immuno Ther Exp 1986;34:451– 60. 25. Luscher TF, Diederich D, Siebenmann R, et al. Difference between endothelium-dependent relaxation in arterial and in venous coronary bypass grafts. N Engl J Med 1988;319:462–7. 26. Yang Z, von Segesser L, Bauer E, et al. Different activation of the endothelial L-arginine and cyclooxygenase pathway in the human internal mammary artery and saphenous vein. Circ Res 1991;68:52– 60. 27. Kojda G, Noack E. Nitric oxide liberating, soluble guanylate cyclase stimulating and vasorelaxing properties of the new nitrate-compound SPM 3672. J Cardiovasc Pharmacol 1993; 22:103–11. 28. Giles GR, Boehmig HJ, Lilly J, et al. Mechanism and modification of rejection of heterografts between divergent species. Transplant Proc 1970;2:522–37. 29. Perper RJ, Najarian JS. Experimental renal heterotransplantation. III. In widely divergent species. Transplantation 1966; 4:377– 88. 30. Kux M, Boehmig HJ, Amemiya H, et al. Modification of hyperacute canine renal homograft and pig-to-dog heterograft rejection by the intra-arterial infusion of citrate. Surgery 1971;70:103–12. 31. Kux M. Hyperakute nierenabstobung mit und ohne antikorpernachweis. Chirurg 1975;46:262–7. 32. Stemberger H, Rauhs R, Thetter O, et al. Beeinflussung der nierenxenotransplantation durch vorperfusion mit spezifischen antikorpern. Wiener Klinische 1978;90:197–201. 33. Mazzoni G, Benichou J, Porters KA, et al. Renal homotransplantation with venous outflow or infusion of antigen into the portal vein of dogs or pigs. Transplantation 1977;24:268 –73. 34. Robson SC, Kopp C, Lesnikoski B. Platelets and xenograft rejection. Xenotransplant 1994;2:38 – 44. 35. Makowka L, Chapman FA, Cramer DV, et al. Plateletactivating factor and hyperacute rejection. The effect of a platelet-activating factor antagonist, SRI63– 441, on rejection of xenografts and allografts in sensitized hosts. Transplantation 1990;50:359 – 65. 36. Coughlan AF, Berndt MC, Dunlop LC, et al. In vivo studies of P-selectin and platelet-activating factor during endotoxemia, accelerated allograft rejection and discordant xenograft rejection. Transplant Proc 1993;25:2930 –1. 37. Robson SC, Siegel JB, Lesnikoski BA, et al. Aggregation of human platelets induced by porcine endothelial cells is dependent upon both activation of complement and thrombin generation. Xenotransplant 1996;3:24 –34.
A Model of Xenograft Hyperacute Rejection Attenuates Endothelial Nitric Oxide Production: A Mechanism for Graft Vasospasm? David G. Cable, MD, Kunikazu Hisamochi, MD, and Hartzell V. Schaff, MD Background: The deposition of complement components within grafts, complement consumption, and prolongation of graft function by complement inactivation imply a pivotal role for complement in xenograft hyperacute rejection. The current investigations examined the endothelial production of vasoactive substances in pulmonary arteries during simulated hyperacute rejection. Methods and Results: Canine pulmonary arteries were suspended in organ chambers and exposed to either autologous canine serum for 90 minutes or heterologous porcine serum for 30, 60, or 90 minutes. Following serum exposure, the vessels were allowed a one-hour equilibration in buffered crystalloid solution. Dose-response curves were obtained with acetylcholine, sodium nitroprusside, and calcium ionophore A23187 following contraction with phenylephrine (1026M) in the presence of indomethacin (1025M). Receptor-dependent, endothelial-dependent relaxations to acetylcholine (1029–1024M) were impaired with 30-, 60-, or 90-minute porcine serum exposure when compared to vessels exposed to autologous canine serum (n 5 10, 7, 9, respectively; p , .05; 2-way ANOVA). Receptor-independent, endothelial-dependent relaxations to calcium inophore (1029–1026M) were significantly impaired at 60- and 90-minute porcine exposures only (n 5 7, 8; p , .05). Endothelial-independent relaxations to sodium nitroprusside (1029–1024M) were not impaired with either canine or porcine serum exposure. Oxyhemoglobin (1026M) abolished acetylcholine-mediated relaxations, indicating that nitric oxide was the predominant mediator. Conclusions: Simulated hyperacute xenograft rejection impairs endothelium-dependent relaxation of canine pulmonary arteries. Both basal and stimulated production of nitric oxide is impaired by heterologous serum exposure and, subsequently, complement activation. Reduced production of nitric oxide may explain, in part, the vasospasm and thrombosis of xenografts during hyperacute rejection. J Heart Lung Transplant 1999;18: 177–184.
