Aortic clamping/unclamping, organ transplantation, and trauma-related surgery represent a growing number of surgical procedures that involve conditions of ischemia-reperfusion (IR).1,2 Circulatory resumption (reperfusion) is frequently associated with remote functional and metabolic repercussions,2 and several attempts have been made to attenuate them.3,4 Xanthine oxidase (XO), either alone or via its generated reactive oxygen species (ROS), is directly linked to such remote organ damage.2,3,5
Circulating XO and ROS interfere with normal vascular function in various species,2,6,7 a process that may lead to hazardous hemodynamic deterioration, especially if the vasculature responds abnormally to vasoactive substances. This process occurs even in the presence of an intact endothelial layer.7 A possible direct repercussion of liver IR on the normal vasculature has not been explored in depth before except for previous studies in which we demonstrated the loss of normal aortic tone and reactivity to phenylephrine (PE). Mannitol and methylene blue effectively mitigated such deleterious effects,8-10 suggesting that ROS are involved in this disturbance of normal vascular tone.
Glutathione in its reduced form (GSH) plays a central role in cellular defense against ROS and also acts extracellularly to scavenge the generated ROS.11 Tissue ischemia depletes intracellular GSH. Maintaining high cellular GSH levels or replenishing them reduces the magnitude of the destructive potentials of ROS.12 GSH intracellular buildup, however, requires that cysteine be supplied from the outside.13 N-acetyl-L-cysteine (NAC) is an exogenous GSH precursor. Its administration has been proven to protect against local post warm or cold IR-induced injury in various organs.12-15 We have recently shown15 that NAC significantly attenuated the development of IR-induced lung reperfusion injury when it was added to the reperfusate of the postischemic liver while traversing the isolated perfused normal lung, indicating that NAC exerted a direct scavenging activity and replenished lung GSH content.
Based on these previous observations, and given that NAC has a vasodilatory potential,16 we now used an ex-vivo double-organ (liver-aorta) model to assess (1) the behavior of the aorta in the presence and absence of incremental NAC doses both during and after its exposure to the perfusate of a post-ischemic liver; (2) the aortic response to PE in the presence or absence of NAC; and (3) whether the effect of NAC on the aorta is GSH-dependent.
MATERIALS AND METHODS
This study was performed in accordance with the PHS policy on Humane Care and Use of Laboratory Animals, the NIH Guide for the Care and Use of Laboratory Animals, and the Animal Welfare Act (7USC et seq). The protocol for animal care was approved by the Institutional Animal Care and Use Committee of Tel Aviv Sourasky Medical Center. The organs were harvested from adult male Wistar rats weighing 380-420 g (n = 160, 80 supplying the livers and 80 the aortas) and anesthetized by intraperitoneal sodium pentobarbital injection (50 mg/kg).
Isolated Liver Preparation
Liver isolation and perfusion were accomplished as previously described.3 Briefly, perfusion cannulas were placed in the portal vein (in) and in the suprahepatic inferior vena cava (out). The liver was then placed in an environmental chamber designed to control its temperature (37.5°C) and minimize evaporation. A pressure transducer (Statham Medical P132284, Mennen Medical, Inc, Clarence, NY), a digital thermometer (NJM-100, Webster Laboratories, Alta Dena, CA), and a bubble trap were set directly in front of the isolated liver, and ports for the collection of aliquots were positioned at the entrance and exit of the perfusate of the liver. All data were continuously recorded on a physiological recorder (Grass Model 7D polygraph, Grass Instruments, Quincy, MA) connected to a hemodynamic monitor (CS3™, Datex-Ohmeda®, Helsinki, Finland).
