INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

NITRIC OXIDE (NO) is a potent vasodilator synthesized from the semiessential amino acid L-arginine by a family of enzymes, NO synthases (NOS) (14). These isoenzymes are generally subclassified into constitutive forms (cNOS), which are continuously expressed basally and mediate normal homeostatic regulation in the organism, and an inducible form (iNOS), which is expressed in response to inflammatory stimuli, such as cytokines and bacterial lipopolysaccharide, and contribute to an organism's response to disease. The inducible isozyme of NOS is expressed in inflammatory diseases such as sepsis and is believed to account, at least in part, for the vascular hypocontractility and loss of vascular tone observed in this syndrome (8). This is supported by investigations using "selective" inhibition of iNOS, which prevents the decrease in blood pressure and the accumulation of the metabolic end products of NO, nitrite and nitrate, in the plasma from septic animals (1, 7, 9, 21, 22). By contrast, the predominant role of NO in the systemic hypotension characteristic of sepsis is not supported by investigations, which have demonstrated sepsis-induced hypotension in knockout mice that do not express the iNOS gene (11). Thus the role of induced production of NO in the hypotension of sepsis remains controversial.

To investigate the role of iNOS in sepsis, it is essential to selectively block this isozyme. The first generation of NOS inhibitors used modified analogs of the natural substrate L-arginine, such as N^G-nitro-L-arginine methyl ester (L-NAME) and N^G-monomethyl-L-arginine, to competitively block enzyme activity. These inhibitors are nonselective for the different isozymes of NOS (17). The next generation of NOS inhibitors are more selective for particular isozymes, such as 7-nitroindazole for the neuronal cNOS, N^G-nitro-L-arginine for the endothelial cNOS, and S-methylisothiourea (18) and aminoguanidine for the inducible isozyme (17). Rather than acting as a competitive inhibitor of NOS, aminoguanidine has been reported to act as a mechanism-based or suicide inhibitor, which forms an enzyme-activity-dependent covalent attachment to the NOS enzyme (19, 20). Aminoguanidine blocks iNOS activity at concentrations ~100 times less than that required to block the constitutive isozymes in vitro (13, 16). We have demonstrated that aminoguanidine reverses the vascular hypocontractility induced by incubation of vascular rings with lipopolysaccharide in vitro (16).

Although there are many reports documenting selective inhibition of iNOS in vitro (13, 19, 21, 22), there has not yet been conclusive evidence of selective iNOS inhibition in vivo. This is important because the selectivity of these inhibitors for iNOS vs. cNOS may vary in different tissues, and these inhibitors often inhibit cNOS at higher doses. Thus we used a clinically relevant model of sepsis, induced by cecal ligation and perforation (CLP) in rats, to examine the hemodynamic effects of aminoguanidine administration in vivo. Furthermore, we investigated the effect of increasing doses of aminoguanidine on constitutive and iNOS activities in lung and vascular tissue from these animals.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CLP model of sepsis. This experimental protocol was approved by the Animal Ethics Review Committee at the University of Western Ontario. After a 7- to 14-day acclimatization period in the animal housing facility, Sprague-Dawley rats (300-450 g) were instrumented for fluid and analgesic administration and blood pressure monitoring. Under general anesthesia (2% halothane with O₂), the rat was shaved in the abdominal region and in the dorsal and anterior cervical regions. A 0.3-cm incision was made in the dorsal cervical region for tunneling of the catheters. A 2-cm longitudinal incision was made in the anterior cervical region to expose the right external jugular vein and carotid artery. A Silastic catheter (0.047-in. OD, Silimed, Montreal, Quebec) was inserted into the superior vena cava via the jugular vein for administration of analgesic (fentanyl citrate, 2 µg/ml in 0.9% saline, 10 ml · kg¹ · h¹). A PE-50 polyethylene catheter (VWR Canlab, Mississauga, ON) was inserted into the carotid artery for monitoring of blood pressure and sampling of arterial blood. Catheters were tunneled through to the dorsal cervical region and passed through a harness-swivel device (Ealing Scientific, St.-Laurent, Quebec) sutured to the skin with 2-0 silk.

