INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Nitric oxide (NO) has been detected in the breath exhaled by many mammalian species, including humans (1), elephants (2), guinea pigs, and rabbits (3). While paranasal sinuses seem to be an important source of exhaled NO (4), it has been reported that exhaled NO is also derived from the lower airways and is synthesized in an airflow-dependent fashion (5). Many hypotheses have been proposed to provide a function for exhaled NO. Exhaled NO may function as an endogenously produced vasodilator that is autoinhaled to produce pulmonary vasodilation (6). Exhaled NO levels have been measured as a marker of endogenous NO production, since NO is produced in large amounts by inflamed airway cells in asthmatic exacerbations (7). NO in airways and sinuses may be bacteriostatic or bacteriocidal and reduce colonization and infection.

Throughout the body, NO is synthesized by nitric oxide synthases (NOS), and the cellular location and function of the three known NOS isoforms (NOS 1 or neuronal, NOS 2 or inducible and NOS 3 or endothelial) have been extensively studied. All of the three NOS isoforms are expressed in cells that have direct contact with or are in proximity to the alveolar surface and the lower and upper airways.

The enzymatic source of exhaled NO is unknown. Since the available pharmacological NOS inhibitors are not completely selective for one NOS isoform (10, 11), we studied mice with a congenital absence of each of the NOS isoforms. We developed a measurement technique using group analysis employing a sensitive chemiluminescence NO analyzer to detect the low NO concentrations exhaled by mice, and studied wild-type mice and mice with a congenital deficiency of either NOS 1, NOS 2, or NOS 3.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

After institutional approval by the Massachusetts General Hospital Subcommittee on Research Animal Care, we studied NOS 1-, 2-, and 3-deficient mice, as well as control B6129/F1 (NOS +/+) mice of both sexes with a body weight of 18 to 35 g and an age of 2 to 5 mo. The generation of the NOS 1-deficient (NOS 1/) and NOS 3-deficient (NOS 3/) mice has been described by Dr. P. L. Huang and colleagues (12, 13). NOS 2-deficient (NOS 2/) mice were generated by Dr. J. D. MacMicking and colleagues (14).

Experimental Setup

Calibration. The NO analyzer (NOA 280; Sievers, Boulder, CO) was calibrated twice at the beginning and again at the end of each experiment to assess the stability of the zero and signal levels. Oxygen (purity 99.999%; BOC Gases, Port Allen, LA) was used to calibrate the zero level. No NO₂ was present in the purified oxygen (measured with an NO_X analyzer CLD 700 AL; Ecophysics, Duernten, Switzerland), suggesting the absence of prior NO contamination. NO (4.06 ppm) in nitrogen (BOC Gases) was diluted 1:50 with NO-free nitrogen to obtain a final concentration of 81.2 ppb NO. This concentration was used for calibration. Serial dilutions of the 4.06 ppm NO-nitrogen gas mixture (1:100, 1:200, and 1:400) were each repeated five times and measured to assess linearity of the analyzer in the range of interest (0-50 ppb). The concentrations measured were 39.9 ± 1.3 ppb (M ± SE; dilution 1:100, goal 40.6 ppb), 19.7 ± 0.9 ppb (1:200, goal 20.3 ppb), and 9.9 ± 0.8 ppb (1:400, goal 10.15 ppb).

The CO₂ analyzer (Model 2200; Traverse Medical Monitors) was calibrated prior to each experiment with 4-6% CO₂. The output signals of the NO and CO₂ analyzers were digitized using an analog-digital converter, displayed on a computer screen, and continously recorded (DI 220; Dataq Instruments, Akron, OH).

NO Measurements

Experimental apparatus (Figure 1). Mice were studied in an airtight 400-ml chamber. To obtain a thermostable environment with no NO contamination, this chamber was placed inside a larger chamber that was continuously flushed with pure oxygen at a rate of 15 L/min during the experiment. The concentration of NO in the outer chamber was < 1 ppb during each experiment; the temperature was stable at 27° C. Two tubes, used for in- and outlet gas flushing, were connected to the chamber and equipped with stopcocks. A 100-ml gas reservoir and the oxygen supply were connected to the inlet via a stopcock; the NO analyzer and the CO₂ analyzer were connected to the outlet.