Livers were perfused via the portal vein with freshly prepared hemoglobin-free, modified Krebs-Henseleit (Krebs) solution. The incoming perfusate had a constant temperature of 37°C and a pH of 7.36-7.42 and was equilibrated with 95%O2-5%CO2. The flow rate was set at 4 mL/min/g liver ideal weight3 and readjusted during stabilization to maintain effluent pH and PCO2 within normal ranges. Liver perfusion pressure during stabilization was kept within known values of this specific model.3,9,10
Isolated Aortic Ring Preparation
As in previous experiments,9,10 the thoraxes of separately anesthetized rats were opened to quickly identify the descending thoracic aorta and excise a part of it. This piece was placed in a beaker containing oxygenated warm Krebs. After being cleaned, the aorta was cut to make a ring ∼2.5 mm wide. It was then mounted onto an electromechanical transducer (FT-03, Grass Instrument Co, Quincy, MA) and placed within a jacketed organ bath of 30 mL working capacity, whereupon changes in the ring's tone were continuously recorded. The buffer was constantly warmed to 37°C and equilibrated with 95%O2-5%CO2.
Specific Study Groups
Following a stabilization phase (30 minutes), the livers (8 replicates/group) were randomly assigned to 1 of 10 protocols. The livers of groups 1-5 (control, C) were perfused for 120 minutes with Krebs in the same manner as during stabilization. The livers of groups 6-10 (ischemic-reperfused, IR) were subjected to 120-minutes of warm (37°C) ischemia. All the livers were then reperfused for 5 minutes (ie, minutes 151-155 of perfusion). The liver reperfusate in both control and IR livers was diverted into the aortic ring's bath via a warmed (37.5°C) oxygenator instead of being discarded, as was done when the liver was perfused alone.3
N-Acetyl-L-cysteine (NAC, BF Goodrich Diamalt GmbH, Raubling, Germany) was administered in the designated groups at 1 of 4 doses: 25 mg/kg rat body weight (making 0.12 mM NAC in Krebs solution), 50 (0.25 mM), 100 (0.5 mM), or 150 mg/kg (0.74 mM). NAC was added to the hepatic perfusate during the reperfusion period following previous data showing that NAC would confer protection to the organs at the time the oxidant stress-generating process (ie, reperfusion) commences.8-10 NAC 0.12, 0.25, 0.5, and 0.74 mM was added to the perfusates of the control livers (groups 2-5) and to those of the ischemic livers (groups 7-10), respectively; groups 1 and 6 were the nontreated control and ischemic groups of livers, respectively. The control rings (attached to liver groups 2-5) were bathed with NAC to discern possible direct effects of the drug on the aortic ring tone independently from the effects of ischemia.
Aortic Ring Study Protocol
Figure 1 illustrates the ring study phases (stabilization, exposure, postexposure) and their correlation with liver reperfusion. Transition from one step to the next was performed after the ring's tone reached a steady state.
- Following its preparation, each ring was left to reach its maximal grade of relaxation.
- The ring was stretched with a weight tension of 2.0 g to standardize its resting (isometric) tone.
- Acetylcholine (6 μM) was added to the bath to ensure endothelial integrity.
- Krebs exchange (3 times the bath volume, ie, 100 mL) with fresh, warm, oxygenated Krebs to enable the ring to reach a new basal tone.
- PE 1 μM was added to the solution to standardize the contracting potency of the rings.
- Rinse with Krebs (100 mL), allowing the ring to return to stabilization tone or attain a new resting tone.
When the above multistep stabilization phase ended, the rings were ready for the double-organ experiment.9
- The ring was exposed to the liver reperfusate for as long as needed to attain a new steady tone state.
- PE at the original dose was added to the bathing solution to test the ring's contraction potency.
- Bath solutions were exchanged (100 mL) with fresh, warm, oxygenated Krebs, and the rings were left to reach a new resting tone.
This part of the experiment investigated the ring's response to PE after the potentially damaging agents were removed from the bath. Following a repeated acetylcholine test to preclude endothelial injury, steps 2-3 in the Exposure phase were repeated. The experiment then terminated.