NOS assay. NOS activity was quantitated in vitro by the conversion of L-[³H]arginine to L-[³H]citrulline as previously described (16). Rat thoracic aorta and whole lung were homogenized in six volumes (wt/vol) of ice-cold homogenization buffer (pH 7.4; 10 mM HEPES, 0.1 mM EDTA, 1 mM dithiothreitol, and 0.32 M sucrose) by using a Brinkmann Polytron (Brinkmann Instruments, Mississauga, ON). Tissue homogenate (50-µl aliquots) was incubated in quadruplicate for 15 min at 37°C in the presence of incubation buffer {100 µl, pH 7.4; concentrations before addition of homogenate: 50 mM potassium phosphate; 60 mM L-valine; 1 mg/ml bovine serum albumin; 1 mM -NADPH; 10 µM FAD; 10 µM tetrahydrobiopterin; 30 µM L-[2,3-³H]arginine (200 counts · min¹ · pmol¹);and 1.2 mM MgCl₂}. When calcium-dependent (cNOS) NOS activity was assayed, 100 nM calmodulin and 2 mM CaCl₂ were added to the incubation buffer. When the calcium-independent (iNOS) activity was assayed, 1.2 mM EDTA and 1 mM EGTA were added to the incubation buffer. The reaction was terminated by the addition of 500 µl of cold stop buffer (4°C; ~pH 5.5; 100 mM HEPES; 12 mM EDTA). Dowex 50W (200-400, 8% cross-linked, Na⁺ form; 2 ml 1:1 wt/vol with water; ~pH 7.0) was added to the solution to remove any excess L-[³H]arginine. The test tubes were then centrifuged for 20 min at 600 g, and 0.5 ml of the supernatant was added to 4.5 ml of liquid scintillation fluid. The radioactivity was determined by liquid scintillation counting (Beckman Scientific Instruments, Mississauga, ON). The protein concentration of rat lung and aortic homogenates was determined by a microplate modification of the Bradford method, with bovine serum albumin as the standard and homogenization buffer as the blank (2).

1

1

Chemicals. Potassium phosphate monobasic was purchased from Fisher Scientific (Mississauga, ON). Cytoscint Environmentally Safe scintillation fluid, 1,4 dithiothreitol, phenylmethylsulfonyl fluoride, and Coomassie brilliant blue G250 were purchased from ICN Biomedical (Mississauga, ON). L-[³H]arginine was purchased from Amersham (Oakville, ON). All other reagents were purchased from Sigma Chemical (Sigma-Aldrich Canada, Mississauga, ON).

Statistical analysis. All data are expressed as means ± SE, and multiple groups were compared by using factorial ANOVA with Fisher's post hoc analysis. P values <0.05 were considered significant. All statistical tests were calculated by using the StatView-4.5 program on a Macintosh computer.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of CLP and aminoguanidine on mean arterial pressure (MAP). The induction of sepsis by CLP significantly decreased MAP at 21 h postsurgery compared with that in Sham animals (P < 0.0001) (Table 1). Aminoguanidine (1.75 and 17.5 mg/kg) did not alter MAP in either Sham or CLP animals [not significant (NS) compared with either animals not receiving aminoguanidine or to the pretreatment values]. Aminoguanidine (175 mg/kg) significantly increased MAP in both Sham and CLP animals (P < 0.05 for each group of animals compared with their preaminoguanidine baseline).

Effect of CLP and aminoguanidine on arterial blood gases, lactate, and leukocytes. Blood gases, arterial lactates, and white blood counts, at 24 h postsurgery, in samples from Sham and CLP animals administered aminoguanidine (0-175 mg/kg) are shown in Table 2. As we have previously reported with this model (4, 5, 12, 23), sepsis resulted in a significant increase in arterial lactate concentration and decrease in leukocyte count. Neither these values nor the blood gases were affected in any way by aminoguanidine treatment (NS compared with baseline samples). None of these parameters was altered in Sham animals administered aminoguanidine (17.5 and 175 mg/kg; data not shown).

Effect of CLP and aminoguanidine on NOS activities. Thoracic aortas from rats made septic by CLP had increased iNOS activity (Fig. 1A) and decreased cNOS activity (Fig. 1B) compared with aortas from Sham rats (P < 0.05 for both iNOS and cNOS activities). A low dose of aminoguanidine (1.75 mg/kg, n = 8) had no effect on cNOS and iNOS activities in CLP rats. By contrast, higher doses of aminoguanidine (17.5 and 175 mg/kg, n = 9 and 8, respectively) inhibited the CLP-induced increase in iNOS activity (P < 0.05 compared with CLP), whereas cNOS activities were unchanged (NS compared with CLP).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Inducible and constitutive nitric oxide synthase [iNOS (A) and cNOS (B), respectively] activities in thoracic aortas from sham-operated (Sham) and cecal ligation and perforation (CLP) rats administered aminoguanidine [AG; Sham: 0 (n = 8), 17.5 (n = 6), or 175 mg/kg (n = 8); CLP: 0 (n = 8), 1.75 (n = 8), 17.5 (n = 9), or 175 mg/kg (n = 8)]. Values are means ± SE. * P < 0.05 compared with 24-h Sham rats not receiving aminoguanidine. # P < 0.05 compared with CLP rats not receiving aminoguanidine (factorial ANOVA).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   iNOS (A) and cNOS (B) activities in lungs from Sham and CLP rats administered aminoguanidine. Doses and groups are as defined in Fig. 1. Values are means ± SE. * P < 0.05 compared with 24-h Sham rats not receiving aminoguanidine. # P < 0.05 compared with CLP rats not receiving aminoguanidine (factorial ANOVA).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results have demonstrated that administration of aminoguanidine in vivo improved MAP in a rat model of sepsis, although only at doses higher than that which selectively inhibits iNOS. We have also shown that sepsis results in a depression of cNOS activity compared with sham-operated animals in all tissues studied. Furthermore, at appropriate doses, aminoguanidine can selectively inhibit the sepsis-induced elevation of iNOS activity in both the thoracic aorta and the lung, with no significant effect on cNOS activities in these tissues.