View larger version (34K): [in this window] [in a new window]   Figure 1.   Experimental setup: Groups of five mice were placed into the inner chamber, which then was closed and flushed with purified oxygen for 3 min. NO and CO2 concentrations during the flushing period were monitored. After flushing, stopcocks A and B were closed toward the chamber for 4 min. The reservoir was filled with 100 ml purified oxygen. After isolation time, the reservoir was opened toward the chamber and gas was withdrawn by the NO and CO2 analyzer to measure NO and CO2 concentrations. Only the first 5 s of withdrawal with a stable NO and CO2 signal was used for calculations, before significant dilution with the reservoir gas occurred. Signals from the analyzers were digitized, recorded, and displayed.

Experimental protocol. After the mice were placed in the inner chamber (see Figure 1), this chamber was flushed with highly purified oxygen at a flow rate of 1 L/min. CO₂ and NO concentration were measured and recorded at the outlet during the flushing period. After flushing, the reservoir was filled with 100 ml O₂ and the system was closed. The flushing period to remove basal NO from the chamber (3 min) and the optimal time of study resulting in a measurable NO signal with minor degrees of ambient hypercarbia were determined to be 4 min in pilot experiments. After exactly 4 min of isolation, gas for NO and CO₂ measurements was sampled from the chamber. The reservoir was then opened into the chamber to allow gas sampling without generating negative pressures within the chamber. Stable gas concentrations were obtained during the last 15 s before isolation of the chamber and during the first 5 s after opening the sampling tubes to the analyzer. Control experiments without mice were performed using an identical time sequence of flushing and airtight closure to assess the presence of NO absorption or release in the chamber system, or a NO leak from the environment into the chamber system. Serial dilutions of 4.06 ppm NO (1:100, 1:200, and 1:400) were measured after being flushed through the chamber system and compared to the values obtained directly from the mixture bag.

Experimental groups. A total of 315 mice were studied. Male and female mice were matched for age, sex, and weight prior to assignment to groups and experiments. NOS +/+ mice (80 mice) were placed in groups of 5 mice (n = 16 groups of five mice) into the chamber and exhaled gas concentrations were measured in separate experiments after isolation in the chamber for 4 min.

Five groups of five NOS +/+ mice received an intraperitoneal injection of L-NAME (1 mg/g [body weight] bw), which was repeated after 1 h. Thirty minutes after the second L-NAME injection, the NO and CO₂ concentrations after 4 min of inner chamber isolation were measured. A second set of five groups of NOS +/+ mice received an intraperitoneal injection of L-arginine (5 mg/g bw), which was repeated after 1 h. Ten minutes after the second L-arginine injection, NO and CO₂ concentrations after 4 min of inner chamber isolation were measured.

NOS 1/ mice (75 mice, n = 15 groups), NOS 2/ mice (70 mice, n = 14 groups), and NOS 3/ mice (80 mice, n = 16 groups) were measured after 4 min of inner chamber isolation. Seven groups of NOS 1/, six groups of NOS 2/, and seven groups of NOS 3/ mice received L-NAME as described above, and NO and CO₂ concentrations were measured. Eight groups of NOS 2/, five groups of NOS 1/, and nine groups of NOS 3/ mice received L-arginine as described above, and NO and CO₂ concentrations were measured after 4 min of isolation.

To determine the contribution of inner and outer surface microbial colonization to NO production, two groups (10 mice total) of germ-free NIHSGF mice were studied, and gas concentrations were measured after isolation in the inner chamber for 4 min.

Statistical Analysis

NO and CO₂ production rates were calculated as moles of gas produced per gram body weight per minute of inner chamber isolation. Differences between groups were determined using a variance analysis with a post hoc Newman-Keuls test. The effects of interventions were examined using a paired t test. A significant difference was assumed with a p value below 0.05. All data are expressed as mean ± standard error.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Flushing serial dilutions through the chamber without mice resulted in the following concentrations: 38.0 ± 1.1 ppb (dilution 1:100 after chamber) versus 39.9 ± 1.3 ppb (dilution 1:100 directly from mixture bag, p = NS), 18.9 ± 1.3 ppb (1:200 dilution after chamber) versus 19.7 ± 0.9 ppb (1:200 dilution from bag, p = NS), and 9.2 ± 0.5 ppb (1:400 dilution after chamber) versus 9.9 ± 0.8 ppb (1:400 dilution from bag, p = NS). No NO was detected after flushing the system with purified grade 5 oxygen.