The total XO activity (the reduced and oxidized enzyme forms) in hepatic reperfusates and in fresh ring tissue was analyzed following the procedure of Hashimoto17 with modifications. Reduced glutathione (GSH) content in the ring tissue was analyzed in all samples using a specific assay kit (Calbiochem 354102, San Diego, CA). The bath's Krebs O2 and CO2 partial pressures and the pH were measured using the AVL® OMNI 8™ modular system (AVL, Graz, Austria) adapted for asanguineous solutions.
Aortic Ring Tone Assessment
Variations in the ring's tone during the various phases of the study were expressed in absolute force units (g). The time required for each ring to regain constant tension levels during each study phase was later calculated as reported previously.9 Briefly, the first derivatives of the polynomial fit equation that expressed the rate of the change (ie, decrease) in the measured tone of the rings in each group were evaluated for each t point (where t expresses time in minutes) for up to 20 minutes. The values thus obtained were then compared among the groups: the larger the number, the faster was the rate of change in the ring's loss of tension (ie, relaxation).
The data of the variables are summarized as mean ± SD. A 2-way analysis of variance (ANOVA) with repeated measures was used to compare changes in a ring tension or in enzyme concentrations in the reperfusates. Post-hoc comparisons were applied at determined time points, using the Student Newman-Keuls test. The regression curves that evaluated the degree of relaxation were also compared by ANOVA. The Student t test was used to compare various indices within the ring's tissues. The significance level was set at P < 0.05.
The pH, and the CO2 and O2 partial pressures in the aortic ring bathing solutions were always within known normal ranges.9Table 1 displays the total activity of XO and that of GSH within the various liver reperfusates just before entering the rings' beakers. Although both enzymes remained unchanged in all nonischemic liver perfusates (groups 1-5), they both changed (P < 0.01) during the reperfusion of the IR livers (groups 6-10). The higher the NAC concentration, the lower was the activity of XO, but not the content of GSH. The XO activity was the highest in the IR-0 solutions, whereas the GSH content was the highest in the IR-0.12 solutions.
Figure 2 summarizes the total XO activities and the GSH levels in the tissues of the various aortic rings. Although XO activity was similarly low in all the rings bathed in the nonischemic perfusates, there was an increase (P < 0.01) in the activity within the IR-exposed rings. NAC was associated with lower XO activities in the IR-treated groups only (P < 0.01), the lowest value being registered in the IR-0.5 and the IR-0.74 rings compared with the IR-0 (group 6) rings. GSH values in all the control rings (groups 1-5), including the NAC-treated ones, were similar, although the C-0.74 ones contained more (P < 0.01) GSH than all other control rings. The levels of GSH in the IR-exposed rings increased to a maximal value in the IR-0.25 group and then decreased as the NAC concentration in the solution increased. The GSH values in the IR-0.5 and IR-0.74 rings were lower (P < 0.01) than those of the IR-0.25 ones. Overall, the XO values were inversely proportionate, albeit not in a linear manner, to the trends seen in the GSH pools and to the NAC doses. There was no correlation between NAC and GSH tissue contents.
Aortic Ring Tone Data
Acetylcholine similarly reduced the aortic tone of all rings (data not shown). They all contracted similarly when standardized with PE and followed an identical spontaneous return to basal tone within the same time frame after rinsing with Krebs solution (Fig. 3).
The incubation of the rings with the various nonischemic effluents demonstrated a dilatory effect of the various NAC concentrations (P < 0.001, Fig. 4). Comparatively, when exposed to the various IR reperfusates, the NAC-containing solutions protected the rings in a dose-dependent manner, which, however, was also associated with NAC dose-dependent vasodilatation. Specifically, IR-0 solution caused an immediate ring spasm followed by a protracted contraction; the generated tension was more intense than that recorded among all the IR- plus NAC-treated rings (Fig. 4). The IR- plus NAC-0.25-treated rings (P < 0.01) reached a plateau tension similar to the values recorded before perfusion or in the nonischemic perfusate-bathed rings (groups 1-5). The IR-0.5 and the IR-0.74 rings displayed an initial and brief contraction only. Approximately 3 minutes later, resting tension started to differ among the groups: the IR-0.5 rings relaxed by ∼50% of the contraction of the IR-0 rings, reaching a plateau within ∼7 minutes that was still (P < 0.01) different from all IR and control ring groups except for the IR-0.74 rings, whose tone restoration was ultimately similar to their control counterparts but different from all other IR rings.
