Exercise-Induced Asthma May Be Associated With Diminished Sweat Secretion Rates in Humans

Original Research ASTHMA

Exercise-Induced Asthma May Be Associated With Diminished Sweat Secretion Rates in Humans* Chan Park, MD; Christopher Stafiord, MD, FCCP; and Warren Lockette, M D

Background: Muscarinic receptor agonists increase water secretion from the acinar cells of

respiratory, sweat, salivary, and lacrimal glands. Mice lacking the gene for aqueous water channel aquaporin (Aqp) 5 exhibit methacholine-inducedbronchiolar hyperreactivity when compared to normal mice. Individuals with asthma also have enhanced airway responsiveness to methacholine and diminished airway hydration. Because Asp5 in humans is also expressed in respiratory, sweat, salivary, and lacrimal glands, we hypothesized that those individuals with exercise-induced asthma and excessive bronchiolar reactivity should also have decreased muscarinic receptordependent sweat, salivary, and tear gland secretions. Methods: Healthy, athletic subjects who are suspected of having exercise-induced bronchospasm were recruited, and FEV, values were determined following provocative airway challenges with methacholine. Measurements of pilocarpine-induced sweat secretion were taken in 56 volunteers, and some additional subjects also had timed collections of saliva and tear production. Results: Subjects manifesting excessive airway reactivity demonstrated by exaggerated methacholineinduced reductions in FEV, also had diminished values for pilocarpine-induced sweat secretion (n = 56; r = - 0.59; p < 0.OOOl).The rate of pilocarpine-stimulated sweat secretion in our subjects correlated highly with salivary flow rate (r = 0.69; p < 0.OOOl)and tearing rate (r = 0.86; p < 0.001). Concluaion: Hyperhidrosis, sialorrhea, and excessive tearing are traits that may indicate a phenotype that predicts resistance to hyperactive airway diseases such as exercise-induced asthma in humans. (CHEST 2008; 134:552558) Key words: aquaporin; exercise-induced asthma; humans; salivary gland; sweat; tearing Abbreviations: Aqp = aquaporin; MCT = methacholine clmllenge test -

c

old, dry air inhaled during exercise can stimulate bronchospasm.l.2When the airway surface water lost during breathing is not quickly replenished, subtle increases in the osmolality of the fluid bathing the trachea and bronchioles stimulate the release of inflammatory mediators that induce bronchospasm. During exercise, the ventilation rate can increase to nearly 150 Urnin, and because of evaporative fluid loss it is predictable that exerciseinduced bronchospasm will be seen most often with exertion during the winter months or in climates where athletes are exposed to cold, dry air. Furthermore, exertion-induced tachypnea enhances drying of the airways.1-4 Because glandular secretions of sodium, chloride, and water contribute much to the osmolality of the epithelial surface,5S6 and hence to bronchiolar smooth muscle reactivity, aberrancies in 552

the control of epithelial water secretion in the respiratory tree may be associated with hyperactive airway disease. Airway hydration is primarily dependent on the secretion of sodium and chloride into the airways from submucosal glands of the respiratory epithelium because water movement through aquaporin channels follows the osmotic force generated by the secretion of these electrolytes. An isoform of these protein pores, aquaporin (Aqp) Fj, is expressed in alveolar type I cells of the lungs and in the plasma membranes of secretory cells of the subepithelial glands of the respiratory tree.7 Mice have been bred whose gene encoding Aqp5 has been disrupted, and they express significantly less Aqp5 when compared to mice normally expressing this gene.S+-'l Compared with wild-type mice, it has been shown that Original Research

