Septic shock is associated with protracted peripheral vasodilation and myocardial depression, both of which can be induced by endotoxin. Indeed, these cardiovascular effects can be reproduced experimentally in vitro by the incubation of vascular rings with endotoxin and in vivo by the intravenous administration of endotoxin in animals (1) and in humans (2). However, the mechanisms underlying these endotoxin-induced cardiovascular alterations are only partially understood.
In the cell, endotoxin ultimately decreases the binding of calcium to the myofilaments, the final common pathway for the modulation of muscular contraction. A decrease in the calcium influx through voltage-dependent channels has been implicated in this process (3–6). Because the current through these channels is tightly tuned by the transmembrane potential, endotoxin may influence the conductance of other ionic channels. Such a hypothesis is supported by the observation that exposure of aortic rings to endotoxin is associated with hyperpolarization (7).
The opening of the ATP-sensitive potassium (KATP ) channels under the influence of endotoxin may be a common pathway linking these electrical alterations with the relaxation of the vascular smooth muscle cell (VSMC) (8–10). Indeed, exposure to endotoxin is followed by the opening of the KATP channels (11,12), probably under the influence of nitric oxide (NO) and/or a decrease in the intracellular concentration in ATP (7,8,13). The contribution of the opening of KATP channels to the endotoxin-related vasodilation can be assessed pharmacologically by selective inhibitors, including sulfonylureas such as glibenclamide. This inhibitory effect may be selective, as the sulfonylurea receptor is spatially and functionally linked to the KATP channels (14).
The aim of the present study was to assess the effects of glibenclamide on the pressor response of isolated aortic rings to catecholamines and on global and regional hemodynamic parameters in a large animal model of endotoxic shock. An increase in the vasopressor effect of glibenclamide in the presence of endotoxin would support a role for the opening of KATP channels in the pathogenesis of the cardiovascular alterations observed in septic shock.
MATERIALS AND METHODS
These studies were approved by the local ethical committee for animal welfare.
Isolated vessels (Study A)
Thoracic aortas from Wistar rats, killed by a 5-min exposure to CO2, were cleared of periadventitial fat and were cut into rings of 3–4 mm length. Endothelium was removed from one-half of the rings (denuded rings) by gentle rubbing of the intimal surface, but was left intact in the others (intact rings). The rings were mounted in organ baths (20 mL) filled with warmed (37°C), oxygenated (95% O2/5% CO2) Krebs' solution consisting of (in millimoles) NaCl 118, KCl 4.7, KH2PO4 1.2, MgSO4 1.2, CaCl2 2.5, NaHCO3 25, and glucose 5.1.
Isometric contraction was measured with isometric transducers. A tension of 1 g was applied and the rings were equilibrated for 60 min. Rings were maximally contracted with potassium chloride (80 mM) and were then precontracted with norepinephrine (NE, 5.10−7 M). The absence of endothelium in the denuded rings was confirmed by the lack of acetylcholine-induced relaxation.
The rings were incubated with (ENDO+) or without (ENDO-) Escherichia coli endotoxin (100 ng/mL) for 4 h. Concentration-response curves to NE (5.10−11 to 5.10−7 M) were then obtained in the absence (GLI0) or in the presence of glibenclamide (50 and 100 μM added 5 min prior to the measurements, GLI50 and GLI100, respectively). The response to NE was recorded after careful washing to remove the previous solution.
A total of 72 rings were examined (six rings/group).
Endotoxic shock (Study B)
After being weighed, 25 mongrel dogs (25 ± 5 kg) were anesthetized with sodium pentobarbital at an initial loading dose of 30 mg/kg, followed by a constant intravenous infusion of 4 mg/kg/h (pump Infusomat; B. Braun, Melsungen, Germany) via an indwelling catheter (Abbocath 16G; Abbott Laboratories, Abbott Park, IL) placed in a right forepaw vein. After endotracheal intubation (internal diameter of the tube was 9 mm; Mallinckrodt, Phillips, NJ), the dogs were mechanically ventilated with room air (Servo ventilator 900B; Siemens-Elema, Solna, Sweden). Controlled ventilation was facilitated with pancuronium bromide given as an initial bolus of 0.15 mg/kg followed by an infusion of 0.075 mg/kg/h. Respiratory rate was set at 12 breaths/min and tidal volume was adjusted to obtain an end-tidal CO2 (Pet CO2, 47210A capnometer; Hewlett-Packard, Waltham, MA) between 28 and 35 mmHg. The ventilator settings were not changed during the experiment. A right femoral arterial catheter was inserted and connected to a pressure transducer for arterial pressure monitoring and blood sampling. A 7-F balloon-tipped pulmonary artery catheter (93A-131; Baxter, Irvine, CA) was inserted through the right external jugular vein under guidance of pressure waveform, as read on a four-channel monitor (Sirecust 302A; Siemens, Erlangen, Germany). Through a midline laparotomy, a splenectomy was performed to prevent autotransfusion of erythrocytes during hypotension. Ultrasonic flow probes were placed around the superior mesenteric (4–6 mm), left renal (3–4 mm), and left femoral (3–4 mm) arteries for simultaneous readings of the corresponding regional blood flows (mesenteric blood flow [MBF], renal blood flow [RBF], and femoral blood flow [FBF]).
