INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

WHEN INHALED AS AN AEROSOL, furosemide offers protection in asthmatic subjects against several indirect bronchoconstrictor stimuli, including exercise (3), ultrasonically nebulized distilled water (20), cold air hyperventilation (11), sodium metabisulfite (16), adenosine (17), and both the early and the late responses to allergens (2). Recently, Tanigaki and colleagues (25) reported that inhaled furosemide rapidly reversed obstruction in patients with severe, acute asthma who were unresponsive to conventional therapy. Because aerosol administration of this agent is expected to produce few side effects, furosemide has promising clinical application in the treatment of this potentially fatal disease.

The antiasthmatic mechanism of furosemide is not well understood, but it is clearly unrelated to the diuretic actions of this drug (11, 17). As the result of an intra-airway thermodynamic analysis of exercise-induced asthma, Gilbert et al. (9) suggested that the antiasthmatic property of furosemide is related to dilation of airway vasculature, which should increase blood flow to the tissue and reduce the thermal gradient for heat exchange across the airway wall. This notion is supported by observations that furosemide dilates rat tracheal arterioles and venules in vivo (5). Inhaled furosemide has no effect against bronchoconstriction induced by histamine (8, 17), methacholine (11), or PGF₂ (24), indirectly suggesting that this agent does not dilate airway smooth muscle. However, in vitro studies produce varying results, with some suggesting that furosemide dilates airway smooth muscle (23) but others reporting that it does not (7, 12).

In the present study, we hypothesized that furosemide dilates airway vasculature but not airway smooth muscle. To test this theory, we isolated bronchial arteries and bronchial smooth muscle from domestic pigs, a species having lung anatomy and physiology closely that resemble that of humans, and measured the in vitro contractile responses of these tissues to furosemide.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Airway Excision

Bronchial Rings

Preparation. Distal bronchi (~2-4 mm OD, 4-9 mm in length) were dissected from the surrounding lung parenchyma. Bronchial rings were suspended in 10-ml organ baths containing warm (37°C) Krebs buffer that was gassed with 95% O₂-5% CO₂, and the rings were then attached to isometric force transducers (Radnoti Glass Technology). Transducer output was recorded on a physiograph (Grass Instruments, Quincy, MA). The bronchial rings were set to a preload of 1.5 g and allowed to equilibrate for at least 60 min before the protocol was begun.

Experimental protocol. Bronchial rings were preconstricted with 10⁵ M acetylcholine, washed, and allowed to stabilize for 30 min. This concentration of acetylcholine produces 20-25% of the maximum constrictor response in porcine bronchial smooth muscle (unpublished observations). This priming sequence was repeated (typically 1-2 times) until sequential tension responses were within 10% of one another. When the priming sequence was completed, 10⁵ M acetylcholine was added, and either furosemide (from 10⁸ to 10³ M) or isoproterenol (from 10⁸ to 3 × 10⁶ M), a nonselective -adrenergic-receptor agonist, was added to the bath in sequential, increasing concentrations. The furosemide vehicle was saline at 10⁸ to 3 × 10⁶ M and dimethylformamide at 10⁵ to 10³ M. The isoproterenol vehicle was saline at all concentrations. Control tissues received only the vehicle.

Isolated Vessels

Preparation. After the airway excision, the tissues were transferred to a refrigerated dissection dish filled with cold (10°C) Krebs-Ringer bicarbonate buffer. Segments of bronchial artery (~100-300 µm in diameter, 2-3 mm in length) were then dissected from the lobar bronchi. The vessels were mounted as rings in a dual-wire myograph (Myograph Systems 3, J.P. Trading, Aarhus, Denmark) and bathed in warm (37°C) Krebs buffer that was continually gassed with 95% O₂-5% CO₂.

Experimental protocol. Vessel rings were preconstricted with either 10⁴ M norepinephrine (NE) or 10⁸ M vasopressin (VP), washed, and allowed to stabilize for 30 min. This priming cycle was repeated (typically 1-2 times) until the tension response was consistent. After the final priming sequence, the preconstrictor was added, and a dose-response relationship with furosemide was determined.

4

8

8

3

8

6

5

3

Data Analysis

All data are expressed as means ± SE of three to six observations. Number of observations (n) indicate the number of tissues, each from a different animal, used in the study. A one-way ANOVA was used to compare treatment groups at all concentrations of furosemide, and P < 0.05 was accepted as the level of statistical significance.

