INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Nitric oxide (NO) is a multifunctional mediator produced by the NO synthase (NOS) family of enzymes present in virtually all mammalian cells (1, 2). NO mediates a wide variety of biologic phenomena in the respiratory system, including pulmonary vasodilatation, bronchodilatation, and antimicrobial and antitumor host defense (3, 4). Abnormalities of endogenous NO synthesis, release and/or reactions with other molecular species have been implicated in the pathophysiology of several respiratory disorders such as asthma and acute lung injury (3, 5, 6).

Support for the putative roles of NO in vivo is based in part on the direct measurement of NO or its oxidative metabolites (e.g., nitrites and nitrates) in various biologic compartments (7). With regard to the direct measurement of respiratory NO, gas-phase NO has been found in the exhaled breath of humans and various animal species (8). Indeed, there is intense research interest in a potential clinical role for the measurement of exhaled NO (E_NO) in human disease. For example, increased E_NO levels have been found in asthmatic patients, whereas patients with cystic fibrosis have decreased E_NO levels (5, 9). These differences in E_NO levels in disease states have led to the suggestion that measurement of E_NO may be of clinical use as a noninvasive biologic marker of lung disease.

Given the potential roles for endogenous NO in human disease, there has been much work on measuring E_NO in many animal species and in various models of human disease. The measurement of E_NO in mice, with increasing availability and use of NOS isoform-deficient and transgenic mouse strains, would complement current research approaches into the roles of endogenous NO. However, in previous work, most investigators have used either larger animal species and/or mechanically ventilated animals to measure E_NO (6, 10, 11). Invasive exhaled gas sampling techniques do not easily allow for repeated measurements of E_NO during the evolution of a disease state, and are subject to the poorly defined effects of anesthesia, surgery, and mechanical ventilation. Thus, in this report, we describe a novel technique for the noninvasive sampling of exhaled gas and the measurement of E_NO in a single, spontaneously breathing mouse.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal Use

Male C57BL/6 mice 10 to 12 wk of age and weighing 21 to 28 g (Charles River Labs, St. Constant, PQ, Canada) were allowed to acclimatize to animal housing quarters for 1 wk with access to food and water ad libitum. The study protocol was approved by the institutional Animal Ethics Committee in accord with the principles of the Canadian Council on Animal Care and was supervised by a veterinarian.

Description of Apparatus

Gas phase NO was measured by a chemiluminescence-based NO analyzer sensitive to 0.1 ppb NO (Model 270B; Sievers Instruments, Boulder, CO) (7). The output from the NO analyzer (mV) was recorded on a chart recorder (Chromatopac C-R1A; Shimadzu Co., Kyoto, Japan). For the measurement of E_NO, a single mouse was placed inside a sealed Plexiglas chamber (85 ml; Cadillac Plastics, London, ON, Canada) perfused with medical air (FI_O2 = 0.21) using high accuracy flow meters (± 1%; Matheson Gas Co., Whitby, ON, Canada) (Figure 1). The outflow from the chamber passed through a Nafion dryer (Sievers) ensuring that dry cylinder calibration gas has the same final humidity as exhaled gas. The sample was then intermittently collected in a Mylar balloon (NO-inert) over 120 s before being introduced into the NO analyzer. Apart from the Plexiglas chamber, all other parts of the calibration and exhaled gas collection system are NO-inert, including Teflon tubing, flow meters (glass, stainless steel, and Teflon components) and stainless steel Swagelok compression fittings (Huron Valve and Fitting, Sarnia, ON, Canada).

View larger version (44K): [in this window] [in a new window]   Figure 1.   Exhaled NO (ENO) analysis apparatus. A single mouse was placed inside a sealed Plexiglas chamber and allowed to acclimatize for 15 min. At the rear of the chamber, a T-junction allows for the mixing of medical air (FIO2 = 0.21) and NO calibration gas, the flows of these gases being controlled by high accuracy flow meters (± 1%, not shown). The outflow from the chamber passes through a hygroscopic perfluminated ion-exchange membrane (Nafion) dryer that allows equilibration of humidity in all samples to a uniform ambient humidity level. The sample is then intermittently collected in a Mylar balloon (NO-inert) over 120 s and subsequently introduced into the NO analyzer. For the comfort of the mice, excess flow of medical air through the chamber is provided and is evacuated through an underwater seal (1 cm H2O) to prevent contamination of gas samples by ambient air.