Intravascular pressures, including mean arterial pressure (MAP), mean pulmonary arterial pressure (MPAP), balloon-occluded pulmonary artery pressure (PAOP), and right atrial pressure (RAP) were determined from a strip-chart recorder (2600; Gould, Cleveland, OH) at end-expiration. Cardiac output was measured by thermodilution (COM-2; Baxter, Irvine, CA) using three to five bolus injections of cold D5W, started at end-inspiration. A temperature probe was used online to control for variations in injectate temperature. Cardiac index (CI) was calculated as the ratio of cardiac output/weight. The regional blood flows were read simultaneously by a previously calibrated flowmeter (T208; Transonic Systems, Ithaca, NY). Fractional blood flows were calculated as measured flow/simultaneous cardiac output.
The study protocol is shown in Figure 1. The dogs were randomly allocated to one of the five experimental groups (five dogs/group). Dogs in groups A, B, and C received a bolus of endotoxin (2 mg/kg over 1 min) immediately after the baseline measurements. Immediately after the 60-min readings, the dogs in groups B through E received a 10-min intravenous infusion of glibenclamide at a dose of 0.15 (groups B and D) or 1.0 mg/kg (groups C and E). Standard therapy included a normal saline infusion at a rate titrated to maintain PAOP constant. At the end of the study, the animals were sacrificed by administration of a bolus of pentobarbital.
Recordings of hemodynamic parameters (intravascular pressures, cardiac output, and regional blood flows), arterial and mixed venous blood gases, and arterial plasma lactate and glucose were taken at baseline (time 0), and 30, 60, 120,180, and 240 min after baseline. Arterial and mixed blood samples were simultaneously withdrawn for immediate determination of blood gases (ABL 500; Radiometer, Copenhagen, Denmark), lactate, and glucose (2300 Stat Plus; Yellow Springs Instruments, Yellow Springs, OH).
Derived hemodynamic parameters (vascular resistance and stroke work) were calculated using the usual formulas (15). Oxygen delivery (DO2) and consumption (VO2) were calculated as the product of CI and arterial and arteriovenous difference content of oxygen, respectively.
Endotoxin and glibenclamide preparation
E. coli endotoxin (055:B5; Difco, Detroit, MI) was dissolved in normal saline at a final concentration of 10 mg/mL. Glibenclamide (Hoechst, Frankfurt, Germany) was dissolved in 1 N NaOH (at 60°C) and in distilled water at a final concentration of 1 mM (study A) or 1 mg/mL (study B), and was passed through a Millipore filter (0.22 μm), according to the recommendations of the manufacturer.
The data are expressed as mean ± SD. Results of the dose-response curves to NE are expressed as a percentage of the maximal contraction to KCl. The data recorded in dogs were analyzed over time and between groups. All data were analyzed by a two-way analysis of variance for repeated measurements and were adjusted for multiple comparisons by the Bonferroni's correction. A P value <0.05 was considered statistically significant.
Effects of endotoxin
Incubation with endotoxin was associated with a right- and downward shift of the concentration-response curve to NE. This effect was less pronounced in the intact (Fig 2, A versus C, IC50 10−8 vs. 5.10−7 M NE) than in the denuded (Fig. 2, B versus D, IC50 5.10−9 vs. 10−8 M NE) vessels.
Effects of glibenclamide
Without endotoxin (Fig. 2, A and B), the addition of glibenclamide did not significantly influence the concentration-response curve to NE, regardless of the state of the endothelium. In contrast, in the presence of endotoxin (Fig. 2, C and D), the addition of glibenclamide was followed by a downward shift of the concentration-response curve to NE in the intact (Fig. 2C, IC50 > 5.10−6 vs. 5.10−7 M NE), but not in the denuded (Fig. 2D, IC50 5.10−7 vs. 10−8 M NE) rings. The decrease in the contractile response of the intact rings was dose dependent.
