Nitric oxide synthase enzymes in the airways of mice exposed to ovalbumin: NOS2 expression is NOS3 dependent

Abstract

Objectives and design: The function of the airway nitric oxide synthase (NOS) isoforms and the lung cell types responsible for its production are not fully understood. We hypothesized that NO homeostasis in the airway is important to control inflammation, which requires upregulation, of NOS2 protein expression by an NOS3-dependent mechanism.

Materials or subjects: Mice from a C57BL/6 wild-type, NOS1(-/-), NOS2(-/-), and NOS3(-/-) genotypes were used. All mice strains were systemically sensitized and exposed to filtered air or ovalbumin (OVA) aerosol for two weeks to create a subchronic model of allergen-induced airway inflammation.

Methods: We measured lung function, lung lavage inflammatory and airway epithelial goblet cell count, exhaled NO, nitrate and nitrite concentration, and airway NOS1, NOS2, and NOS3 protein content.

Results: Deletion of NOS1 or NOS3 increases NOS2 protein present in the airway epithelium and smooth muscle of air-exposed animals. Exposure to allergen significantly reduced the expression of NOS2 protein in the airway epithelium and smooth muscle of the NOS3(-/-) strain only. This reduction in NOS2 expression was not due to the replacement of epithelial cells with goblet cells as remaining epithelial cells did not express NOS2. NOS1(-/-) animals had significantly reduced goblet cell metaplasia compared to C57Bl/6 wt, NOS2(-/-), and NOS3(-/-) allergen-exposed mice.

Conclusion: The airway epithelial and smooth muscle cells maintain a stable airway NO concentration under noninflammatory conditions. This "homeostatic" mechanism is unable to distinguish between NOS derived from the different constitutive NOS isoforms. NOS3 is essential for the expression of NOS2 under inflammatory conditions, while NOS1 expression contributes to allergen-induced goblet cell metaplasia.

Figure 1

Total cells recovered by lung lavage from strains of mice exposed to filtered air (a), or 2 weeks of OVA aerosol, (b) and total eosinophils from mice exposed to 2 weeks of OVA aerosol (c). Of the filtered air-exposed mice (a), the NOS2^−/− mice had a greater number of cells present in their lavage (8.65 ± 6.90 × 10⁴ (n = 16), P < .01) compared to the other genotypes of mice examined (NOS1^−/−  2.78 ± 1.71 × 10⁴ (n = 12), NOS3^−/−3.03 ± 1.68 × 10⁴ (n = 11), C57Bl/6 3.85 ± 2.98 × 10⁴ (n = 27)). Of the OVA-exposed mice, the NOS2^−/− mice had 15.65 ± 9.93 × 10⁵ (P < .05, n = 17) cells (b), significantly more than either NOS1 (6.20 ± 5.05 × 10⁵ (n = 12)), NOS3 (6.7 ± 1.4 × 10⁵ (n = 11)), or C57BL/6 (7.86 ± 8.02 × 10⁵ (n = 21)) mice. Eosinophils comprised a significant proportion of the cells present in the lavage of mice exposed to OVA. NOS2^−/− mice exposed to OVA had 11.9 ± 1.76 × 10⁵ eosinophils (c) in lavage, which was significantly more than in NOS1 (2.4 ± 0.6 × 10⁵), NOS3 (5.3 ± 0.45 × 10⁵), or C57BL/6 (2.9 ± 3.0 × 10⁵) mice exposed to OVA. Data are presented as mean values ± SEM. * denotes P < .05; **P < .01 by ANOVA.

Figure 2

Exhaled NO (a) and lung lavage NO_x (b) concentrations in strains of mice exposed to filtered air or 2 weeks of OVA. NOS2^−/− mice exposed to filtered air have a lower exhaled NO concentration compared to C57Bl/6 mice exposed to filtered air (2.24 ± 1.7 (n = 15) versus 5.12 ± 4.2 (n = 26) ppb, P = .02). Air-exposed NOS2^−/− mice have lower exhaled NO levels compared to air-exposed C57Bl/6 mice (5.10 ± 0.41 (n = 8) versus 9.1 ± 0.7 (n = 8) ppb, P < .001). Air-exposed NOS1^−/− and NOS3^−/− also had significantly lower exhaled NO levels compared to the C57Bl/6 mice (P < .05). After exposure to OVA, there were no significant increases in the exhaled NO levels in any strain compared to their respective air-exposed group. C57Bl/6 mice exposed to OVA had a significant decrease in exhaled NO compared to their filtered air controls in this experiment. In contrast, lung lavage NO_x concentration (b) from NOS1^−/− and NOS2^−/− mice exposed to OVA was significantly less than C57Bl/6 mice exposed to OVA. Data are presented as means ± SEM. * denotes P < .05; ***P < .001 by ANOVA.

Figure 3

Correlation between the concentrations of NO_x measured in lung lavage and intensity of staining for NOS2 protein in airways of NOS3^−/− mice exposed to both air and OVA. Data are presented as both raw data and best fit line with 95% confidence intervals; (m = 18.02 ± 6.17, P = .01).