Exhaled NO Analysis System Calibration

A calibration curve for NO concentration versus mV signal was established daily. Medical air (< 1.0 ppb NO; Praxair, London, ON, Canada) was used to define the zero level, and serial dilutions of a calibration NO gas (640 ppb in N₂; Matheson) with medical air resulted in standard gases of known NO concentration of 128, 64, 32, and 10 ppb. Calibration signals for each [NO] were recorded in triplicate. To determine if our system had any effect on [NO], the signals generated by standard concentrations of NO were compared when these gases (1) passed through the animal chamber/balloon collection system, and (2) bypassed the animal chamber and were introduced directly into the NO analyzer.

To determine the reaction chamber pressure that allows for an optimal sensitivity (mV signal/ppb NO) of the NO analyzer, 640 ppb NO gas were analyzed at various reaction chamber pressures (5 to 11 mm Hg) and the resulting signals compared.

E_NO Samples from Mice

A single mouse was placed in the animal chamber, which was flushed with medical air for 15 min (resulting in a stable signal) before sampling occurred. In all experiments, E_NO for each animal was sampled in triplicate and is reported in ppb (FE_NO). Excretion of NO in exhaled gas (VE_NO) was calculated as the product of FE_NO and chamber flow rate, and it is reported in mol/min/g body weight.

Effect of animal chamber flow rate on E_NO. In six mice, we determined the animal chamber flow rate (60, 100, 200 or 300 ml/min) of Medical Air that yielded the greatest signal while the mouse remained comfortable. For all subsequent measurements of mouse E_NO, the optimal animal chamber flow rate (60 ml/min) and NO analyzer reaction chamber pressure setting (8.0 mm Hg) were used.

Within-day and between-day variability of E_NO. Within-day variation was assessed by analyzing E_NO from 12 mice both in the morning (7:00 to 11:00 A.M.) and in the afternoon (1:00 to 5:00 P.M.). Between-day variability was assessed by measuring E_NO in six mice daily (between 7:00 A.M. and 5:00 P.M.) for 7 d.

Effect of food intake on E_NO. E_NO from six mice was sampled (1) after fasting for 14 h and (2) 4 h after food (PicoLab Mouse Diet 20; PMI Nutrition, Brentwood, MO) was reintroduced. Animals were allowed access to water throughout this experiment.

Effect of anesthesia and tracheostomy on E_NO. E_NO from six mice was measured before and 30 min after an intraperitoneal injection of ketamine/xylazine (1.2 and 1 mg/g body weight, respectively, in 0.9% saline). To determine the contribution of the upper airways to the E_NO signal, a 21-gauge silastic catheter was inserted into the lower trachea through a midcervical tracheotomy (under the same anesthesia as above). Thus, upper airway contact with exhaled gas was eliminated.

Biologic Responsiveness of E_NO

Effects of NOS inhibition on E_NO. After the baseline measurement of E_NO in five mice, 1 mg/g body weight of N^G-nitro-L-arginine methyl ester (L-NAME; Sigma Chemical Co., St. Louis, MO) was administered intraperitoneally twice, 1 h apart (12). Thirty minutes after the second injection, E_NO was again measured.

Effects of lipopolysaccharide-induced lung injury on E_NO. In six mice, E_NO was measured before and 10 h after an intraperitoneal injection of 12.5 µg/g body weight bacterial lipopolysaccharide (0.5 ml, LPS, E. coli serotype 011:B4; Sigma). Six sham mice received an intraperitoneal injection of saline (0.5 ml). All animals were killed (sodium pentobarbitol, 60 mg/kg given intraperitoneally) after the final E_NO measurement. Arterial blood was obtained through cardiac puncture for measurement of nitrites/nitrates (NO₂ + NO₃ = NO_x) in plasma samples by chemiluminescence (13). Lung form LPS- and saline-treated mice were excised and frozen in liquid nitrogen, and pulmonary cNOS and iNOS activity was quantified by the conversion of [³H]L-arginine to [³H]L-citrulline (14). Enzyme activities are expressed as pmol [³H]L-citrulline evolved/min/mg protein. As a marker of LPS-induced lung injury, the lung wet-dry weight ratio was calculated.

