Hypoxic pulmonary blood flow redistribution and arterial oxygenation in endotoxin-challenged NOS2-deficient mice

Sepsis and endotoxemia impair hypoxic pulmonary vasoconstriction (HPV), thereby reducing arterial oxygenation and enhancing hypoxemia. Endotoxin induces nitric oxide (NO) production by NO synthase 2 (NOS2). To assess the role of NO and NOS2 in the impairment of HPV during endotoxemia, we measured in vivo the distribution of total pulmonary blood flow (QPA) between the right (QRPA) and left (QLPA) pulmonary arteries before and after left mainstem bronchus occlusion (LMBO) in mice with and without a congenital deficiency of NOS2. LMBO reduced QLPA/QPA equally in saline-treated wild-type and NOS2-deficient mice. However, prior challenge with Escherichia coli endotoxin markedly impaired the ability of LMBO to reduce QLPA/QPA in wild-type, but not in NOS2-deficient, mice. After endotoxin challenge and LMBO, systemic oxygenation was impaired to a greater extent in wild-type than in NOS2-deficient mice. When administered shortly after endotoxin treatment, the selective NOS2 inhibitor L-NIL preserved HPV in wild-type mice. High concentrations of inhaled NO attenuated HPV in NOS2-deficient mice challenged with endotoxin. These findings demonstrate that increased pulmonary NO levels (produced by NOS2 or inhaled at high levels from exogenous sources) are necessary during the septic process to impair HPV, ventilation/perfusion matching and arterial oxygenation in a murine sepsis model.

J. Clin. Invest.104:1421–1429 (1999).

The normal pulmonary vasculature constricts in response to alveolar hypoxia. In contrast, systemic vessels vasodilate in response to hypoxia. Hypoxic pulmonary vasoconstriction (HPV) preserves systemic arterial oxygenation by redistributing blood flow away from collapsed or poorly ventilated lung regions toward well-ventilated lung regions, thereby matching pulmonary ventilation (V) with perfusion (Q).

The mechanisms responsible for HPV are intrinsic to pulmonary vascular smooth muscle cells: HPV does not appear to require the presence of an intact endothelium (1). Vasoactive mediators released from a variety of pulmonary sources, including the endothelium and circulating cells, are important modulators of the pulmonary vascular response to hypoxia (2). Nitric oxide (NO) was identified as the endothelium-derived relaxing factor (EDRF) released from arteries and veins (3). NO synthases (NOSs) produce NO by conversion of L-arginine to L-citrulline. Constitutive NOSs, NOS1 and NOS3 (initially isolated from neurons and endothelial cells, respectively), are calcium/calmodulin dependent and produce low levels of NO that regulate homeostatic processes and contribute to the low pulmonary vascular resistance (PVR) (4, 5). Acute inhibition of endogenous NO production has been shown to enhance HPV in normal animals (6), and inhaled NO reverses human HPV (7).

In addition to its vasoregulatory function, NO participates in the response to inflammatory stimuli (8). Exposure to cytokines, such as IL-1β and TNF-α, or microbial products, such as endotoxin, stimulates cells to express an inducible NOS (NOS2) that generates greater amounts of NO than the constitutive isoforms (9). It is well established that the same inflammatory mediators that induce NOS2 also attenuate HPV, resulting in increased blood flow to poorly ventilated lung regions (V/Q mismatching), augmented right-to-left shunting of venous blood, and reduced arterial oxygenation (10).

The role of NO in the impairment of HPV during endotoxemia and sepsis is incompletely understood. The existing evidence is controversial, with reports suggesting that HPV is either increased (11, 12) or unchanged (13, 14) after acute inhibition of NO formation with L-arginine analogs during sepsis. After an endotoxin challenge, suppression of NOS2 expression with dexamethasone and inhibition of NOS2 activity by administration of L-arginine analogs restore vascular reactivity to vasoconstrictors in both systemic (15) and pulmonary vessels (16). However, the use of L-arginine analogs or dexamethasone to evaluate the role of NOS2 in the endotoxin-induced impairment of HPV is limited by the incomplete specificity of these agents for NOS2. To identify a role for NOS2 in the endotoxin-induced impairment of HPV, we compared the ability to redistribute pulmonary blood flow after left mainstem bronchus occlusion (LMBO) in wild-type mice and in mice with a congenital deficiency of NOS2. We report that NOS2 deficiency protects mice from endotoxin-induced impairment of HPV. In endotoxin-treated NOS2-deficient mice, replacement of molecular NO in the lung by breathing 40 parts per million (ppm) NO for 22 hours impaired HPV when measured 1 hour after discontinuing NO inhalation, but did not effect LMBO-induced pulmonary blood flow redistribution in endotoxin-treated wild-type mice or untreated mice of either strain.