From the Cardiac Surgical Research Center and the Section of Cardiovascular Surgery, Mayo Clinic and Mayo Foundation, Rochester, Minn., USA Submitted April 10, 1998; accepted July 10, 1998. Supported in part by The Mayo Foundation. Performed during DGC tenure as a Clinical Investigator Research Fellow. Presented at the International Society for Heart and Lung Transplantation Eighteenth Annual Meeting and Scientific
Sessions, Chicago, Illinois, April 18, 1998. Reprint requests: Hartzell V. Schaff, MD, Section of Cardiovascular Surgery, Mayo Clinic and Mayo Foundation, 200 First Street SW, Rochester, MN 55905. Copyright © 1998 by the Mayo Foundation. Copyright © 1999 by the International Society for Heart and Lung Transplantation. 1053-2498/99/$–see front matter PII S1053-2498(98)00023-0
177
178
C
Cable et al.
omplement activation, upon the recognition of recipient natural antibodies to graft endothelial aGal epitopes, induces hyperacute rejection of discordant xenografts. The rapid reduction in complement component levels after revascularization of a xenograft suggests a central role of complement in hyperacute rejection.1 Xenografts implanted in recipients with a congenital deficiency in complement components function longer than grafts implanted in normal recipients.2,3 Moreover, recipient complement is deposited within the xenograft as evidenced by immunohistology.4 While the pivotal role of complement activation in hyperacute rejection appears firmly established, the precise mechanism by which complement generates the sequel of hyperacute rejection is not clearly explained. The endothelium is injured during complement activation via several mechanisms. Membrane attack complexes generated form transmembrane channels.5,6 An upregulation of membrane P-selectin densities, an adherence protein for leukocytes, is concomitantly demonstrable.7,8 Endothelial release of heparan sulfate has been shown to occur during complement activation,9 impairing endothelial adhesion to the extracellular matrix.10 Less-well established is the role of endotheliumderived relaxing factors in the pathophysiology of hyperacute rejection. Endothelium-dependent relaxations can result from cyclo-oxygenase metabolites of arachidonic acid, nitric oxide synthase stimulation, and production of the endothelium-derived hyperpolarizing factor (EDHF).11 Levels of 6-keto-prostaglandin F1a, the stable metabolite of prostaglandin I2, are not elevated during hyperacute rejection.12 Therefore, the prostanoids produced by the endothelium, as a group, do not appear to be involved in hyperacute rejection. Nitric oxide is a potent endogenous vasodilator produced by the vascular endothelium, which inhibits platelet aggregation and promotes platelet disaggregation, attenuates leukocyte adhesion, and serves as a superoxide radical scavenger.13–18 We have previously demonstrated reduced endothelial nitric oxide production in canine coronary arteries during simulated hyperacute xenograft rejection is mediated by complement activation.19 The objective of the current investigation was to determine whether endothelial production of nitric oxide was impaired during simulated hyperacute rejection in pulmonary arteries, and compare this to our previously published results in coronary arteries. Further, the present study evaluated whether im-
The Journal of Heart and Lung Transplantation March 1999
pairment of endothelial production of nitric oxide occurred through receptor-dependent or independent pathway.
MATERIAL AND METHODS Humane care was provided in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH Publication No. 85-23, revised 1985). Purpose-bred, heart worm-free,20 adult mongrel dogs weighing 20 –30 kg were anesthetized with intravenous sodium pentobarbital (30 mg/kg; Fort Dodge Laboratories, Fort Dodge, IA) and rapidly exsanguinated. The lungs were excised and placed in cool, oxygenated modified Krebs–Ringer (crystalloid solution) solution of the following composition (mmol/L): NaCl, 118.3; KCl, 4.7; MgSO4, 1.2; KH2PO4, 1.2; CaCl2, 2.5; NaHCO3, 25; glucose, 11.1; pH 5 7.4.
