Reactive oxygen species from NADPH oxidase contribute to altered pulmonary vascular responses in piglets with chronic hypoxia-induced pulmonary hypertension

Abstract

Our main objective was to determine whether reactive oxygen species (ROS), such as superoxide (O₂⁻) and hydrogen peroxide (H₂O₂), contribute to altered pulmonary vascular responses in piglets with chronic hypoxia-induced pulmonary hypertension. Piglets were raised in either room air (control) or hypoxia for 3 days. The effect of the cell-permeable superoxide dismutase mimetic (SOD; M40403) and/or PEG-catalase (PEG-CAT) on responses to acetylcholine (ACh) was measured in endothelium-intact and denuded pulmonary resistance arteries (PRAs; 90-to-300-μm diameter). To determine whether NADPH oxidase is an enzymatic source of ROS, PRA responses to ACh were measured in the presence and absence of a NADPH oxidase inhibitor, apocynin (APO). A Western blot technique was used to assess expression of the NADPH oxidase subunit, p67phox. A lucigenin-derived chemiluminescence technique was used to measure ROS production stimulated by the NADPH oxidase substrate, NADPH. ACh responses, which were dilation in intact control arteries but constriction in both intact and denuded hypoxic arteries, were diminished by M40403, PEG-CAT, the combination of M40403 plus PEG-CAT, as well as by APO. Although total amounts were not different, membrane-associated p67phox was greater in PRAs from hypoxic compared with control piglets. NADPH-stimulated lucigenin luminescence was nearly doubled in PRAs from hypoxic vs. control piglets. We conclude that ROS generated by NADPH oxidase contribute to the aberrant pulmonary arterial responses in piglets exposed to 3 days of hypoxia.

there is increasing appreciation that reactive oxygen species (ROS), such as O₂⁻ and H₂O₂, may act as essential participants in normal cell signaling and be involved in regulation of tone and reactivity in a number of vascular beds including the pulmonary circulation (8, 29, 33, 42, 49, 50). NADPH oxidases are enzymes that have been shown to be prominent sources of ROS in many vascular beds (5, 21, 29). Moreover, evidence is accumulating that ROS, derived at least in part from NADPH oxidases, play a role in abnormal cell signaling and reactivity and thereby contribute to a number of vascular diseases, including pulmonary hypertension (5, 7, 21, 29, 37, 43, 47). For example, there is evidence that ROS produced by NADPH oxidase are involved in the development of chronic hypoxia-induced pulmonary hypertension in adult animals (20, 31, 37, 38, 53). However, the potential contribution of ROS derived from NADPH oxidase to the altered regulation of pulmonary vascular tone in newborns with chronic hypoxia-induced pulmonary hypertension is not yet known. Awareness of the fundamental differences in the regulation of pulmonary vascular tone in newborns and adults limits the ability to extrapolate to the newborn findings on the role of ROS in altered cell signaling in adult lungs (41, 55).

Regulation of pulmonary vascular tone is also known to differ between large, conduit level arteries and smaller, resistance level arteries (1, 2, 45). The vast majority of studies to date have evaluated the potential contribution from ROS to abnormal reactivity in larger, conduit level vessels (20, 37, 38, 53). Because of their critical role in regulating pulmonary vascular tone, it is important to identify derangements in ROS in resistance level pulmonary arteries. We previously showed that resistance level pulmonary arterial responses to acetylcholine (ACh) are altered in newborn piglets with pulmonary hypertension induced by 3 days' exposure to hypoxia (17–19). The major purpose of this study was to test the hypothesis that ROS are involved with the aberrant responses to ACh that develop in resistance pulmonary arteries (PRAs) at this early stage of pulmonary hypertension in newborn piglets. In addition, we wanted to determine whether NADPH oxidase is an enzymatic source of ROS in the PRAs of piglets with chronic hypoxia-induced pulmonary hypertension.

METHODS

Animals.