After 4 min of inner chamber isolation, NOS +/+ mice produced 362 ± 35 × 10¹⁵ mol g¹ min¹ NO and 2.0 ± 0.1 × 10⁶ mol g¹ min¹ CO₂. NOS 1/ mice produced 592 ± 74 × 10¹⁵ mol g¹ min¹ NO (p < 0.05 versus NOS +/+ and versus NOS 2/), NOS 2/ mice produced 330 ± 34 × 10¹⁵ mol g¹ min¹ NO and NOS 3/ mice produced 766 ± 101 × 10¹⁵ mol g¹ min¹ NO (p < 0.001 versus NOS +/+ and versus NOS 2/) (see Table 1, Figure 2). The CO₂ production rate during 4 min of chamber isolation was not significantly different between any of the groups (see Table 1).

View larger version (19K): [in this window] [in a new window]   Figure 2.   NO production by NOS 3+/+, NOS 1/, NOS 2/, and NOS 3/ mice after 4 min of chamber isolation. NOS 1/ mice and NOS 3/ mice produced significantly more NO than NOS +/+ or NOS 2/ mice. *p < 0.05, ***p < 0.001 versus NOS +/+ mice; #p < 0.05, ###p < 0.001 versus NOS 2/ mice.

After the intraperitoneal administration of L-NAME or L-arginine, all mice appeared less physically active and calmer. No mouse died as a result of the intraperitoneal medication. The administration of L-NAME (see Figure 3) significantly reduced NO production in both the NOS +/+ mice and the NOS 3/ mice (p < 0.05 versus before L-NAME). NO production by NOS 1/ and NOS 2/ mice were not affected by L-NAME administration. CO₂ production was reduced by L-NAME in all mice (p < 0.01 versus before L-NAME, see Figure 3). The administration of L-arginine (see Figure 4) significantly increased the NO production rate in NOS +/+ mice (p < 0.01 versus before L-arginine) and NOS 3/ mice (p < 0.05 versus before L-arginine), whereas NOS 1/ mice (p = 0.06 versus before L-arginine) and NOS 2/ were not significantly affected. CO₂ production was significantly decreased by L-arginine in all mice (p < 0.05 in NOS +/+ versus before L-arginine, and p < 0.01 in NOS 1/, NOS 2/ and NOS 3/ mice versus before L-arginine, see Figure 4).

View larger version (15K): [in this window] [in a new window]   Figure 3.   Effects of intraperitoneal injections of L-NAME (two injections at 1 mg/g bw within 1 h) on NO production in mice. Note that L-NAME reduced NO production in NOS +/+ and NOS 3/ mice. *p < 0.05 versus preadministration, **p < 0.01 versus preadministration.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Effects of intraperitoneal injections of L-arginine (two injections at 5 mg/g bw within 1 h) on NO production in mice. Note that L-arginine increased NO production in NOS +/+ and NOS 1/ mice. *p < 0.05 versus preadministration, **p < 0.01 versus preadministration.

NIHGF germ-free mice produced 371 ± 40 × 10¹⁵ mol g¹ min¹ NO and 1.54 ± 0.28 × 10⁶ mol g¹ min¹ CO₂ during 4 min in chamber isolation, which was not different from the results obtained with NOS +/+ mice.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We learned that genetically intact mice that express all three isoforms of NOS exhale NO in measurable concentrations into a small closed chamber. The exhaled NO production rate can be inhibited by treatment with the competitive NOS inhibitor L-NAME and can be stimulated by challenging with the NOS substrate, L-arginine. Congenital absence of the NOS 2 isoform did not reduce baseline exhaled NO production as compared to NOS +/+ mice, and the basal exhaled NO production rate by NOS 2/ mice was not responsive to either inhibition or stimulation of NOS activity by L-NAME or L-arginine, respectively. The absence of NOS 1 or NOS 3 resulted in an increased baseline exhaled NO production rate, as compared to NOS +/+ or NOS 2/ mice, suggesting the up-regulation of an NOS-isoform may be responsible for the gaseous NO production.

