INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bianco and colleagues showed that inhaled furosemide inhibits bronchoconstriction induced in asthmatics by exercise (1) and nebulized water (2). This inhibitory effect has since been confirmed for a range of bronchoconstrictor stimuli (3). Inhaled furosemide has also been shown to counteract experimentally induced dyspnea (6).

The mechanism responsible for these effects of furosemide remains to be established. It has been postulated that inhaled furosemide acts indirectly on sensory receptors in the airway epithelium and its vicinity, and it has been shown to inhibit the discharge of laryngeal irritant receptors (7). However, the actions of inhaled furosemide on receptors in the tracheobronchial tree have not been studied and thus its effects on pulmonary vagal receptors remain elusive. The aim of this study was to elucidate the effects of inhaled furosemide on tracheobronchial receptors by recording the discharge of pulmonary slowly adapting stretch receptors (SARs) and pulmonary rapidly adapting receptors (RARs) in anesthetized rats.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present experiments were performed under the "Guiding Principle for the Care and Use of Animals in the Field of Physiological Sciences" recommended by the Physiological Society of Japan. Experiments were conducted on 20 male Wistar rats (220 to 550 g) anesthetized by intraperitoneal injection of sodium pentobarbital (60 mg/ kg). Supplements of one-third of the initial dose were given as needed by the same route. Adequacy of anesthesia was regularly assessed in all animals by absence of withdrawal reflexes or change in respiratory rate in response to a noxious stimulus. The animals was tracheotomized below the larynx with polyethylene tubing attached to a T-piece. The humidified and hyperoxic gas mixture (fraction of inspired oxygen [FI_O2] = 0.40) was delivered to one side at 1.0 to 1.5 L/min and the expiratory side monitored with a pressure transducer for airway pressure (Paw). Throughout the experiment body temperature was maintained at 37.5 to 38.5° C with a heat lamp. To prevent atelectasis and keep pulmonary compliance constant, the lung was inflated every 30 min to an end-expiratory pressure of 20 cm H₂O by immersion of the expiratory outlet.

The right vagus nerve was exposed in the neck and freed from surrounding tissue, cut in the vicinity of the nodose ganglion, and placed on a parafilm platform, where the distal end was desheathed and soaked in 0.25% collagenase (Sigma, St. Louis, MO) in saline to facilitate dissection. The nerve trunk was then divided into seven to 15 fine filaments, which were laid across a platinum/iridium recording electrode and kept under mineral oil. Action potentials were filtered (0.1 to 3 kHz), amplified, and displayed on an oscilloscope. The output of the oscilloscope was relayed to a time-amplitude window discriminator (MET-1100; Nihon Koden, Tokyo) and an audiometer. The signal from the window discriminator was fed into a pulse counter with a bin width of 0.1 s (MET-1100; Nihon Koden), a thermal array recorder (Omniace RT3200; NEC-Sanei, Tokyo), and an analogue-to-digital converter. Individual action potentials could be distinguished when fewer than three units were active simultaneously. The contralateral vagus was left intact.

To determine its conduction velocity (CV), in three animals stimulating electrodes were placed on the vagus nerve at its exit from the thoracic cage approximately 20 mm from the recording electrode and suprathreshold stimuli (duration, 50 µs) delivered at 1 Hz. CV was calculated as the conduction distance between recording and stimulating electrodes divided by the time between the arrival of the stimulus artifact and action potentials at the recording electrode.

Three types of inflation test were carried out by submerging the expiratory outlet at specific depths under water, after hyperventilation to suppress spontaneous breathing. Although SAR activity was stimulated during hyperventilation, the activity returned to the precontrol level within 3 to 4 breaths. Baseline spike frequency (sf) was evaluated for 0.1 s at the end of spontaneous deflation of the lungs, when Paw reaches zero. The tests were as follows: (1) Sudden increase in Paw to 10 to 20 cm H₂O for at least 5 s to measure adaptation of receptors. An adaptation index (AI) based on that described by Widdicombe (8) was used for all units, and was calculated as the peak sf less the average sf during the second second of inflation to 10 cm H₂O as a percentage of peak sf. (2) A stepped increase of Paw to 2.5, 5, 7.5, and 10 cm H₂O for at least 2 s each. (3) A ramp increase in Paw from 0 to 10 cm H₂O over 6 to 12 s to produce a continuous curve of Paw versus sf. Because there was no significant difference in the Paw versus sf relationship determined by the ramp and stepwise methods, and because the relationship between Paw and sf between 0 and 10 cm H₂O appeared linear, a linear regression technique was used so that sf = S × Paw + sfbase, where S represents sf sensitivity to change in Paw (sf/Paw) and sfbase is the calculated value of sf at Paw = 0.

