INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

LUNG DISEASES, SUCH AS ASTHMA (10), chronic obstructive pulmonary disease (12), endotoxin shock (33), and vascular microembolism (50), or inhalation of oxidant irritants, such as toxic smoke (35), cigarette smoke (35), and ozone (45), may cause increased pulmonary production of reactive oxygen species from endogenous and/or exogenous sources. The major reactive oxygen species are the superoxide anion radical, hydrogen peroxide (H₂O₂), and the hydroxyl radical (· OH) (5). The superoxide anion radical dismutates to form H₂O₂, which, in the presence of iron, can further react to form · OH, a more reactive oxygen radical, via the Fenton reaction (5). Vagal sensory receptors are known to play an important role in detecting the onset of pathophysiological conditions and are responsible for triggering defensive or protective airway reflexes (8, 29, 39). While abundant information suggests the importance of reactive oxygen species in producing pulmonary pathophysiological consequences (4), very few studies have investigated their role and mechanism in eliciting airway reflexes.

The concept that reactive oxygen species may stimulate pulmonary sensory receptors and play a vital role in eliciting airway reflexes is considerably new (29). This concept is indirectly supported from findings that vagally mediated airway reflexes evoked by inhaled cigarette smoke (27), inhaled wood smoke (22), or pulmonary air embolism (6) and responses of vagal sensory receptors to the latter two insults (7, 24, 25) are greatly attenuated by pretreatment with · OH scavengers. The evidence is not limited to sensory receptors located in the lower airways, because airway reflexes evoked by wood smoke-induced stimulation of nasal (17) and laryngeal (30) afferents are also inhibited by pretreatment with · OH scavengers. However, questions still remain as to whether the suppressive effects of · OH scavengers in these studies result from the prevention of stimulation of airway sensory receptors by increased · OH and/or from the removal of the basal function of · OH, leading to diminished receptor sensitivity to these airway insults. This concept is directly supported by results obtained from two recent studies (42, 52) in which direct injections of H₂O₂ into lung parenchyma produced increases in the rate and the amplitude of phrenic bursts. These H₂O₂-induced phrenic responses were totally abolished by bilateral vagotomy, suggesting that they are respiratory reflexes mediated through lung vagal afferents (42). Surprisingly, pretreatment with a · OH scavenger failed to inhibit H₂O₂-induced phrenic responses (42). Both the specific types of lung vagal sensory receptors and the oxygen radical mechanism participating in the elicitation of H₂O₂-evoked respiratory reflexes are still unclear.

There are three major types of vagal sensory receptors serving the major afferent system of the airways and lungs: slowly adapting receptors, rapidly adapting receptors (RARs), and C-fiber nerve endings (C fibers) (8, 51). Afferent activity arising from the first two types is conducted by myelinated fibers, whereas activity from C fibers is conveyed by unmyelinated fibers (8, 51). Owing to their high sensitivities to various chemical stimuli, the latter two types of pulmonary receptors have been suggested to play important roles in eliciting airway reflexogenic responses to inhaled irritants or chemical mediators (29, 39). To differentiate the roles of lung C fibers and RARs in evoking airway reflexes, several techniques for producing differential vagal blockages have been employed. For instance, perineural application of capsaicin, a chemical extracted from hot peppers, to both cervical vagi has been shown to selectively block the airway reflexes resulting from stimulation of lung C fibers (23, 26, 32). On the other hand, the cooling of both cervical vagi to 6-7°C has been demonstrated to differentially block airway reflexes originating from stimulation of lung myelinated afferents (23, 28).

The present study was undertaken in anesthetized rats to investigate 1) the dose-response relationship of acute ventilatory responses to inhalation of aerosolized H₂O₂, 2) the roles of lung vagal C fibers and RARs in eliciting these ventilatory responses by using techniques of differential vagal blockade, and 3) the oxygen radical mechanism underlying these ventilatory responses by using various antioxidant pretreatments.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animal preparation. Sprague-Dawley rats of either sex were anesthetized with an intraperitoneal injection of chloralose (100 mg/kg; Sigma Chemical) and urethane (500 mg/kg; Sigma Chemical). A polyethylene catheter was inserted into the jugular vein and advanced until the tip was close to the right atrium for intravenous administration of pharmacological agents. The right femoral artery was cannulated for measuring arterial blood pressure (ABP). During the course of the experiments, supplemental doses of chloralose (20 mg · kg¹ · h¹) and urethane (100 mg · kg¹ · h¹) were administered to maintain abolition of the corneal reflex and pain reflexes induced by pinching the animal's tail. Body temperature was maintained at ~36°C throughout the experiment by means of a servo-controlled heating blanket.