isolated perfused lungs from Aqp5-deficient mice have a 10-fold reduction in airspace-capillarypermeability induced by changes in the relative osmolality between the alveoli and the vasculature.10 In addition to the respiratory epithelia, A 9 5 is heavily expressed in the fluid-secreting portion of sweat, salivary, and tear We were intrigued that mice deficient in Aqp5 not only had increased bronchiolar hyperreactivity,g but also had decreased pilocarpine-stimulated sweat secretion9 and salivary gland output.12 We reasoned that if airway hydration contributes to bronchiolar smooth muscle tone, the increased bronchiolar smooth muscle sensitivity to inuscarinic agonists found in patients with hyperactive airways disease should be associated with coiicomitant abnormalities of pilocarpine-stimulated sweat gland secretion, and sweat, salivary, and tear secretion rates should correlate highly among individuals. To test this hypothesis, we measured sweat, salivary gland, and lacrimal secretion rates in subjects with, and without, hyperactive airways, which was demonstrated by having exaggerated responses to inhaled methacholine.

MATERIALSAND METHODS These studies were approved by our institutional review board, and all volunteers gave written informed consent. We recruited 56 subjects who had been referred to our medical center for the evaluation of signs and symptoms suggestive of new-onset, exercise-induced asthma. Subjects were otherwise healthy, young, male and female US Navy and Marine Corps personnel between 18 and 32 years of age. This evaluation included the measurement of bronchiolar reactivity to methacholine. No patients were enrolled if they met the guidelines for exclusion from undergoing a methacholine challenge test (MCT) according to the recommendations of the American Thoracic Society,l 3 smoked, had received therapy with inhaled or systemic steroids, or had a history of respiratory infection within the previous 4 weeks. Baseline spirometry measurements were made in duplicate, and we recorded the maximum volume of air that subjects could forcefully exhale (ie, FEV,). Because acute bronchospasm *From the Departments of Emergency Medicine (Dr. Park) and Clinical Investigation (Dr. Lockette), and the Division of Pulmonary and Critical Care Medicine (Dr. Stafford), Naval Medical Center, Sail Diego, CA. The comments expressed represent the personal opinions of the authors and the do not necessarily reflect the views of the Department of txe Navy or the Department of Defense. The authors have reported to the ACCP that no significant conflicts of interest exist with any com anies/organizationswhose products or services may be discussegin this article. Manuscript received February 7 , 2007; revision accepted April 7, 2008.

Reproductionof this article is rohibited without written pennission from the American College o Chest Physicians (www.chestjoumd. org/misc/re rints.shtm1) C o ~ r e s p o n ~ ~ nto:c eWarren Lockette, MD, Clinical Inuestigations Department, Naval Medical Center, Sun LXego, 34800 Bob Wilson Dr, Mail Code KCA, Sari Diego, C A 92134; e-mail: [email protected] DOI: 10.1378/chest.08-0366

P

www.chestjournal.org

obstructs air flow, the reduction in FEV, is a sensitive indicator of bronchiolar hyperreactivity. Subjects were challenged with the muscarinic agonist methacholine chloride (Provocholine; Metlrapharm; Coral Springs, FL) at increasing concentrations (O.Ofi, 0.25,1.00,4.00,and 16.00mg) in a futed volume of saline solution until they experienced a 20% drop in FEV, from baseline. Nebulized methacholine was delivered with a dosimeter (KoKo Dosimeter with automatic actuation, model 414400; Ferraris; Louisville, CO). Spirometry measurements were obtained (Med Graphics Elite Series; Med Graphics Corp; St. Paul, MN), and the data were processed with appropriate software (Breeze, version 6.0A; Med Graphics Corp). Upon completion of the MCT, we measured sweating responses to the iontophoretic application of another muscarinic agonist, pilocarpine, in these subjects. The volar aspect of the dominant forearm was washed with sterile water. After the skin had dried, bipolar electrodes covering agar discs impregnated with 0.1% pilocarpine (Pliogel discs; Westcor Inc; Salt Lake City, UT) were strapped to the skin. A constant current was applied for 5 min using a commercially available iontophoresis device (Westcor Iontophoresis Apparatus; Westcor Inc). After 5 min of' current, the pilocarpine discs were removed, the skin was flushed with ultrapure water, dried, and a 1-cm2 collection device was strapped to the forearm. Sweat secretions were collected over the next 20 min. The sweat samples were then transferred to a microfuge tube and weighed using an ultrasensitive laboratory balance (Mettler-Toledo Co; Columbus, OH). Sweat sodium concentrations were analyzed using atomic absorption spectrophotometry, and a commercial osniometer measured the freezing point of depression. To assess the relationship between sweat and salivary flow rates, we recruited an additional 58 healthy subjects. To collect saliva, each subject dried their mouth with cotton gauze, and then preweighed cotton dental plugs were placed under the volunteer's tongue for 5 min. Subsequently, the samples were weighed for the determination of saliva output. Lacrimal tear formation was also assessed in some of these subjects using a standard Schirmer test. Sweat, saliva, and tear samples were obtained with the investigator and subject blinded to the results of the MCTs. Statistical analyses were accomplished using a statistical software package (Prizm; Graphpad; San Diego, CA). The means of the values obtained between cohorts were compared with an unpaired t test. If there were significant differences in the variances of the values obtained for each population, a Welch correction was performed; if there were multiple comparisons, a Bonferroni adjustment in the p value was made. A Pearson test was used to evaluate parametric correlations, and a p value of 0.05 was considered to he statistically significant.

