The FASEB Journal express article 10.1096/fj.02-0668fje. Published online February 5, 2003.

Protection against lipopolysaccharide-induced endothelial dysfunction in resistance and conduit vasculature of iNOS knockout mice Sharmila D. Chauhan,* Gloria Seggara,† Phuong A. Vo,* Raymond J. MacAllister,‡ Adrian J. Hobbs,¶ and Amrita Ahluwalia* *Clinical Pharmacology, Barts and The London, Queen Mary’s School of Medicine, Charterhouse Square, London EC1M 6BQ UK; †Department of Physiology, University of Valencia,Valencia, Spain; ‡Centre for Clinical Pharmacology, University College London, 5 University Street, London WC1E 6JJ UK; ¶Wolfson Institute for Biomedical Research, University College London, Gower Street, London WC1, UK Corresponding author: Amrita Ahluwalia, Clinical Pharmacology, Barts and The London, Queen Mary’s School of Medicine, Charterhouse Square, London EC1M 6BQ UK. E-mail: [email protected] ABSTRACT Endothelial dysfunction is a characteristic of, and may be pathogenic in, inflammatory cardiovascular diseases, including sepsis. The mechanism underlying inflammation-induced endothelial dysfunction may be related to the expression and activity of inducible nitric oxide synthase (iNOS). This possibility was investigated in isolated resistance (mesenteric) and conduit (aorta) arteries taken from lipopolysaccharide (LPS)-treated (12.5 mg/kg i.v.) or salinetreated iNOS knockout (KO) and wild-type (WT) mice. LPS pretreatment (for 15 h, but not 4 h) profoundly suppressed responses to acetylcholine (ACh) and significantly reduced sensitivity to the NO donor spermine-NONOate (SPER-NO) in aorta and mesenteric arteries of WT mice. This effect was temporally associated with iNOS protein expression in both conduit and resistance arteries and with a 10-fold increase in plasma NOx levels. In contrast, no elevation of plasma NOx was observed in LPS-treated iNOS KO animals, and arteries dissected from these animals did not express iNOS or display hyporeactivity to ACh or SPER-NO. The mechanism underlying this phenomenon may be suppression of eNOS expression, as observed in arteries of WT animals, that was absent in arteries of iNOS KO animals. These results clearly demonstrate that iNOS induction plays an integral role in mediation of the endothelial dysfunction associated with sepsis in both resistance and conduit arteries. Key words: nitric oxide • resistance arteries • inflammation • vasorelaxation • sepsis

B

acterial sepsis is a systemic inflammatory state characterized by vascular smooth muscle dysfunction, leading to hypotension, inadequate tissue perfusion, and organ failure (1). The impaired contractile function of the smooth muscle and resultant decreased blood pressure have been attributed to increased NO-mediated dilatation, secondary to the induction and activity of the inducible isoform of nitric oxide synthase (iNOS) (2). Indeed, inflammation-induced iNOS expression has been identified in several species, including humans

(3). In addition, inhibition of iNOS activity restores blood pressure in experimental models of sepsis and reverses hypotension in human sepsis (4). Evidence also exists for endothelial dysfunction in sepsis and other inflammatory cardiovascular diseases (5, 6). Given the role of the endothelium in the regulation of local blood flow, hemostasis, and vascular permeability, aberrant endothelial function may also play a significant part in the pathogenesis of inflammatory cardiovascular disease. However, the mechanisms involved in inflammation-induced endothelial dysfunction are not fully characterized. One possibility is that iNOS-generated NO not only mediates smooth muscle hyporesponsiveness but is also responsible for endothelial dysfunction. The lack of selective inhibitors of iNOS has prevented the specific role of this isoform of NOS from being clearly determined. However, the availability of iNOS knockout (KO) mice presents an opportunity to investigate the role of this isoform in endotoxemia with greater specificity. In animal models, sepsis has been mimicked by administration of bacterial lipolysaccharide (LPS), a component of the cell wall of gramnegative bacteria. Two recent studies have demonstrated that the absence of iNOS in these animals protects against LPS-induced contractile hyporeactivity of both conduit and resistance arteries (7, 8); however, the impact of iNOS deletion on endothelial dysfunction has not been established. In the present study, we investigated the direct effects of LPS administration on conduit and resistance arteries of wild type (WT) and iNOS KO animals to clarify the role of iNOS in mediating endothelial dysfunction in endotoxemia. MATERIALS AND METHODS All compounds used were obtained from Sigma (Poole, UK) except U-46619 (Biomol) and LPS (Difco, Detroit, MI). U-46619 was dissolved in ethanol and then diluted in saline; all other drugs were prepared using saline (0.9%). The eNOS antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA), iNOS antibody from Becton and Dickenson (Worthing, UK), and COX-2 antibody from Cayman (Nottingham, UK). Animals and tissue collection Male (25–30 g) WT (C57BL/6J) and iNOS KO mice (9) were bred in-house. All experiments were conducted according to the Animals (Scientific Procedures) Act 1986, UK. Mice were pretreated with saline or LPS (12.5 mg/kg, i.v.) 4 or 15 h before they were killed by cervical dislocation. The thoracic aorta and mesentery were removed and placed in cold physiological salt solution (PSS) of the following composition (mmol/l): NaCl 119, KCl 4.7, CaCl2 2.5, MgSO4 1.2, NaHCO3 25, KH2PO4 1.2, and glucose 5.5. Third-order arteries were cleaned of surrounding fat and mounted in a isometric tension myograph (10), and aortic rings were suspended in 25-ml organ baths, both containing PSS maintained at 37°C and gassed with 5% CO2 in O2. Functional studies Following a 60-min wash period, mesenteric arteries were normalized using standard procedures (11), and aortic rings were maintained at basal tension of 0.3 g for 60 min. Subsequently, both resistance and conduit arteries were repeatedly contracted with KCl (125 mmol/l and 48 mmol/l, respectively) until responses were reproducible. Relaxation of U-46619 (1 µmol/l) precontracted