Figure 5 illustrates the rings' response to PE while they were under the effect of the hepatic reperfusates. The control-0 rings (group 1) contracted to the same magnitude recorded during stabilization, whereas the NAC-treated control rings (groups 2-5) responded proportionately less. In contrast, PE generated a NAC-dose-proportionate reduction in contractility among the IR rings (groups 6-10), first observed 5-7 minutes after the addition of PE. PE effect in the IR-0 rings (group 6) was only ∼18% of that obtained by the C-0 (group 1) aortic rings. The IR-0.12 and the IR-0.25 rings' reactions were similar, and both contracted better (P < 0.001) than the IR-0 rings. The PE-generated contraction in the IR-0.5 and IR-0.74 rings was (P < 0.01) less than in all other groups of rings, including IR-0 ones.
Overall, the time necessary for the IR-0 to regain maximal relaxation at the end of exposure to liver solution and to PE was the fastest compared with all other IR rings. In the IR-0 and the IR-0.12 rings, the time-dependent rates of tension readjustments were characteristically devoid of the gradual process of relaxation observed in all the controls and partly in the IR- plus NAC-0.25- or IR-0.5 mM-treated rings. The first derivatives of the polynomial fit equation of the process of relaxation during minutes 5-10 after the test solutions were replaced with fresh Krebs displayed values that were the highest in the IR-0 group, followed by marginal differences in the IR-0.12 rings (0.35, -0.25, -0.18, and -0.23, -0.17, -0.12, respectively, for minutes 6, 7, and 8). Conversely, the calculated values for the IR-0.5 and IR-0.74 rings were low and almost indistinguishable from those of the controls (-0.12, -0.099, -0.077 and -0.11, -0.09, and -0.081, respectively, for minutes 6, 7 and 8, P < 0.001 versus IR-0 and IR-0.12 rings), representing a slow, control-like rate of progressive relaxation. The IR-0.25 values were between the 2 slopes of relaxation progression. The abrupt decrease in the tone (IR-0, IR-0.12) continued during the first 15 minutes of the recovery from the exposure phase. Thus, the first derivatives of the polynomial fit equation of the process of relaxation for the IR-0 rings continued to be the highest: -0.142, -0.137, -0.064, 0.126, and 0.63 (minutes 9, 10, 11, 12, and 15 of relaxation) compared with -0.055, -0.043, -0.037, 0.038, and 0.063 (P < 0.001) in the IR-0.5 rings that actually relaxed 10 times slower.
During the postexposure study phase of the experiment, the control rings previously exposed to Krebs plus NAC demonstrated a residual dilatory tendency on addition of PE, whereas IR aortal rings still displayed an abnormal tone regulation, albeit associated with partial recovery. Specifically, the control rings (groups 1-5) now exposed to plain (ie, no NAC) Krebs and PE contracted better (P = 0.001) than during the exposure phase and in dose-related linearity: by +6% (C-0), +51% (C-0.12), +70% (C-0.25), +183% (C-0.5), and +450% (C-0.74). IR rings now also generated (P < 0.001) more force in response to PE (+31% [IR-0], +44% [IR-0.12], +92% [IR-0.25], +869% [IR-0.5], and +590% [IR-0.74]) than in the previous phase; however, these new values were (P < 0.001) lower (by ∼35%) than their control counterparts (IR-0.25 > IR-0.5 > IR-0.74 > IR-0.12 > IR-0).