Experimental treatments of vagus nerves. The techniques of perivagal capsaicin treatment (PCT) and vagal cooling (VC) were adopted from those reported previously (26). In brief, a segment (~3 mm) of each cervical vagus nerve was wrapped in a cotton strip presoaked in a capsaicin solution (0.25 mg/ml). After 15-20 min, when the reflex responses to an intravenous injection of capsaicin (1 µg/kg) (see RESULTS) had been abolished, the cotton strips were removed. This dose of capsaicin injection is known to stimulate lung C fibers in rats (16, 25). The blocking effect of PCT on the reflex responses to capsaicin injection has been shown to last for 8-120 min (23). Solutions of capsaicin were made daily from a refrigerated stock solution (250 mg/ml, Sigma Chemical), which was prepared by dissolving capsaicin in a solution containing 0.5% Tween 80 (Sigma Chemical), 0.5% ethanol, and 99% saline. To perform VC, both cervical vagus nerves were placed in the grooves (5 mm long) of a pair of copper radiators. The temperature of the coolant (antifreeze) circulating through the radiators was lowered to 7°C, a temperature that eliminates the reflex apneic response to lung inflation (Ptr = 10 cmH₂O) (see RESULTS). Our preliminary results showed that a higher concentration of capsaicin solution of PCT and a lower temperature of VC were required to block the reflex responses to a higher dose of capsaicin injection and to a larger volume of lung inflation. The experimental conditions of PCT and VC used in this study were chosen to avoid nonspecific blocking effects on the reflex responses mediated via vagal afferents (26) and to avoid possible damaging effects on these afferents. To perform sham treatment of the vagus nerves, a segment (~3 mm) of each cervical vagus nerve was wrapped in a cotton strip that was presoaked in the vehicle of the capsaicin solution for 20 min. Furthermore, during the test with H₂O₂ challenge, both cervical vagus nerves were placed in the grooves of the copper radiators with circulating coolant maintained at 36°C. To perform vagotomy, both vagus nerves were sectioned at the cervical level.

Generation and delivery of aerosolized H₂O₂. Various concentrations (0.1, 0.2, and 0.4%) of a H₂O₂ solution were prepared just before each generation by mixing 35% H₂O₂ (Shimakyu, Osaka, Japan) with PBS to the desired concentration, with the pH adjusted to 7.4. H₂O₂ aerosol was generated by delivering an airstream (15 ml/s) through the cup of an active ultrasonic nebulizer (ULTRA-NEB 99, DeVilbiss) containing the H₂O₂ solution. The particle sizes of the aerosol generated by this nebulizer ranged from 0.5 to 5 µm. Airway exposure to aerosolized H₂O₂ was achieved by turning a three-way stopcock for a 30-s period; this three-way stopcock communicates with the distal end of the pneumotachograph and a side arm of the tubing (8 mm ID) attached to the outlet of the nebulizer. Through this connection, H₂O₂ aerosol was spontaneously inhaled into the lower airways and lungs. With the use of a dye tracer, the time lag between the onset of challenge and the arrival of the aerosolized tracer in the airways was estimated to be <1 s. This estimation was based on postmortem checks of the presence of the dye tracer in the airways in 10 animals whose tracheal tubes were quickly disconnected from the circuit delivering aerosolized dye tracer 1 s after the onset of challenge.

Administration of antioxidant. Catalase (Cat; Sigma Chemical) was dissolved in PBS to a concentration of 750,000 IU/ml. Heat-inactivated Cat (HiCat) was prepared by heating the Cat solution to 100°C for 15 min. Aerosol of Cat or HiCat was generated by using an ultrasonic nebulizer (ULTRA-NEB 99) and was spontaneously inhaled into the lower airways for a period of 5 min by using a circuit similar to that for delivery of H₂O₂ aerosol. Both dimethylthiourea (DMTU; Sigma Chemical) and deferoxamine (Def; Sigma Chemical) were dissolved in saline. Iron-saturated Def (Def+Fe) was prepared by adding 98 mg of FeCl₃6H₂O (Fluka) to 1 ml Def (250 mg/ml) for 1 h at room temperature, as described previously (43). DMTU (1.5 g/kg), saline, Def (15 mg/kg), or Def+Fe was slowly injected into the vein for 1 min. Cat is an enzyme that catalyzes the breakdown of H₂O₂ into O₂ and H₂O (5). Def is an iron chelator that prevents the formation of · OH derived from H₂O₂ via the Fenton reaction (15), whereas DMTU is an · OH scavenger (13). The doses of Cat, DMTU, and Def were determined by preliminary results from a pilot study. In that pilot study, the doses of the latter two drugs were found to attenuate the H₂O₂-evoked airway reflexes, and doubling the doses did not improve their suppressive effects on these airway reflexes. The doses of DMTU and Def were within the range reported previously in studies of · OH-related respiratory responses (11, 22).