RESULTS Fifty-six otherwise healthy subjects with signs and symptoins suggestive of exercise-induced bronchospasin volunteered for this study. Twenty-two subjects were classified as having a positive MCT result using standard clinical criteria4.13 (ie, they had at least a 20%fall in their FEV, over baseline measurements following methacholine challenge), and 34 subjects had a minimal response, if any, to methacholine challenge. As shown in Figure 1, there were wide-ranging differences in the sensitivity to this muscarinic receptor agonist, and the mean fall in FEV, in those subjects who had bronchiolar hyperCHEST / 134 / 3 / SEPTEMBER, 2008

553

p
..

. ..

. .".. .

L".

I . .

0'

Negative

Positive

Negative

(n=w

(n=22)

(n-3)

Positive

FIGURE 1. Muscarinic agonist-induced reductions in FEV, in healthy patients undergoing an MCT. Subjects inhaled cumulative increases in methacholine until their FEV, decreased by 20%or they reached a dose of 16 mg. Those subjects who had a positive MCT test result had significantly greater reductions in their FEV, when com ared to subjects who were relative1 unresponsive to methacfoline (ie,a negative test). The mean ( % SEM) in FEV, was 27.9 C 1.6% vs 9.0 t 1.1%,respectively (n = 56; p < 0.0001).

FIGURE 2. The difference in mean sweat secretion rates between MCT-negative and MCT-positive subjects. Following the iontophoretic application of a 0.1% pilocarpine solution, sweat was dlected continuously for 20 min. The sweat volumes collected were sigdicantly higher amon those subjects who had a negative MCT result when compared to kose volunteerswho had MCT results that were su estive of hyperactive airway disease. The mean ( 2 SEM) sweat vo umes were 59 -e 3 vs 37 2 3 mg sweat per cm2 of skin every 20 min, respectively (n = 56; p < 0.0001).

reactivity, as manifested by a positive MCT result, was significantly higher than in those subjects who were deemed to have a negative test result (ie,mean [ ? SEMI fall in FEV,, 27.9 t 1.6 vs 9.0 +- 1.1, respectively; p < 0.0001; n = 56). We next measured sweat volumes following the administration of the muscarinic agonist pilocarpine in these subjects. Those individuals who were most sensitive to methacholine (ie,those with the greatest fall in FEV,) were also the volunteers who were the least sensitive to pilocarpine-induced sweat secretion. The mean sweat rates were significantly higher among those subjects who had a negative MCT result when compared to those volunteers who had MCT results suggestive of hyperactive airway disease (59 2 3 vs 37 3 mg of sweat per 20 min, respectively; p < 0.0001; n = 56) [Fig 21. The subjects' fall in FEVl following the inhalation of this muscarinic agonist had a highly significant, negative correlation with their sweat volumes (n = 56; r = -0.59; p < 0.0001) [Fig 31. Those individuals who had the most hyperactive airways tended to sweat the least. In humans, water is secreted from the secretory coil of serous sweat glands in a manner that maintains osmolality comparable to that which is found in the plasma and interstitial fluidl4; reabsorption of