The present study demonstrates that NAC affords protection from acute reperfusion dysfunction in aortic rings in a dose-dependent fashion. NAC attenuates the prolonged aortic spasm and allowed for a more normal vascular response to PE not only during the ring's exposure to the postischemic liver reperfusate but also after its removal. The amounts of NAC in increasing doses are inversely related to XO activity in the circulation and in the aorta tissues but not to GSH content, indicating a major oxidant-quenching effect rather than a GSH precursory role. The altered response of the IR-0 aortas to PE after the damaging (IR) solution was removed (postexposure phase), albeit higher by an overall 31% than during the exposure itself, points to a prolonged but slowly reversible humoral inhibitory process that is attenuated by NAC.
It has been previously demonstrated that hepatic IR impairs aortic tone3,5,8 and that, despite its vasoconstrictor capability,18 methylene blue provides the aorta with dose-dependent protection not only during the phase of exposure to the liver reperfusate but also after it.9,18 Our hypothesis is that the oxidative process that had been initiated and maintained active by the high circulating XO would induce aortic tone impairment, possibly by its generated ROS byproducts in the presence of abundantly circulating purine substrates and free molecular oxygen.3,5,7,19 We now reasoned that NAC could protect the aortic tone via 1 of 2 modalities (or possibly a combination of them): (1) direct oxidant quenching, (2) GSH replenishment. Because NAC can contribute to nitric oxide (NO)-dependent vasodilation, we opted for doses lower (by 35%-50%) than those used in previous liver-lung studies15,20 and also used control rings this time.
The present data support the contention that the major beneficial effect of NAC in this IR double-organ model is via oxidant scavenging. XO directly correlated with NAC doses in an inverse manner, both in the solutions and in the rings' tissues. Despite the relatively higher XO activity in the IR-0.12 and IR-0.25 rings compared with the other 2 IR-treated rings, and despite the relatively weak vasodilatory effect that was registered in the control-0.12 and -0.25 rings, NAC-0.25 and -0.5 mM reduced to a minimum the abnormal contraction under the influence of the IR reperfusates and enabled an appreciable reaction to PE. These beneficial effects could reflect the quenching effect of NAC of the XO-generated oxidants in the circulation, as was suggested previously.15,20 The demonstrated lack of linearity between the NAC concentrations and GSH replenishment in the presence of oxidative stimuli is in agreement with previous data,21 where the velocity of contraction, of relaxation, and coronary flow were contained in the IR-treated hearts because of the protection afforded to the hearts by the oxidant-scavenging capacity of NAC in the circulation rather than because of GSH replenishment.
XO and GSH levels in the various solutions further clarify the role of NAC in protecting the aorta. The control-treated rings (except for C-0.74) had similar GSH contents. This can be explained by a maximal level of biosynthesis of GSH in the presence of NAC under normal conditions.15,20 Alternatively, the overall higher levels of GSH in the IR-treated groups could represent both GSH replenishment plus “cross-contamination” of GSH that originated from the damaged hepatocytes whose contents poured into the circulation, as was observed for XO or lactate aminotransferase.3 On the other hand, as was shown previously, the somewhat lower GSH in the IR-paired lungs compared with the control ones15,20 could be caused by the circulating perfusion system that was used in previous studies, whereas the present rings were bathed with the damaging solutions in a noncirculating medium. The exceptionally low content of GSH in the IR-0.74 rings corroborates the above contention: because of a lower circulating oxidant activity12 (and possibly less GSH leakage) within the solution (because of the intense quenching activity of NAC), this group of rings was less damaged (and dilated to a lesser extent), thus leading to a lesser grade of absorption of circulating compounds. These views explain the lack of direct NAC-to-GSH correlation.