Measurements of wet-to-dry weight ratio of pulmonary tissues. Because H₂O₂ has been reported to induce pulmonary edema (34), the wet-to-dry weight ratio of the airway and lung tissues was assessed. For this purpose, animals were killed by an overdose of anesthetics, and the airways and lungs were excised and separated into left and right portions. The right portion of the tissues was bled and placed in an oven (80°C, 48 h) to dry. The wet-to-dry weight ratios of tissues were then determined.

Experimental procedures. In this study, 128 rats (weight 392 ± 8 g) were randomly and evenly divided into 16 groups of animals. In groups 1-4, ventilatory responses to the challenge of 0.1, 0.2, or 0.4% H₂O₂ or PBS were studied. In groups 5-8, ventilatory responses to the challenge of 0.2% H₂O₂ were studied before and 20 min after sham nerve treatment, PCT, or vagotomy or before and during VC. In groups 9-14, ventilatory responses to the challenge of 0.2% H₂O₂ were studied before and 10 min after pretreatment with Cat or HiCat, or before and 30 min after pretreatment with DMTU, saline (the vehicle of DMTU), Def, or Def+Fe. In groups 15 and 16, animals were subjected to two challenges of 0.2% H₂O₂ or PBS separated by 60 min. Immediately after the second challenge, the animals were killed, and airways and lungs were excised to measure the wet-to-dry weight ratio. To determine whether functions of unmyelinated and myelinated fibers were affected, the reflex apneic responses induced by an intravenous injection of capsaicin (1 µg/kg) and by inflating the lungs to a value of Ptr of 10 cmH₂O were also compared before and after various treatments of vagus nerves or antioxidants. Before each test of the H₂O₂ challenge, the animal's lungs were hyperinflated (Ptr > 25 cmH₂O) to establish a constant volume history. Based on the results of our preliminary study, at least 60 min were allowed to elapse between the two H₂O₂ challenges to avoid possible tachyphylaxis.

Data analysis and statistics. Respiratory frequency (f), VT, mean ABP, and heart rate were measured at 3-s intervals. An interval of at least 2 min before and 3 min after the H₂O₂ challenge was measured. Baseline data were calculated as the mean over the reading 2 min immediately before the H₂O₂ or PBS challenge. Because we found that changes in f induced by the challenge exhibited a biphasic pattern, responses occurring during 0-21 s and during 21-90 s after the onset of challenge were defined as the initial and delayed responses, respectively. The peak response was defined as the value averaged over 3 s after the H₂O₂ or PBS challenge. These physiological parameters were analyzed by using a computer equipped with an analog-to-digital converter (DASA 4600, Gould) and software (BioCybernatics, 1.0; Taipei, Taiwan). Results obtained from the computer analysis were routinely checked with those obtained by manual calculations for accuracy. Data of cardiopulmonary parameters and wet-to-dry weight ratios were compared by using the paired t-test, two-way repeated-measures ANOVA, or two-way mixed-factorial ANOVA followed by Fishers least significant difference procedure, when appropriate. Data of wet-to-dry weight ratios were compared by using the unpaired t-test. A value of P < 0.05 was considered significant. All data are presented as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Ventilatory responses to H₂O₂. Inhalation of 0.2% aerosolized H₂O₂ consistently evoked bradypnea followed by tachypnea (Fig. 1A) in all eight rats studied. The initial decrease in f began at 7.4 ± 0.7 (range of 5-10) s after onset of the challenge and lasted for 13.9 ± 2.4 (range of 9-18) s (Fig. 2A). Subsequently, f quickly increased and reached its peak at 35.6 ± 1.9 (range of 30-45) s after onset of the challenge (Fig. 2A). This delayed tachypnea lasted for 91.5 ± 18.6 (range of 30-174) s. As a group, a general reduction in VT was found during both the initial bradypneic and delayed tachypneic periods (Fig. 2B), although some breaths had a slight increase in VT. At the onset of or during the delayed tachypneic period, augmented inspiration (3) was also evoked in five of the eight rats studied (Fig. 3A). The augmented inspiration is characterized by a two-step inspiratory flow, resulting in an exceedingly large VT (3): the first step resembles the inspiratory flow profile of the preceding breaths, but the second step took place near the end of the inspiration. On average, both the lowest and peak f values during the initial and delayed periods, respectively, significantly differed from the baseline f (Fig. 2C). Additionally, average VT values during these two periods were significantly smaller than the baseline VT (Fig. 2D). In contrast to these effects, inhalation of aerosolized PBS caused no significant changes in breathing patterns (Fig. 2).