Effects of unilateral alveolar hypoxia on pulmonary blood flow. To assess HPV in vivo, we developed a murine model in which differential pulmonary blood flow measurements could be obtained before and during left lung unilateral alveolar hypoxia (Figure 1). Because occlusion of the right mainstem bronchus caused severe hypoxemia and hemodynamic instability, we elected to occlude the left mainstem bronchus (LMBO), which produced a stable model of unilateral lung collapse. Differential pulmonary blood flow was measured at thoracotomy using an ultrasonic flow probe placed around the right pulmonary artery, and the flow distribution between the right and left lung was assessed by transiently occluding the left pulmonary artery (Figure 1). Transient occlusion of left pulmonary artery flow did not produce any systemic hemodynamic effect (data not shown). HPV was assessed as the percent decrease in QLPA/QPA at 5 minutes after LMBO. Measurements of pulmonary blood flow distribution using a flow probe correlated closely with measurements using intravenous injections of fluorescent microspheres (r² = 0.97; Figure 2a).

Figure 1

Representative tracings of pulmonary and systemic hemodynamic measurements before and during LMBO with and without transient left pulmonary artery (LPA) occlusion. Effects of unilateral alveolar hypoxia induced by LMBO on central hemodynamics before (a) and during LMBO (b and c) are portrayed. Total alveolar collapse occurred about 1 minute after LMBO. Online recordings of mean flow through the right pulmonary artery (QRPA), mean pulmonary artery pressure (P_PA), mean systemic arterial pressure (P_SA), and mean airway pressure (P_ALV) in a saline-treated wild-type mouse at baseline and after LMBO. To assess blood flow distribution between right and left pulmonary arteries, the left pulmonary artery was transiently occluded, at which point QRPA = QPA (see a and c). The difference of QPA – QRPA equals QLPA. Measurements were taken at baseline and 5 minutes after LMBO. Arrows indicate occlusion (90 seconds) and release of left pulmonary artery.

Figure 2

(a) Correlation of percent pulmonary blood flow to the left or right lung assessed by intravenous injection of fluorescent microspheres (15 μm) and by simultaneous measurement of QRPA with an ultrasonic flow probe. In the latter method, differential blood flow distribution between the right and left pulmonary artery was assessed by transient occlusion of the left pulmonary artery, as described in the legend to Figure 1. Values are expressed as the fractional flow to the right or left lung before and after LMBO (saline-treated wild-type mice; n = 4). Note the close agreement between the 2 methods (r² = 0.967). (b) The left lung pulmonary flow-pressure relationship before (baseline) and after 5 minutes of LMBO in saline-treated wild-type mice (n = 6). Note the significant increase of the slope, which represents an increased incremental left lung pulmonary vascular resistance induced by LMBO. Left pulmonary artery flow (QLPA) was measured by an ultrasonic flow probe, and the slopes were generated by reducing QPA with a transient occlusion of the inferior vena cava (P < 0.05, slope differs versus baseline). (c) Changes in total left lung pulmonary vascular resistance (TLPVR) during re-expansion of the collapsed left lung in saline-treated wild-type mice (n = 3). The left lung was inflated by a continuous injection of 5% CO₂ in N₂ up to a peak inspiratory pressure of 30 cm H₂O. Values are expressed as a linear regression of all data points with the respective 95% confidence intervals.

Before LMBO, hemodynamic parameters did not differ between wild-type mice and NOS2-deficient mice. Five minutes after LMBO, QLPA/QPA was reduced by 46 ± 5% in saline-treated wild-type mice and by 50 ± 7% in saline-treated NOS2-deficient mice (Figure 3). In supplemental studies in which QLPA was measured directly using a flow probe around the left pulmonary artery, we observed that the LMBO-induced pulmonary blood flow redistribution was reflected by an increase in incremental left pulmonary vascular resistance from 102 ± 18 mmHg•min•g•mL^–1 to 210 ± 38 mmHg•min•g•mL^–1 (Figure 2b). LMBO-induced pulmonary blood flow redistribution was not attributable to mechanical factors associated with left lung collapse, as left pulmonary vascular resistance (LPVR) did not change when the collapsed lung was reinflated to a normal lung volume under direct vision with 5% CO₂ in N₂ (n = 3; Figure 2c).