Endotoxin administration was followed by significant decreases in MAP (113 ± 15 mmHg to 56 ± 20 mmHg, P < 0.01), PAOP (5.0 ± 2.7 mmHg to 2.9 ± 2.5 mmHg, P < 0.05), and cardiac output (225 ± 69 mL/min/kg to 90 ± 31 mL/min/kg, P < 0.01). SVR and pulmonary vascular resistance increased (1655 ± 368 dynes/s/cm−5 to 2123 ± 922 dynes/s/cm−5 and 131 ± 56 to 347 dynes/s/cm−5 ± 216 dynes/s/cm−5, respectively, both P < 0.05), whereas LVSW and RVSW decreased (51.2 ± 14.5 g to 11.2 ± 8.3 g and 4.7 ± 2.3 g to 1.9 ± 1.1 g, respectively, both P < 0.05). After 30 min of fluid resuscitation, MAP remained low (67 ± 13 mmHg, P < 0.05 versus baseline), whereas PAOP and cardiac output were restored to baseline values (4.3 ± 2.8 mmHg and 230 ± 88 mL/min/kg, respectively, not significant versus baseline). Accordingly, SVR decreased (1037 ± 565 dynes/s/cm−5, P < 0.05 versus baseline), LVSW and RVSW were partially and fully restored by the 30-min fluid treatment, respectively (31.6 ± 13.4 g and 5.7 ± 2.9 g, P < 0.05 and not significant versus baseline, respectively). The amount of normal saline required to keep PAOP constant throughout the experiment (5.0 ± 1.1 mmHg) averaged 13.2 ± 8.0 mL/kg. In the absence of treatment other than fluid administration (group A), these parameters remained constant throughout the study (Fig. 3, group A).
Regional blood flow
Endotoxin administration was followed by decreases in mesenteric, renal, and femoral blood flows (groups A-C pooled, all P < 0.05 versus baseline). The baseline values slightly differed, probably as a result of the small number of animals/group and differences in the weights. Blood flow distribution was altered, as fractional mesenteric and renal blood flows significantly increased after endotoxin (groups A-C pooled, 7.7% ± 3.3% and 3.8% ± 1.9% to 10.3% ± 5.0%, and 6.1% ± 2.3%, respectively, both P < 0.05 versus baseline), but fractional femoral blood flow was unchanged (from 2.4% ± 0.8% to 2.0% ± 0.9%, not significant). After the 30-min fluid resuscitation, the values of these parameters were restored to baseline and remained stable (Table 1, group A).
Oxygen-derived and metabolic parameters
Oxygen delivery decreased after endotoxin administration (groups A-C pooled, P < 0.05 versus baseline) and was restored to baseline after fluid administration. Oxygen consumption was unaffected by endotoxin and fluids. However, plasma lactate concentration increased early after endotoxin (from 1.5 ± 0.6 to 3.5 ± 1.8 mEq/dL, P < 0.05) and remained elevated despite fluid therapy in group A (Table 2).
Blood glucose concentration remained stable during the first hour after endotoxin administration and progressively declined in group A.
Effects of glibenclamide
The cardiovascular effects of glibenclamide differed in endotoxic shock and in control conditions. After 0.15 mg/kg glibenclamide, MAP remained stable, regardless of whether endotoxin was present (group B, from 59 ± 11 mmHg to 72 ± 1 mmHg; group D, from 99 ± 7 mmHg to 92 ± 16 mmHg), but cardiac output decreased in the absence of endotoxin (group D). After 1 mg/kg glibenclamide, MAP increased in the presence of endotoxin (group C, from 71 ± 1 mmHg to 88 ± 3 mmHg, P < 0.05), but decreased in the absence of endotoxin (group E, from 118 ± 13 mmHg to 92 ± 27 mmHg, P < 0.05). Cardiac output decreased in both groups (C and E). Hence, SVR increased in all groups except group B, whereas LVSW increased in group B and decreased in the other groups. Glibenclamide had no consistent effect on either pulmonary vascular resistance or on RVSW (data not shown).
Regional blood flows
Following glibenclamide, mesenteric blood flow decreased in groups C through E, but fractional mesenteric blood flow did not change. Renal blood flow decreased in the absence of endotoxin (groups D and E), but fractional renal blood flow was unchanged. Femoral blood flow decreased in group C and was unaltered in the others. Fractional femoral blood flow was unchanged.