Newborn pigs (2–3 days old) were placed in a hypoxic normobaric chamber for 3 days. Normobaric hypoxia was produced by delivering compressed air and N₂ to an incubator (Thermocare). Oxygen content was regulated at 10–12% O₂ (Po₂ 64–78 Torr), and CO₂ was maintained at 3–6 Torr by absorption with soda lime. The chamber was opened two times per day for cleaning and to weigh the piglets. The animals were fed ad libitum with an artificial sow milk replacer from a feeding device attached to the chamber. We have previously found no differences in vascular responses between piglets raised in a room-air environment for 3 days and piglets raised on a farm (15, 19). Therefore, for this study, the control piglets were studied on the day of arrival from the farm at 5–7 days of age, i.e., the same postnatal age as the hypoxic piglets on the day of study. All experimental protocols were performed in adherence with the NIH guidelines for the use of experimental animals and approved by the Animal Care and Use Committee of Vanderbilt University Medical Center. The animal resource facility is fully accredited by the AALAC.

Cannulated artery preparation.

On the day of study, the piglets were preanesthetized with ketamine (30 mg/kg im) and acepromazine (2 mg/kg im) and then anesthetized with pentobarbital (10 mg/kg iv). All animals were given heparin (1,000 IU/kg iv) and then exsanguinated. The thorax was opened, and the lungs were removed and placed in cold (4°C) physiological saline solution (PSS) until use. The PSS had the following composition (in mM): 141 Na⁺, 4.7 K⁺, 125 Cl⁻, 2.5 Ca²⁺, 0.72 Mg²⁺, 1.7 H₂PO₄⁻, 25 HCO₃⁻, and 11 glucose. Immediately before use, segments of 90- to 300-μm diameter pulmonary arteries were dissected from a lung lobe.

The system used to study cannulated arteries has been described in detail previously (19). Briefly, it consists of a water-jacketed plastic chamber in which proximal (inflow) and distal (outflow) cannulas were mounted. An arterial segment was threaded onto the proximal cannula and tied in place with a 22-μm nylon suture. The distal end of the artery was then tied onto the distal cannula, the artery was filled with PSS, and large side branches were tied off. The distance between the cannula tips was adjusted with a micrometer connected to the proximal cannula so that the slack was taken out of the artery. The exterior of the artery was suffused with PSS from a reservoir at 37°C and aerated with a gas mixture containing O₂⁻, CO₂, and N₂, giving a Po₂ of 140 Torr, a Pco₂ of 38 Torr, and a pH of 7.37. The arterial lumen was filled from a syringe containing PSS, aerated with the same gas mixture as the reservoir, and connected to the cannula with polyethylene tubing.

Inflow pressure was adjusted by changing the height of the infusion syringe. Pressure transducers were placed on both the inflow side between the syringe and the artery and at the outflow end of the system. The artery was discarded if the pressures were not equal (indicating a leak in vessel). The diameter of the artery was observed continuously with a video system containing a color camera (Panasonic 5000) and television monitor. Vessel diameters were measured with a video scaler (FORA IV). The video scaler was calibrated with a micrometer scale.

Cannulated artery protocols.

Each artery was allowed to equilibrate for 30–60 min to establish basal tone. The control arteries were equilibrated at a transmural pressure of 15 cmH₂O, and the hypoxic arteries were equilibrated at a transmural pressure of 25 cmH₂O. These pressures were chosen as they represent in vivo pressures (15, 16). We have previously shown no effect from these transmural pressures on pulmonary arterial responses to ACh (19). Following establishment of basal tone, all arteries were tested for viability by contraction to the thromboxane A₂-mimetic U46619 (10⁻⁸ M). To check for a functional endothelium in control arteries, responses to ACh (10⁻⁶ M) were evaluated. We previously found that hypoxic arteries constricted to ACh but dilated to another endothelium-dependent agent, the calcium ionophore A-23187 (19). Therefore, responses to A-23187 were used to check for a functional endothelium in hypoxic arteries.

In one series of studies, we evaluated the contribution of endogenous O₂⁻ and H₂O₂ on ACh responses in control and hypoxic arteries. To do this, changes in vessel diameter to ACh (10⁻⁸ to 10⁻⁵ M) were measured before and after adding a cell-permeable superoxide dismutase (SOD) mimetic, which dismutates O₂⁻ to H₂O₂, and/or an H₂O₂ decomposing enzyme, which converts H₂O₂ to H₂O. We used either M40401 or M40403 (3 μg/ml) as the SOD mimetic, according to availability. PEG-catalase (PEG-CAT), 250 U/ml, was used as the H₂O₂ decomposing enzyme. For all these studies, after assessing for viability and a functional endothelium, changes in vessel diameter were measured in response to cumulative doses of ACh (10⁻⁸ to 10⁻⁵ M). Next, the vessels were washed with PSS, and the cell-permeable O₂⁻ and/or H₂O₂ decomposing enzyme(s) were added to the reservoir. Twenty minutes after adding the ROS decomposing enzyme(s), dose responses to ACh (10⁻⁸ to 10⁻⁵ M) were repeated.