Figure 3

Change in fractional blood flow to the left lung (QLPA/QPA) 5 minutes after LMBO in wild-type mice (+/+) and NOS2-deficient mice (–/–) treated with saline (control, open bars; n = 7 each), endotoxin (endotoxin, closed bars; n = 6 each), or endotoxin with 5 mg/kg L-NIL intraperitoneally (endotoxin with L-NIL, striped bar; n = 5) 3 hours after the endotoxin challenge. (*P < 0.05, endotoxin versus control; ^‡P < 0.01, wild-type versus NOS2-deficient; ^#P < 0.01 endotoxin + L-NIL versus endotoxin alone). Measurements were obtained following 1 hour after discontinuation of breathing air or 40 ppm NO in air for 22 h (^§P < 0.05 NO versus without NO). Endotoxin challenge caused a marked reduction of pulmonary blood flow redistribution after LMBO in wild-type, but not in NOS2-deficient mice. After prolonged inhalation of 40 ppm NO, endotoxin-treated NOS2-deficient mice, measured 1 hour after discontinuation of NO inhalation, showed the same loss of HPV as endotoxin-treated wild-type mice. Saline-treated NOS2-deficient mice and wild-type mice breathing 40 ppm NO for 22 hours retained their ability to redistribute pulmonary blood flow after LMBO, when measured 1 hour after discontinuing NO inhalation.

Effect of endotoxemia on pulmonary and systemic hemodynamics during unilateral alveolar hypoxia. After challenge with intraperitoneal injection of 10 mg/kg E. coli endotoxin, mice appeared lethargic with piloerection and diarrhea. Approximately 50% of mice died within 1 week after endotoxin challenge, and there were no differences in mortality between wild-type and NOS2-deficient mice (data not shown). Before LMBO, hemodynamic parameters did not differ between saline-treated mice and endotoxin-treated mice (Table 1), but QPA tended to be higher in endotoxin-treated wild-type mice than in NOS2-deficient mice (230 ± 35 vs. 120 ± 32 μL•min^–1•g^–1 bw), although this difference did not reach statistical significance. Comparison of QPA for all studied endotoxin-challenged wild-type and NOS2-deficient mice confirmed that there was no difference between the two genotypes (wild-type mice 180 ± 20 μL/min per gram; NOS2-deficient mice 140 ± 15 μl/min per gram; n = 21 and 15, respectively). Because differences in QPA potentially affect pulmonary blood flow redistribution, we tested whether changes in QLPA/QPA depended on changes in QPA. Linear regression analysis revealed no correlation between QPA and QLPA/QPA during LMBO (r²=0.03, both for all mice and for all endotoxin-treated mice). Five minutes after LMBO, the fraction of blood flow directed to the hypoxic left lung (QLPA/QPA) was reduced by 46 ± 5% in saline-treated wild-type mice and by 18 ± 5% in endotoxin-treated wild-type mice (P < 0.001; Figure 3). In contrast, LMBO reduced QLPA/QPA by 50 ± 7% in saline-treated NOS2-deficient mice and by 51 ± 6% in endotoxin-treated NOS2-deficient mice (Figure 3). There was no difference in the response to endotoxin between SV129, SV129B6F1, or C57BL/6 wild-type strains or between NOS2-deficient mice with SV129/B6F1-hybrid background and NOS2-deficient mice backcrossed for 10 generations on a C57BL/6 background (data not shown).

Table 1

Hemodynamic measurements

Effect of endotoxemia on PaO₂ and PvO₂ during unilateral alveolar hypoxia. To determine whether the differences in LMBO-induced pulmonary blood flow redistribution noted in wild-type and NOS2-deficient mice after endotoxin challenge are reflected by differences in LMBO-induced changes in arterial oxygenation, the PaO₂ was continuously measured before and during LMBO. In separate endotoxin-treated wild-type and NOS2-deficient mice, the effect of LMBO on PvO₂ was also assessed. In saline-treated wild-type mice (n = 4), breathing at FiO₂ 1.0, LMBO for 5 minutes decreased the PaO₂ from 432 ± 7 mmHg to 225 ± 13 mmHg. The PaO₂ was 342 ± 23 mmHg in endotoxin-treated wild-type mice and 347 ± 33 mmHg in endotoxin-treated NOS2-deficient mice. After 5 minutes of LMBO, the PaO₂ decreased more in endotoxin-treated wild-type mice (145 ± 9 mmHg) than in endotoxin-treated NOS2-deficient mice (230 ± 18 mmHg; P < 0.05; Figure 4).