In Vitro Studies Third-generation pulmonary arteries were delicately dissected from the lung and the overlying connective tissue and removed without injury to the intima. The arteries were divided into segments (5– 6 mm in length), and the endothelium was mechanically denuded in half the rings as previously described.21 Arterial segments, with and without endothelium, were studied in parallel organ chambers containing crystalloid solution (pH 7.4, 37° C) aerated with 95% O2–5% CO2.11,22 The blood vessel segments were suspended between two stainless steel clips connected to a strain gauge for the measurement of isometric force (Statham UC2; Gould, Cleveland, OH). The rings were placed at the optimal point of their length-tension relation by progressive stretching until contraction to potassium chloride (20 mM) was maximal.23 The presence or absence of endothelium was confirmed by response to acetylcholine (1026M) during contraction with phenylephrine (1026M). After confirmation of the presence or absence of endothelium, the segments were allowed to equilibrate for 30 minutes in the presence of indomethacin (1025M) to block endogenous cyclo-oxygenase activity, as the study was designed to specifically evaluate the nitric oxide synthase pathway.
The Journal of Heart and Lung Transplantation Volume 18, Number 3
Groups Serum was prepared as previously described,19 resulting in an acellular solution. Serum preparation significantly reduced the content of all formed blood elements. Vessels were exposed to either canine serum for 90 minutes or porcine serum for 30, 60, or 90 minutes at 37° C, aerated with 95% O2–5% CO2. Following serum exposure, the organ bath was rinsed with crystalloid solution and the segments equilibrated for 60 minutes. Dose-response curves were then performed after contraction with phenylephrine (1026M) in the presence of indomethacin (1025M). A maximum of three dose-response curves were performed on any one vessel, and we allowed a 30-minute equilibration period between dose-response curves; no curves were obtained after administration of calcium ionophore. In separate experiments, aliquots of canine and porcine sera were treated by heat-inactivation or by immunoadsorption of complement. Heat-inactivation was accomplished by raising the temperature of the serum to 56° C for 30 minutes and the subsequent cooling to 37° C prior to incubation of the vessels.24 Immunoadsorption was accomplished with 1:16 dilution of goat anti-C3 fractionated antiserum (Sigma, St. Louis, MO) at 37° C.19 Vessels were exposed to inactivated serum for 60 minutes because the prior timecourse studies determined that both acetylcholine and calcium ionophore (receptor-independent, endothelial-dependent) relaxations were impaired by this length of exposure.
Drugs The following drugs (Sigma, St. Louis, MO) were used: acetylcholine chloride, sodium nitroprusside, calcium ionophore A23187, indomethacin, phenylephrine, and goat anti-human C3 antiserum. All drugs were prepared daily with distilled water except for indomethacin (dissolved in 1025M sodium carbonate) and calcium ionophore A23187 (dissolved in 0.5% DMSO). These doses of sodium carbonate and DMSO have been previously documented to evoke no vascular effects.25 Oxyhemoglobin was prepared as previously described.15 Concentrations of all drugs are expressed as final molar concentration in the organ chamber.
Data Analysis Results are expressed as means 6 standard error of the mean (SEM). In all experiments, n refers to the number of animals from which blood vessels were utilized. Basal release of nitric oxide was calculated
Cable et al.
179
as described by Luscher.15 Responses are expressed as percent change from the baseline contracted state. Statistical analysis of the data utilized 2-way ANOVA evaluation with post-hoc Newman–Keuls test and critical ranges where appropriate. Values of p , .05 were considered statistically significant.
RESULTS Scanning electron microscopy performed on canine pulmonary arteries following a one-hour exposure to either canine or porcine serum showed similar structure with no areas of endothelial denudement. Thus, reduced production of endothelium-derived relaxing factors could not be accounted for by reduction in endothelial cell volume, or mass. Mild cellular swelling was noted in both groups, possibly representing artifact in preparation.