To determine the influence of the endothelium, air was infused into arteries of control and hypoxic piglets (2, 22). Functional disruption of the endothelium was verified by loss of dilation to ACh and/or A-23187 in the control arteries and to A-23187 in the hypoxic arteries. Reactivity to U46619 was used to confirm viability of the arteries. Then, responses to ACh (10⁻⁸ to 10⁻⁵ M) were measured in vessels both before and after adding the SOD mimetic and PEG-CAT either separately or combined.

In another series of studies, we evaluated the contribution from NADPH oxidase to changes in vessel diameter in response to ACh in both endothelium-intact vessels and endothelium-disrupted vessels. For all these studies, changes in vessel diameter were continuously monitored while cumulative doses of ACh were added (10⁻⁸ to 10⁻⁵ M) before and then 20 min after the addition of a NADPH oxidase inhibitor, apocynin (APO; 10⁻⁶ M).

We previously found that thromboxane synthase inhibition altered ACh responses in hypoxic arteries (19). To evaluate the effect of combined antagonism of thromboxane and NADPH oxidase on ACh responses in hypoxic arteries, we measured changes in vessel diameter in response to ACh (10⁻⁸ to 10⁻⁵ M) before and after addition of the thromboxane synthase inhibitor, dazoxiben (10⁻⁵ M), plus APO (10⁻⁶ M).

For all of the above studies, vessel responses to the vehicle used for solubilization of the ROS decomposing enzymes or NADPH oxidase inhibitors were evaluated.

Immunoblot analyses of the NADPH oxidase subunit p67phox.

Pulmonary arteries (20- to 600-μm diameter) were dissected from lungs of control and hypoxic piglets, frozen in liquid nitrogen, and stored at −80°C until use for immunoblot analysis.

We performed preliminary studies with different amounts of total protein to determine the dynamic range of the immunoblot analysis. An amount of protein that was within the dynamic range of the immunoblot analysis (5 μg of protein for total p67phox; 15 μg of protein for membrane fraction p67phox) was then used to compare protein abundances between homogenates of small pulmonary arteries from control and hypoxic piglets described below.

Frozen samples of small pulmonary arteries (20–600 μm diameter) from control (n = 5) and hypoxic (n = 5) piglets were crushed under liquid N₂ in a prechilled mortar and pestle into a fine powder, transferred to a tube containing homogenate buffer, and then sonicated using three 15-s pulses, taking care not to foam the sample. Vessel homogenates were then centrifuged at 9,000 g for 10 min at 4°C, and some of the supernatant was stored at −80°C as “total homogenate.” The remainder of the supernatant was centrifuged at 100,000 g for 2 h at 4°C, and the pellet was resuspended in the homogenization buffer and stored at −80°C as “membrane fraction.” Protein concentrations for total homogenate and membrane fractions were determined by protein assay (Bradford). All total vessel homogenate and membrane fractions were diluted with PBS to obtain a protein concentration of 1 mg/ml. Aliquots of the protein solutions were solubilized in an equal volume of denaturing, reducing sample buffer (20 mM Tris·HCl, 2.5 mM EDTA, 0.5% Triton X-100, 0.05% SDS, 100 mM NaCl, 1 mM PMSF, 10 ng/ml leupeptin, and 10 μg/ml pepstatin), heated to 80°C for 15 min, and centrifuged for 3 min at 5,600 g in a microfuge. Equal volumes of these supernatants were applied to Tris-glycine precast 8% polyacrylamide gels (Novex) so that equal amounts of protein were loaded. Electrophoresis was carried out in 25 mM Tris, 192 mM glycine, and 0.1% SDS (pH 8.3) at 125 V for 1.7 h. The proteins were transferred from the gel to a nitrocellulose membrane (Novex) using a Bio-Rad transfer box at 100 V for 1 h in 25 mM Tris, 192 mM glycine, and 20% methanol (pH 8.3). The membrane was incubated overnight at 4°C in PBS containing 10% nonfat dried milk and 0.1% Tween 20 to block nonspecific protein binding. To detect p67phox, the nitrocellulose membrane was incubated overnight at 4°C with the primary antibody (1:500 for total p67phox, 1:1,000 for membrane fraction p67phox; BD Biosciences) diluted in PBS containing 0.1% Tween 20 and 1% nonfat dried milk (carrier buffer), followed by incubation for 1 h at room temperature with a horseradish peroxidase-conjugated secondary antibody (Zymed) diluted in the carrier buffer (1:2,500). The nitrocellulose membrane was washed three times between the first two incubations with the carrier buffer and three times with the carrier buffer plus one time with PBS containing 0.1% Tween 20 following the final incubation. To visualize the antibody, the membranes were developed using enhanced chemiluminescence reagents (ECL, Amersham), and the chemiluminescent signal was captured on X-ray film (ECL Hyperfilm, Kodak). Similar procedures were followed to reprobe the membranes for actin (Sigma). The bands for each protein were quantified using densitometry.