Figure 4

Effects of regional hypoxia induced by LMBO on the systemic arterial partial pressure of oxygen (PaO₂) in endotoxin-treated wild-type mice (thick line) and endotoxin-treated NOS2-deficient mice (narrow line). Endotoxin was administered 22 hours before study by an intraperitoneal injection of 10 mg/kg E. coli endotoxin. Continuous recordings of PaO₂ were obtained with a Clark-type oxygen electrode located in the aortic arch. Data are the mean of independent experiments with wild-type mice (n = 5) and NOS2-deficient mice (n = 3). Note that LMBO decreased the PaO₂ in both groups, but the decrease of PaO₂ was more marked in wild-type mice than in NOS2-deficient mice (P < 0.05; 5 minutes after LMBO).

PvO₂ did not differ in endotoxin-treated wild-type mice and endotoxin-treated NOS2-deficient mice (before LMBO: 39 ± 5 and 44 ± 8 mmHg, respectively; after LMBO: 34 ± 4 and 43 ± 7 mmHg, respectively; n = 3 for both groups).

Prolonged pharmacologic inhibition of NOS2 activity. To determine whether prolonged pharmacologic inhibition of NOS2 preserves HPV, wild-type mice (n = 5) were injected intraperitoneally with 10 mg/kg endotoxin, and 3 hours later were treated with L-NIL (5 mg/kg intraperitoneally), a selective inhibitor of NOS2 activity. This dose of L-NIL was sufficient to prevent the increase in pulmonary cGMP levels in wild-type mice 7 hours after endotoxin challenge (data not shown). Pulmonary blood flow studies were performed 22 hours after the endotoxin challenge. In endotoxin-exposed mice treated with L-NIL, we measured no differences in systemic and pulmonary hemodynamics before LMBO, compared with saline-treated or endotoxin-treated wild-type mice. The reduction of QLPA/QPA was greater in endotoxin-challenged wild-type mice treated with L-NIL than in wild-type mice treated with endotoxin alone (53 ± 10% vs. 18 ± 5%, respectively; P < 0.01; Figure 3).

Acute pharmacologic inhibition of NOS2 with L-NIL. To determine whether acute inhibition of NOS2 enzyme activity augmented HPV during LMBO, pulmonary blood flow studies were performed in saline-treated and endotoxin-treated wild-type mice during which L-NIL (5 mg/kg intravenously) was administered after LMBO. In saline-treated wild-type mice (n = 5), acute L-NIL administration did not further reduce QLPA/QPA during LMBO (49 ± 3% before L-NIL; 56 ± 3% after L-NIL). All other hemodynamic parameters did not change after L-NIL injection. In wild-type mice treated with endotoxin 22 hours earlier (n = 4), acute L-NIL administration did not further reduce QLPA/QPA during LMBO (22 ± 11% reduction before L-NIL; 24 ± 18% reduction after L-NIL).

Prolonged inhalation of 4 or 40 ppm NO. To learn whether increased pulmonary levels of NO, rather than another NOS2 product, contribute to the endotoxin-induced impairment of HPV, saline-treated and endotoxin-challenged wild-type and NOS2-deficient mice breathed 40 ppm NO in air for 22 hours. The hemodynamic studies were performed 1 hour later, allowing ample time for any potential vasodilator action of inhaled NO to dissipate. Inhalation of 40 ppm NO for 22 hours did not impair HPV in saline-treated wild-type mice and NOS2-deficient mice: reduction of QLPA/QPA after LMBO was 51 ± 10% in saline-treated wild-type mice and 50 ± 11% in saline-treated NOS2-deficient mice after breathing 40 ppm NO for 22 hours (Figure 3). Except for a modest difference in heart rate, no other differences in pre-LMBO hemodynamics were found between endotoxin-treated and saline-treated mice exposed to prolonged inhalation of 40 ppm NO for 22 hours (Table 1). After prolonged inhalation of 40 ppm NO, endotoxin-treated NOS2-deficient mice demonstrated a marked attenuation of HPV compared with endotoxin-treated NOS2-deficient mice that did not breathe supplemental NO (reduction of QLPA/QPA: 20 ± 7% vs. 50 ± 11%, respectively; P < 0.05; Figure 3). In contrast, prolonged NO inhalation did not affect the endotoxin-induced impairment of HPV in wild-type mice (reduction of QLPA/QPA: 19 ± 10% in endotoxin-treated wild-type mice breathing 40 ppm NO for 22 hours versus 18 ± 5% in endotoxin-treated wild-type mice not exposed to NO; Figure 3). In contrast, breathing 4 ppm NO for 22 hours did not impair HPV in endotoxin-treated NOS2-deficient mice (reduction of QLPA/QPA: 62 ± 2%; n = 4).