The current findings could also point to the possibility of NAC's attenuating aortic tone impairment by interfering with ROS-mediated changes on the NO system. ROS were previously shown to reduce the availability of NO, thus leading to aortic endothelial dysfunction in an isolated aorta model preparation.22 This mechanism of tone impairment is supported by the recent demonstration that reduced synthesis of NO was involved in mediating rat mesenteric artery hyporeactivity23 and by the report that superoxide radicals added to the medium caused loss of endothelium-dependent arterial relaxation in a canine model.24 These findings are in accordance with the presently documented variations in the aortic tone and across-the-board proportionate regain of normal relaxation, especially in the IR-0.5 rings, irrespective of the GSH levels. Beckman et al25 also implicated a peroxynitrite radical-mediated mechanism for endothelial damage after ischemia-reperfusion. The suggestion of NAC averting ROS-induced NO-inhibited dilatory mechanism of action had been brought up in 2 previous studies15,20 and explains how, despite the lack of linearity of the GSH content in the various IR-treated rings, ring protection mixed with dilation correlated with the increasing doses of NAC. Noteworthy, it is our suggestion that in the presence of high oxidant levels, high NAC concentrations (as in IR-0.74 group) could associate with NO-induced peroxynitrite radical,25 which, consequently, would lead to GSH consumption and thus reduce its tissue pool.
As NAC scavenged circulating XO and reduced ROS generation, it might have also attenuated the contractile response to PE. Indeed, the minimal contraction of the IR-0.5 and the IR-0.74 rings when tested with PE during the exposure phase and the dilation obtained in those groups later when Krebs was refreshed would indicate an overriding vasodilatory effect over the PE contractile effect. Also, the fact that all NAC-treated rings, both controls and IRs, exhibited a short-lasting contraction when the exposure-phase solutions were refreshed with plain Krebs reflects an annulment of the residual effect of NAC-induced NO stimulation. Moreover, these appear to explain the differences between this NAC-ROS effect on NO compared with previously demonstrated methylene blue-ROS effects.8,9
A limitation of this study needs to be addressed: because a direct and precise measurement of free radicals in vivo, especially the hydroxyls, is very difficult to obtain because of their extremely short half-lives, specific ROS were not measured; thus, their direct interaction with NAC is lacking. We and others documented that an increase in XO activity would faithfully represent ROS damaging activity, so that XO reduction and GSH changes, such as those observed following the administration of antioxidants, would point to direct toxic effects of ROS.3,8,26,27
In conclusion, acute tone impairment of the rat aorta and an aberrant reaction to PE occur when exposed to the reperfusate of an ischemic liver. These effects did not completely reverse even after the toxic agents were removed from the circulation; they were, however, all reduced by NAC in a dose-specific manner. The mechanism(s) by which NAC protects the vascular tone against oxidative stress reperfusion injury seem to be oxidant quenching rather than a direct replenishment of GSH.
N-Acetyl-L-cysteine was kindly donated by BF Goodrich Diamalt GmbH, Raubling, Germany. Esther Eshkol is thanked for editorial assistance; my thanks to Dr. Perla Ekstein for her critical analyses and most valuable input on the composition and content of this paper.