Lucigenin-enhanced chemiluminescence.

Pulmonary arteries (90- to 300-μm diameter) were dissected from lungs of control and hypoxic piglets and placed into wells containing PBS. Either APO (10⁻⁶ M) or M40403 (3 μg/ml) was added to some wells, after which all wells were incubated for 20 min at 37°C. Then, NADPH (10⁻⁴ M) was added to some wells, and the plate was placed into a luminometer (BMG Fluostar Optima) and allowed to equilibrate for 20 min at 37°C. For each well, after injecting lucigenin ([9,9′-bis(N-methylacridinium nitrate)], 5 μmol/l), scintillation counts, i.e., relative light units (RLUs), were obtained for 500 s. Some wells contained no vessels so that the RLUs could be background corrected. In addition, at the end of the assay, the vessels were dried so that the RLUs could be normalized to vessel dry weight.

Statistics.

Data are presented as means ± SD. For cannulated artery studies, the change from baseline diameter in response to each dose of ACh was calculated for all vessels. To estimate the relationship between ACh and percent change in diameter in endothelium-intact or air-embolized vessels treated with vehicle (untreated) or inhibitors in arteries from control or hypoxic piglets, we used a linear mixed effects model. Each artery was exposed to multiple doses of ACh, so we included a random intercept to control for the correlation arising from taking repeated observations on the same artery. Results are presented as average change in diameter over ACh dose with 95% confidence intervals (CI) or as P values from asymptotically valid Wald X² significance tests of parameters in the mixed effects models (26). For the lucigenin studies, lucigenin-derived chemiluminescence was compared between treated and untreated vessels for control and hypoxic groups using ANOVA with Fisher's protected least significant differences post hoc comparison test (36). To compare p67phox total and membrane fraction amounts between control and hypoxic arteries, an unpaired t-test was used. P < 0.05 was considered significant.

Materials.

Concentrations for each drug listed in cannulated artery protocols and the lucigenin-derived chemiluminescence were expressed as final molar concentrations in the vessel bath or wells, respectively. ACh, A-23187, and APO were obtained from Sigma Chemical. M40403 and M40401 were generous gifts from Activbiotics (Lexington, MA) and Metaphore Pharmaceuticals (St. Louis, MO), respectively. Dazoxiben was a generous gift from Pfizer (Groton, CT). Dazoxiben and PEG-CAT were solubilized in distilled H₂O. M40403 was solubilized in 26 mM NaHCO₃ buffer. M40401 was solubilized in buffered saline. ACh was solubilized in saline. APO was solubilized in DMSO.

RESULTS

For the cannulated artery studies, the mean diameter of vessels used for all studies was 203 ± 41 μm for control arteries and 215 ± 44 μm for hypoxic arteries. None of the vehicles significantly changed arterial diameter in the concentrations used for solubilization of any of the agents.