Impairment of EndotheliumDependent Relaxations Receptor-dependent relaxations Canine pulmonary arteries were precontracted with phenylephrine (1026M) in the presence of indomethacin (1025M). Addition of acetylcholine (1029– 1024M), a receptor-dependent agonist, produced progressive relaxations in pulmonary arteries with an intact endothelium (Figure 1A). In pulmonary arteries in which the endothelium had been mechanically denuded, no significant relaxations were noted with increasing concentrations of acetylcholine, confirming no direct action of this agent on vascular smooth muscle. Exposure of pulmonary arteries to heterologous porcine serum significantly inhibited the endothelium-dependent relaxations to acetylcholine. This was first noted after only 30 minutes of xenoserum exposure (p 5 .001 vs autologous serum). Similarly, xenoserum exposure of 60 and 90 minutes significantly inhibited relaxations to acetylcholine (p 5 .0005 and p 5 .0008 vs autologous serum, respectively). There was no statistically significant difference between the response of pulmonary arteries exposed to heterologous serum for any period of time. The reduced relaxations of pulmonary arteries to acetylcholine was associated with a reduced sensitivity to the drug. The calculated dose, which produced a 50% maximum effect (EC50) in the autologous group was 27.24 6 0.06 logM. After only 30 minutes of heterologous serum exposure, the EC50 was significantly altered, 26.80 6 0.07 logM (p 5 .003 vs autologous). A similar shift in the doseresponse curve to acetylcholine was noted with
180
Cable et al.
The Journal of Heart and Lung Transplantation March 1999
tylcholine upon heterologous serum exposure suggests an impairment in receptor/G-protein function. Oxyhemoglobin (1026M) abolished relaxation to acetylcholine in pulmonary arteries exposed to either autologous or heterologous serum (Figure 1B).
FIGURE 1 (A) Following precontraction with
phenylephrine (1026M), increasing concentrations of acetylcholine produced no significant relaxations in pulmonary arteries denuded of endothelium. In arteries with an intact endothelium, increasing concentrations of acetylcholine was associated with progressive relaxations. Relaxations to acetylcholine were significantly reduced in canine pulmonary arteries with endothelium after 30, 60, or 90 minutes of exposure to porcine serum compared to vessels exposed to canine serum for 90 minutes. (B) In endothelium-intact coronary arteries exposed to either canine or porcine serum for 60 minutes, oxyhemoglobin (1026M) abolished the relaxations to increasing concentrations of acetylcholine after precontraction with phenylephrine (1026M) in the presence of indomethacin (1025M).
60-minute (EC50 26.74 6 0.08 logM, p 5 .0001 vs autologous) and 90-minute (EC50 26.83 6 0.13 logM, p 5 .007 vs autologous) exposure to heterologous serum. There was no significant difference between the EC50 for pulmonary arteries exposed to heterologous serum for 30, 60, or 90 minutes. The change in sensitivity of pulmonary arteries to ace-
Receptor-independent relaxations Addition of increasing concentrations of calcium ionophore A23187 (1029–1026M) to precontracted pulmonary arteries with an intact endothelium caused progressive relaxations (Figure 2A). As noted with vessels exposed to acetylcholine, pulmonary arteries in which the endothelium had been mechanically denuded exhibited no significant relaxations to increasing concentrations of calcium ionophore, confirming the lack of a direct action on vascular smooth muscle. Endothelium-dependent relaxations to calcium ionophore were significantly inhibited upon exposure of pulmonary arteries to heterologous porcine serum. In contrast to acetylcholine dose-response curves, though, this was not significant with 30 minutes of xenoserum exposure (p 5 0.12 vs autologous serum). Only with exposures of 60- and 90-minute durations was calcium ionophore-mediated relaxations significantly impaired (p 5 .0008 and p 5 .001 vs autologous serum, respectively). Also, in contrast to the results noted with acetylcholine, there were significant differences between the response of pulmonary arteries exposed to heterologous serum for 30 minutes compared to those exposed for 60 minutes (p 5 .01 vs 30 min xenoserum) and 90 (p 5 .04 vs 30 min xenoserum). This suggests that although receptor-mediated release of nitric oxide is impaired, receptor-independent release is preserved early in the course of simulated hyperacute rejection. A reduced sensitivity to calcium ionophore was also noted, similar to the results obtained with acetylcholine. The EC50 in the autologous group was 27.19 6 0.10 logM. After 30 minutes of heterologous serum exposure, the EC50 was not significantly changed, 26.96 6 0.08 logM (p 5 0.10 vs autologous). A shift in the dose-response curve to calcium ionophore was noted, however, with 60-minute (EC50 26.74 6 0.06 logM, p 5 .003 vs autologous) and 90-minute (EC50 26.74 6 .07 logM, p 5 .002 vs autologous) exposure to heterologous serum. Oxyhemoglobin (1026M) abolished relaxation to calcium ionophore in pulmonary arteries exposed to either autologous or heterologous serum (Figure 2B). This suggested that the nitric oxide synthase pathway mediated the entirety of these relaxations
The Journal of Heart and Lung Transplantation Volume 18, Number 3
Cable et al.