1. O'Leary AM, Veall G, Butler P, et al. Acute pulmonary oedema after tourniquet release. Can J Anaesth
2. Gelman S. The pathophysiology of aortic cross-clamping and unclamping. Anesthesiology
3. Weinbroum A, Nielsen VG, Tan S, et al. Liver ischemia-reperfusion increases pulmonary permeability in rat: role of circulating xanthine oxidase. Am J Physiol
4. Kretzschmar M, Klein U, Palutke M, et al. Reduction of ischemia-reperfusion syndrome after abdominal aortic aneurysmectomy by N-acetylcysteine but not mannitol. Acta Anaesthesiol Scand
5. Weinbroum AA, Hochhauser E, Rudick V, et al. Direct induction of acute lung and myocardial dysfunction by liver ischemia-reperfusion. J Trauma
6. White CR, Darley-Usmar V, Berrington WR, et al. Circulating plasma xanthine oxidase contributes to vascular dysfunction in hypercholesterolemic rabbits. Proc Natl Acad Sci USA
7. Dowell FJ, Hamilton CA, McMurray J, et al. Effects of a xanthine oxidase/hypoxanthine free radical and reactive oxygen species generating system on endothelial function in New Zealand white rabbit aortic rings. J Cardiovasc Pharmacol
8. Weinbroum AA, Hochhauser E, Rudick V, et al. Multiple organ dysfunction following remote circulatory arrest: common pathway of radical oxygen species? J Trauma
9. Weinbroum AA, Kluger Y, Shapira I, et al. Methylene blue abolishes aortal tone impairment induced by liver ischemia-reperfusion in a dose response manner: an isolated-perfused double-organ rat model study. Shock
10. Weinbroum AA, Kluger Y, Rudick V. Impairment of aortal tone by no flow-reflow conditions and its partial amelioration by mannitol. Ann Thorac Surg
11. Stein HJ, Oosthuizen MM, Hinder RA, et al. Oxygen free radicals and glutathione in hepatic ischemia/reperfusion injury. J Surg Res
12. Fukuzawa K, Emre S, Senyuz O, et al. N-Acetylcysteine ameliorates reperfusion injury after warm hepatic ischemia. Transplantation
13. Nakano H, Boudjema K, Alexandre E, et al. Protective effects ofN-acetylcysteine on hypothermic ischemia-reperfusion injury of rat liver. Hepatology
14. Qiu Y, Bernier M, Hearse DJ. The influence of N-acetylcysteine on cardiac function and rhythm disorders during ischemia and reperfusion. Cardioscience
15. Weinbroum AA, Rudick V, Ben-Abraham R, et al. N-Acetyl-L-cysteine for preventing lung reperfusion injury after liver ischemia-reperfusion: a possible dual protective mechanism in a dose-response study. Transplantation
16. Harrison PM, Wendon JA, Gimson AE, et al. Improvement by acetylcysteine of hemodynamic and oxygen transport in fulminant hepatic liver. N Engl J Med
17. Hashimoto S. A new spectrophotometric assay method of xanthine oxidase in crude tissue homogenate. Anal Biochem
18. Bharadwaj LA, Prasad K. Mechanism of hydroxyl radical-induced modulation of vascular tone. Free Radic Biol Med
19. Barnard ML, Matalon S. Mechanisms of extracellular reactive oxygen species injury to the pulmonary microvasculature. J Appl Physiol
20. Weinbroum AA, Kluger Y, Abraham RB, et al. Lung preconditioning with N-acetyl-L-cysteine prevents reperfusion injury after liver no flow-reflow: a dose-response study. Transplantation
21. Weinbroum AA, Kidron A, Hochhauser E, et al. Liver glutathione level influences myocardial reperfusion injury following liver ischemia-reperfusion. Med Sci Monit
22. Ota Y, Kugiyama K, Sugiyama S, et al. Impairment of endothelium-dependent relaxation of rabbit aortas by cigarette smoke extract-role of free radicals and attenuation by captopril. Atherosclerosis
23. O'Brien AJ, Singer M, Clapp AH. Effect of endotoxin on vasoconstrictor responsiveness in rat mesenteric artery: role of inducible nitric oxide synthase. Intensive Care Med
. 1999;25(Suppl 1):S149.
24. Lawson DL, Mehta JL, Nichols WW. Coronary reperfusion in dogs inhibits ENDO dependent relaxation: role of superoxide radicals. Free Radic Biol Med
25. Beckman JS, Beckman TW, Chen J, et al. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci USA
26. Weinbroum AA, Goldin I, Kluger Y, et al. Methylene blue in preventing hemodynamic and metabolic derangement following superior mesenteric artery clamping/unclamping: an intratracheal vs. intraperitoneal dose-response study. Shock
27. Terada LS, Dormish JJ, Shanley PF, et al. Circulating xanthine oxidase mediates lung neutrophil sequestration after intestinal ischemia-reperfusion. Am J Physiol