Statistical Analysis

All values reported are mean ± SE. Statistical significance was accepted at p values less than 0.05. The effect of the time of day, L-NAME, and LPS administration on E_NO was analyzed using Student's paired t test. Effects of anesthesia and tracheostomy, as well as stability of E_NO over 7 d were analyzed by a repeated-measures ANOVA.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Exhaled NO Analysis System Calibration

There was a high degree of linearity between [NO] (0, 10, 32, 64, 128 ppb) in standard gas samples passed through our animal chamber/sample collection system and the NO analyzer mV signal (r² > 0.999). Moreover, for the repeated analysis of these standard gases daily over a 1-wk period, the CoV was < 4% (data not shown). The concentration of NO in a gas sample was not affected positively or negatively by passage through our animal chamber and sample collection system. The mV signals generated from standard samples of known NO concentration varied by < 1% when passed through the chamber/ collection system versus when passed directly into the NO analyzer (data not shown). When a gas of known [NO] (640 ppb) was analyzed at various NO analyzer reaction chamber pressures (5 to 11 mm Hg), an inverted U-shaped curve of sensitivity (mV signal/ppb NO) was found. Optimal NO analyzer sensitivity was identified at a reaction chamber pressure of 8 mm Hg. This operating pressure was used in all subsequent measurements of E_NO.

E_NO Samples from Mice

Effect of animal chamber flow rates on mouse E_NO. All animals adapted quickly to the confines of the chamber, and after an initial period of exploratory behavior, appeared comfortable and quiet during exhaled gas collection and analysis; at no time was any mouse observed to be sleeping. A decreasing flow rate (300, 200, 100, 60 ml/min) of medical air into the animal chamber was associated with an increasing signal, such that the highest FE_NO signal for each animal was recorded at a flow rate of 60 ml/min (Figure 2). The mean VE_NO for six mice tested at various flow rates were not significantly different (p = NS). A flow rate of 20 ml/min generated a higher FE_NO than 60 ml/min (data not shown), but the animals had increased respiratory effort and appeared agitated. Thus, the optimal flow rate of 60 ml/min was used in all subsequent experiments both for the comfort of the animal and the reliability of the data. Under these conditions, mean FE_NO for 25 mice was 10.1 ± 1.1 ppb (range, 5.9 to 16.1), for a calculated VE_NO of 1,061 ± 110 × 10¹⁵ mol/min/g (range, 728 to 1,643 × 10¹⁵).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Effect of varying flow rate of medical air on the concentration of exhaled NO (FENO) in mice (n = 6). There was an inverse exponential relationship between chamber flow rate and FENO.

Within-day and between-day variability of E_NO. The time of day at which E_NO samples were collected from mice and analyzed had no effects on E_NO levels. Mean FE_NO in 12 mice in the A.M. (7:00 to 11:00) and P.M. (1:00 to 5:00) time periods was 7.9 ± 0.5 and 8.0 ± 0.5 ppb, respectively (p = NS). In individual mice, the change in E_NO (from A.M. to P.M.) ranged from 1.5 ppb to +1.4 ppb. Moreover, when measured daily for 7 d in six mice, E_NO varied over a maximal range of ± 2.1 ppb from each mouse's mean E_NO. The mean CoV for E_NO in individual mice over 1 wk was 16 ± 1%.

Effect of food intake on E_NO. Mice noticeably increased their food intake upon its reintroduction after the fasting period. However, E_NO levels were unaffected by food intake (10.6 ± 0.5, fasting versus 11.3 ± 0.4 ppb, nonfasting, p = NS).

Effects of anesthesia and tracheostomy on E_NO. After anesthesia, mice remained unresponsive to pain for at least 40 min, and they had decreased but stable respiratory effect and rate. Anesthesia significantly decreased FE_NO levels by 42 ± 7% (Figure 3). Tracheostomy in anesthetized mice to eliminate upper airway contribution to E_NO signal had no additional effect on the reduced FE_NO levels after anesthesia.

View larger version (13K): [in this window] [in a new window]   Figure 3.   Effect of anesthesia and tracheostomy on the concentration of exhaled NO (FENO) in mice (n = 6) (*p < 0.01 versus Baseline). Whereas anesthesia-reduced ENO tracheotomy had no additional effect on ENO in anesthetized mice.

Biologic Responsiveness of E_NO Signal

Intraperitoneal administration of the nonisoform selective NOS inhibitor L-NAME significantly decreased FE_NO by 51 ± 6% (p < 0.01) in five mice (Figure 4). Systemic administration of LPS significantly increased FE_NO by 30 ± 7% (p < 0.05) in six mice (Figure 5). There was no effect of intraperitoneal saline injection on FE_NO in sham mice (data not shown). Mice injected with LPS had significantly higher plasma NO_x concentrations than did sham mice 667 ± 71 versus 29 ± 3 µM, respectively, p < 0.01). Animals given LPS also exhibited higher pulmonary iNOS activity (3.0 ± 0.2 versus 0.8 ± 0.2 pmol L-citrulline/min/mg protein, p < 0.01) and lower pulmonary cNOS activity (0.01 ± 0.01 versus 0.8 ± 0.2, p < 0.01) than did sham mice. The lung wet-dry weight ratio, a marker of lung injury, was significantly higher in LPS-injected mice than in sham mice (5.5 ± 0.3 versus 4.2 ± 0.1, p < 0.01).