these electrolytes through the water-impermeable sweat duct reduces the final osmolality of the secreted sweat.15 As a result of this arrangement, net water secretion is determined primarily by sodium excretion, and we postulated that net sodium excretion should correspond to net sweat secretion in our cohort. Indeed, we found that net sweat fluid excretion correlated highly with net sodium excretion (Fig 4).We did not measure other electrolytes, as the net excretion rate of osmotically active solutes was dependent, for the most part, on the net sodium excretion (r = 0.78; n = 46; p < 0.0001) [Fig 51. As expected, sweat sodium excretion rates were higher in subjects who were relatively unresponsive to inhaled methacholine when compared to those subjects who had a fall in FEV, of > 20% (2.78 0.48 vs 1.52 t 0.22 p,Eq/cm2 skid20 min, respectively; p < 0.05) [Fig 61. Although there were significant differences in the absolute sodium excretion rates between those subjects who had a positive MCT result and those individuals who did not, there were no significant differences in the mean sodium concentration of the secreted sweat among those subjects having a positive MCT result (39.7 t 6.4 mEq/L; subjects who had a negative MCT result, 42.0 4.5 mEq/L; p = not sigr&cant [data not

fd

*

554

T

_+

+_

Original Research

q=56

2501 I

pcO.Oo0l r= -0.59

15~1

,

> c .-

-a0

1001 i

200

. .

n

L

m

....

. *

..

rn

V/

I

I

I

I

1

10

20

30

40

50

0

shown]). We next tested our hypothesis that the degree of an individual's sweat excretion would correlate with their salivary excretion rate independent of their responsiveness to inhaled methacholine. Indeed, such was the case; we found a highly significant

150

50

75

100

FIGURE5. Sweat osmolality is predicted by the sweat sodium concentration. Because water passively follows electrolytes from the sweat secretory coil, net sweat osmolality is de endent on the secretion and reabsorption of sodium and c hlorii by the sweat land. Net sweat osmolality was dependent on sodium excretion n = 46; r = 0.78; p < 0.0001).

B

correlation between pilocarpine-stimulated sweat secretion and unstimulated salivary gland flow rates (n = 34; r = 0.69; p C 0.0001) [Fig 71. Initially, sweat fluid secretion was sluggish, but, over time, the rate of sweating increased. We re-

41

a

4-

a

L

p<0*05

1

./

0-

2.5

5.0

7.5

Negative (n=l2)

10.0

Net sodium secretion (in mEq/20min) FIGURE 4. Net sodium secretion drives sweat sodium secretion in sweat. The net amount of sweat fluid collected mrrelated highly with net sodium excretion rate (n = 79; r = 0.76; p < 0.0001). www.chestjournal.org

1

25

Swat sodium concentration (in mM)

FIGURE3. Total sweat volumes correlate with the maximum reduction in FEV, induced by methacholine. The subjects' fall in FEV, following the inhalation of this muscarinic agonist had a highly significant, negative correlation with their sweat volumes (n = 56; r = -0.59; p < O.OOO1). Those individuals who had bronchiolar hyperreactivity tended to have much lower sweat excretion rates.

0 0.0

r= 0.78

0

Maxfall in FEW (in percent)

*

DCOS3001

50i

I

01 0

.-

Positive (n=9)

FIGURE 6. Sweat sodium excretion rates were hi her in subjects who were relatively unresponsive to inhaled metfmcholine when compared to those subjects who had a fall in FEV, of > 20%. The mean (tSEM) sweat sodium excretion rates were 2.78 + 0.48 vs 1.52 0.22 pEq/cm2 of skin per 20 min, respectively (p < 0.05).