View larger version (61K): [in this window] [in a new window]   Fig. 1.   Experimental records illustrating acute responses to inhalation of 0.2% aerosolized H2O2 in an anesthetized and spontaneously breathing rat. A: control responses. B: responses after perivagal capsaicin treatment (PCT). R, respiratory flow; VT, tidal volume; ABP, arterial blood pressure. Horizontal bars indicate the duration (30 s) of H2O2 challenge. Between the two H2O2 challenges, 60 min were allowed to elapse. Note that the H2O2 challenge evoked bradypnea followed by tachypnea during the control and that the former response was prevented after PCT.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Mean acute ventilatory responses to inhalation of aerosolized PBS, and 0.1, 0.2, or 0.4% aerosolized H2O2 in 4 groups of rats. A and B: periods between the 2 vertical dashed lines indicate the duration of aerosol challenge for respiratory frequency and VT, respectively. Data are values averaged over a 3-s period. C: data of respiratory frequency in the initial and delayed responses were measured as the lowest values during 0-21 s and peak values during 21-90 s, respectively, postchallenge. D: data of VT in the initial and delayed responses were measured as values of the same breaths measured for responses of respiratory frequency. Data in each group are means ± SE from 8 rats. * Significantly different from the corresponding baseline in the same group; a significantly different from the corresponding response to PBS; b significantly different from the corresponding response to 0.1% H2O2; c significantly different from the corresponding response to 0.2% H2O2: P < 0.05.

View larger version (49K): [in this window] [in a new window]   Fig. 3.   Experimental records illustrating acute responses to inhalation of 0.2% aerosolized H2O2 in an anesthetized and spontaneously breathing rat. A: control responses. B: responses during vagal cooling (VC). Horizontal bars indicate the duration (30 s) of the H2O2 challenge. Note that the H2O2 challenge evoked bradypnea followed by tachypnea and augmented inspiration during the control and that the latter 2 responses were prevented during VC. See legend to Fig. 1 for further explanations.

Effects of nerve treatments on ventilatory responses to H₂O₂. Neither PCT nor sham nerve treatment significantly altered baseline f and VT values, whereas both VC and vagotomy significantly decreased baseline f and increased baseline VT values (Figs. 1, 3, and 4). The reflex apnea evoked by the intravenous capsaicin injection was completely abolished by PCT or vagotomy, but was not significantly affected by VC or sham nerve treatment (Table 1). Intravenous capsaicin injection also produced bradycardia [change in () heart rate = 100.4 ± 18.8 beats/min] and hypotension (ABP = 22.5 ± 2.1 mmHg), both of which were nearly abolished by PCT (heart rate = 19.5 ± 4.4 beats/min and ABP = 5.9 ± 0.9 mmHg after PCT; n = 8). However, these cardiovascular responses were unaffected by sham nerve treatment (heart rate = 106.1 ± 16.4 beats/min and ABP = 20.3 ± 2.4 mmHg at control; heart rate = 107.3 ± 8.8 beats/min and ABP = 21.4 ± 1.9 mmHg after treatment; n = 8). The reflex apnea induced by lung inflation was totally eliminated by VC or vagotomy but was not significantly affected by PCT or sham nerve treatment (Table 1).

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Mean acute ventilatory responses to inhalation of 0.2% aerosolized H2O2 before and after various nerve treatments in 4 groups of rats: PCT (A), VC (B), bilateral vagotomy (VG; C), and sham nerve treatment (SNT; D). Data in each group are means ± SE from 8 rats. * Significantly different from the corresponding baseline; a significantly different from the corresponding response before nerve treatment: P < 0.05. See legend to Fig. 2 for further explanations.