181
FIGURE 3 Increasing concentrations of sodium
nitroprusside evoked relaxations after precontraction with phenylephrine (1026M) in the presence of indomethacin (1025M) in both endothelium-intact and endothelium-denuded vessels. No significant difference in relaxation to increasing concentrations of sodium nitroprusside was noted.
FIGURE 2 (A) Following precontraction with
phenylephrine (1026M), increasing concentrations of calcium ionophore produced no significant relaxations in pulmonary arteries denuded of endothelium. In endothelium-intact vessels, increasing concentrations of calcium ionophore A23187 evoked progressive relaxations. Relaxations to calcium ionophore A23187 were significantly reduced in canine pulmonary arteries with endothelium after 60 or 90 minutes of exposure to porcine serum compared to vessels exposed to canine serum for 90 minutes. (B) In endothelium-intact coronary arteries exposed to either canine or porcine serum for 60 minutes, oxyhemoglobin (1026M) abolished the relaxations to low concentrations of calcium ionophore A23187 (1029–1026M) after precontraction with phenylephrine (1026M) in the presence of indomethacin (1025M).
to calcium ionophore, as it had with acetylcholinemediated relaxations.
Preservation of EndotheliumIndependent Relaxations Impaired relaxations to acetylcholine and calcium ionophore following heterologous serum exposure
could be caused by reduced production of nitric oxide from the endothelium or impairment in the ability of the vascular smooth muscle to respond. To exclude the latter, pulmonary arteries precontracted with phenylephrine (1026M), were stimulated with sodium nitroprusside, a spontaneous donor of nitric oxide. Arterial segments exposed to either autologous or heterologous serum relaxed similarly with increasing concentrations of sodium nitroprusside (1029–1024M), suggesting that, in fact, the ability of the smooth muscle to respond to nitric oxide was intact following heterologous serum exposure (Figure 3). Stated differently, the guanylate cyclase/ cGMP pathway was intact in all arterial segments.
Effect of Complement Inactivation Pulmonary artery segments exposed to heat-inactivated heterologous serum for one hour exhibited normal relaxations to acetylcholine (Figure 4A). Similarly, heterologous serum in which the complement cascade was blocked by immunoadsorption with anti-C3 antiserum had no effect on pulmonary artery relaxations to acetylcholine (Figure 4B).
Basal Production of Nitric Oxide The previous experiments demonstrated a reduction in the endothelial production of nitric oxide following heterologous serum exposure, an observation that could not be explained by structural degradation of the endothelium. This finding correlates with
182
Cable et al.
The Journal of Heart and Lung Transplantation March 1999
DISCUSSION
FIGURE 4 (A) Heat-inactivated porcine serum had no
effect on relaxations of endothelium intact vessels to increasing concentrations of acetylcholine. (B) Similarly, canine coronary artery rings responded normally to increasing concentrations of acetylcholine after 60 minutes of exposure to C3-inactivated porcine serum.
stimulated production of nitric oxide. To determine if basal production was also affected, we utilized an indirect determination of basal nitric oxide production as previously described.26,27 The addition of oxyhemoglobin (1026M) to the organ bath following a one-hour serum exposure resulted in a relative contraction of the pulmonary arteries. Subsequent contraction of these arteries with phenylephrine (1026M) resulted in a tension of 5.2 6 0.1 gm in the autologous serum group and 4.8 6 0.2 gm in the heterologous serum group (n 5 5, p 5 .17). The calculated basal release of nitric oxide was 12.5 6 1.9% in the autologous serum group and 7.9 6 1.7% in the heterologous serum group (n 5 5, p 5 .005); this represented a 38 6 8% reduction of basal nitric oxide production with one hour of xenoserum exposure.