vessels to acetylcholine (ACh; 10 µmol/l) was used to determine endothelial integrity (vessels that relaxed by at least 50% were deemed endothelium-intact). Following this, cumulative concentration-response curves were constructed to either the thromboxane A2-mimetic, U-46619 (0.001–3 µmol/l), or phenylephrine (PE; 0.001–30 µmol/l) to determine the impact of LPS treatment on vasoconstrictor activity. Arteries were then precontracted with PE (aorta) or U46619 (mesenteric arteries) to produce a level of contraction equivalent to 80% of the maximal response to KCl, and cumulative concentration-response curves were constructed to ACh (0.001–10 µmol/l) or spermine-NONOate (SPER-NO; 0.001–3 µmol/l). Western blotting Aortae and mesenteries removed from animals treated as previously described were snap frozen in liquid nitrogen, pulverized at –80°C using a stainless steel pestle and mortar, and then resuspended in phosphate-buffered saline (PBS) containing EDTA (1 mmol/l), phenylmethylsulfonyl fluoride (1 mmol/l), leupeptin (0.2 mmol/l), and pepstatin A (0.05 mmol/l). Samples were centrifuged at 13,000 rpm for 5 min. The supernatant was removed and mixed in an equivalent volume of sample buffer (containing Tris-HCl, 20 mmol/l; EDTA, 2 mmol/l; sodium dodecyl sulfate (SDS), 20%; mercaptoethanol, 10%; glycerol, 20%; and bromophenol blue, 0.025%) and reduced by boiling for 4 min. Samples were then subjected to SDS-polyacrylamide gel electrophoresis (PAGE) (10%) followed by electrotransfer onto nitrocellulose membranes. Membranes were then incubated overnight with rabbit polyclonal iNOS antibody (1:2000 dilution), rabbit polyclonal eNOS antibody (1:3000 dilution), or rabbit cyclooxygenase 2 (COX-2) polyclonal antibody (1:3000 dilution). Following incubation with the appropriate horseradish peroxidase-linked second antibody, signals were detected using the electrochemiluminescence (ECL) detection system and autoradiographic film. Densitometric analysis was performed on scanned images (Hewlett Packard Scanjet) and analyzed using Psion software (National Institutes of Health, Bethesda, MD). Plasma NO measurement Blood samples were collected by intracardiac puncture from saline- and LPS-treated (12.5 mg/kg, i.v., 15 h) WT and iNOS KO animals. Samples were centrifuged at 13,000 rpm for 10 min, and the plasma was collected. Samples were analyzed for nitrite (NO2– ) and nitrate (NO3–) by using chemiluminescence as described previously (12). In brief, samples and standards containing NO2– and NO3– were first reduced to NO, which was then quantified after reaction with ozone, using a nitric oxide analyzer (NOA 280, Sievers, Boulder, CO). To determine total NO2– and NO3– concentrations, collectively termed “NOx,” we added samples to 0.1 M vanadium(III) chloride in 1 M hydrochloric acid refluxing at 90°C under nitrogen. NO2– concentrations were determined by adding samples to 1.5% potassium iodide in glacial acetic acid under nitrogen at room temperature. Concentrations of NO3– were calculated by subtraction of NO2– from NOx values. Calculations and statistics Tension is expressed as absolute tension in either milliNewtons (mN) or grams (g). ACh- and SPER-NO-induced relaxations are expressed as percentage reversal of induced tone. The potency of agonists is expressed as the negative logarithm of the EC50 (pEC50). All data are

shown as mean ±SE. Tests of significance between curves were conducted using two-way ANOVA for multiple comparisons or a Students t test for differences between two data groups, where P0.05) and KO animals (10.8±0.7 mN vs. 9.2±0.6 mN, n=7–31; P>0.05). Similarly, a 15-h pretreatment also had no effect on arteries of WT (8.7±0.5mN vs. 8.6±0.5 mN, respectively; n=32–34; P>0.05) and KO (10.8±0.7 mN vs. 9.8±0.4 mN; n=31–35; P>0.05) animals. Both PE and U-46619 produced concentration-dependent contraction of mesenteric arteries from saline-treated WT animals that was similar to the responses of arteries from KO animals. LPS treatment (both 4 and 15 h) reduced the potency and maximum contraction to PE (Fig. 1A) but had no effect on the response to U-46619 (Table 1; Fig. 1C) in vessels from WT animals. In contrast, LPS treatment did not alter the responses of vessels from KO animals to PE or U-46619 (Table 1; Fig. 1B and 1D). Vasodilator responses in mesenteric resistance arteries Both ACh and SPER-NO produced concentration-dependent relaxation of arteries from salinetreated WT animals that was similar to the responses of arteries from KO animals. LPS treatment (both 4 and 15 h) reduced the potency and maximum relaxation to ACh (Fig. 2A) and reduced the potency of SPER-NO (Fig. 2C, 15 h only) in vessels from WT animals (Table 1). In contrast, LPS treatment did not alter the responses of vessels from KO animals to ACh or SPER-NO (Table 1; Fig. 2B and D). Vasoconstrictor responses of conduit arteries LPS treatment for 4 h had no effect on contractile responses to potassium in arteries of WT (saline-treated: max=0.70±0.07 g; LPS-treated: 0.81±0.2 g, n=18) or KO (max=0.78±0.1 g, max=0.69±0.07 g, n=12–19; P>0.05) animals. However, unlike the resistance arteries, LPS treatment at 15 h resulted in a small but significant (P