Pulmonary vasoreactivity to angiotensin II. To determine whether the impact of endotoxin on HPV was attributable to a nonspecific effect of endotoxin on pulmonary vascular contractile function, we measured the pulmonary vasoconstrictor response to increasing intravenous doses of angiotensin II in saline-treated wild-type mice and wild-type mice 22 hours after endotoxin challenge. PVR increased from 74 ± 24 at baseline to 184 ± 25 mmHg•min•g•mL^–1 at 5.0 μg/kg per minute angiotensin II in saline-treated wild-type mice without endotoxin (P < 0.01) and from 55 ± 11 to 174 ± 40 mmHg•min•g•mL^–1 in endotoxin-challenged wild-type mice (P < 0.001). At any angiotensin II infusion dose level, there was no difference in the PVR of endotoxin-treated and saline-treated mice. P_LA at baseline did not differ between saline-treated wild-type mice (6 ± 1 mmHg) and endotoxin-treated wild-type mice (5 ± 1 mmHg), and P_LA did not change in response to angiotensin II infusion in either group.

Lung wet/dry weight ratios. Wet/dry lung weight ratios did not differ between endotoxin-challenged wild-type (wet weight/dry weight: 5.0 ± 0.5) and NOS2-deficient mice (4.6 ± 0.2), or saline-treated wild-type (4.6 ± 0.1) and saline-treated NOS2-deficient mice (4.5 ± 0.1). Breathing 4 or 40 ppm NO for 22 hours did not alter the wet/dry lung weight ratios.

To investigate the role of NO and NOS2 in the impairment of HPV in sepsis, we developed an in vivo mouse model to assess the redistribution of pulmonary blood flow in response to unilateral alveolar hypoxia produced by LMBO. In mice not exposed to endotoxin, LMBO doubled left lung PVR, thereby diverting 50% of left pulmonary blood flow to the right lung and causing a modest decrease in PaO₂. Changes in murine pulmonary blood flow distribution measured using ultrasonic flow probes at thoracotomy were closely correlated with measurements obtained using intravenous injection of fluorescent microspheres (see Figure 2a) and were similar to those reported by investigators studying large animal models (24, 25). Pulmonary blood flow redistribution induced by bronchial occlusion has been attributed to mechanical factors associated with pulmonary collapse (21), as well as to HPV. In contrast, several in vivo animal studies have demonstrated that the pulmonary blood flow redistribution in response to bronchial occlusion was primarily dependent on changes in PvO₂, but not on changes in intrabronchial pressure (22, 26). We observed that slow re-expansion of the atelectatic left mouse lung with nitrogen did not reduce the left lung’s PVR (Figure 2c), suggesting that in our murine model, the pulmonary vasoconstrictor response to LMBO is also attributable to HPV rather than mechanical factors.

In wild-type mice 22 hours after an endotoxin challenge, we observed that the ability to redistribute pulmonary blood flow in response to regional lung hypoxia was severely impaired, leading to a marked deterioration in systemic arterial oxygenation. This endotoxin-induced impairment of pulmonary blood flow redistribution in the mouse is similar to that observed in awake sheep after intravenous challenge with live bacteria (11) or endotoxin (10). We do not believe that the endotoxin-induced impairment of HPV was attributable to nonspecific dysfunction of the vasomotor contractile apparatus because we found that the ability of angiotensin II to vasoconstrict the pulmonary vasculature did not differ in endotoxin-challenged and saline-treated mice. Moreover, the deleterious effect of endotoxin on the murine pulmonary vasculature was reversible with restoration of HPV at 14 days after endotoxin administration (data not shown).

In NOS2-deficient mice challenged with endotoxin 22 hours earlier, HPV was not impaired and there was preservation of systemic arterial oxygenation during LMBO. Similarly, HPV was preserved in wild-type mice treated with L-NIL, a specific inhibitor of NOS2 enzyme activity, 3 hours after endotoxin challenge. These results demonstrate that NOS2 enzyme activity is critical to produce the endotoxin-induced impairment of HPV.