The purpose of this investigation was to determine if nitric oxide production of pulmonary arteries is impaired during simulated hyperacute xenograft rejection and if this impairment is attributable to complement activation. The major findings of the present study were: (1) exposure of canine pulmonary arterial segments to porcine serum for 60 and 90 minutes significantly attenuated endotheliumdependent vasodilatation; (2) exposure of canine pulmonary arteries to porcine serum did not alter endothelium-independent vasodilatation; and (3) inhibition of complement with either heat-inactivation or immunoadsorption prevented impairment of endothelium-dependent vasodilatation. The current study utilized an animal combination that has previously been shown to cause complement activation during hyperacute rejection.28 –32 Mazzoni et al demonstrated capillary thrombosis within 20 minutes in canine-to-porcine renal xenografts.33 Furthermore, we have previously described reproducible impairment of endothelial nitric oxide production attributable to complement activation in canine coronary arteries treated with porcine serum.19 We recognize that data from this animal combination may not be transferable to humans, but the model is useful in laboratory studies of hyperacute rejection. The present investigations demonstrated the reduction of endothelium-dependent relaxations to acetylcholine following exposure to heterologous porcine serum. Relaxations to sodium nitroprusside, a nitric oxide donor, were unchanged, suggesting that the function of the vascular smooth muscle was intact; ie, the ability of the vascular smooth muscle to relax to nitric oxide was not impaired by exposure to heterologous serum. This fact argues against nonspecific activation of membrane complexes and, subsequent, ubiquitous vascular injury. Rather, the failure of calcium ionophore to fully correct this impairment in endothelium-dependent relaxation signals a defect in the nitric oxide synthase pathway directly during simulated hyperacute rejection. Several pharmacologic tools are available to illustrate action of nitric oxide in vascular relaxation. Many are familiar with the L-arginine analogues, but two factors precluded their use in the present study. First, many have nonspecific actions, such as inhibition of L-arginine transport, which are difficult to predict and compound the experimental variables. Secondly, and more importantly for this study, they require a finite period of time to inhibit nitric oxide
The Journal of Heart and Lung Transplantation Volume 18, Number 3
production, averaging 45 minutes to 1 hour. As the current investigations demonstrated a time-dependent phenomena, it required a pharmacologic agent which would have immediate action. Oxyhemaglobin served this purpose. It allowed rapid inhibition of nitric oxide following serum exposure. It might be argued that reduced production of nitric oxide during hyperacute rejection might represent nonspecific cellular injury. Complement activation generates membrane attack complexes; these create pores, cellular swelling ensues, and nitric oxide production would be expected to decrease as the cell undergoes death. Thus, impaired nitric oxide generation could be a marker of cellular dysfunction, but this hypothesis is not supported by our data. Electron microscopic examination of rings after exposure to heterologous serum demonstrated only mild endothelial cell swelling without significant lysis. Basal production of nitric oxide was only partially reduced with heterologous serum. Furthermore, vascular smooth muscle function, as assessed by sodium nitroprusside response, was unchanged. Another explanation is that impaired nitric oxide production in pulmonary arteries following heterologous serum exposure was the result of selective damage to the nitric oxide pathway. It should be emphasized that in the current study we specifically excluded the effects of prostanoids by the use of indomethacin; Hammer et al previously demonstrated that stable metabolite of prostaglandin I2 is not elevated during hyperacute rejection.12 More importantly, though, we eliminated in our preparation the interaction of platelets to the endothelium. It is known that stimulation of endothelial cells during hyperacute rejection increases generation of platelet-activating factor, von Willebrand Factor, and glycoprotein GPIb,34 –36 as well as thrombin release, and this may promote platelet aggregation.37 It was not our intention in this study to evaluate all aspects of hyperacute rejection, but rather we sought to simplify our preparation so that we could specifically evaluate the endothelial production of nitric oxide. Differences were noted in the present findings compared to our earlier work with canine coronary arteries.19 First, the pulmonary endothelium appears more susceptible to the injury induced by heterologous serum exposure because a shorter duration of exposure produced significant impairment of endothelium-dependent relaxations in pulmonary compared to coronary arteries. Pulmonary arteries were impaired after only 30 minutes of exposure, but coronary arteries required 60 minutes
Cable et al.