Solutions and Drugs

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of Furosemide on Bronchial Smooth Muscle

8

3

5

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Effect of furosemide on acetylcholine-preconstricted bronchial smooth muscle. Values are means ± SE. Responses to furosemide are normalized to the initial tension induced by 105 M acetylcholine. , Effect of furosemide (n = 3); , effect of the vehicle (n = 3). For concentrations of furosemide <105 M, the vehicle was saline. For 105 M and higher concentrations, N,N-dimethylformamide was the vehicle. No relaxant effect was observed with furosemide at any dose, and a small increase in tension occurred in both control and furosemide-treated tissues due to the vehicle.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Effect of isoproterenol on acetylcholine-preconstricted bronchial smooth muscle. Values are means ± SE. Responses to isoproterenol are normalized to the initial tension induced by 105 M acetylcholine. , Effect of isoproterenol (n = 5);  effect of the vehicle (saline; n = 5). Isoproterenol significantly decreased tension in bronchi, demonstrating that these tissues are very responsive to a bronchodilator. * Significant difference from control responses, P < 0.05.

Effect of Furosemide on Bronchial Artery Smooth Muscle

4

4

3

3

3

8

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of furosemide on norepinephrine-preconstricted bronchial arteries. Values are means ± SE. Responses to furosemide are normalized to the initial tension induced by 104 M norepinephrine. , Effect of furosemide (n = 5); , effect of the vehicle (n = 5). For concentrations of furosemide <105 M, the vehicle was saline. For 105 M and higher concentrations, N,N-dimethylformamide was the vehicle. Significant relaxation of bronchial vascular smooth muscle occurred at furosemide concentrations 104 M. A small decrease in tension was observed in control tissues due to the vehicle. * Significant difference from control responses, P < 0.05.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effect of furosemide on vasopressin-preconstricted bronchial arteries. Values are means ± SE. Responses to furosemide are normalized to the initial tension induced by 108 M vasopressin. , Effect of furosemide (n = 5); , effect of the vehicle (n = 5). For concentrations of furosemide <105 M, the vehicle was saline. For 105 M and higher concentrations, N,N-dimethylformamide was the vehicle. Significant relaxation of bronchial vascular smooth muscle occurred at furosemide concentrations 103 M. Small decreases in tension observed in control tissues are due to the vehicle. * Significant difference from control responses, P < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The aim of this study was to determine whether furosemide exerts a differential effect on contractility of bronchial smooth muscle and bronchial arterial smooth muscle. We observed that furosemide relaxed bronchial arterial smooth muscle, supporting the previous supposition that the antiasthmatic properties of furosemide are related to vasodilation and increased blood flow to the airways (9). Conversely, we found that porcine bronchial smooth muscle was unresponsive to furosemide but relaxed in response to the -adrenergic-receptor agonist isoproterenol. We conclude that furosemide, under the experimental conditions of this study, does not cause relaxation of preconstricted pig bronchi. These data therefore suggest that the antiasthmatic activity of furosemide is probably not mediated by airway dilation.

In a previous study, we showed that furosemide, when topically applied in vivo to the adventitial surface of the rat trachea, dilates tracheal arterioles and venules by a nitric oxide- and cyclooxygenase-independent mechanism (5). This technique is powerful, but it has shortcomings inherent to in vivo preparations. For instance, when judging the effects of an exogenously applied vasoactive substance on vessel diameter, it is difficult to assess the importance of peripheral influences such as alterations in intravascular pressure and endogenously released autocoids such as neurotransmitters, hormones, and inflammatory mediators. In the present study, furosemide was applied to isolated bronchial vessels to minimize the influence of these peripheral factors. These results confirmed our previous in vivo findings (5) that furosemide is indeed a dilator of the airway vasculature.

The precise mechanism by which furosemide protects the airways against asthmogenic challenges is unknown. Because this agent is effective against a variety of indirect bronchoconstrictor stimuli such as allergens, nebulized distilled water, exercise, cold air, sodium metabisulfite, AMP, and bradykinin, the antiasthmatic properties of loop diuretics must depend on a mechanism common to all stimuli. Gilbert and co-workers (9) first demonstrated that furosemide reduced the transairway thermal exchange gradient that develops during hyperpnea and suggested that this agent increased blood flow to the airways by dilating the airway vasculature. This action of furosemide would reduce airway smooth muscle reactivity by minimizing the thermal insult associated with hyperpnea. Unfortunately, this theory does not explain how furosemide protects against allergenic and chemical insults. Conversely, furosemide-induced dilation could enhance airway blood flow and increase the clearance of exogenously administered bronchoprovocants and endogenously released inflammatory mediators. This hypothesis is supported by the study of Csete et al. (6), which showed that the antigen-induced airway obstruction in allergic sheep was influenced by tracheal blood flow. The bronchoconstrictor mediators released by mast cells after stimuli such as exercise, distilled water, and allergens, as well as inflammatory substances released through sensory nerve stimulation, could therefore be rapidly cleared by the increased microcirculatory flow, minimizing their effects on airway smooth muscle.