It is unlikely that impairment of HPV in wild-type mice 22 hours after endotoxin challenge is attributable to excess pulmonary NO levels, as acute administration of L-NIL immediately after LMBO did not restore HPV. These results differ from the findings of Griffiths et al. (12), who observed that 4 hours after rats were challenged with endotoxin, NOS2 inhibitors augmented the pulmonary vasoconstrictor response to both hypoxia and angiotensin II. A possible explanation for these differing results is that Griffiths et al. studied animals at an earlier time after endotoxin challenge, when pulmonary NOS2 is abundant, whereas we studied mice 22 hours after endotoxin challenge, when pulmonary NOS2 enzyme levels are no longer abundant (refs. 12 and 27, unpublished observations). For example, we found that pulmonary cGMP levels were markedly increased 7 hours after endotoxin challenge but had returned to baseline after 22 hours. These results demonstrate that endotoxin-induced pulmonary NOS2 expression is necessary to impair HPV but that continued NOS2 activity is not required for the impairment of HPV measured 22 hours after the endotoxin challenge.

Under certain conditions, NOS2 can produce superoxide, as well as NO (28). We examined whether NO produced by NOS2 contributed to the endotoxin-induced impairment of HPV by replenishing pulmonary NO levels via inhalation in endotoxin-challenged NOS2-deficient mice. NOS2-deficient mice were challenged with endotoxin and placed in an exposure chamber containing varying concentrations of NO for 22 hours. Pulmonary blood flow redistribution in response to LMBO was measured 1 hour after removal from the chamber, by which time pulmonary NO levels would not be expected to be elevated. HPV was impaired in endotoxin-challenged NOS2-deficient mice that breathed 40 ppm NO (see Figure 3). In contrast, breathing 4 ppm NO for 22 hours did not impair HPV in endotoxin-treated NOS2-deficient mice. These data suggest that markedly elevated pulmonary NO levels (either produced endogenously by NOS2 or inhaled) are required to impair HPV. If our observations in mice are extrapolated to human beings, an important clinical implication of our studies is that the administration of high concentrations of inhaled NO to patients with pulmonary inflammation may attenuate HPV, leading to a paradoxical decrease in arterial oxygenation. Moreover, these findings support the current clinical practice of using the lowest effective concentration of inhaled NO when treating patients with acute respiratory failure (29).

In saline-treated wild-type and NOS2-deficient mice, inhalation of 40 ppm NO for 22 hours did not impair HPV when the animals were studied 1 hour after removal from the exposure chamber (see Figure 3). These results suggest that sustained increases in pulmonary NO levels alone are insufficient to cause lasting impairment of HPV. In addition, breathing 40 ppm NO for 22 hours did not further impair HPV in endotoxin-challenged wild-type mice. Thus, it appears that inhaled NO and endotoxin-induced pulmonary NO production are not additive with respect to impairing HPV.

Our studies suggest that the lasting impairment of HPV after endotoxin challenge requires both increased pulmonary NO levels and additional endotoxin-induced inflammatory products. For example, stimulation of cells by endotoxin or cytokines results in the generation of large amounts of superoxide (O₂^–), which can react with NO to form peroxynitrite (ONOO^–). Peroxynitrite is a potent oxidant that can nitrate tyrosine residues in numerous proteins (30, 31). High levels of nitrotyrosine have been found in pulmonary cells of endotoxin-challenged animals and human beings with sepsis (27, 31). Kristof et al. recently reported that NOS2-deficient mice were less susceptible to endotoxin-induced acute lung injury, associated with a marked decrease in endotoxin-induced pulmonary nitrotyrosine immunoreactivity (27). Although the mechanisms responsible for hypoxia-induced contraction of pulmonary vascular smooth muscle cells are incompletely characterized, it is possible that formation of peroxynitrite from NO and superoxide interferes with the HPV sensor or effector mechanisms by inducing lipid peroxidation, inhibiting mitochondrial electron transport, stimulating tyrosine phosphorylation, or interfering with hypoxia-responsive potassium channels (31).

Sepsis, a major cause of mortality in intensive care units, is the most common cause of acute respiratory distress syndrome (ARDS). Sepsis-induced impairment of HPV leads to increased V/Q mismatching and right-to-left shunting of venous blood, producing arterial hypoxemia. In this study, we observed that NO synthesis by NOS2, as well as additional endotoxin-induced products, is required to impair HPV in endotoxin-treated mice. These findings suggest at least 2 strategies for the prevention of the sepsis-induced impairment of HPV. One method would be to prevent the loss of HPV by inhibiting pulmonary NO production by NOS2. The discovery of new, highly selective inhibitors of NOS2 would offer the advantage of blocking excess NO production by NOS2 with concomitant preservation of regional blood flow regulation by NOS3. A second therapeutic method would be to reduce the levels of the yet unknown septic factors that combine with NO to impair HPV.