Effects of antioxidants on ventilatory responses to H₂O₂. Pretreatment with Cat, HiCat, DMTU, the saline vehicle, Def, or Def+Fe significantly altered neither baseline f nor VT values (Figs. 5-7), nor did it significantly affect the reflex apnea evoked by intravenous capsaicin injection or by lung inflation (Table 1). After pretreatment with Cat, both the initial bradypneic and the delayed tachypneic responses to H₂O₂ were entirely prevented (Figs. 5 and 6A). On average, the initial lowest f value, the delayed peak f value, and the accompanying VT value for the two periods did not significantly differ from their respective baseline values (Fig. 6A). After pretreatment with DMTU or Def, the initial bradypneic response to H₂O₂ was completely abolished, and the delayed tachypneic response was partly attenuated. On average, neither the initial lowest f value nor the accompanying value significantly differed from their respective baseline values (Fig. 7, A and C). Furthermore, the delayed peak f value and the accompanying VT value were significantly smaller and greater, respectively, than the control responses (Fig. 7, A and C). In sharp contrast, pretreatment with HiCat, the saline vehicle, or Def+Fe failed to affect either the initial bradypneic or the delayed tachypneic responses to H₂O₂. On average, the initial lowest f value, the delayed peak f value, and the accompanying VT value of the two periods did not significantly differ from their corresponding control responses (Figs. 6B and 7, B and D).

View larger version (51K): [in this window] [in a new window]   Fig. 5.   Experimental records illustrating acute responses to inhalation of 0.2% aerosolized H2O2 in an anesthetized and spontaneously breathing rat. A: control responses. B: responses after pretreatment with catalase (Cat). Horizontal bars indicate the duration (30 s) of the H2O2 challenge. Note that the H2O2 challenge evoked bradypnea followed by tachypnea and augmented inspiration during the control and that all responses were prevented after pretreatment. See legend to Fig. 1 for further explanations.

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Mean acute ventilatory responses to inhalation of 0.2% aerosolized H2O2 before and after pharmacological pretreatment in 2 groups of rats: pretreatment with Cat (A) or heat-inactivated Cat (HiCat; B). Data in each group are means ± SE from 8 rats. * Significantly different from the corresponding baseline; a significantly different from the corresponding response before Cat or HiCat: P < 0.05. See legend to Fig. 2 for further explanations.

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Mean acute ventilatory responses to inhalation of 0.2% aerosolized H2O2 before and after pharmacological pretreatment in 4 groups of rats: pretreatment with dimethylthiourea (DMTU; A), saline vehicle of DMTU (Veh; B), deferoxamine (Def; C), or iron-saturated Def (Def+Fe; D). Data in each group are means ± SE from 8 rats. * Significantly different from the corresponding baseline; a significantly different from the corresponding response before pretreatment: P < 0.05. See legend to Fig. 2 for further explanations.

Responses of ABP to H₂O₂. Inhalation of 0.2% aerosolized H₂O₂ generally produced an increase in ABP during the initial period and a decrease in ABP during the delayed period (Figs. 1 and 3). As a group, the value of ABP changed from a baseline of 103.7 ± 1.3 mmHg (n = 88) to an initial peak value of 113.6 ± 1.4 mmHg and to a delayed lowest value of 92.5 ± 1.5 mmHg under control conditions. The initial increase in ABP evoked by H₂O₂ challenge persisted after either PCT, sham nerve treatment, or pretreatment with HiCat, DMTU, the saline vehicle, Def, or Def+Fe and was prevented either during VC, after vagotomy, or after pretreatment with Cat (Table 2). The delayed decrease in ABP evoked by the H₂O₂ challenge persisted either during VC, after sham nerve treatment, or after pretreatment with HiCat, the saline vehicle, or Def+Fe and was prevented after PCT, vagotomy, or pretreatment with Cat, DMTU, or Def (Table 2).

Wet-to-dry weight ratio of pulmonary tissues. In the group of rats receiving two challenges with 0.2% aerosolized H₂O₂, the average wet-to-dry weight ratio of the airway and lung tissues was 4.84 ± 0.03 (n = 8), which did not significantly differ from that (4.81 ± 0.03; n = 8) measured in the group of rats receiving two challenges with aerosolized PBS.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Results of this study demonstrate that inhalation of 0.2% aerosolized H₂O₂ consistently evoked initial bradypnea followed by delayed tachypnea. Furthermore, during the tachypneic period, augmented inspiration was also evoked in the majority of rats (64.8%) studied. These ventilatory responses were due to the effects of H₂O₂, because inhalation of aerosolized PBS did not alter the breathing patterns, and they were airway reflexes mediated through lung vagal afferents, because bilateral cervical vagotomy totally prevented these responses. Because the overall ventilatory responses to H₂O₂ challenge were complex, we believe that more than one neural mechanism may be involved in eliciting these responses. We further demonstrate that the initial bradypnea was totally abolished after PCT but still persisted during VC. In contrast, the delayed tachypnea and delayed augmented inspiration were completely eliminated by VC but were preserved after PCT. It is well established that PCT and VC differentially block the conduction of unmyelinated C and myelinated fibers, respectively (2, 14). Indeed, we also observed that PCT abolished the reflex apnea resulting from stimulation of lung C fibers by capsaicin (8, 29) but did not affect the lung inflation-induced reflex apnea mediated through vagal myelinated afferents (29). On the other hand, VC completely eliminated the reflex apnea in response to lung inflation but did not affect the reflex apnea in response to capsaicin. In contrast to these effects of PCT or VC, sham nerve treatment failed to influence ventilatory responses to H₂O₂ challenge. Taken together, these results suggest that the initial bradypnea is mediated through vagal unmyelinated afferents, whereas the delayed tachypnea and augmented inspiration are mediated through vagal myelinated afferents.