For endothelium-intact arteries (Fig. 1, A–C), each 10-fold increase in ACh, from 10⁻⁸ to 10⁻⁵ M, caused, on average, a −5.71% change [95% CI = (−6.64%, −4.78%)] in diameter (i.e., constriction) in untreated hypoxic arteries and a 3.65% increase [95% CI = (2.87%, 4.43%)] in diameter (i.e., dilation) in untreated control arteries. In contrast, vessel diameter did not change with increasing ACh doses for either hypoxic or control arteries treated with an SOD mimetic, M40401 or M40403 (Fig. 1A, P = 0.61, hypoxic; P = 0.47, control), PEG-CAT (Fig. 1B, P = 0.56, hypoxic; P = 0.40, control), or a combination of an SOD mimetic and PEG-CAT (Fig. 1C, P = 0.72, hypoxic; P = 0.18, control). Notably, the dose-response curves for the SOD mimetic-treated arteries (Fig. 1A) were significantly different from untreated arteries for both hypoxic (P < 0.001) and control (P < 0.001) groups. Similarly, the dose-response curves for both control and hypoxic arteries treated with either PEG-CAT (Fig. 1B) or a combination of an SOD mimetic and PEG-CAT (Fig. 1C) differed from the dose-response curves of the respective group of untreated arteries (for PEG-CAT, P < 0.001, hypoxic; P < 0.001, control; for an SOD mimetic + PEG-CAT, P < 0.001, hypoxic; P < 0.001, control). Thus, separate and combined treatment with an SOD mimetic and/or PEG-CAT abolished the dilator response to ACh in control arteries and abolished the constrictor response to ACh in hypoxic arteries.

Fig. 1.

ACh-induced changes in diameter in untreated control (n = 43 arteries from 41 piglets) and hypoxic (n = 30 arteries from 29 piglets) arteries and in control and hypoxic arteries treated with the SOD mimetic, M40401 or M40403 (A; treated control arteries: n = 12 arteries from 11 piglets; treated hypoxic arteries: n = 12 arteries from 11 piglets); the H₂O₂ decomposing agent, PEG-catalase (PEG-CAT) (B; treated control arteries: n = 12 arteries from 11 piglets; treated hypoxic arteries: n = 10 arteries from 10 piglets); or combined treatment with either M40401 or M40403 and PEG-CAT (C; treated control arteries: n = 19 arteries from 19 piglets; treated hypoxic arteries: n = 8 arteries from 8 piglets). Please note that the same untreated controls and the same untreated hypoxic arteries are used in A–C. All values are means ± SD. *Diameter changes with increase in ACh from 10⁻⁸ to 10⁻⁵ M, P < 0.05, linear mixed effects model; +dose-response curve different from respective group of untreated arteries, P < 0.05, linear mixed effects model.

After air infusion to disrupt the endothelium, an increase in ACh from 10⁻⁸ to 10⁻⁵ M caused a significant decrease in diameter (i.e., constriction) for both untreated hypoxic (−7.14% change on average for each 10-fold increase in ACh, 95% CI = [−8.24%, −6.04%]) and untreated control arteries (−4.29% change on average for each 10-fold increase in ACh, 95% CI = [−5.36%, −3.21%]) (Fig. 2, A–C). Both separate and combined treatment with an SOD mimetic (M40401 or M40403) and PEG-CAT reduced the constriction due to ACh in endothelium-disrupted arteries of both groups of piglets (Fig. 2A, SOD mimetic, P < 0.001, hypoxic; P < 0.001, control; Fig. 2B, PEG-CAT, P < 0.001, hypoxic; P < 0.001, control; Fig. 2C, SOD mimetic + PEG-CAT, P < 0.001, hypoxic; P < 0.001, control).

Fig. 2.

ACh-induced changes in diameter of air-embolized arteries from untreated control (n = 19 arteries from 19 piglets) and hypoxic piglets (n = 20 arteries from 17 piglets) and from control and hypoxic arteries treated with the SOD mimetic, M40401 or M40403 (A; treated control arteries: n = 8 arteries from 8 piglets; treated hypoxic arteries: n = 10 arteries from 7 piglets); the H₂O₂ decomposing agent, PEG-CAT (B; treated control arteries: n = 5 arteries from 5 piglets; treated hypoxic arteries: n = 5 arteries from 5 piglets); or combined treatment with either M40401 or M40403 and PEG-CAT (C; treated control arteries: n = 6 arteries from 6 piglets; treated hypoxic arteries: n = 5 arteries from 5 piglets). Please note that the same untreated controls and the same untreated hypoxic arteries are used in A–C. All values are means ± SD. *Diameter changes with increase in ACh from 10⁻⁸ to 10⁻⁵ M, P < 0.05, linear mixed effects model; +dose-response curve different from respective group of untreated arteries, P < 0.05, linear mixed effects model.