The authors thank J. Hromi for technical assistance, Y. Chiao Chang for statistical advice, and M. Takata for helpful comments. This study was supported by United States Public Health Service (grants HL-42397 and HL 55377 to W.M. Zapol and K.D. Bloch). R. Ullrich is supported by a grant from the Max Kade Foundation. K.D. Bloch is an Established Investigator of the American Heart Association. The Massachusetts General Hospital has licensed patents on the use of inhaled NO in respiratory failure, and the authors have a right to receive royalties.

1. Rodman, DM, Yamaguchi, T, Hasunuma, K, O’Brien, RF, McMurtry, IF. Effects of hypoxia on endothelium-dependent relaxation of rat pulmonary artery. Am J Physiol 1990. 258:L207-L214.
   View this article via: PubMed Google Scholar
2. Voelkel, NF. Mechanisms of hypoxic pulmonary vasoconstriction. Am Rev Respir Dis 1986. 133:1186-1195.
   View this article via: PubMed Google Scholar
3. Ignarro, LJ, Buga, GM, Wood, KS, Byrns, RE, Chaudhuri, G. Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide. Proc Natl Acad Sci USA 1987. 84:9265-9269.
   View this article via: PubMed CrossRef Google Scholar
4. Cremona, G, Wood, AM, Hall, LW, Bower, EA, Higenbottam, T. Effect of inhibitors of nitric oxide release and action on vascular tone in isolated lungs of pig, sheep, dog and man. J Physiol (Lond) 1994. 481:185-195.
   View this article via: PubMed Google Scholar
5. Stamler, JS, Loh, E, Roddy, MA, Currie, KE, Creager, MA. Nitric oxide regulates basal systemic and pulmonary vascular resistance in healthy humans. Circulation 1994. 89:2035-2040.
   View this article via: PubMed Google Scholar
6. Archer, SL, Tolins, JP, Raij, L, Weir, EK. Hypoxic pulmonary vasoconstriction is enhanced by inhibition of the synthesis of an endothelium derived relaxing factor. Biochem Biophys Res Commun 1989. 164:1198-1205.
   View this article via: PubMed CrossRef Google Scholar
7. Frostell, CG, Blomqvist, H, Hedenstierna, G, Lundberg, J, Zapol, WM. Inhaled nitric oxide selectively reverses human hypoxic pulmonary vasoconstriction without causing systemic vasodilation. Anesthesiology 1993. 78:427-435.
   View this article via: PubMed Google Scholar
8. Nathan, C, Xie, QW. Nitric oxide synthases: roles, tolls, and controls. Cell 1994. 78:915-918.
   View this article via: PubMed CrossRef Google Scholar
9. Nussler, AK, et al. Stimulation of the nitric oxide synthase pathway in human hepatocytes by cytokines and endotoxin. J Exp Med 1992. 176:261-264.
   View this article via: PubMed CrossRef Google Scholar
10. Hutchison, AA, Ogletree, ML, Snapper, JR, Brigham, KL. Effect of endotoxemia on hypoxic pulmonary vasoconstriction in unanesthetized sheep. J Appl Physiol 1985. 58:1463-1468.
    View this article via: PubMed Google Scholar
11. Fischer, SR, et al. Nitric oxide synthase inhibition restores hypoxic pulmonary vasoconstriction in sepsis. Am J Respir Crit Care Med 1997. 156:833-839.
    View this article via: PubMed Google Scholar
12. Griffiths, MJ, Curzen, NP, Mitchell, JA, Evans, TW. In vivo treatment with endotoxin increases rat pulmonary vascular contractility despite NOS induction. Am J Respir Crit Care Med 1997. 156:1-8.
    View this article via: PubMed Google Scholar
13. Stevens, T, Morris, K, McMurtry, IF, Zamora, M, Tucker, A. Pulmonary and systemic vascular responsiveness to TNF-alpha in conscious rats. J Appl Physiol 1993. 74:1905-1910.
    View this article via: PubMed CrossRef Google Scholar
14. Fox, GA, Paterson, NA, McCormack, DG. Effect of inhibition of NO synthase on vascular reactivity in a rat model of hyperdynamic sepsis. Am J Physiol 1994. 267:H1377-H1382.
    View this article via: PubMed Google Scholar
