INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Endotoxin-derived lipopolysaccharide (LPS) is an initiator of sepsis, causing systemic hypotension and acute respiratory distress syndrome (ARDS) in approximately 25% of patients (1). Pulmonary vascular resistance (PVR) and pulmonary arterial pressure (Ppa) are elevated in patients with ARDS, after correction for arterial hypoxemia (2). Similarly, Ppa is raised in animal models of acute lung injury (3), and treatment with LPS may abolish hypoxic pulmonary vasoconstriction (HPV) (4).

Nitric oxide (NO) is an endothelium-derived relaxant factor (5). In blood vessels, constitutive NO synthase (cNOS) is confined to the endothelium and is calcium-dependent (5). Endotoxin and certain cytokines induce calcium-independent NO Synthase (NOS) in endothelium (6) and vascular smooth muscle (6). Induction of NOS following administration of LPS is associated with overproduction of NO and vascular hyporesponsiveness in isolated rat thoracic aorta (7) and main pulmonary artery (PA) (8). Mice lacking the inducible NOS (iNOS) gene are protected against the hypotensive and lethal effects of LPS (9). Evidence implicating overproduction of NO in the pathogenesis of sepsis has led to the use of NOS synthase inhibitors in vivo in animal models and more recently in patients with septic shock. Hence, N^G-monomethyl L-arginine (L-NMMA) and N^W-nitro-L-arginine elevate systemic blood pressure by causing vasoconstriction, leading to decreased cardiac output and oxygen delivery (10). In patients with ARDS, inhaled NO decreases Ppa, PVR, and shunt fraction, without affecting cardiac output or systemic pressures (11).

Using isolated perfused lungs (IPL) from sham- and LPS-treated rats and the selective iNOS inhibitor aminoguanidine (AG) (12), we investigated the effects of NOS induction on pulmonary vascular reactivity. In order to demonstrate increased NOS activity in the peripheral lung following LPS administration and dose-dependent suppression of this activity by AG, NOS activity was assessed in whole lung tissue. The effects of LPS treatment and iNOS inhibition on pulmonary vasoconstriction induced by hypoxia, AII, and PGF₂ were examined.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue Preparation for the IPL

Male Wistar rats (300-350 g) were anesthetized with diazepam (0.6 mg/kg, intraperitoneally) and hypnorm (fentanyl 0.315 mg/ml and fluanisone 10 mg/ml, intramuscularly). The isolated, blood-perfused, and ventilated in situ rat lung preparation (13) was used with modifications (vide infra). LPS-treated rats were injected slowly with sodium heparin made up to a volume of 2 ml with Krebs' solution containing 4% bovine serum albumin (BSA). This supplemental fluid was used to compensate for the LPS-induced dilation of capacitance vessels; this volume was used as it enabled approximately the same blood volume (10 ml) to be obtained from sham- and LPS-treated rats. The animal was ventilated via a tracheotomy with normoxic gas (21% O₂, 5% CO₂, 74% N₂) by a fixed-volume, small-animal pump (Harvard, Kent, UK) with a tidal volume of 4 ml and frequency of 16 breaths/ min, maintaining the blood perfusate within a physiological pH range of 7.34-7.42. The preparation was perfused with 20 ml of mixed blood and KH containing 4% BSA. Blood gas tensions and hemoglobin concentrations were measured 10 min after the perfusion was established (Corning 178 pH/blood gas analyzer and Co-oximeter, Essex, UK) and were adjusted between 7.34-7.42 by adding small volumes of NaHCO₃ to the perfusate. Hemoglobin concentrations in the sham group were higher than those in the LPS group (6.35 ± 0.09 [n = 41] versus 5.93 ± 0.12 g/dl [n = 32], respectively, p = 0.007). After the institution of the perfusion circuit and adjustment of the pH according to blood gas analysis, a NOS inhibitor (L-NMMA or AG) or saline vehicle was added to the reservoir, after which a 45-min equilibration period was allowed to establish a stable baseline perfusion pressure. A flow rate of 18 ml/min was used to produce a Ppa within the range of in vivo measurements in the conscious rat (14), and at which the pressure-flow relationship is linear (data not shown). The partial pressure of oxygen in the perfusate and the baseline Ppa for the four groups are shown in Table 1. Oxygen tensions did not differ among the groups. Baseline Ppa was significantly elevated in the LPS compared with sham groups, both in the presence of vehicle and AG. AG caused a significant increase in baseline pressure only in preparations from LPS-treated rats (p = 0.0009).