Oxygen-derived and metabolic parameters
After glibenclamide, oxygen delivery progressively fell whether endotoxin was present or not. However, oxygen consumption fell only in the groups without endotoxin (groups D and E). Blood lactate concentrations returned to baseline following endotoxin and glibenclamide (groups B and C). Blood glucose levels decreased significantly in groups C and E.
The present study evaluated the effects of glibenclamide, an inhibitor of KATP channels, after exposure to endotoxin in vitro and in vivo. Interestingly, in vitro, glibenclamide decreased the contractile response to NE, when the endothelium was intact and endotoxin was present. In contrast, in vivo, glibenclamide (1 mg/kg) exerted vasopressor and negative inotropic effects that were not influenced by endotoxin; importantly, the vasopressor and negative inotropic effects of the low dose of glibenclamide (0.15 mg/kg) were attenuated and even reversed in the presence of endotoxin.
The hyporesponsiveness of isolated vascular rings to NE following exposure to endotoxin was of the same magnitude as has previously been described (16). An inhibitory effect of glibenclamide on the contraction of endothelium-denuded rat aortic rings to phenylephrine has been described, both in the presence and in the absence of endotoxin (17). Our findings are different because the vasorelaxant effect of glibenclamide required the integrity of the endothelium and the presence of endotoxin. As the concentrations of glibenclamide were similar in both studies, differences are likely due to different times of exposure to endotoxin. In another recent study using isometric force measurement, Chan et al. (18) reported that glibenclamide had a vasorelaxant effect in rat aortas in nonendotoxic conditions. These authors also noted that glibenclamide decreased the contractile response of isolated rat mesenteric arteries to phenylephrine (19). In these studies, these relaxing effects of glibenclamide were blunted after removal of the endothelium, and were prevented by NO inhibition. Similarly, in the present study, the presence of endotoxin unmasked a vasorelaxant effect of glibenclamide, possibly via an increased production of NO by the endothelium (20,21). Another mechanism for the endothelium-dependent nature of these effects might be the inhibition of an endothelium-derived hyperpolarizing factor, released under the influence of endotoxin (7).
Our in vivo model of endotoxin administration followed by fluid resuscitation reproduced the typical alterations of septic shock, including vasodilation, myocardial dysfunction, lactic acidosis despite maintained oxygen delivery, and hypoglycemia. There was a redistribution of blood flow favoring the mesenteric and renal vascular beds, which was reversible after fluid loading. We paid special attention to the maintenance of preload, as estimated by PAOP, to minimize the effects of a decrease in effective blood volume on cardiac output. In these conditions, glibenclamide administration was followed by an increase in blood pressure only during endotoxemia. This observation is in agreement with other studies that showed an increase in blood pressure following glibenclamide in endotoxemic animals, but did not explore the respective contributions of myocardial and vascular effects (22–26).
The present study investigated further the underlying mechanisms of the cardiovascular effects of glibenclamide during endotoxin shock. At a dose of 1 mg/kg, the increase in MAP was associated with an increase in SVR, indicating an increase in vascular tone. In contrast, myocardial performance, assessed by cardiac output and LVSW, decreased. In nonendotoxic conditions, the drop in MAP was primarily due to a decrease in myocardial performance, as reflected by decreases in cardiac output and LVSW. Glibenclamide has been shown to exert a negative inotropic effect secondary to myocardial ischemia in anesthetized dogs (22).
The decrease in mesenteric and renal blood flows that we observed following glibenclamide is consistent with other findings in conscious rats (27–29) and argues for the involvement of the KATP channels in the regulation of blood flow to these areas. Interestingly, the renal vasculature was less sensitive to the effects of glibenclamide when endotoxin was present than without endotoxin. However, the distribution of blood flow was probably minimally affected by glibenclamide, as the fractional flows were unchanged.
The lower blood lactate concentration following endotoxin and glibenclamide may be related to a better matching of oxygen cellular needs because oxygen consumption was maintained despite a decrease in oxygen delivery. This finding may also have been due to the hypoglycemia, which would have decreased the substrate availability for lactate formation. In this regard, the effects of endotoxin and glibenclamide on blood glucose concentrations were additive, suggesting that the pathways involved are different.
