Increased nitric oxide (NO) production is the cause of hypotension and shock during sepsis. In the present experiments, we have measured the contribution of endothelial (e) and inducible (i) nitric oxide synthase (NOS) to systemic NO production in mice under baseline conditions and upon LPS treatment (100 μg/10 g ip LPS). NO synthesis was measured by the rate of conversion of l-[guanidino-15N2]arginine to l-[ureido-15N]citrulline, and the contribution of the specific NOS isoforms was evaluated by comparing NO production in eNOS-deficient [(–/–)] and iNOS(–/–) mice with that in wild-type (WT) mice. Under baseline conditions, NO production was similar in WT and iNOS(–/–) mice but lower in eNOS(–/–) mice [WT: 1.2 ± 0.2; iNOS(–/–): 1.2 ± 0.2; eNOS(–/–): 0.6 ± 0.3 nmol · 10 g body wt–1 · min–1]. In response to the challenge with LPS (5 h), systemic NO production increased in WT and eNOS(–/–) mice but fell in iNOS(–/–) mice [WT: 2.7 ± 0.3; eNOS(–/–): 2.2 ± 0.6; iNOS(–/–): 0.7 ± 0.1 nmol · 10 g body wt–1 · min–1]. After 5 h of LPS treatment, blood pressure had dropped 14 mmHg in WT but not in iNOS(–/–) mice. The present findings provide firm evidence that, upon treatment with bacterial LPS, the increase of NO production is solely dependent on iNOS, whereas that mediated by cNOS is reduced. Furthermore, the data show that the LPS-induced blood pressure response is dependent on iNOS.
- nitric oxide
- stable isotope
- constitutive nitric oxide synthase
- inducible nitric oxide synthase
nitric oxide (NO) is one of the major mediators of the cardiovascular collapse that is associated with septic shock. Under baseline conditions, NO is produced mainly by constitutive nitric oxide synthases [neuronal NOS (nNOS) and endothelial NOS (eNOS); see Ref. 18], although inducible NOS (iNOS) is also known to be locally active in physiological conditions (20, 32). Treatment with lipopolysaccharide (LPS) induces iNOS (NOS2) in many tissues, including macrophages, vascular smooth muscle cells, and endothelial cells. iNOS is considered to be the high-capacity NO-producing enzyme that causes the fall in blood pressure during septic shock. However, this conclusion is based on the use of NOS inhibitors, because direct measurements of iNOS-mediated NO production were not available (12, 31).
A few years ago, Castillo et al. (3) determined the production of NO in vivo with an elegant stable isotope technique (3). This technique is based on the fact that NOS produces citrulline and NO in stoichiometric amounts, with each containing a nitrogen atom derived from the guanidino moiety of arginine. By infusing a l-[guanidino-15N2]arginine stable isotope, it is possible to measure the actual NO synthesis by measuring l-[ureido-15N]citrulline appearance (Fig. 1). Since then, this method has been used to measure NO production in several studies (4, 13, 29). In the present experiments, we aimed to measure the contribution of eNOS (NOS3) and iNOS to whole body NO production under baseline conditions and upon treatment with LPS, using the technique of Castillo et al. The respective contributions of iNOS and eNOS were deduced from the comparison of wild-type (WT) mice with mice engineered specifically to lack iNOS [iNOS(–/–) (22)] or eNOS [eNOS(–/–) (10)]. In addition, we assessed changes in blood pressure and heart rate upon LPS treatment in iNOS(–/–) mice.
Animals. Studies were performed in male WT (C57BL6/J), eNOS(–/–), and iNOS(–/–) mice (22–26 g, 2–3 mo old). The NOS knockouts were produced in a 129Sv background and crossed back six times with C57BL6/J mice. All strains were originally obtained from Jackson Laboratories, bred in the animal facility of the Academic Medical Center of the University of Amsterdam (Amsterdam, The Netherlands), and transported to the Centralized Animal Facilities of Maastricht University. They were kept in the new environment for at least 1 wk before experimentation. The mice were fed standard laboratory chow, had free access to water, and were kept at 25°C with a 12:12-h light-dark cycle (7:30 AM to 7:30 PM). Experiments were performed in accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals (28) and were approved by the Committee for Animal Research of Maastricht University.