Although activities of both RARs and slowly adapting receptors are conducted by myelinated fibers, it is known that the former type is more sensitive to chemical stimuli, whereas the latter type is relatively insensitive (8, 51). Furthermore, it is well established that lung C fibers and RARs play important roles in evoking respiratory reflexogenic responses to chemical irritants and that their stimulation produces inhibitory and excitatory effects, respectively, on breathing (8, 29, 39). Hence, it is very plausible that the initial bradypnea may result from stimulation of lung C fibers, whereas the delayed tachypnea and augmented inspiration are due to activation of RARs. In fact, preliminary results from our electrophysiological study in anesthetized rats revealed that delivery of aerosolized H₂O₂ by a respirator stimulated lung C fibers and RARs, whereas it had little effect on pulmonary slowly adapting receptors (37).

When the initial bradypnea evoked by 0.2% H₂O₂ challenge was prevented by PCT, the delayed tachypnea emerged earlier. Conversely, when the delayed tachypnea was prevented by VC, the initial bradypnea was prolonged. Furthermore, five of eight rats displayed H₂O₂-evoked delayed augmented inspiration before PCT, but all eight rats exhibited this response after PCT. Thus it appears that 0.2% H₂O₂ challenge evoked both C-fiber-mediated inhibitory (bradypnea) and RAR-mediated excitatory (tachypnea and augmented inspiration) ventilatory responses within the same time period in each intact animal, resulting in functional antagonism between the inhibitory and excitatory responses. As a result, when one type of response was blocked by experimental intervention, the other was unrestrained or unmasked. The functional antagonism between the two reflex responses has previously been reported in the study of other inhaled chemical irritants. For example, inhaled acrolein evokes a dominant bradypnea in intact rats, which overrides the submissive augmented inspiration; the latter response is revealed after PCT (26). Inhaled wood smoke triggers bradypnea in some intact rats and augmented inspiration in others; opposite displayed responses occur during VC or after PCT (23). The functional antagonism could also be observed when the concentration of H₂O₂ used for the challenge was altered in this study. Lowering the concentration to 0.1% produced a reduced delayed tachypnea, which allowed the initial bradypnea to be prolonged. On the other hand, increasing the concentration to 0.4% triggered an amplified delayed tachypnea, which overrode the initial bradypnea. The reduced tachypnea to 0.1% and amplified tachypnea to 0.4% H₂O₂ might imply that lung C fibers are more sensitive to the stimulus than RARs. Thus it seems that animals are more likely to display inhibitory and excitatory reflex responses with challenge using lower and higher concentrations of H₂O₂, respectively, and the dominance of the displayed responses depends on the concentration of the H₂O₂ challenge. It is assumed that the preference of the central controller may contribute to differences in this dominance of displayed responses at low and high concentrations. Consequently, a simple, displayed response does not mean that only a simple neural mechanism is involved. Yu and co-workers (42, 52) demonstrated that direct injections of one concentration of H₂O₂ into lung parenchyma of rabbits produced increases in the rate and amplitude of phrenic bursts. These phrenic responses were completely abolished by bilateral vagotomy, suggesting that they are mediated through lung vagal afferents. The increase in the burst rate is similar to tachypnea, yet the increase in the burst amplitude is contradictory to the reduction in VT observed in this study. Differences in the method of H₂O₂ challenge and animal species may also account for this discrepancy. The different pattern of response reported by Yu and co-workers may also be explained by the entrainment of their phrenic discharge to the artificial ventilation, whereas the rats in the present study were spontaneously breathing.