Using an ultrasonic nebulizer (NE-U12, Omron, Tokyo; 90% of the aerosol particle size ranged between 1 and 8 µm), 60 to 100 mg of furosemide (Hoechst, Frankfurt, Germany; 10 mg/ml, n = 29) or vehicle (Hoechst, 6 to 10 ml, n = 9) was delivered over 5 min via the tracheal cannula. Within 5 to 30 min after inhalation of furosemide or vehicle, the inflation tests were repeated. As a control, intravenous furosemide (Hoechst, 10 mg/kg, n = 5) was given and the Paw versus sf relationship measured 15 min after injection.

All data are presented as means ± SEM. Statistical significance of the data was assessed using the paired t test. Comparisons among groups (furosemide or vehicle inhalation and intravenous furosemide) were assessed using analysis of variance (ANOVA) and post hoc tests (Bonferroni). Differences in mean values of variables were judged to be significant if p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Forty-three SARs (84%, Figure 1A) and eight RARs (16%, Figure 1B) were recorded. A sudden increase in Paw from zero to 10 cm H₂O elicited Breuer-Hering reflex apnea lasting for 1 to 4 s. The mean AI of SARs and RARs were 9.6 ± 1.1% (range, 0 to 25%; n = 42) and 65.7 ± 4.5% (range, 50 to 89%; n = 8), respectively (Figure 1C): hence the two groups of receptor were clearly distinguishable. Application of negative Paw (10 to 15 cm H₂O) inhibited activity in 16 of 17 SARs tested and activated one unit. All RARs tested were activated by negative Paw (n = 5). All SARs (18.5 ± 2.6 m/s, n = 8) and RARs (19 and 10 m/s, n = 2) examined had CV greater than 6 m/s, indicative of myelinated fibers.

View larger version (25K): [in this window] [in a new window]   Figure 1.   Typical responses of (A) SAR and (B) a RAR to maintained inflation of the lungs. From top to bottom: unit activity, sf, and Paw. Firing of the SAR continued during lung inflation whereas that of the RAR decayed steeply. (C) Distribution of AI in tracheobronchial receptors. The distribution could be separated into SARs (AI  25%) and RARs (AI  50%).

sf of SARs fluctuated over the respiratory cycle. Twenty-eight SARs fired exclusively during inspiration (I unit; 65%), 14 receptors during both inspiration and expiration (IE unit; 33%) and one predominantly during the postinspiratory phase (PI unit; 2%).

SAR Activity and sf Responses to Change in Paw

During spontaneous breathing the transitional sf of SARs at zero Paw was 10.3 ± 2.7 impulses/s (range; zero to 60 impulses/s; n = 43: Table 1) and peak sf was 87.7 ± 5.8 impulses/s (range; 20 to 208 impulses/s; n = 42).

In the present study a ramp method was used to evaluate the effects of lung inflation on sf, because there was no significant difference from a stepwise inflation method. In most SAR units sf increased in response to a ramp increase in Paw from zero to 10 cm H₂O (e.g., Figure 2A). In some SARs, sf was strongly activated in the lower range of Paw (zero to 4 cm H₂O) and reached a plateau at higher levels of Paw.

View larger version (23K): [in this window] [in a new window]   Figure 2.   Effects of inhaled furosemide on SAR activity induced by ramp increases in Paw. Traces as in Figure 1. Note the increased sf response 30 min after inhaled furosemide (B) compared with control (A). (C) Group data for the effect of inhaled furosemide on induced SAR activity with ramp inflation of the lungs. Inhaled furosemide shifted the Paw versus sf response curve upward. *p < 0.01.

Effects of Inhaled Furosemide on Baseline Activity of SARs and on the Relationship between Paw and sf

Inhalation of furosemide increased significantly both transitional sf (from 13.2 ± 3.6 to 31.7 ± 5.6 impulses/s, p < 0.01; n = 29) and peak sf (from 83.4 ± 8.1 to 150 ± 16 impulses/s, p < 0.01, n = 29) (Table 1). Inhalation of vehicle or intravenous injection of furosemide did not induce significant changes in transitional or peak sf (p > 0.1, n = 9 and n = 5; Table 1). Micturition was noted to occur more than 60 min after inhalation of furosemide, but within 5 min of injection. Increases in transitional and peak sf were significantly greater with inhaled furosemide than after inhaled vehicle or intravenous furosemide (p < 0.01 or p < 0.05, Table 1). In some units, inhalation of furosemide changed the firing pattern from I to IE.