Compared with untreated arteries, treatment with the NADPH oxidase inhibitor, APO, altered the dose-response curve to ACh in both endothelium-intact (Fig. 3A; P < 0.001, hypoxic; P < 0.001, control) and endothelium-disrupted arteries (Fig. 3B; P < 0.001, hypoxic; P < 0.001, control) of both hypoxic and control groups of piglets. The combination of APO and the thromboxane synthase antagonist, dazoxiben, also altered the dose-response curves to ACh in endothelium-intact hypoxic arteries (P < 0.001, Fig. 3A). Notably, dose-response curves to ACh did not differ between hypoxic arteries treated solely with APO and hypoxic arteries treated with the combination of APO and dazoxiben (P = 0.13, Fig. 3A). Of interest, unlike the separate and combined effects of treatment with ROS decomposing agents (Fig. 1, A–C), hypoxic arteries treated solely with APO or with the combination of APO and dazoxiben showed a significant decrease in diameter with increasing doses of ACh [Fig. 3A, for APO-treated hypoxic arteries, −1.63% change in diameter on average for each 10-fold change in ACh, 95% CI = (−2.97%, −0.28%); for APO + dazoxiben-treated hypoxic arteries, −3.29% change in diameter on average for each 10-fold change in ACh, 95% CI = (−4.69%, −1.89%)]. That is, treatment with either APO or the combination of APO and dazoxiben reduced but did not abolish constriction due to increasing doses of ACh in hypoxic arteries.

Fig. 3.

A: ACh-induced changes in diameter in untreated control (n = 15 arteries from 12 piglets) and hypoxic (n = 25 arteries from 21 piglets) arteries and in control and hypoxic arteries treated with either the NADPH oxidase inhibitor apocynin (treated control arteries: n = 15 arteries from 12 piglets; treated hypoxic arteries: n = 13 arteries from 9 piglets) or combined treatment with the NADPH oxidase inhibitor apocynin plus the thromboxane synthase inhibitor dazoxiben (treated hypoxic arteries: n = 12 arteries from 12 piglets). All values are means ± SD. *Diameter changes with increase in ACh from 10⁻⁸ to 10⁻⁵ M, P < 0.05, linear mixed effects model; +dose-response curve different from respective group of untreated arteries, P < 0.05, linear mixed effects model. B: ACh-induced changes in diameter of air-embolized arteries from control and hypoxic piglets in the presence and absence (untreated) of the NADPH oxidase inhibitor apocynin (control arteries: n = 8 arteries from 8 piglets; hypoxic arteries: n = 7 arteries from 7 piglets). All values are means ± SD. *Diameter changes with increase in ACh from 10⁻⁸ to 10⁻⁵ M, P < 0.05, linear mixed effects model; +dose-response curve different from respective group of untreated arteries, P < 0.05, linear mixed effects model.

Immunoblot analyses of total and membrane fractions of p67phox in small pulmonary artery homogenates from control and hypoxic piglets are shown in Fig. 4, A and B. As determined by densitometry, there was no difference in the mean data for total p67phox protein abundance in homogenates of small pulmonary arteries of both groups (Fig. 4A). In contrast, the membrane fraction of p67phox was increased in homogenates of small pulmonary arteries from hypoxic compared with control piglets (Fig. 4B). In the absence of NADPH, lucigenin-derived chemiluminescence was undetectable for either group of arteries. In contrast, as shown in Fig. 5, in the presence of NADPH, lucigenin-derived chemiluminescence increased nearly twofold in PRAs from hypoxic piglets compared with those from control piglets. Furthermore, both APO and M40403 reduced lucigenin-derived chemiluminescence in PRAS from both groups of arteries. In fact, lucigenin-derived chemiluminescence was similar in APO-treated hypoxic arteries and untreated control arteries.

Fig. 4.

A: immunoblot results and corresponding densitometry for total p67phox relative to actin in resistance level pulmonary artery homogenates from control and hypoxic piglets. B: immunoblot results and corresponding densitometry for the membrane fraction of p67phox relative to actin in resistance level pulmonary artery homogenates from control and hypoxic piglets. *Different from control arteries, P < 0.05, unpaired t-test.

Fig. 5.