+

CHEST I 134 / 3 I SEPTEMBER, 2008

555

301

el3 w.002 r=0.73

8

”, 0

o

.

I

I

I

I

25

50

75

100

Sveatwl (in ug)

o 0.0

2.5

5.0

7.5

10.0

Salivary secretion rate (in rng/5min) FIGURE7 . Sweat secretion rates correlated hi hlv with salivary secretion rates; those individuals who secrete the most sweat tended to have a high degree of basal salivary flow rates (n = 34; r = 0.69; p < 0.0001).

FIGURE8. Thirteen of our subjects were randomly selected to have the rate of tear fonnation measured. In each of these volunteers, fdter pa er was placed over the conjunctiva of one eye for 5 min. In tgis representative sample, those individuals who had greater pilocarpine-induced Sweat secretion also had more tear formation (T = 0.73; p = 0.002).

$i

corded the time that was needed for an individual to secrete 2.5, 5.0, and 19.0 pL of sweat per cm2 of skin. Following the iontophoresis of pilocarpine, there was a lag in time before maximal sweating occurred. We also noted that those individuals who were the most sensitive to inhaled methacholine, and who secreted the least amount of sweat, tended to reach their maximum sweat output later than those individuals who did not have a fall in FEVl following the inhalation of inethacholine (data not shown). Finally, in a subset of subjects who were randomly selected, we assessed lacrimal tear secretion using a standard Schirmer test. We found that the degree of sweat secretion also correlated with lacrimal tear secretion (n = 13; r = 0.73; p = 0.002) [Fig 81.

DISCUSSION It has been shown that targeted disruption of the gene encoding Aqp5 results in mice with decreased osmotically dependent water movement into, and out of, alveoli, but the loss of Aqp5 did not modify hydrostatically driven lung edema or active alveolar fluid reabsorption. Furthermore, lower airway humidification, as determined by the moisture content of the expired air, was reduced by only 3 to 4% in mice that lacked both Aqp5 and Aqpl.lOJl On the basis of these two reports,lO.ll one would infer an ambiguous role for Aqp5 in lung physiology. However, pilocarpine-stimulated fluid secretion from 556

submucosal glands of the upper airways showed a 57% reduction in the fluid secretion rate in the mice .~ laborawith targeted disruption of A ~ p 5 Another tory demonstrated9 that the targeted disruption of Aqp5 showed significantly increased concentrationdependent bronchoconstriction in response to IV administered acetylcholine, and this bronchiolar hyperreactivity was not due to differences in the contractility of isolated tracheal smooth muscle. The investigatorswho bred those mice posited that because the local increase in the osmolarityof the airway surface liquid induces bronchoconstriction in asthmatic patients, the absence of Aqp5 at the luminal surface of tracheal and bronchial murine epithelial cells could lead to altered extracellular ionic composition that could affect the release of bronchoconstrictors in their mice. However, studying the methacholine-induced secretion of fluid froin the respiratory tree of humans is difficult. We postulated that inuscarinic agonist-induced sweat secretion could serve as a surrogate marker of respiratory “sweat.” Respiratory gland epithelium and sweat gland epithelium share identical mechanisms of muscarinic receptor (m3 subtype)-dependent sodium, chloride, and water secretion.l”l9 A common genetic aberration in one of these responses induced by the activation of the muscarinic receptor should be manifested in measurements of both respiratory function and sweat gland activity. Specifically, we hypothesized that, in humans, if hyperactive airway diseases results from abnormalities in Aqp5-mediated fluid secretion by the epithelial glands of the respiratory tree, the increased bronchiolar smooth muscle sensitivity to muscarinic agonists found in patients with hyperacOriginal Research