Inhalation of furosemide resulted in an upward shift of the Paw versus sf relationship (Figure 2C). The sf sensitivity to Paw (S; sf/Paw) increased by 72.1% (from 8.6 ± 0.6 to 14.8 ± 1.4 Hz/cm H₂O, n = 29, p < 0.01; Figure 3, left) accompanied by an increase in the calculated sf at zero Paw (sfbase) from 18.0 ± 4.5 to 49.5 ± 8.1 Hz, n = 29, p < 0.01; Figure 3, right). Inhalation of vehicle decreased sf sensitivity slightly (n = 9, p < 0.01) without change in sfbase, but injection of furosemide did not elicit any appreciable changes in Paw-induced SAR activity. The increase in sf sensitivity with furosemide inhalation was significantly greater than that after vehicle or intravenous furosemide (p < 0.01, Figure 3).

View larger version (23K): [in this window] [in a new window]   Figure 3.   Changes in sensitivity to lung inflation (left) and sf (right) in SARs before (B) and after (A) application of inhaled furosemide [furo (inh)], inhaled vehicle [veh (inh)], and intravenous furosemide [furo (iv)]. S-ratio of sf changes (sf) to changes in Paw (Paw); sfbase-baseline sf before inflation tests. Inhaled furosemide but not vehicle or intravenous furosemide increased sfbase and response to lung inflation.

Effect of Inhaled Furosemide on RAR Activity

In spontaneously breathing animals RARs were inactive except at Paw  5 cm H₂O. Inhalation of furosemide suppressed the peak sf of RARs from 80 ± 10 to 47 ± 5 Hz (n = 6, p < 0.05; Figures 4 and 5) but did not change the AI induced by 10 to 20 cm H₂O of Paw (65.7 ± 4.5% to 67.5 ± 4.2%, n = 6, p > 0.1).

View larger version (26K): [in this window] [in a new window]   Figure 4.   Effects of inhaled furosemide on RAR activity induced by a step increase in Paw (A) before and (B) 30 min after inhalation of furosemide. Traces as in Figure 1. Note the decrease in sf response to lung inflation after furosemide inhalation.

View larger version (35K): [in this window] [in a new window]   Figure 5.   Group data showing spike frequency response of RARs to lung inflation before (shaded column) and after (open column) inhalation of furosemide. Peak sf was significantly smaller in RARs after furosemide. *p < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have demonstrated in the current study that inhalation of furosemide elicits enhanced SAR activity during both spontaneous breathing and lung inflation. In addition, inhaled furosemide suppresses RAR activity.

Identification of Tracheobronchial Receptor Types

There have been discrepancies (9) in the definition of SARs and RARs in rats, depending on experimental conditions such as spontaneous versus artificial ventilation, and closed versus open chest. Criteria for their differentiation have included the AI, nerve fiber CV, responses to inflation or deflation, and responses to pharmacological agents (e.g., intravenous veratridine, intravenous histamine, and inhalation of diethyl ether).

Because in the current study the AI exhibited a distinctly bimodal distribution with a gap between 25% and 50%, we selected this index to differentiate SARs and RARs. We regarded units as SARs for AI  25% and RARs for AI  50%. For RARs these values differ from those originally given for cats by Knowlton and Larrabee (AI  80%) (13) and Widdicombe (AI  70%) (8) and for rats by Bergren and Peterson (AI  80%) (10) and Davies and coworkers (AI = 100%) (11), although the latter used a different definition. Our criterion for SARs was, however, similar to that of Lai and Kou (12) who regarded SARs as having an AI  50% in rats.

"Deflation-sensitive," "collapse," and "irritant" receptors, described in various mammalian species, are probably all identical to RARs (14). Deflation inhibited nearly all SARs tested in our study (16 of 17; 94%) and activated al RARs (5 of 5). As we did not test all SARs with deflation, it is possible that more were "deflation-sensitive." The RARs described in this study resemble the "irritant like" receptors reported by Tsubone (9) with an average AI of 59%. However, "irritant like" receptors may be better defined as deflation sensitive SARs, as advocated by Bergren and Peterson (10). Using this criterion, deflation-sensitive units would be categorized as RARs. We found only one deflation-sensitive unit out of 17 SARs (6%), a proportion smaller than that of Bergren and Peterson (22%) (10) or Tsubone (9). The ratio of RARs to total tracheobronchial receptors was 16% in this study, comparable with that described in rats by Bergren and Peterson (9%) (10) but smaller than that of Davies and coworkers (30%) (11). Differences in criteria for the AI or in experimental conditions might account for these differences. Although there has been controversy regarding the validity of the AI for defining RARs, the SARs identified in this study (AI  25%) correspond well to those described by other investigators (8).