15. Wu, CC, Chen, SJ, Yen, MH. Nitric oxide-independent activation of soluble guanylyl cyclase contributes to endotoxin shock in rats. Am J Physiol 1998. 275:H1148-H1157.
    View this article via: PubMed Google Scholar
16. Villamor, E, et al. Group B streptococcus and E. coli LPS-induced NO-dependent hyporesponsiveness to noradrenaline in isolated intrapulmonary arteries of neonatal piglets. Br J Pharmacol 1995. 115:261-266.
    View this article via: PubMed Google Scholar
17. MacMicking, JD, et al. Altered responses to bacterial infection and endotoxic shock in mice lacking inducible nitric oxide synthase. Cell 1995. 81:641-650.
    View this article via: PubMed CrossRef Google Scholar
18. Laubach, VE, Shesley, EG, Smithies, O, Sherman, PA. Mice lacking inducible nitric oxide synthase are not resistant to lipopolysaccharide-induced death. Proc Natl Acad Sci USA 1995. 92:10688-10692.
    View this article via: PubMed CrossRef Google Scholar
19. Schwartz, D, et al. Inhibition of constitutive nitric oxide synthase (NOS) by nitric oxide generated by inducible NOS after lipopolysaccharide administration provokes renal dysfunction in rats. J Clin Invest 1997. 100:439-448.
    View this article via: JCI PubMed Google Scholar
20. Steudel, W, et al. Pulmonary vasoconstriction and hypertension in mice with targeted disruption of the endothelial nitric oxide synthase (NOS 3) gene. Circ Res 1997. 81:34-41.
    View this article via: PubMed Google Scholar
21. Burton, AC, Patel, DJ. Effect on pulmonary vascular resistance of inflation of the rabbit lungs. J Appl Physiol 1958. 12:239-244.
    View this article via: PubMed Google Scholar
22. Barer, GR, Howard, P, McCurrie, JR, Shaw, JW. Changes in the pulmonary circulation after bronchial occlusion in anesthetized dogs and cats. Circ Res 1969. 25:747-764.
    View this article via: PubMed Google Scholar
23. Glenny, RW, Bernard, S, Brinkley, M. Validation of fluorescent-labeled microspheres for measurement of regional organ perfusion. J Appl Physiol 1993. 74:2585-2597.
    View this article via: PubMed Google Scholar
24. Domino, KB, et al. Time course and responses of sustained hypoxic pulmonary vasoconstriction in the dog. Anesthesiology 1984. 60:562-566.
    View this article via: PubMed Google Scholar
25. Sprague, RS, Thiemermann, C, Vane, JR. Endogenous endothelium-derived relaxing factor opposes hypoxic pulmonary vasoconstriction and supports blood flow to hypoxic alveoli in anesthetized rabbits. Proc Natl Acad Sci USA 1992. 89:8711-8715.
    View this article via: PubMed CrossRef Google Scholar
26. Pirlo, AF, Benumof, JL, Trousdale, FR. Atelectatic lobe blood flow: open vs. closed chest, positive pressure vs. spontaneous ventilation. J Appl Physiol 1981. 50:1022-1026.
    View this article via: PubMed Google Scholar
27. Kristof, AS, Goldberg, P, Laubach, V, Hussain, SN. Role of inducible nitric oxide synthase in endotoxin-induced acute lung injury. Am J Respir Crit Care Med 1998. 158:1883-1889.
    View this article via: PubMed Google Scholar
28. Xia, Y, Zweier, JL. Superoxide and peroxynitrite generation from inducible nitric oxide synthase in macrophages. Proc Natl Acad Sci USA 1997. 94:6954-6958.
    View this article via: PubMed CrossRef Google Scholar
29. Zapol, WM. Minidose inhaled nitric oxide: less is better. Intensive Care Med 1993. 19:433-434.
    View this article via: PubMed CrossRef Google Scholar
30. Haddad, IY, et al. Quantitation of nitrotyrosine levels in lung sections of patients and animals with acute lung injury. J Clin Invest 1994. 94:2407-2413.
    View this article via: JCI PubMed Google Scholar
31. Beckman, JS, Koppenol, WH. Nitric oxide, superoxide, and peroxynitrite: the good, the bad, and ugly. Am J Physiol 1996. 271:C1424-C1437.
    View this article via: PubMed Google Scholar