The finding that selective inhibition of iNOS is possible in vivo with aminoguanidine is supported by previous investigations using models of sepsis induced by endotoxin. With these models, investigators have reported a decrease in plasma nitrate and nitrite levels after administration of doses of aminoguanidine similar to those used in the present study (15, 18, 22) and have demonstrated selective iNOS inhibition in vitro by using different tissues (18). Furthermore, Alden et al. (1) have recently demonstrated that aminoguanidine decreased plasma concentration of NO metabolites in a rat model of peritoneal sepsis. However, our work is the first to measure both cNOS and iNOS activities in tissues from septic and control animals after administration of aminoguanidine. These data indicate that aminoguanidine selectively inhibits iNOS in the thoracic aorta and lung to varying degrees. The reason for this variability may be related to regional heterogeneity in the distribution, uptake, and/or metabolism of the drug.

In septic animals, we found decreased cNOS activity in all tissues studied. This depression of cNOS activity in sepsis is supported by previous work by us (16) and others (10), in which endotoxin produced an upregulation of iNOS activity and mRNA, with a concomitant downregulation of cNOS activity and mRNA. This is also supported by the previous demonstration that cytokines, important in the induction of iNOS, also produce an instability of cNOS mRNA (24). Previous work using short-term (i.e., 5-h in vitro incubation) endotoxin stimulation of rat thoracic aorta has suggested that the resulting depression of cNOS activity is due to reversible inhibition of the constitutive isozymes by the elevated production of NO by iNOS (16). This has been supported by evidence indicating that cNOS isozymes are more susceptible than is the iNOS isozyme to NO "self-inhibition" by NO binding to the heme-iron core of the enzyme (3, 6). In contrast, investigations using the endotoxin model in vivo have shown that the elevation in iNOS expression is associated with a concomitant reduction in cNOS mRNA (10), hence a depression of cNOS activity that is not restorable through pharmacological means (i.e., through iNOS inhibition or by increasing substrate availability). Thus the finding in the present study, that 24 h of sepsis resulted in increased iNOS activity and that the depressed cNOS activity was not restored after selective inhibition of iNOS, is consistent with the previous in vivo observation (10) that there is a downregulation of cNOS mRNA and, consequently, of functional protein in sepsis. Thus it appears that the chronic (i.e., >12 h) in vivo models of sepsis produce downregulation of cNOS, whereas the acute in vitro model results in a reversible depression of cNOS activities.

Septic animals exhibited a decreased MAP compared with sham-operated animals. Similar to other investigators (1), a moderate dose of aminoguanidine [17.5 mg/kg in the present study; 25 mg/kg in the case of Alden et al. (1)] did not alter MAP in either septic or control rats. An improvement of MAP in septic animals was only observed after administration of a high dose (175 mg/kg) of aminoguanidine, much higher than the dose found to selectively inhibit iNOS in the thoracic aorta and lung. The lack of an effect on blood pressure of the moderate dose of aminoguanidine, which selectively inhibited iNOS activity in thoracic aortas and lung in vitro, would indicate either 1) that the aminoguanidine is not affecting the smaller resistance vessels that are responsible for the regulation of systemic blood pressure; 2) that the drug is not reaching or effective in the vascular beds of significance in the sepsis-induced hypotension; or, more likely, 3) that factors in addition to NO, such as vasodilator prostanoids or other vasoactive mediators, contribute to the vascular hypocontractility and resultant hypotension of sepsis. This latter hypothesis is supported by studies using iNOS knockout mice, which have provided compelling evidence that mechanisms in addition to iNOS induction contribute to the hypotension in sepsis (11).

In summary, our data show that aminoguanidine can selectively inhibit iNOS activity in vivo. The degree of this inhibition may be variable among tissues investigated, although further investigations are needed to examine this phenomenon. Although aminoguanidine did raise blood pressure in this model of sepsis, this was only achieved at a dose higher than that used to selectively inhibit iNOS activity. These findings further support the hypothesis that factors in addition to the induction of iNOS may contribute to the loss of vascular tone in sepsis.