tive airways disease should be associated with concomitant abnormalities of pilocarpine-stiddated sweat gland secretion. We did not measure fluid secretion in the respiratory trees of our subjects, and our results demonstrate only an association rather than cause and effect. However, all of our data are consistent with our hypothesis. Those individuals who were the most sensitive to methacholine-induced bronchospasm were more likely to have markedly diminished sweat output. The measurement of pilocarpine-induced sweat secretion may be a surrogate marker for defective respiratory gland water secretion. Unlike contrasting sodium excretion rates between subjects with high and low pulmonary sensitivity to inhaled methacholine, skin surface sweat sodium concentration and osmolality did not vary as a function of airway responsiveness. The reabsorption of sodium and chloride in the sweat duct is rate limited (ie, higher sweating rates result in higher sweat sodium and chloride concentrations)20 because the capacity of the sweat duct to reabsorb electrolytes can be saturated. Accordingly, it was not surprising that the sweat fluid secretion was dependent on the absolute secretion rate of sodium and chloride. Through a common mechanism utilizing the heterotrimeric G proteins Gaq and G a i l l , and phospholipase C p l , m3 muscarinic receptor activation stimulates the secretion of sodium and chloride in the secretory coil of the subepithelial respiratory and serous sweat glands. The in3 muscarinic receptor concomitantly stimulates the translocation of AqpS to the apical membrane of these cells, and water follows the osmotic gradient generated by the movement of these e l e c t r ~ l y t e s .Direct ~~~l~ cannulation of the secretory coil of sweat glands has shown that sweat is secreted iso-osmotically,14 and the hypotonicity of surface sweat results from the reabsorption of sodium and chloride as these ions muve along the water-impermeable sweat ducts. Therefore, any differences in secreted sweat concentrations are due to contrasts in electrolyte reabsorption in the sweat duct. Because we found no differences in final sweat concentrations or osmolality among subjects of varying respiratory methacholine sensitivity, it is likely that the differences in sweating rates were directly due to contrasting responses to pilocarpine-stimulated sodium, chloride, and water excretion by the sweat secretory coil. These differences are likely due to individual variation in m3 muscarinic receptor-dependent sodium, chloride, or water secretion mediated by Caq, G a i l l , or phospholipase C p l . Others21 have tried unsuccessfully to link asthma with respiratory gland secretion abnormalities. Those investigators used a surrogate marker of airway surface electrolytes (ie, by measuring nasal potential differences). However, www.chestjournal.org

these measurements of nasal potential differences are dependent on the concentration of airway surface electrolytes, which, according to our model, would not be different among patients with and without hyperactive airways disease, rather than the absolute secretion rates of sodium and chloride. Alternatively, our findings could represent individual variation in m3 muscarinic receptor signal transduction or in the control of Aqp5 expression or trafficking. It should be noted, however, that a decreased expression of Aqp5 has been correlated with mucus overproduction and hyperactive airways in COPD patients.22 We present other evidence that a common abnormality in m3 muscarinic receptor signal transduction may be responsible for the association between methacholine-induced bronchoconstriction and diminished pilocarpine-stimulated sweat secretion. There is high basal muscarinic-dependent saliva secretion in humans; atropine is frequently used to reduce respiratory gland and salivary gland secretions preoperatively. We next postulated that pilocarpinestimulated sweat secretion would be associated with a high degree of unstimulated saliva production; such was the case. From these observations, it can be suggested that there exists a phenotype of hyperreactive airways, diminished sweat secretion, reduced tearing, and having a relatively “dry” mouth. Furthermore, physical training is associated with a reduction of sweat sodium levels and sweat fluid I o s s . ~ ~It. remains ~~ to be determined whether or not the increased frequency of exercise-induced asthma among elite athletes is due to concomitant reductions in fluid secretion from the respiratory gland epithelium much in the way that sweat sodium and water loss are minimized with physical training. Our observations may also explain why respiratory smooth muscle is hyperresponsive to muscarinic agonists in vivo, but not in vitro, in many animal models of allergyzs; exercise-induced asthma may not represent an intrinsic disorder of bronchiolar smooth muscle. Instead, differences in the composition of airway surface liquid among individuals with variations in respiratory gland secretion rates may affect the release of inflammatory mediators that subsequently induce bronchoconstriction in vivo. In addition, the expression of AqpS has been shown to decline with ages; this Ending may explain the higher prevalence of exercise-induced bronchospasm in older athletes. It is also possible that the salutary benefit of p-adrenergic bronchodilators may stem not only from their direct effect on bronchiolar smooth muscle, but also because they may facilitate the cyclic adenosine monophosphate-dependent translocation of the AqpS protein to the cell meinbrane.27 Finally, we have identified a phenotype that suggests a correlation between muscarinic receptorCHEST 1 134 1 3 1SEPTEMBER, 2008