In order to elicit HPV, the animals were ventilated with a gas mixture of 5% CO₂, 3% O₂, 92% N₂. The size of the pressor response to hypoxia varied with the number of challenges; the first was followed by three slightly larger responses that do not significantly differ from each other in size (data not shown). Changes in Ppa induced by hypoxia or vasoconstrictors were expressed as the increase over baseline Ppa.

The NOS Assay

The IPL was prepared from sham- and LPS-treated rats as described above. Following a 30-min equilibration period, vehicle (200 µl distilled water), AG (100 µM or 1 mM) or L-NMMA (1 mM) was added to the perfusate. After a further 45 min, change in Ppa was recorded. The left lung was excised, frozen in liquid nitrogen, and stored at 70°  C until NOS activity was measured as described previously (15).

Functional Studies

Effects of LPS treatment and AG on angiotensin II and prostaglandin F₂-induced vasoconstriction. Following a 30-min equilibration period, vehicle (200 µl distilled water) or AG (1 mM) was added to the perfusate reservoir. After a further 45 min, changes in Ppa induced by AII (0.5, 1 and 2 µg) were recorded with at least 10 min between each recording. In separate preparations, PGF₂ (0.1 µM to 0.1 mM) was added consecutively.

Drugs

The following drugs were used: aminoguanidine hemisulphate, angiotensin II, bovine serum albumin, calmodulin, calcium chloride, Dowex-50W (sodium form), HEPES buffer, L-arginine hydrochloride, lipopolysaccharide from Salmonella enteritidis (code number L6011), L-valine, N^G-monomethyl-L-arginine acetate from Sigma (Poole, Dorset, UK); [³H]-L-arginine from Amersham International (Bucks, UK); ethylene-diamine-tetra-acetic acid from BDH Chemicals Ltd (Dagenham, Essex, UK); hypnorm (fentanyl 0.315 mg/ml and fluanisone 10 mg/ml) from Janssen (Wantage, UK); midazolam hydrochloride from Roche Products Ltd (Welwyn Garden City, UK); and prostaglandin F₂ from Upjohn (Crawley, UK).

Statistics

Results are expressed as mean ± standard error of the mean. Comparisons between means are made using Student's unpaired t test. p Value < 0.05 was considered to be significant for all tests.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of LPS Treatment and NOS Inhibitors on NOS Activity in Whole Lung

The results of the NOS assay are shown in Figure 1a and b. Levels of total and calcium-independent NOS activity were low in lungs from sham-treated rats but were elevated almost tenfold after LPS. AG caused dose-dependent suppression of LPS-induced NOS activity; AG and L-NMMA (1 mg) depressed NOS activity in LPS-treated specimens compared with that seen in sham-treated groups.

View larger version (15K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Figure 1.   Effects of LPS treatment and NOS inhibitors on total and calcium-independent NOS activity in whole rat lung. (A) Total NOS activity measured by [H3]-L-citrulline production from [H3]-L-arginine in whole lung from sham (open bars) and endotoxin (LPS) treated (hatched bars) rats perfused with blood in the absence (Control) or presence of the NOS inhibitors, aminoguanidine (AG), or NG-monomethyl-L-arginine (L-NMMA). Results are expressed as mean ± SEM of n observations; **p < 0.01, comparing preparations in the absence and presence of an NOS inhibitor. (B) Calcium-independent NOS activity in whole lung from sham (open bars) and LPS treated (hatched bars) rats perfused with blood in the absence (Control) or presence of the NOS inhibitors, aminoguanidine (AG), or NG-monomethyl-L-arginine (L-NMMA).

Effects of LPS Treatment and AG on HPV

The HPV response in preparations from LPS-treated rats was greater than that in sham-treated animals (p < 0.001 comparing sham versus LPS with and without AG) (Figure 2). AG caused an insignificant (p = 0.07) increase in HPV in preparations from sham-treated animals, but a very marked increase in the LPS-treated groups (p < 0.001), suggesting that LPS- induced NO attenuates, but does not overcome, an increased HPV response.