Because there is an overlap in the CVs of SARs and RARs (10, 14), that criterion does not provide a useful distinction between these receptors. Histamine has been used to identify irritant receptors, and it has been reported that intravenous injection of a relatively high dose of histamine affected blood pressure but not RARs (10). Whether or not the receptors defined as RARs in this study can respond to irritant chemicals such as ammonia vapor and cigarette smoke, or hypotonic and/or low chloride solutions was not explored in this study. The further characterization of the RARs would provide sufficient information on stimulation of the receptors.

Effect of Inhaled Furosemide on Tracheobronchial Receptors

It is generally accepted that RARs are located in epithelium, close to the tight junctions and more basally whereas SARs are located deeper within the smooth muscle (14). Thus, it is possible that SARs are influenced less than RARs in response to inhalation of furosemide. However, our finding that not only RARs but also SARs are consistently influenced by inhaled furosemide suggests that there may be no barrier for penetration of furosemide across the epithelium deep into the submucosal site and that furosemide can easily reach the SARs. Direct action of inhaled furosemide on receptors is likely to be responsible for activation of stretch receptor afferents, because intravenous injection of furosemide did not change spontaneous or lung inflation-induced activity of SARs. Assuming that there may be direct action of furosemide on airway sensory nerves, it is likely that effective concentrations of furosemide are not achieved by intravenous route.

The cellular mechanisms responsible for the inhibition of RARs and the activation of SARs in response to inhalation of furosemide are unknown. However, the loop diuretic furosemide has long been considered a specific inhibitor of the Na⁺-K⁺-2Cl cotransporter in the basolateral membrane of tracheobronchial mucosa (15). Blockage of this mechanism may increase both [Na⁺] and [Cl] in the submucosal extracellular fluid within the vicinity of sensory nerve receptors ([Na⁺]ecf and [Cl]ecf, respectively) (15). An increase in [Na⁺] within the vicinity of SARs may enlarge the [Na⁺] gradient between the extracellular and intracellular fluids of the nerve endings, which in turn may increase the action potentials during depolarization of the receptor membrane. Matsumoto and coworkers showed that veratridine (Na⁺ channel opener) inhibits SAR activity in ouabain (Na⁺-K⁺ ATPase inhibitor)- treated animals and suggested that SAR stimulation is related to the [Na⁺] gradient, which is regulated by Na⁺ pump activity (16). Hence an increase in [Na⁺]ecf might activate nerve endings of SARs. On the other hand, an increase in [Cl]ecf caused by furosemide could inhibit RAR activity, because decreased [Cl]ecf is thought to depolarize the nerve endings of RARs (17, 18). Both RARs and SARs are affected by airway smooth muscle tone (19). However, change in smooth muscle tone cannot explain the difference in behaviors between SARs and RARs because the two different types of receptors responded to inhaled furosemide in completely opposite direction. Furthermore, it has been shown in in vitro study that furosemide lacks the effect on airway smooth muscle (20).

Clinical Relevance

Our findings are relevant for certain problems in respiratory medicine. Inhalation of furosemide has been shown to have an inhibitory effect on cough induced by inhalation of low-chloride solution in normal subjects (3) and to prevent bronchoconstriction in patients with asthma (2). In that pulmonary reflexes relax airway smooth muscle through stretch receptors and mediate cough and bronchoconstriction through irritant receptors (1, 2, 4, 5), we hypothesized that activation of stretch receptors and inhibition of irritant receptors might be the mechanism of inhibition of cough and bronchoconstriction produced by inhalation of furosemide. Our finding that inhaled furosemide activates pulmonary stretch receptors and inhibits the activity of pulmonary irritant receptors in anesthetized rats strongly supports this hypothesis. We reported recently that inhaled furosemide alleviates experimentally induced dyspnea (6). Vagal afferent fibers may also play an important role in mediating this sensation (21). Thus, activation of pulmonary stretch receptors and inhibition of vagal irritant receptors seems to be the most plausible explanation for the observed reduction in subjective and objective measures of respiratory embarrassment after inhalation of furosemide.