183
of exposure. Secondly, the present study demonstrated a reduction in both receptor-dependent and receptor-independent relaxations following heterologous serum exposure; canine coronary arteries demonstrated only a significant impairment of receptor-dependent relaxations. It is not clear whether this represents an initial attack on the receptor/Gprotein complex by complement activation or as suggested in our previous communication, a significant contribution of endothelium-derived hyperpolarizing factor (EDHF) to relaxations in coronary arteries. If it is true that EDHF relaxations are unaffected by complement activation, as previous work has demonstrated,19 then the nitric oxide synthase pathway would be the only endotheliumderived relaxing factor impaired during hyperacute rejection. In conclusion, basal release of nitric oxide in pulmonary arteries is reduced following xenoserum exposure, and stimulated endothelium-dependent relaxations are attenuated. By electron microscopy, there were minimal structural alterations of endothelial cells. Attenuation of nitric oxide release would further enhance the prothrombotic state of hyperacute rejection because nitric oxide promotes platelet disaggregation and inhibits platelet aggregation. Moreover, a reduction in nitric oxide during hyperacute rejection would be expected to lead to vasoconstriction, leukocyte adhesion, and augmentation of reperfusion injury. The authors wish to thank Marilyn Oeltjen for devoted technical support and Donna Stucky for the diligent preparation of this manuscript.
REFERENCES 1. Gewurz H, Clark DS, Finstad J, et al. Role of the complement system in graft rejections in experimental animals and man. Ann NY Acad Sci 1966;129:673. 2. Zhow XJ, Niesen N, Pawlowski I, et al. Prolongation of survival of discordant kidney xenografts by C6 deficiency. Transplantation 1990;50:896. 3. Brauer RB, Baldwin III WM, Daha MR, et al. Use of C6-deficient rats to evaluate the mechanism of hyperacute rejection of discordant cardiac xenografts. J Immunol 1993; 151:7240. 4. Platt JL, Fischel RJ, Matas AJ, et al. Immunopathology of hyperacute xenograft rejection in a swine-to-primate model. Transplantation 1991;52:214. 5. Weisman HF, Bartow T, Leppo MK, et al. Soluble human complement receptor type 1: in vivo inhibitor of complement suppressing post-ischemic myocardial inflammation and necrosis. Science 1990;249:146 –51. 6. Rus HG, Niculescu F, Vlaicu R. Presence of C5b-9 complement complex and S-protein in human myocardial areas with necrosis and sclerosis. Immunol Lett 1987;16:15–20. 7. Mulligan MS, Poley MJ, Bayer RJ, et al. Neutrophil-depen-
184
8.
9.
10.
11.
12.
13.
14.
15.
16. 17.
18.
19.
20.
21.
22.
Cable et al.
dent acute lung injury: requirement for P-selectin (GMP140). J Clin Invest 1992;90:1600 –7. Ma X, Weyrich AS, Lefer DJ, et al. Diminished basal nitric oxide release after myocardial ischemia and reperfusion promotes neutrophil adherence to coronary endothelium. Circ Res 1993;72:403–12. Platt JL, Vercellotti GM, Lindman BJ, et al. Release of heparan sulfate from endothelial cells: implications for pathogenesis of hyperacute rejection. J Exp Med 1990;171: 1363– 8. Platt JF, Dalmasso AP, Lindman BJ, et al. The role of C5a and antibody in the release of heparan sulfate from endothelial cells. Eur J Immunol 1991;16:15–20. Vanhoutte PM, Boulanger CM, Mombouli JV. Endotheliumderived relaxing factors and converting enzyme inhibition. Am J Cardiol 1995;76:3E–12E. Hammer C, Schutz A, Pratschke J, et al. Bridging to transplant: allogeneic heart transplantation after xenografting. J Heart Lung Transplant 1992;11:S182– 8. Ignarro LF, Buga GM, Wood KS, et al. Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide. Proc Natl Acad Sci USA 1987;84:9265. Radomski MW, Palmer RMJ, Moncada S. Comparative pharmacology of endothelium-derived relaxing factor, nitric oxide and prostacyclin in platelets. Br J Pharmacol 1987;92:181. Rodomski MW, Palmer RMJ, Moncada S. The anti-aggregating properties of vascular endothelium: interactions between prostacyclin and nitric oxide. Br J Pharmacol 1987;92:639. Schafer AI, Alexander RW, Handin RI. Inhibition of platelet function by organic nitrate vasodilators. Blood 1980;55:649. Gryglewski RJ, Plamer RM, Moncada S. Superoxide anion is involved in the breakdown of endothelium-derived relaxing factor. Nature 1986;320:454 – 6. Rubanyi GM, Vanhoutte PM. Superoxide anions and hyperoxia inactivate endothelium-derived relaxing factors. Am J Physiol 1986;250:H822–7. Cable DG, Hisamochi K, Schaff HV. Complement mediates attenuation of endothelium-dependent relaxations in canine coronary arteries following porcine serum exposure: a mechanism for vascular thrombosis in xenograft hyperacute rejection. Circulation 1997;96:II58 – 64. Kaiser L, Tithof PK, Williams JF. Depression of endothelium-dependent relaxation by filarial parasite products. Am J Physiol 1990;259:H648 –52. Pearson PJ, Schaff HV, Vanhoutte PM. Long-term impairment of endothelium-dependent relaxations to aggregating platelets after reperfusion injury in canine coronary arteries. Circulation 1990;81:1921–7. Pearson PJ, Schaff HV, Vahoutte PM. Acute impairment of endothelium-dependent relaxations to aggregating platelets following reperfusion injury in canine coronary arteries. Circ Res 1990;67:385–93.