NADPH-stimulated chemiluminescence in resistance level pulmonary arteries from control and hypoxic piglets in the presence and absence (untreated) of either the NADPH oxidase inhibitor apocynin or the SOD mimetic M40403 (control piglets: n = 40 untreated arteries, n = 10 apocynin-treated arteries, n = 6 M40403 treated arteries; hypoxic piglets: n = 44 untreated arteries, n = 12 apocynin-treated arteries, n = 7 M40403-treated arteries). All values are means ± SD. *Different from untreated hypoxic arteries; +different from untreated control arteries, P < 0.05, ANOVA with Fisher's protected least significant differences post hoc comparison test.

DISCUSSION

We have previously shown that newborn piglets exposed to 3 days of chronic hypoxia develop pulmonary hypertension and exhibit altered pulmonary arterial responses to ACh, an endothelium-dependent vasodilator (15). An important new finding in this study is that both separate and combined treatments with a cell-permeable SOD mimetic, M40401 or M40403, and an H₂O₂ decomposing enzyme, PEG-CAT, abolished constriction to ACh in resistance level pulmonary arteries from hypoxic piglets. This finding suggests that the ROS, O₂⁻, and H₂O₂ play a prominent role in mediating the aberrant pulmonary vascular responses to ACh that develop in newborn piglets exposed to 3 days of chronic in vivo hypoxia.

Another important finding in this study is that treatment with a NADPH oxidase inhibitor, APO, also reduced aberrant ACh responses in hypoxic arteries suggesting that ROS, derived at least in part from NADPH oxidase, mediate ACh-induced constriction in hypoxic arteries. Of note, unlike treatment with agents that decompose ROS, i.e., M40401 or M40403 and/or PEG CAT, the constrictor response to ACh was not abolished by treatment with a NADPH oxidase inhibitor. Thus, sources of ROS in addition to NADPH oxidase are likely to contribute to the aberrant response to ACh in hypoxic vessels.

In support of a role for ROS derived from NADPH oxidase is our finding that the amount of the NADPH oxidase subunit, p67phox, is increased in the membrane fraction of resistance pulmonary arteries from hypoxic piglets. Activation of NADPH oxidase and generation of ROS require the translocation of cytosolic subunits, including p67phox, to the membrane. It is important to realize that by itself, translocation of p67phox to the membrane is not sufficient to activate NADPH oxidase and cause release of ROS (44). This is because complete assembly of all the membrane-linked and cytosolic subunits on the plasma membrane is required for ROS generation (30, 44). Instead of having more activated, i.e., completely preassembled NADPH oxidase, it is possible that during exposure to chronic hypoxia more NADPH oxidase has been partially assembled (30). Our findings are consistent with the notion that NADPH oxidase has been primed during exposure to chronic hypoxia such that, dependent on the amount of available substrate and the stimulus (30), hypoxic arteries have the potential to generate more NADPH oxidase-derived ROS than do control arteries.

Relevant to our findings, it is known that ROS can stimulate release of arachidonic acid (6, 40), the substrate used for production of all arachidonic acid metabolites, including the potent constrictor, thromboxane. Therefore, one mechanism by which ROS could mediate the impaired ACh responses observed in hypoxic pulmonary arteries is by stimulating production of thromboxane. This possibility is supported by our previous findings showing that thromboxane contributes to the aberrant ACh responses in pulmonary arteries from hypoxic piglets (19).

It should also be considered that ROS per se may act as constrictors (3). This possibility has been suggested by others and is supported by recent studies showing that both O₂⁻ and H₂O₂ activate Rho/Rho kinase in the rat aorta (25). However, particularly in the case of H₂O₂, there is also evidence that ROS act as dilators (12, 32, 35). Another possibility that warrants consideration is that endogenous levels or activities of antioxidants, such as SOD, could be altered and contribute to impaired ACh-induced responses in newborn hypoxic arteries.

Other enzymatic sources of ROS, such as uncoupled nitric oxide synthase (NOS), should also be considered. However, we have previously reported that NOS antagonists impaired ACh responses in both control and hypoxic arteries of newborn piglets (14). Thus, we have no evidence that ROS derived from uncoupled NOS contributes to impaired ACh responses after 3 days of hypoxia. Nonetheless, it has been shown in a variety of vascular disorders, including other models of pulmonary hypertension, that removing O₂⁻ allows for greater bioavailability of nitric oxide and thereby facilitates dilation (4, 7, 10, 27, 28).