The average VT values were generally reduced during both the initial bradypneic and delayed tachypneic periods after challenge with 0.2% H₂O₂. This reduction in VT was not seen either after PCT, during VC, or after vagotomy, suggesting that it is also mediated through a vagal mechanism. This reduction in VT may possibly be related to the inhibitory effect of lung C fibers on VT when they are activated (8, 29) and the decrease in inspiratory time during tachypnea. Whatever the causes, blocking the function of either unmyelinated or myelinated vagal afferents virtually prevented this reduction in VT. In fact, inhaling aerosolized H₂O₂ conversely evoked an increase in VT under these three experimental conditions. The H₂O₂-induced increase in VT observed in vagotomized animals indicates that it was mediated through a nonvagal mechanism. This nonvagal response probably was overridden by the dominant inhibitory vagal influence in intact animals and could be revealed only after the function of either C-fiber afferents or myelinated afferents had been blocked. Again, these results demonstrate that another functional antagonism exists between an inhibitory vagal and an excitatory nonvagal influence on VT after H₂O₂ challenge. The nonvagal mechanism responsible for this increase in VT is not clear at present, but lung sympathetic afferents may be a possible origin. Afferent activity influenced by various respiratory maneuvers has been recorded from high-thoracic white rami communications (20). A functional role of lung sympathetic afferents in eliciting respiratory responses to stimuli has been suggested previously (9). Alternatively, central effects are also possible.

That H₂O₂ can activate visceral vagal or sympathetic afferents and trigger reflexes is not unique to the airways and lungs. Topical application of H₂O₂ to the gastrointestinal tract in cats (44), to the heart in cats (18), and to the heart in rats (40, 41, 47, 48) has been shown to stimulate abdominal sympathetic afferents, cardiac sympathetic afferents, and cardiac vagal afferents, respectively. Stimulation of the former two types of sympathetic afferents by H₂O₂ leads to a pressor reflex, whereas activation of the latter vagal afferents by H₂O₂ results in a depressor reflex (19, 44). Similar to the airway reflexes, there is functional antagonism between the pressor and depressor reflexes after local application of H₂O₂ to the cat heart (19).

Although the evidence above seems to support our hypothesis, the exact mechanisms by which H₂O₂ stimulates lung C fibers and RARs and triggers the resultant airway reflexive responses are still not fully understood. One plausible mechanism is that these two types of pulmonary receptors are chemically activated by H₂O₂ and/or H₂O₂-related stimuli. Direct application of H₂O₂ has been shown to depolarize nerve terminals isolated from guinea pig brain (46) and myenteric neurons in the guinea pig distal colon (49). In this study, pretreatment with Cat totally abolished all airway reflexive responses to H₂O₂ challenge, whereas pretreatment with HiCat failed to do so, suggesting that these reflexive responses are consequences resulting from specific actions of H₂O₂. However, the latency of ventilatory responses to inhalation of H₂O₂ aerosol observed in this study was much longer than those (within 1 or 2 s) to inhalation of other chemical irritants, such as cigarette (27, 28) and wood smoke (22, 23), in the same rat model; the effects of these two types of smoke are thought to be mediated through an oxygen radical mechanism. The long latency caused us to doubt whether lung C fibers and RARs are chemically stimulated by H₂O₂ itself. In this regard, we found that pretreatment with DMTU or Def prevented the initial bradypneic response, significantly attenuated the delayed tachypneic response, and largely reduced the occurrence of augmented inspiration evoked by H₂O_2. In contrast, pretreatment with HiCat, saline (the DMTU vehicle), or Def+Fe failed to produce these suppressive effects. The suppressive effects of Cat, DMTU, and Def on airway reflexive responses were unlikely due to anesthetic or deleterious influence on lung C fibers and RARs, because, after these pretreatments, the reflex apnea could still be evoked in response to capsaicin injection or lung inflation. It is conceivable that Def inhibits the formation of · OH derived from H₂O₂ by chelating the catalyzing iron (15), whereas DMTU scavenges · OH after their formation (13). Consequently, these airway reflexive responses were prevented or attenuated by lowering the · OH burden. These observations suggest that the reflex effects of H₂O₂ are, in part, mediated through the action of · OH. This notion is supported by the preliminary result obtained from a followup study of ours, in which pretreatment with DMTU greatly attenuated the afferent responses of lung C fibers and RARs to H₂O₂ challenge in rats (21). It appears that the C-fiber-mediated inhibitory response was more vulnerable to blockades by DMTU or Def compared with the RAR-mediated excitatory ventilatory responses. It is possible that a radical mechanism other than · OH may participate in triggering the RAR-mediated excitatory ventilatory responses. Alternatively, it is possible that differences in the receptor sensitivity to · OH and/or the accessibility of antioxidants to the receptor sites may contribute to this dissimilar vulnerability. Several studies have shown that · OH is the key reactive oxygen species involved in the stimulation of abdominal sympathetic afferents, cardiac sympathetic afferents, and cardiac vagal afferents by H₂O₂ (18, 40, 44, 48). Our laboratory previously reported that · OH scavengers attenuated airway reflexive responses or vagal afferent responses to inhaled wood smoke (22, 24, 25) or pulmonary air embolism (6, 7). These observations lead to the notion that reactive oxygen species may stimulate pulmonary sensory receptors and evoke airway reflexes. In these studies, we, however, could not exclude the possibility that the suppressive effects of · OH scavengers may result from the removal of the basal function of · OH, leading to diminished receptor sensitivity to these airway insults. This notion now gains strong support from the present findings that direct challenges of H₂O₂ trigger vagally mediated airway reflexes, which can be inhibited by antioxidants for H₂O₂ and · OH.