View larger version (22K): [in this window] [in a new window]   Figure 2.   Effects of AG on hypoxic pulmonary vasoconstriction (HPV) in isolated, perfused lungs from sham- and LPS-treated rats. HPV is measured as the increase in pulmonary pressure above baseline (Ppa mm Hg) in preparations from sham (open bars) and endotoxin (LPS, hatched bars) treated rats in the presence and absence of aminoguanidine (AG). Results are expressed as mean ± SEM of at least eight observations; **p < 0.01, comparing the groups in the absence and presence of aminoguanidine.

Effects of LPS Treatment and AG on All and PGF₂-induced Vasoconstriction

As with HPV, the AII-induced increase in Ppa in preparations from LPS-treated rats was greater than those from sham-treated animals (p < 0.001 comparing sham versus LPS with and without AG) (Figure 3). AG (1 mM) caused an insignificant (p = 0.07) increase in the response to AII in preparations from sham-treated animals, but a very marked increase in the LPS-treated groups. The response to AII in the LPS/AG group was not assessed because the very high Ppa induced by AII (1 µg) caused pulmonary edema. These data suggest that following LPS administration iNOS activity attenuates, but does not overcome, an increased sensitivity to AII.

View larger version (6K): [in this window] [in a new window]   Figure 3.   Effects of AG on angiotensin II-induced vasoconstriction in isolated, perfused lungs from sham- and LPS-treated rats. The response to angiotensin II (AII) is measured as the increase in pulmonary pressure above baseline (Ppa mm Hg) in sham (open symbols) and endotoxin (LPS: closed symbols) treated preparations in the absence (open square, closed square) and presence (open triangle, closed triangle) of aminoguanidine (AG). Results are expressed as mean ± SEM of at least eight observations; **p < 0.01, comparing the groups in the absence and presence of aminoguanidine.

PGF₂ dose-response relationships in the four groups are demonstrated in Figure 4. Responses in sham and LPS groups were not significantly different. AG (1 mM) enhanced PGF₂-induced contractions in both sham- and LPS-treated groups.

View larger version (10K): [in this window] [in a new window]   Figure 4.   Effects of AG on prostaglandin F2-induced vasoconstriction in isolated, perfused lungs from sham- and LPS-treated rats. The response to prostaglandin F2 (PGF2) is measured as the increase in pulmonary pressure above baseline (Ppa mm Hg) in sham (open symbols) and endotoxin (LPS: closed symbols) treated preparations in the absence (open square, closed square) and presence (open triangle, closed triangle) of aminoguanidine (AG). Results are expressed as mean ± SEM of at least eight observations; *p < 0.05, **p < 0.01, comparing the groups in the absence and presence of aminoguanidine. Tension produced by phenylephrine (1 µM).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Comparing sham- and LPS-treated rats, increased calcium- independent NOS activity in whole lung and AG-enhanced vasopressor responses suggest that iNOS is active in the pulmonary resistance vessels 4 h after LPS. These findings are consistent with our previous (9) demonstration of NOS induction in main PA. Despite evidence of increased NO synthesis, increased sensitivity to AII and hypoxia was demonstrated, as has been reported in the mesenteric circulation of LPS-treated rats (16). This finding contrasts with the hyporesponsiveness demonstrated by isolated main PA (9) and thoracic aorta from identically treated rats (8), the pulmonary circulation in conscious rats that have undergone caecal ligation and puncture (14), and in the systemic circulations of a variety of animal models of sepsis. By contrast with previous studies (4), enhanced HPV has recently been reported in isolated PA exposed in vitro to LPS (17) and in the isolated blood-perfused rat lung after low-dose TNF (18). In common with the above results, the latter authors also reported that PGF₂-induced vasocontraction was unchanged following exposure to LPS.

Inhibition of iNOS with AG caused a small increase in baseline pressure and a large increase in sensitivity to AII and hypoxia. These findings are in contrast to those in which nonselective NOS inhibition failed to augment HPV in the conscious rat after caecal ligation and puncture (14), abolished the augmentation of HPV in isolated LPS-exposed PA (17), and selectively augmented HPV without affecting AII-induced contraction in the untreated, isolated, perfused rat lung (18).

NOS induction protects against LPS-induced PHT but also impairs HPV, thereby having the potential for both beneficial and deleterious effects in patients with ARDS. The consequences of using NOS inhibitors in patients might require the simultaneous administration of inhaled NO.