The measurement of gaseous NO from the respiratory tract of a mouse in the low parts per billion (ppb) range provides a challenge for chemiluminescence NO measurement technology and the experimental design. Mask breathing (4, 15) and nasal cannulation (16) are commonly used in larger mammal species and humans to obtain NO from the upper respiratory tract, and are not feasible for the mouse. Endotracheal intubation in the mouse is possible and lower respiratory tract NO levels have been measured between 4 and 6 ppb NO (17). However, tracheal intubation limits the detection of NO sources to the tracheobronchial tree and the lungs, and excludes the upper airways and paranasal sinuses. Because environmental NO pollution can easily lead to NO concentrations that far exceed the NO concentrations produced by the mouse, it was particulary important to create an experimental design in which the baseline NO concentration was known to be zero. To amplify the low NO concentrations produced by a mouse, we studied groups of five mice.

In addition to production of gaseous NO from the upper and lower respiratory tract, the NO we collected may partially have been derived from NOS activity of the skin (18, 19) and of the upper or lower gastrointestinal tract (20, 21). The fractional contribution of non-airway NO to the total NO concentration that we measured is not known. It is unlikely that NO is derived from microbial colonization of skin and internal mucosa, as germ-free mice produced amounts of NO similar to wild-type mice with a normal microbial flora.

The absolute levels of exhaled NO concentrations varies considerably among mammalian species. A spectrum of animal data is available, showing a broad variation of exhaled and nasally sampled NO concentrations among species. The comparison to our data is difficult as it was not feasible in mice to obtain separate nasal and lower respiratory tract gas sampling as it is in larger animals. We therefore refer in this study to "exhaled gas" as all gas derived from both the lower and the upper respiratory tract including nasal sources. Most animal and human studies have reported measuring continuous exhaled concentrations, whereas the technical conditions of our study only allowed the measurement of the mean exhaled NO production rate per unit body weight over 4 min.

NO was not detectable in the nasally sampled gas of cats, dogs, cows, horses, a yak, a brown bear (22), and Weddell seals. Gas obtained from the nasal cavity of pigs contained 16- 17 ppb NO, 180-650 ppb NO in a chimpanzee, 210-460 ppb NO in rhesus monkeys, and below 10 ppb in giraffes (22). Gas sampled from the nasal trunk of elephants contained 30 ppb NO (2). Expired gas from mechanically ventilated animals following a tracheostomy contained 13 ppb NO in rabbits, 6 ppb in guinea pigs, and 1 ppb in rats (23). It appears that the exhaled NO levels in mice are in a range comparable to NO concentrations exhaled by other mammals.

In healthy humans, the majority of exhaled NO is released from the epithelium of the upper respiratory tract, especially the nasal cavity. Gas from paranasal sinuses has a high NO concentration, and the majority of nasally sampled NO derives from the paranasal mucosa (24). Gas samples obtained via a tracheostomy or endotracheal tube excluding NO production from sources above the vocal cords were found to be in the range of 1-5 ppb, whereas nasally sampled gas contained as much as 400 ppb NO in healthy individuals (25).

All three NOS isoforms are expressed in tissues in close proximity to respiratory gases, in the nasal cavity (26), the conducting airways (31, 32), and resident and migratory cells of the alveolar surface (33). Exhaled NO produced by healthy individuals seems to be primarily synthetized by the constitutive isoform of the paranasal sinuses (24). Immunohistochemical and mRNA in situ hybridization studies by Lundberg and coworkers showed that an NOS resembling NOS 2 was constitutively expressed apically in the sinus epithelium of the paranasal sinuses (24), suggesting that NOS 2 is the primary source of nasal and paranasal NO. In preliminary experiments, we found evidence that NOS 2 immunoreactivity in mice is predominantly present in the single cell layer of the nasal mucosa, but rarely expressed in trachea and lung (no unpublished results).

In the presence of inflammatory airway and pulmonary disease, NO production from the lower respiratory tract increases substantially and significantly contributes to exhaled NO concentrations (7, 34). A congenital absence of NOS 2 has been associated with decreased allergic airway inflammation in mice, suggesting that NOS 2 is a key inflammatory factor (37), although the significance of this finding remains controversial (38).

In our study, NOS 2-deficient mice produced a similar quantity of NO as compared with wild-type mice at baseline (see Table 1). It was not possible for us to suppress baseline NO exhalation in NOS 2-deficient mice by treating them acutely with a nonselective NOS inhibitor, L-NAME (see Figure 3). In contrast, L-NAME reduced exhaled NO production in wild-type mice, suggesting that NOS 2 in wild-type mice responded to L-NAME with decreased NO production. Healthy volunteers who received L-NAME by nebulization subsequently exhaled less NO (39). Decreased NO exhalation after L-NAME has been reported in ventilated guinea pigs and rabbits (23).