Taken together, the findings of the present study suggest that vascular tone can be regulated by glibenclamide during endotoxemia, presumably by counteracting the opening of KATP channels. The opening of KATP channels after exposure to endotoxin has been reported in a patch-clamp study (12), provided that arginine was present in the incubation bath. Importantly, the channel opening was reversed by NO and guanylate cyclase inhibition. In support of the concept that NO-mediated vascular relaxation involves the opening of KATP channels, glibenclamide can inhibit the vasodilation induced by peroxynitrite and acetylcholine (10,30).
The involvement of guanylate cyclase may also explain our findings, as in the same model, the administration of two inhibitors of guanylate cyclase, methylene blue (31) and oxadiazoloquinoxalinone (J. C. Preiser, Q. Sun, D. Madj-Sadok, and J. L. Vincent, unpublished data), induced a larger increase in blood pressure and cardiac output in endotoxic than in nonendotoxic conditions. The activity of guanylate cyclase might be increased by mediators other than NO, as the time between endotoxin and glibenclamide administration was too short to allow the activation of the inducible NO synthase (32,33).
The reasons for the contrasting observations in vitro and in vivo are of special interest. Differences in the polarization of the cells or the absence of neurally mediated regulation of the vascular tone in vitro could be proposed as explanations. Other possibilities may include interspecies differences in the responsiveness to endotoxin or to glibenclamide, or differences in the tissue concentrations of both substances.
In conclusion, the administration of glibenclamide exerted dramatically different effects in vitro and in vivo, suggesting that the effects of the opening of KATP channels during endotoxic shock are either masked by compensatory mechanisms or are of limited physiopathological importance.
Glibenclamide was a gift from Hoechst AG (Aventis), Frankfurt, Germany. The in vitro studies were paid for by FRSM (grant no. 3.4567.00).
The authors thank Pascale Jespers and Suzanne Foulon for their help with the in vitro studies.
1. Fink MP, Heard SO: Laboratory models of sepsis and septic shock
. J Surg Res 49:186–196, 1990.
2. Suffredini AF, Fromm RE, Parker MM, Brenner M, Kovacs JA, Wesley RA, Parrillo JE: The cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 321:280–287, 1989.
3. Bigaud M, Julou-Schaeffer G, Parratt JR, Stoclet JC: Endotoxin-induced impairment of vascular smooth muscle contractions elicited by different mechanisms. Eur J Pharmacol 190:185–192, 1990.
4. Ives N, King JW, Chernow B, Roth BL: BAY k 8644, a calcium channel agonist, reverses hypotension in endotoxin-shocked rats. Eur J Pharmacol 130:169–175, 1986.
5. Preiser JC, Moulart D, Cosyns B, Vincent JL: Administration of the calcium agonist BAY K 8644 in endotoxic shock. Circ Shock 35:199–206, 1991.
6. Wakabayashi I, Hatake K, Kakishita E, Nagai K: Diminution of contractile response of the aorta from endotoxin-injected rats. Eur J Pharmacol 141:117–122, 1987.
7. Chen SJ, Wu CC, Yang SN, Lin CI, Yen MH: Abnormal activation of K+
channels in aortic smooth muscle of rats with endotoxic shock: electrophysiological and functional evidence. Br J Pharmacol 131:213–222, 2000.
8. Standen NB, Quayle JM, Davies NW, Brayden JE, Huang Y, Nelson MT: Hyperpolarizing vasodilators activate ATP-sensitive K+
channels in arterial smooth muscle. Science 245:177–180, 1989.
9. Tare M, Parkington HC, Coleman HA, Neild TO, Dusting GJ: Hyperpolarization and relaxation of arterial smooth muscle caused by nitric oxide
derived from the endothelium
. Nature 346:69–71, 1990.
10. Wei EP, Kontos HA, Beckman JS: Mechanisms of cerebral vasodilation by superoxide, hydrogen peroxide, and peroxynitrite. Am J Physiol 271:H1262–H1266, 1996.
11. Colden-Stanfield M, Gallin EK: Modulation of K+
currents in monocytes by VCAM-1 and E-selectin on activated human endothelium
. Am J Physiol 275:C267–C277, 1998.
12. Miyoshi H, Nakaya Y, Moritoki H: Nonendothelial-derived nitric oxide
activates the ATP-sensitive K+
channel of vascular smooth muscle cells. FEBS Lett 345:47–49, 1994.
13. Philipson LH, Steiner DF: Pas de deux or more: the sulfonylurea receptor and K+
channels. Science 268:372–373, 1995.
14. Ammala C, Moorhouse A, Gribble F, Ashfield R, Proks P, Smith PA, Sakura H, Coles B, Ashcroft SJ, Ashcroft FM: Promiscuous coupling between the sulphonylurea receptor and inwardly rectifying potassium channels. Nature 379:545–548, 1996.