Experiment 1: NO production in response to LPS treatment in WT, eNOS(–/–), and iNOS(–/–) mice. The experimental procedure is depicted in Fig. 2. At time = 0 h, food was withdrawn, and WT, iNOS(–/–), and eNOS(–/–) mice were randomly assigned to receive LPS (Escherichia coli O55:B5, 100 μg/10 g body wt ip) or a corresponding volume of saline (0.2 ml/10 g body wt ip). After LPS treatment (5 h), mice were anesthetized with ketamine and medetomidine, as described previously (8). Catheterization of the jugular vein and carotid artery was as described previously (8). During the surgical procedures, the mice were kept at 37°C by a temperature controller (Technical Service, Maastricht University) and heating pads.
A primed-constant infusion of l-[guanidino-15N2]arginine and l-[13C-ureido-5,5,2H2]citrulline (Mass Trace, Woburn, MA) was given in the jugular vein (per mouse: a priming dose of 850 nmol l-[guanidino-15N2]arginine and 215 nmol l-[13C-aureido-5,5,2H2]citrulline, followed by infusion of 1,700 nmol/h l-[guanidino-15N2]arginine and 430 nmol/h l-[13C-ureido-5,5,2H2]citrulline). When the steady state of tracers and metabolites was reached (at 20 min), blood (200 μl) was sampled from the carotid artery. Blood was centrifuged to obtain plasma, because we recently showed that plasma sampling is required in metabolic organ-balance studies with amino acid tracers that do not equilibrate well with blood cells (6). For determination of amino acid concentrations and tracer-to-tracee ratios (TTR; for definition, see Calculation of arginine and citrulline flux and NO production), 80 μl plasma were added to 7 mg dry sulfosalicylic acid, mixed, frozen in liquid nitrogen, and stored at –80°C. Plasma amino acid concentrations were determined as described previously (25). Amino acid TTRs were measured using a fully automated liquid chromatography-mass spectrometry system (26), with precolumn derivatization with 9-fluorenylmethylchloroformiate (27). Plasma nitrate was measured as described previously (2). In short, 100 μl acetonitrile were added to 50 μl plasma and mixed and centrifuged. Clear supernatant (60 μl) was added to 140 μl Super-Q water. This solution (100 μl) was injected in an IC-PackHR column (75 mm × 4.6 mm; Waters, Etten-Leur, The Netherlands). The column effluent was monitored at 205 nm. Nitrate was isocratically eluted from the column within 20 min with 6 mM NaCl and 1 mM KH2PO4 (pH = 6).
Calculation of arginine and citrulline flux and NO production. TTR is the ratio of labeled to unlabeled substrate and an equivalent of specific activity (30). Therefore, formulas were derived from metabolic studies using radioactive tracers. Plasma arginine and citrulline fluxes were calculated from the arterial TTR values of l-[guanidino-15N2]arginine and l-[13C-ureido-5,5,2H2]citrulline, respectively, using the steady-state isotope dilution equation (30) where Q is the flux and I is the rate of infusion of the tracer. NO synthesis was measured as the conversion of l-[guanidino-15N2]arginine to l-[ureido-15N]citrulline, as described by Castillo and coworkers (3, 4) and us (2, 7) where QCit is the plasma citrulline flux (nmol · 10 g–1 · min–1), estimated from the primed constant infusion of the l-[13C-ureido-5,5,2H2]citrulline tracer, TTRCit(M+1) is the ratio of l-[ureido-15N]citrulline to unlabeled citrulline, and TTRArg(M+2) is the ratio of l-[guanidino-15N2]arginine to unlabeled arginine (also see Fig. 1).