In addition to its actions on vascular smooth muscle, it is possible that the antiasthmatic properties of furosemide are related to inhibition of liquid and mucus secretion from submucosal glands. Inflammatory autocoids, such as neurokinins and bradykinin, stimulate secretion of glandular mucus and liquid (22), and this process could contribute to airway obstruction in asthma. Furosemide, which blocks transepithelial Cl secretion through inhibition of Na⁺-K⁺-2Cl cotransport, could block a large fraction of this glandular volume secretion and thereby reduce the severity of airway obstruction. Indeed, bumetanide, a related loop diuretic, has been shown to inhibit acetylcholine-induced liquid secretion in porcine bronchi by 70% (26).

It has been suggested that the capacity of inhaled furosemide to reduce the airway response to indirect bronchoconstrictor stimuli may result indirectly through release of bronchodilator prostaglandins from airway epithelium (13) and renal tubular epithelium (14). The airway epithelium is an important source of PGE₂ (21), an agent that has been shown to protect against nebulized distilled water and allergen-induced bronchoconstriction in asthmatic patients (18, 19). However, as suggested by Chung and Barnes (4), it is difficult to conceive why PGE₂ would have a preferential protective effect against indirect challenges such as allergens, hyperpnea, and nebulized distilled water but not against direct airway constrictors such as methacholine. Additionally, the vasodilation response to furosemide in airway (5) and pulmonary (10) circulations has been shown to be cyclooxygenase and nitric oxide independent.

If the antiasthmatic properties of furosemide are related to dilation of airway vasculature, then furosemide must reach concentrations that approach 10⁴ M within the airway interstitium. The inhaled furosemide aerosol must deposit in the airway surface liquid, diffuse across the airway epithelium, and distribute into the interstitial space, which contains the vasculature. Because information on bioavailability of inhaled loop diuretics is currently lacking, it is not possible to estimate the interstitial concentrations of this drug. In an in vitro study of the bioelectric properties of canine tracheal epithelium, 10- to 100-fold higher luminal doses of furosemide were required to produce equivalent effects to adventitial administration, suggesting that the airway epithelium provides a significant barrier to the diffusion of these drugs (27). However, Gilbert and co-workers (9) report that furosemide aerosols increase the delivery of heat to the airways, providing indirect evidence that the inhalation of this drug is sufficient to increase airway blood flow. The restrictive barrier properties of the airway epithelium could contribute to the differential potency of loop diuretics on protection against asthma and explain why aerosolized bumetanide, a more potent loop diuretic, is less protective than furosemide against asthma (17).

Endothelia and epithelia are known to release factors that influence smooth muscle tone. However, we do not believe that such factors play a significant role in the response to furosemide. Our previous studies with rat tracheal arterioles demonstrated that the dilatory responses to furosemide were preserved in the presence of nitro-L-arginine methyl ester, an inhibitor of nitric oxide synthesis, and in the presence of indomethacin, a cyclooxygenase inhibitor (5). Although we cannot discount the possibility of species differences, these findings suggest that furosemide does not induce dilation through nitric oxide or prostaglandin release from endothelial cells. One could argue that the absence of a response to furosemide was related to damage to the bronchial epithelium, which might be required to release a relaxing factor. This is unlikely to be the case. In a previous study of pig bronchi, our laboratory found that the surface epithelium of this tissue was very resistant to abrasion and was completely removed only after vigorous rubbing with a rough, wooden ream (1). Therefore, the epithelium of the tissues used in the present study was certainly intact. Furosemide might have induced an observable dilator response in bronchi if it was added before, rather than after, acetylcholine. We believe that this is unlikely, however, because furosemide did not cause any measurable effect on airway tension at even supramaximal (1 mM) concentrations.

In summary, we provide evidence from this study on pigs that furosemide is a selective dilator of vascular but not airway smooth muscle. Although the precise mechanism by which furosemide protects against an asthmatic challenge remains to be determined, these data support previous contentions that the antiasthmatic properties of this agent are related to changes in airway vascular hemodynamics (9).