15. Zhang H, Rogiers P, Smail N, Cabral A, Preiser JC, Peny MO, Vincent JL: Effects of NO on blood flow distribution and oxygen extraction capabilities during endotoxic shock. J Appl Physiol 83:1164–1173, 1997.
16. Fleming I, Gray GA, Stoclet JC: Influence of endothelium
on induction of the l -arginine-nitric oxide
pathway in rat aortas. Am J Physiol 264:H1200–H1207, 1993.
17. Sorrentino R, d'Emmanuele DV, Lippolis L, Sorrentino L, Autore G, Pinto A: Involvement of ATP-sensitive potassium channels in a model of a delayed vascular hyporeactivity induced by lipopolysaccharide in rats. Br J Pharmacol 127:1447–1453, 1999.
18. Chan W, Yao X, Ko W, Huang Y: Nitric oxide
-dependent relaxation induced by glibenclamide in rat isolated aorta. Cardiovasc Res 46:180–187, 2000.
19. Huang Y, Chan NW: Involvement of endothelium
in relaxant action of glibenclamide on the rat mesenteric artery. Eur J Pharmacol 343:27–33, 1998.
20. Fleming I, Dambacher T, Busse R: Endothelium
-derived kinins account for the immediate response of endothelial cells to bacterial lipopolysaccharide. J Cardiovasc Pharmacol 20:S135–S138, 1992.
21. Szabo C, Mitchell JA, Thiemermann C, Vane JR: Nitric oxide
-mediated hyporeactivity to noradrenaline precedes the induction of nitric oxide
synthase in endotoxin shock. Br J Pharmacol 108:786–792, 1993.
22. Geisen K, Vegh A, Krause E, Papp JG: Cardiovascular effects of conventional sulfonylureas and glimepiride. Horm Metab Res 28:496–507, 1996.
23. Landry DW, Oliver JA: The ATP-sensitive K+
channel mediates hypotension in endotoxemia and hypoxic lactic acidosis in dog. J Clin Invest 89:2071–2074, 1992.
24. Vanelli G, Hussain SN, Aguggini G: Glibenclamide, a blocker of ATP-sensitive potassium channels, reverses endotoxin-induced hypotension in pig. Exp Physiol 80:167–170, 1995.
25. Vanelli G, Hussain SN, Dimori M, Aguggini G: Cardiovascular responses to glibenclamide during endotoxaemia in the pig. Vet Res Commun 21:187–200, 1997.
26. Wu CC, Thiemermann C, Vane JR: Glibenclamide-induced inhibition of the expression of inducible nitric oxide
synthase in cultured macrophages and in the anaesthetized rat. Br J Pharmacol 114:1273–1281, 1995.
27. Gardiner SM, Kemp PA, March JE, Fallgren B, Bennett T: Effects of glibenclamide on the regional haemodynamic actions of α-trinositol and its influence on responses to vasodilators in conscious rats. Br J Pharmacol 117:507–515, 1996.
28. Gardiner SM, Kemp PA, March JE, Bennett T: Regional haemodynamic responses to infusion of lipopolysaccharide in conscious rats: effects of pre- or post-treatment with glibenclamide. Br J Pharmacol 128:1772–1778, 1999.
29. Moreau R, Komeichi H, Kirstetter P, Yang S, Aupetit-Faisant B, Cailmail S, Lebrec D: Effects of glibenclamide on systemic and splanchnic haemodynamics in conscious rats. Br J Pharmacol 112:649–653, 1994.
30. Ma XL, Johnson III, G Lefer AM: Mechanisms of inhibition of nitric oxide
production in a murine model of splanchnic artery occlusion shock. Arch Int Pharmacodyn Ther 311:89–103, 1991.
31. Zhang H, Rogiers P, Preiser JC, Spapen H, Manikis P, Vincent JL: Effects of methylene blue on oxygen availability and regional blood flow during endotoxic shock. Crit Care Med 23:1711–1721, 1995.
32. Preiser JC, Zhang H, Vray B, Hrabak A, Vincent JL: Time course of inducible nitric oxide
synthase activity following endotoxin administration in dogs. Nitric Oxide
33. Preiser JC, Zhang H, Wachel D, Boeynaems JM, Buurman W, Vincent JL: Is the endotoxin-induced hypotension related to nitric oxide
formation? Eur Surg Res 26:10–18, 1994.