On the other hand, lung C fibers and RARs may also be chemically stimulated by chemical mediators released in response to H₂O₂ challenge. H₂O₂ or its derived oxygen radicals may be involved in the release of other chemical mediators, such as prostaglandin (38) and histamine (31), which stimulate lung C fibers and RARs (8). In this connection, our laboratory previously reported that both · OH and cyclooxygenase metabolites participate in the stimulation of lung C fibers and RARs by inhaled wood smoke (24, 25) and in the activation of lung C fibers by pulmonary air embolism (7), and that these two contributory factors are interrelated.

The other plausible mechanism underlying the elicitation of the observed airway reflexive responses is that lung C fibers and RARs are mechanically stimulated after H₂O₂ challenge. Lung C fibers are relatively insensitive to mechanical stimuli (8, 29). However, both lung C fibers and RARs have been shown to be stimulated by pulmonary edema (36), a pathophysiological consequence that may possibly be induced by H₂O₂ (34). In this study, 0.2% H₂O₂ challenges caused no pulmonary edema. Thus this mechanism is questionable, at least in the case of responses to 0.2% H₂O₂ challenge. Unlike lung C fibers, RARs can be mechanically activated by several factors, including bronchoconstriction and changes in lung mechanics (8, 39, 51), results that may also be induced by H₂O₂ (1).

In addition to the ventilatory responses, inhalation of 0.2% aerosolized H₂O₂ initially produced an increase in ABP and subsequently caused a decrease in ABP. These changes in ABP appear to be mediated through vagal mechanisms and specific actions of H₂O₂, because they were totally abolished by vagotomy and pretreatment with Cat, respectively. The fact that the initial increase in ABP persisted after PCT but was prevented during VC suggests that it may be a reflex resulting from stimulating RARs. Conversely, the delayed decrease in ABP persisted during VC, but was prevented after PCT, indicating that it may be a reflex resulting from activation of lung C fibers. These notions also gain support from the observations that, when the C-fiber-mediated inhibitory ventilatory response to H₂O₂ was totally prevented by DMTU or Def, the delayed decrease in ABP vanished. Conversely, whereas the RAR-mediated excitatory ventilatory responses to H₂O₂ were only partially reduced by DMTU or Def, the initial increase in ABP persisted. It is well documented that stimulation of lung C fibers results in hypotension (8, 29). The cardiovascular reflexes originating from activation of RARs, however, have not been well defined (51). It is interesting to note that the reflex effect of RARs preceded the reflex effect of lung C fibers for ABP responses, whereas the reverse sequence was true for airway responses. Evidently, central processing of lung C-fiber and RAR afferent information in evoking airway reflexes greatly differs from that in evoking cardiovascular reflexes.

In summary, airway insult by H₂O₂, a reactive oxygen species, has profound reflex effects on breathing, which are, in part, mediated through the action of · OH. The H₂O₂-evoked initial bradypnea may be a reflex resulting from stimulation of lung C fibers, whereas the H₂O₂-evoked delayed tachypnea and delayed augmented inspiration probably result from excitation of RARs. In each intact animal, H₂O₂ concomitantly stimulated these two types of pulmonary receptors, resulting in functional antagonism between C-fiber-mediated inhibitory and RAR-mediated excitatory airway reflexes. The concentration of H₂O₂ is known to be elevated in exhaled air condensate of patients in several inflammatory disorders of the lungs, including asthma and chronic obstructive pulmonary disease (10, 12). Excitation of both lung C fibers and RARs has been postulated to cause other airway reflexes, including coughing, changes in airway smooth muscle tone, and an increase in mucus secretion (8, 29, 39). The respective roles of lung vagal afferents in these airway reflexes induced by oxygen reactive species are largely unknown and thus require further investigation.