Experiment 2: Heart rate and blood pressure upon LPS treatment of WT and iNOS(–/–) mice. The experimental procedure is depicted in Fig. 2. Under anesthesia, a catheter was implanted in the femoral artery of seven WT and seven iNOS(–/–) mice, as described previously (11). Animals were allowed to recover from surgery for 2 days. On the 3rd day, food was withdrawn, and blood pressure and heart rate were recorded for 1.5 h in awake animals (11). Subsequently, LPS (E. coli O55:B5, 100 μg/10 g ip; Sigma, St. Louis, MO) was administered, and heart rate and blood pressure measurements continued for an additional 6 h. To maintain body temperature, the LPS-treated mice were kept under a thermostated heating lamp.
Statistical analysis. Comparisons were made between NOS knockout and WT mice under baseline conditions. In addition, the effect of LPS treatment was tested. The Mann-Whitney U nonparametric test was used to test for differences. For the hemodynamic data, an ANOVA with time as repeated factor and genotype as between factor was performed. An interaction between time and genotype on mean arterial pressure was present. Therefore, the difference between the genotypes was determined for each time point using Student's t-test. P < 0.05 was considered significant. Data are presented as means ± SE.
Contribution of eNOS- and iNOS-mediated NO production. TTR of l-[guanidino-15N2]arginine, l-[ureido-15N]citrulline, and l-[13C-ureido-5,5,2H2]citrulline are listed in Table 1. Under basal conditions, NO production was 1.2 ± 0.2 nmol · 10 g body wt–1 · min–1 in WT mice (Fig. 3). Baseline NO production was reduced to 50% in eNOS(–/–) mice, but iNOS deficiency did not change systemic NO production (Fig. 3). In response to LPS administration, systemic NO production increased more than twofold to 2.7 ± 0.3 nmol · 10 g body wt–1 · min–1 (Fig. 3). A similar increase in NO production was observed in eNOS(–/–) mice, but no increase in NO production was seen in iNOS(–/–) mice upon LPS treatment, indicating that the increased NO production during experimental endotoxemia is solely dependent on iNOS. In fact, LPS treatment decreased NO production in iNOS(–/–) mice to 50% of baseline values, suggesting that cNOS-mediated NO production was downregulated.
In WT and eNOS(–/–) mice, circulating arginine levels were similar under baseline conditions and increased after LPS challenge (Table 2). Resting arginine levels were higher in iNOS(–/–) mice and did not change in response to LPS. Citrulline levels were increased in LPS-challenged WT mice but not in eNOS(–/–) or iNOS(–/–) mice. Under baseline conditions, the plasma nitrate concentration was 36 ± 6, 28 ± 9, and 16 ± 1 μM in WT, eNOS(–/–), and iNOS(–/–) mice, respectively, and was not correlated with NO synthesis (Fig. 3). In response to treatment with LPS, the plasma nitrate concentration increased 3.2-fold and 3.5-fold, and NO production 2.3-fold and 3.6-fold in WT and eNOS(–/–) mice, respectively. In iNOS(–/–) mice, plasma nitrate was unchanged and NO production reduced to 50% in response to LPS treatment.
Under physiological conditions, NO production corresponded with 2.6% of arginine turnover in WT mice (Table 3). After LPS treatment, NO production was 5.7% of arginine turnover. Citrulline turnover was reduced in iNOS(–/–) mice challenged with LPS.
NO production and circulatory parameters. We then measured mean arterial pressure and heart rate in WT and iNOS(–/–) mice to determine whether NO production from iNOS affects circulatory parameters. Under basal conditions, blood pressure and heart rate were similar in WT and iNOS(–/–) mice (Fig. 4), in agreement with a similar NO production. After LPS challenge (5 h), blood pressure had decreased slightly but significantly in WT mice (14 mmHg; P < 0.05) but not in iNOS(–/–) mice (Fig. 4A). These data correspond with the difference in iNOS-dependent NO production in WT and iNOS(–/–) mice upon LPS treatment. Heart rate was not affected by iNOS gene disruption or LPS treatment (Fig. 4B).