The Journal of Heart and Lung Transplantation March 1999 23. Cohen RA, Shepherd JT, Vanhoutte PM. Inhibitory role of the endothelium in the response of isolated coronary arteries to platelets. Science 1983;221:273– 4. 24. Jonas W, Stankiewicz M. Sheep serum complement sensitisation of sheep erythrocyte-rabbit antibody complexes for hemolysis by guinea-pig complement plus EDTA or MgEGTA. Arch Immuno Ther Exp 1986;34:451– 60. 25. Luscher TF, Diederich D, Siebenmann R, et al. Difference between endothelium-dependent relaxation in arterial and in venous coronary bypass grafts. N Engl J Med 1988;319:462–7. 26. Yang Z, von Segesser L, Bauer E, et al. Different activation of the endothelial L-arginine and cyclooxygenase pathway in the human internal mammary artery and saphenous vein. Circ Res 1991;68:52– 60. 27. Kojda G, Noack E. Nitric oxide liberating, soluble guanylate cyclase stimulating and vasorelaxing properties of the new nitrate-compound SPM 3672. J Cardiovasc Pharmacol 1993; 22:103–11. 28. Giles GR, Boehmig HJ, Lilly J, et al. Mechanism and modification of rejection of heterografts between divergent species. Transplant Proc 1970;2:522–37. 29. Perper RJ, Najarian JS. Experimental renal heterotransplantation. III. In widely divergent species. Transplantation 1966; 4:377– 88. 30. Kux M, Boehmig HJ, Amemiya H, et al. Modification of hyperacute canine renal homograft and pig-to-dog heterograft rejection by the intra-arterial infusion of citrate. Surgery 1971;70:103–12. 31. Kux M. Hyperakute nierenabstobung mit und ohne antikorpernachweis. Chirurg 1975;46:262–7. 32. Stemberger H, Rauhs R, Thetter O, et al. Beeinflussung der nierenxenotransplantation durch vorperfusion mit spezifischen antikorpern. Wiener Klinische 1978;90:197–201. 33. Mazzoni G, Benichou J, Porters KA, et al. Renal homotransplantation with venous outflow or infusion of antigen into the portal vein of dogs or pigs. Transplantation 1977;24:268 –73. 34. Robson SC, Kopp C, Lesnikoski B. Platelets and xenograft rejection. Xenotransplant 1994;2:38 – 44. 35. Makowka L, Chapman FA, Cramer DV, et al. Plateletactivating factor and hyperacute rejection. The effect of a platelet-activating factor antagonist, SRI63– 441, on rejection of xenografts and allografts in sensitized hosts. Transplantation 1990;50:359 – 65. 36. Coughlan AF, Berndt MC, Dunlop LC, et al. In vivo studies of P-selectin and platelet-activating factor during endotoxemia, accelerated allograft rejection and discordant xenograft rejection. Transplant Proc 1993;25:2930 –1. 37. Robson SC, Siegel JB, Lesnikoski BA, et al. Aggregation of human platelets induced by porcine endothelial cells is dependent upon both activation of complement and thrombin generation. Xenotransplant 1996;3:24 –34.