557

dependent bronchospasm, sweating, salivary, and tear secretion rates. Multiple, independent laboratories have reported2830 a potential linkage of the asthma phenotype with chromosome 12q in humans; a search of genome databases indicated that the allele for Aqp5 that encodes the water channel mediating bronchiolar responsiveness, sweating, and salivation is also found at this locus.31 Functional polymorphisms in this gene locus, or in linkage disequilibrium with this locus, may ultimately contribute to this phenotype. Finally, it should be noted that elderly individuals have a greater propensity for heat stroke, and that the mechanisms responsible for age-related declines in sweating may also contribute to age-related decrements in salivation and the phenomenon of “dry eyes.” And, we also note that during their training athletes may sweat, drool, or cry; at least they will not gasp. ACKNOWLEDGMENT: We thank Justin Clark, MD, from the University of Michigan, for providin the measurements of sweat sodium concentrations. Also, the fo%owing individuals from the Naval Medical Center San Diego contributed to our work: Ryan Woodman for drawing our gra hs; Debra ames,Waine Macahter, and Suzana Hazeldon for $e editori assistance; and Robert Riffenburgh, PhD, for statistical analyses. Dr. Lockette has had full access to a l l of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

J

REFERENCES 1 Anderson SD, Schoeffel RE, Follet R, et al. Sensitivityto heat and water loss at rest and during exercise in asthmatic patients. Eur J Respir Dis 1982; 63:459-471 2 Bundgaarad A, Ingemann-Hansen T, Schmidt A, et al. Influence of temperature and relative humidity of inhaled gas on exercise-induced asthma. Eur J Respir Dis 1982; 63:239-244 3 Moloiiey E, O’Sullivan S, Hogan T, et al. Airway dehydration: a therapeutic target in asthma. Chest 2002; 121:1806-1811 4 Rundell KW, Jenkinson DM. Exercise-induced bronchospasm in the elite athlete. Sports Med 2002; 32583-600 5 Song Y, Verlanan AS. Aquaporind dependent fluid secretion in airway submucosal glands. J Biol Chem 2001; 276:41288-41292 6 Hoffert JD, Leitch V, Agre P, et al. Hypertonic induction of aquaporin-5 expression through an ERK-dependent pathway. J Biol Chem 2000; 275:9070-9077 7 Nielson S, Ling LS, Christensen BM, et al. Aquaporins in complex tissues: iI Subcellular distribution in respiratory and glandular tissues of the rat. Am J Physiol 1997: 273C154941561 8 Krane CM, Fortner CN, Hand AR. Aquaporin-5 deficient mouse lungs are hyperresponsive to cholinergic stimulation. Proc Natl Acad Sci U S A 2001: 98:14114-14119 9 Nejsum LN, Kwon TH, Jensen UB. Functional requirement of aquaporin-5 in plasma membranes of sweat glands. Proc Natl Acad Sci U S A 2002; 99:511516 10 Ma T, Fukuda N, Song Y. Lung fluid transport in aquaporind knockout mice. J Clin Invest 2000; 10593-100 11 Song Y, Jayaraman S, Yang B, et al. Role of aquaporin water channels in airway fluid transport, humidification, and surface liquid hydration. J Gen Physiol 2001; 117:573-582 12 Ma T, Song Y, Gillespie A, et d. Defective secretion of saliva in transgenic mice lacking aquaporin-5 water channels. J Biol Chem 1999; 274:20071-20074