View larger version (11K): [in this window] [in a new window]   Figure 4.   Effect on intraperitoneal L-NAME injection (2 mg/g body weight) on exhaled NO concentrations (FENO) in individual mice (n = 5) (*p < 0.01).

View larger version (11K): [in this window] [in a new window]   Figure 5.   Effect of intraperitoneal LPS injection (12.5 µg/g body weight) on exhaled NO concentrations (FENO) in individual mice (n = 6) (*p < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This is the first report of noninvasively measured E_NO in a single, spontaneously breathing mouse. Under standardized operating conditions, E_NO was reproducible and stable over 1 wk, was unaffected by food intake or time of day, but decreased after anesthesia. E_NO in mice was reduced by the NOS inhibitor, L-NAME, and was increased in association with increased pulmonary iNOS activity in LPS-induced lung injury.

The measurement of E_NO in humans and animals is very dependent on the conditions under which exhaled gas is collected and analyzed (15, 16). In our study, the measurement of mouse E_NO, especially FE_NO, was highly dependent on the animal chamber flow rate; the fall in FE_NO with increasing flow was consistent with increasing dilution of the mouse's constant VE_NO. Our technique also required the presence of biased flow in order to keep the mouse comfortable and to prevent ambient air contamination of the animal chamber/gas sampling system. Thus, some dilution and underestimation of the "true" FE_NO of a spontaneously breathing mouse is possible. However, an abstract has reported very similar data from the noninvasive sampling of exhaled gas from a group of five wild-type B6129/F1 mice (12). The only other report of E_NO in mice found FE_NO in mechanically ventilated, anesthetized, and tracheotomized wild-type B6/SV129 mice to be 6.3 ± 0.9 ppb, similar to our postanesthesia data (11).

A limitation of our noninvasive sampling technique is the inability to separate upper and lower airway contribution to E_NO in mice. However, the lack of an independent effect of tracheostomy on the E_NO signal postanesthesia indicates that the upper airway contributes minimally to E_NO in spontaneously breathing mice. This is in sharp contrast to the high concentration of NO in the upper airways of humans, which may contribute significantly to E_NO (17). The exact tissue origin of the E_NO signal in spontaneously breathing mice remains uncertain. Noninvasively measured E_NO from the animal chamber could be derived in part from nonrespiratory sources of NOS such as the intestinal mucosa and skin (18, 19). Bacterial skin flora are unlikely to contribute to E_NO in mice, as E_NO was similar in normal and germ-free mice (12). Finally, NOS-independent production of NO occurs in tissues under conditions such as ischemia or acidosis, caused in part by the release of NO from decomposition of endogenously present S-nitrosothiols (20). Nevertheless, given the decrease in E_NO after L-NAME, a significant portion of our mouse E_NO signal is clearly NOS-dependent. Similar effects of NOS inhibitors on E_NO in humans have previously been reported (23).

The contributions of specific NOS isoforms to E_NO in health and disease remain uncertain. Given a reduced signal in nNOS / mice, it has been suggested that nNOS contributes approximately 45% of the E_NO signal in ventilated healthy mice (11). In the present study, the LPS-induced increase in mouse E_NO reflected increased plasma nitrites/nitrates and increased pulmonary iNOS activity. Although increased E_NO and increased iNOS activity were also found in LPS-exposed, mechanically ventilated rats (10), LPS-induced increased E_NO in pigs was due to increased pulmonary cNOS activity, not to increased iNOS (6). Thus, there are significant species and model differences in the relative contributions of the different NOS isoforms to both basal levels and the increased levels of E_NO in the setting of lung disease. Furthermore, E_NO levels in health and disease likely depend also on local conditions, including the pH, redox potential, the presence of NO-binding molecules (e.g., thiol groups), and the presence of reactive oxidant and antioxidant species (24).

In summary, we report the noninvasive measurement of E_NO from a single, spontaneously breathing mouse. This novel system provides a reproducible, stable, and biologically responsive measure of E_NO in mice. Our noninvasive technique permits E_NO measurement in mice with induced pulmonary disease states in the absence of confounding variables such as anesthesia and mechanical ventilation. In addition, repeated E_NO analysis over time in the same animal is possible such as during evolution of an injury. We expect this technique will be of use in further elucidating the cellular and isoform sources of E_NO, as well as the role of endogenous NO in lung disease.