Figure 4

Total lung compliance in (a) C57BL/6, (b) NOS1^−/−, (c) NOS2^−/−, and (d) NOS3^−/− mice exposed to either filtered air or 2 weeks of OVA. Symbols: (blue) open triangles for filtered air exposure, (black) circles for OVA exposure. Lung compliance was measured at baseline and following serial doses (0, 0.5, 1.0, and 2.0 mg/mL) of nebulized methacholine (MCh). The slope of the MCh response for NOS2^−/− mice exposed to OVA (−0.0028 ± 0.00042, F = 43.56) was significantly different from that of the NOS2^−/− mice exposed to air (−0.0003 ± 0.00017, F = 3.0; P = .005). For NOS3^−/− mice, the slopes of the best fit lines for the MCh response curve for OVA-exposed mice (−0.0005 ± 0.0002, F = 4.627) and filtered air-exposed mice (−0.003 ± 0.0003, F = 95.07) were also different (P = .004). NOS3^−/− mice (both air and OVA exposed) had significantly lower lung compliance at baseline compared to all of the other strains and were most different from the NOS1^−/− (Cdyn: 0.017 ± 0.002 (n = 12) versus 0.031 ± 0.0008 (n = 8) mL/cmH₂O, resp., P < .0001) and NOS2^−/− (Cdyn: 0.017 ± 0.002 versus 0.029 ± 0.0008 (n = 24) mL/cmH₂O, resp., P = .0002).

Figure 5

Total lung resistance in (a) C57BL/6, (b) NOS1^−/−, (C) NOS2^−/−, and (D) NOS3^−/− mice exposed to either filtered air or 2 weeks of OVA. Symbols: (blue) open triangles for filtered air exposure, (black) circles for OVA exposure. Lung resistance was measured at baseline and following serial doses (0.5–2.0 mg/mL) of nebulized methacholine. NOS1^−/− mice exposed to OVA (slope = 0.266 ± 0.027, F = 94.09) had a significantly greater increase in lung resistance compared to the same strain of mice exposed to filtered air (slope = 0.079 ± 0.008, F = 93.15; P = .002). Air- and OVA-exposed NOS2^−/− mice had similar increases in lung resistance (slopes = 0.12 ± 0.01 and .015 ± 0.02, resp.), but their intercepts were significantly different (y intercept 1.09 ± 0.02 versus 1.3 ± .0.06, resp., P = .0001).

Figure 6

Relative band intensity of NOS2 protein by Western blot in (a) NOS1^−/− and (b) NOS3^−/− mice. NOS1^−/− animals showed an upregulation of inducible NOS2 in air control animals and maintained the pattern of NOS2 upregulation in response to OVA treatment. NOS3^−/−animals also showed a significant upregulation of NOS2 in filtered air-exposed animals, but no change in NOS2 in response to OVA treatment. Data are expressed as mean values ± SEM (n = 4). * denotes P < .05; **P < .01 by ANOVA.

Figure 7

Relative band intensity of staining of (a) NOS1 protein and (b) NOS3 protein by Western blot in NOS2^−/− mice. No significant change is in the protein levels of either NOS1 or NOS3 in airways of mice exposed to filtered air or OVA. Data are expressed as mean values ± SEM (n = 4 each).

Figure 8

Immunohistochemical stain of NOS2 protein in 5 μm thick left lung sections from C57Bl/6 (a) filtered air-exposed and (e) OVA-exposed mice, NOS1^−/− (b) filtered air-exposed and (f) OVA-exposed mice, NOS2^−/− (c) filtered air-exposed mice, and NOS3^−/− (d) filtered air-exposed and (g) OVA-exposed mice. Airway smooth muscle layer is indicated by arrow. Images were taken at 400× magnification.

Figure 9

Semiquantitative analysis of NOS2 staining intensity from NOS1^−/− and NOS3^−/− mouse strains exposed to filtered air or OVA. Immunohistochemical staining intensity and consistency of stain in the (a) airway epithelium, (b) smooth muscle of the airways and vasculature, and (c) macrophage populations were scored on a scale of 0–10. Filtered air-exposed NOS1^−/− and NOS3^−/− mice displayed uniform increases in NOS2 protein staining limited to the smooth muscle of airways and vasculature and the airway epithelium with no change in macrophages. OVA-exposed NOS3^−/− mice had a significant reduction in NOS2 staining in the airway epithelium and smooth muscle compared to their filtered air-exposed counterparts. In contrast, OVA-exposed NOS1^−/− mice maintained NOS2 protein staining in the airway epithelium and a significant increase in NOS2 in the macrophage population. Data are presented as mean values ± SEM (n = 5-6); * denotes P < .05; **P < .01 by ANOVA.

Figure 10

PAS staining of 5 μm-thick left lobe lung sections from (a) C57Bl/6 filtered air-exposed, (b) C57Bl/6 OVA-exposed, (c) NOS1^−/− OVA-exposed, and (d) NOS3^−/− OVA-exposed mice. Images were taken at 400× magnification.

Figure 11

Percentage PAS positive cells present in the airway epithelium from C57Bl/6, NOS1^−/−, and NOS3^−/− genotype mice exposed to filtered air or OVA. Filtered air-exposed mice displayed an average of 0.45% PAS positive cells with no difference between the three groups. OVA-exposed NOS1^−/− mice had significantly fewer PAS positive-stained cells compared to the OVA-exposed C57Bl/6. Data are presented as mean values±SEM; ** denotes P < .01 by ANOVA.