Reasoning further from our evidence, NOS 2-deficient mice and wild-type mice responded differently to substrate stimulation with L-arginine (see Figure 4). NOS 2-deficient mice had low and unchanged exhaled NO production after L-arginine injection, whereas wild-type mice exhaled signficantly more NO. This result suggests that wild-type NOS 2 responds to intraperitoneal administration of L-arginine with increased NO synthesis and subsequent exhalation, whereas NOS 2-deficient mice lack this response. The response of wild-type mice to intraperitoneal L-arginine administration resembles the response in humans who exhale more NO after L-arginine nebulization (40). Similarly, tracheally intubated and ventilated guinea pigs and rabbits increase NO exhalation after L-arginine stimulation (23).

The murine exhaled NO responses to L-NAME and L-arginine administration support the assumption that NOS 2 is responsible for NO exhalation in a healthy state. However, it is unclear why a low exhaled NO production rate is preserved in mice with congenital NOS 2 absence (Table 1). It is also possible that a low remaining NOS 2 activity is preserved in NOS 2-deficient mice.

Exhaled NO may be derived from other NOS isoforms in proximity to the airway. NOS 1 is expressed in neuronal tissue in the nasal mucosa as well as in neurons of the airways. As recently shown by DeSanctis and coworkers, tracheotomized NOS 1-deficient mice had lower expired NO concentrations than wild-type mice (17), suggesting that pulmonary NOS 1 contributes to exhaled NO. The authors report that NO concentrations in mixed expiratory gas samples collected over 5 min were 6.3 ± 0.9 ppb in wild-type mice versus 3.9 ± 0.4 ppb in NOS 1-deficient mice. These results suggest that pulmonary and tracheal NOS 1 may contribute to NO exhalation. In our experiments, however, the NO production rate in NOS 1-deficient mice was elevated as compared to wild-type mice. One possible explanation for these differences is the different experimental approaches: DeSanctis' method excluded all upper airway NO and nasal sources, whereas our method included these upper airway sources.

Nitric oxide may also have been synthesized by NOS 3 in nasal and airway epithelial and endothelial cells, where it is ubiquitously expressed, and then diffused into the airspace (33). However, if NOS 1 and/or 3 were responsible for baseline NO production, an effect of NOS inhibitors or NOS substrate on exhaled NO levels would have been expected in NOS 2-deficient mice; this was not observed (Figures 3 and 4)

Nonenzymatic production of NO should be considered as another possible source of exhaled NO. NO can be produced from nitrite in biological systems under certain conditions, e.g., acidosis or ischemia (41, 42). It can be generated by either direct disproportionation or reduction of nitrite to NO in an acidic and highly reduced biological environment. Nonenzymatic NO formation is not blocked by NOS inhibitors and with long periods of ischemia progressing to necrosis, this mechanism of NO formation may predominate (43). It appears unlikely, however, that nonenzymatic NO formation contributed to NO exhalation in our model since none of the required conditions (e.g., tissue ischemia or necrosis) was present.

Both mice treated with L-NAME and L-arginine significantly decreased their CO₂ production rate, independent of the direction that the NO production rate changed. It appears likely that the reduced CO₂ production was secondary to a decreased total metabolic rate, as we observed decreased physical activity of the mice after medication. Whether this was due to sedative effects of the NOS substrate or inhibitor (e.g., L-NAME is known to facilitate sedation and to reduce the threshold for isoflurane anesthesia [44]) or due to "protective" physical inactivity after painful peritoneal stimulation by the injection is not known.

Interestingly, NOS 1- and NOS 3-deficient mice had an increased baseline NO production, as compared to wild-type or NOS 2-deficient mice (see Table 1). Stimulation of NOS with L-arginine in NOS 1- and NOS 3-deficient mice resulted in substantially increased exhaled NO concentrations. L-NAME inhibition considerably diminished the levels of NO exhalation in NOS 1- and NOS 3-deficient mice. Enhanced expression or activity of a NOS isoform, which is present in both NOS 1- and NOS 3-deficient mice, might be responsible for the increased exhaled NO quantities.