Our findings with endothelium-disrupted hypoxic arteries suggest a nonendothelial cellular source of NADPH oxidase-dependent ROS production. In fact, our findings are most consistent with the notion that both endothelial and nonendothelial cell-derived ROS act as and/or mediate production of constrictors (3, 39, 54) and that NADPH oxidase is an important, but perhaps not the sole, enzymatic source of ROS in resistance pulmonary arteries of piglets exposed to 3 days of hypoxia.

Findings in pulmonary arteries from control piglets also merit comment. Our findings indicate that ROS mediate ACh responses in control arteries. One mechanism by which ROS could mediate the ACh-induced responses observed in endothelium-intact control pulmonary arteries is by stimulating production of dilator arachidonic acid metabolites. Consistent with this possibility, we previously found that indomethacin, an inhibitor of the COX pathway of arachidonic acid metabolism, reduced ACh-induced dilation in control pulmonary arteries (19).

Similar to hypoxic arteries, our findings indicate that NADPH oxidase is a prominent enzymatic source of ACh-induced ROS production in control pulmonary arteries and implicate nonendothelial cellular sources of NADPH oxidase. As shown by others, the nonendothelial cellular sources of NADPH oxidase-derived ROS for both control and hypoxic pulmonary arteries could include smooth muscle cells (34, 46) and cells in the adventitia (52), including invading neutrophils, macrophages, or monocytes (11).

It is important to note that our findings do not imply that both O₂⁻ and H₂O₂ act as and/or mediate production of dilators in pulmonary arteries from control piglets. In fact, the findings in endothelium-intact control pulmonary arteries with the H₂O₂ decomposing agent, PEG-CAT, either separately or when combined with an SOD mimetic, are consistent with the view that upon stimulation with ACh, O₂⁻ could mediate constriction but serves as substrate for synthesis of H₂O₂, which induces relaxation. However, if O₂⁻ mediates constriction while H₂O₂ mediates dilation in control arteries, explanation for the diminished dilation in the presence of an SOD mimetic by itself is less clear. This is because one would expect the O₂⁻ dismutated by the SOD mimetic to H₂O₂ to cause dilation. Although the explanation remains unclear, it is of interest that similar to our findings, others have found that ROS decomposing agents diminish agonist-induced dilation in pulmonary arteries from normal, control animals (20, 28).

Our findings add to an accumulating body of evidence that ROS contribute to the disrupted regulation of tone found in a number of vascular diseases (3, 29, 39, 54). Moreover, our findings in the newborn pulmonary circulation implicate NADPH oxidase in the pathogenesis of chronic hypoxia-induced pulmonary hypertension (20, 31, 37, 38, 53). Of note, many of these other studies suggest that an elevation in ROS, derived in part from NADPH oxidase, is part of the pathogenesis of the vascular disease. Our findings do not necessitate elevated basal or agonist-stimulated production of ROS to explain the disrupted response to ACh in hypoxic arteries. Rather, it is possible that ROS signal through pathways that are disrupted by hypoxia, such as arachidonic acid metabolite pathways (13, 39, 51).

Limitations of our study should be mentioned. The use of APO as a specific NADPH oxidase inhibitor has been questioned (9). Moreover, it has recently been suggested that in vascular cells, APO has antioxidant effects that are separate from any potential impact on NADPH oxidase (23). Thus, our findings with APO may reflect contribution from ROS other than NADPH oxidase, such as uncoupled NOS. Likewise, findings with the lucigenin-enhanced chemiluminescence technique must be interpreted with caution (24, 48). This is partly because regardless of the concentration used, lucigenin may undergo redox cycling (24).

Despite these limitations, when taken together, our findings indicate that the vasoactive effects of ACh are mediated at least in part by ROS derived from NADPH oxidase in resistance level pulmonary arteries of both control and hypoxic piglets. Of importance, our findings highlight a need for future investigations that will identify the precise signaling mechanisms by which NADPH oxidase and ROS mediate disparate responses in control and hypoxic arteries. Furthermore, findings in this study identify important new pharmacological targets for infants with pulmonary hypertension due to conditions associated with chronic hypoxia.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants RO1-HL-68572 (C. D. Fike) and RO1-HL-075511 (J. L. Aschner).