558

13 Wubbel C, Asmus MJ, Stevens G, et al. Methacholine challenge testing: comparison of the two American Thoracic Society-recommended methods. Chest 2004; 125:453-458 14 Schulz IJ. Micropuncture studies of sweat formation in cystic fibrosis patients. J Clin Invest 1969: 48:1470-1477 15 Sato K, Kang WH, Saga K, et al. Biology of sweat glands and their disorders; normal sweat gland function. J Am Acad Dermatol 1989; 20:537562 16 Cummins MM, O’Mullane LM, Barden JA, et al. Antisense co-suppression of Gaq and G,,, demonstrates that both isoforms mediate M,-receptor activated Ca2+ signaling in intact epithelial cells. Pflugers Arch 2002; 444:644-653 17 Stummann TC, Poulsen JH, Hay-Schmidt A, et al. Pharmacological investigation of the role of ion channels in salivary secretion. Pflugers Arch 2003; 446:78-87 18 Ishikawa Y, Eguchi T, Skowronski MT, Ishida H. Acetylcholine acts on M 3 muscarinic receptors and induces the translocation of aquaporin-5 water channel via cytosolic Ca2+ elevation in rat parotid glands. Biochem Biophys Res Comm 1998; 245:835-840 19 Ishikawa Y, Skowronski MT, Ishida H. Persistent increase in the amount of aquaporind in the apical plasma membrane of rat parotid acinar cells induced by a muscarinic agonist SNI-2011. FEBS Lett 2000; 477:253-257 20 Tanaka H, Osaka Y, Obara S. Changes in the concentrations of Na+, K+, and C1- in secretion from the skin during progressive increases in exercise intensity. Eur J Appl Physiol 1992; 64:557561 21 Chung NC, Illek B, Widdicombe JH, et al. Measurement of nasal potential difference in mild asthmatics. Chest 2003; 123:1467-1471 22 Wang K, Feng YL, Wen FQ, et al. Decreased expression of human aquaporin-5 correlated with mucus overproduction in airways of chronic obstructive pulmonary disease. Acta Pharmacol Sin 2007; 28:1166-1174 23 Locke W, Talbot NB, Jones HS, et al. Studies on the combined use of measurements of sweat electrolyte composition and rate of sweating as an index of adrenal cortical activity. J Clin Invest 1951; 30:325-336 24 Kirby CR, Convertino VA. Plasma aldosterone and sweat sodium concentrations after exercise and heat acclimation. J Appl Physiol 1986; 61:967-970 25 Turner DJ, Noble PB, Lucas MP, et al. Decreased airway narrowing and smooth muscle contraction in hyperresponsive pigs. J Appl Physiol2002; 93:1296-1300 26 Inoue N, Iida H, Yuan Z, et al. Age-related decreases in the response of aquaporind to acetylcholinein rat parotid glands. J Dent Res 2003; 82:476-480 27 Yang F, Kawedia JD, Menon AG. Cyclic AMP regulates aquaporin-5 expression at both transcriptional and posttranscriptional levels through a protein kinase A pathway. J Biol Chem 2003; 278:3217342180 28 Raby BA, Silverman EK, Lazarun R, et al. Chromosome 12q harbors multiple genetic loci related to asthma and asthmarelated phenotypes. Hum Mol Genet 2003; 1231973-1’379 29 Heinzmann A, Grotherr P, Jerkic SP. Studies on linkage and association of atopy with chromosomal region 12q13-24. Clin Exp Allergy 2000; 30:1555-1561 30 Barnes KC, Freidhoff LR, Nickel R, et al. Dense mapping of chromosome 12q13.12q23.3and linkage to asthma and atopy. J Allergy Clin Immunol 1999; 104:485-491 31 Lee MD, Bhakta KY, Rainai S, et al. Human Aquaporin-5 gene: molecular characterization and chromosomal localization. J Biol Chem 1996; 271:8599-8604

Original Research