The present study was undertaken to measure the contribution of iNOS and eNOS to systemic NO production under basal and LPS-stimulated conditions. The data support the theory that iNOS does not contribute substantially to systemic NO production under basal conditions. In addition, we showed that increased NO production upon treatment with bacterial LPS depends exclusively on iNOS and that cNOS-dependent NO production is downregulated upon LPS treatment. In agreement with earlier reports (23), blood pressure and heart rate were not different in iNOS(–/–) and WT mice under baseline conditions. Under these conditions, blood pressure was increased 15–20% (9, 23), and heart rate decreased 6%, in eNOS(–/–) mice (23). We did not measure blood pressure and heart rate in eNOS(–/–) mice, because NO production in these mice increased to a similar extent as in the WT group. The decreased blood pressure in septic shock appears to be mediated by the increased production of NO. Indeed, and as reported previously (17), arterial blood pressure decreased in WT mice but not in iNOS(–/–) mice upon LPS administration. The findings support earlier data (19) suggesting that pharmacological inhibition of NO synthesis prevents LPS-induced hypotension in rats.
We found that systemic NO production was similar in WT and iNOS(–/–) mice under basal conditions, indicating a marginal, if any, contribution of iNOS to systemic NO production in the basal unstimulated condition. This may be because of the SEs of the study. Baseline NO production was decreased to 50% in eNOS(–/–) mice compared with WT or iNOS(–/–) mice. In response to LPS treatment, systemic NO production was increased. The present study firmly establishes that the increase in systemic NO production was caused solely by an increase in iNOS-mediated NO production, since NO production was not increased in LPS-treated iNOS(–/–) mice. In fact, iNOS(–/–) mice, which are capable of NO production via eNOS and nNOS, had a significantly lower NO production rate than saline-treated iNOS(–/–) mice, indicating down-regulation of eNOS/nNOS-mediated NO production in response to LPS treatment. Our study therefore supports earlier studies that reported a reduction of eNOS/nNOS expression after treatment with LPS (14, 15). In addition to NO synthesis, systemic citrulline flux was reduced by LPS treatment in iNOS(–/–) mice. Systemic citrulline flux is determined partly by glutamine conversion to citrulline in the small intestine. Preliminary data from our laboratory indicate reduced glutamine uptake by the intestine in LPS-treated iNOS(–/–) mice (5), which may explain reduced systemic citrulline flux in LPS-treated iNOS(–/–) mice.
Under baseline conditions, plasma nitrate was not correlated with NO production. After LPS treatment, NO production was increased, and so was plasma nitrate. The increase in plasma nitrate was, however, not quantitatively related to the increase in NO production, suggesting that other factors besides NO production influence plasma nitrate. Plasma nitrate is determined by production (via NOS, but also bacterial synthesis) and excretion by the kidney. Renal function, which is likely affected by LPS treatment, correlates strongly with plasma nitrate (16). Kinetics between Hb and NO do not result in a straightforward stoichiometric production of nitrate (21) and further complicate the measurement of nitrate as an indicator of NO production.
In conclusion, the present findings provide firm evidence that, upon treatment with bacterial LPS, the increase of NO production is solely dependent on iNOS, whereas that mediated by cNOS is reduced. Furthermore, the data show that the LPS-induced blood pressure response is also dependent on iNOS.
This study was supported by Grants 902–23–098 and 902–23–239 from the Netherlands Organization for Scientific Research.
We thank J. L. J. M. Scheyen, H. M. H. van Eijk, and G. A. M. Ten Have for valuable assistance.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 by American Physiological Society