Exercise-induced alterations in phospholipid hydrolysis, airway surfactant, and eicosanoids and their role in airway hyperresponsiveness in asthma

Abstract

The mechanisms responsible for driving endogenous airway hyperresponsiveness (AHR) in the form of exercise-induced bronchoconstriction (EIB) are not fully understood. We examined alterations in airway phospholipid hydrolysis, surfactant degradation, and lipid mediator release in relation to AHR severity and changes induced by exercise challenge. Paired induced sputum (n = 18) and bronchoalveolar lavage (BAL) fluid (n = 11) were obtained before and after exercise challenge in asthmatic subjects. Samples were analyzed for phospholipid structure, surfactant function, and levels of eicosanoids and secreted phospholipase A₂ group 10 (sPLA₂-X). A primary epithelial cell culture model was used to model effects of osmotic stress on sPLA₂-X. Exercise challenge resulted in increased surfactant degradation, phospholipase activity, and eicosanoid production in sputum samples of all patients. Subjects with EIB had higher levels of surfactant degradation and phospholipase activity in BAL fluid. Higher basal sputum levels of cysteinyl leukotrienes (CysLTs) and prostaglandin D₂ (PGD₂) were associated with direct AHR, and both the postexercise and absolute change in CysLTs and PGD₂ levels were associated with EIB severity. Surfactant function either was abnormal at baseline or became abnormal after exercise challenge. Baseline levels of sPLA₂-X in sputum and the absolute change in amount of sPLA₂-X with exercise were positively correlated with EIB severity. Osmotic stress ex vivo resulted in movement of water and release of sPLA₂-X to the apical surface. In summary, exercise challenge promotes changes in phospholipid structure and eicosanoid release in asthma, providing two mechanisms that promote bronchoconstriction, particularly in individuals with EIB who have higher basal levels of phospholipid turnover.

INTRODUCTION

Indirect or endogenous airway hyperresponsiveness (AHR) is a specific feature of asthma in which the release of mediators that lead to bronchoconstriction occurs endogenously from sources within the airways without the administration of an exogenous bronchoconstrictor (as in the case of direct AHR to methacholine challenge) (1–5). Exercise-induced bronchoconstriction (EIB) is a specific form of endogenous AHR that is measured via dry air exercise challenge testing. This test is a strong and highly standardized stimulus for EIB during which an individual achieves a high level of ventilation that serves to transfer both water and heat from the airways leading to a transient shift in airway osmolarity, which is then rapidly corrected by the osmotically sensitive airway epithelium.

Although EIB is well recognized as a specific feature of asthma and is found in ∼30%–50% of subjects with asthma (6, 7), the specific mechanisms responsible for driving and promoting endogenous AHR in the form of EIB remain unclear. Previous studies have demonstrated that either a period of hyperventilation or challenge with an osmotic stimulus leads to the sustained generation of key eicosanoids, including prostaglandin D₂ (PGD₂) and the cysteinyl leukotrienes (CysLTs, LTC₄, D₄, E₄), in the airways of susceptible individuals that are known to promote bronchoconstriction (8–10). However, the mechanism leading to the generation of these mediators is not fully understood. Similarly, prior studies have suggested a role for surfactant dysfunction as an additional mechanism to promote bronchoconstriction in EIB by directly contributing to the propensity of the airways to close (11–13), but there have been no previous studies directly measuring airway phospholipid levels, phospholipid hydrolysis, surfactant composition, or surfactant function in this population.

Prior studies from our laboratory have revealed that dysregulated eicosanoid synthesis is related to alterations in the family of enzymes that regulates the initiation of eicosanoid synthesis (14–17). Cellular membrane phospholipids and phospholipids from other sources such as airway surfactant are cleaved by phospholipase A₂s (PLA₂) at the sn-2 position, which results in the liberation of free fatty acids and lysophospholipid (LysoPL) species (18). In the case of cellular membranes, the sn-2 position is often substituted with polyunsaturated fatty acids (PUFA), including arachidonic acid (AA) that serves as the initiator of eicosanoid synthesis (19, 20). Secreted PLA₂s (sPLA₂) are potential mediators of this dysregulated eicosanoid synthesis and are attractive targets because they have multiple roles including acting in conjunction with group IVA cytosolic PLA₂-α (cPLA₂α), binding specific C-type lectin receptors, and cleaving both membrane and surfactant phospholipids (21–24). Our work has subsequently focused on sPLA₂-X because this enzyme has high activity on the outer cell membrane of mammalian cells and the secreted levels of this enzyme are increased in the airway lumen in subjects with asthma (15).

This study was designed to further explore the mechanisms driving endogenous AHR with a focus on airway phospholipid hydrolysis and the role of sPLA₂-X in this process. We obtained induced sputum and bronchoalveolar lavage (BAL) fluid samples at baseline and following exercise challenge from subjects with asthma who had been rigorously characterized for direct and endogenous AHR. We evaluated these samples for direct and indirect markers of phospholipid hydrolysis, changes in surfactant composition and function, liberation of selected eicosanoids, and changes in sPLA₂-X. We further modeled dry air challenge ex vivo using a primary human airway epithelial cell model in organotypic culture to examine the release of sPLA₂-X in response to osmotic stress.

METHODS

Below is a brief summary of the methods used in this study. A comprehensive description of the methods is available in the Supplemental Data (all Supplemental material is available at https://doi.org/10.6084/m9.figshare.13428419.v1).

Study Subjects and Protocol

The University of Washington Institutional Review Board (IRB) approved the study protocol, and written, informed consent was obtained from all participants (IRB Application Number: 38553, IRB Application Title: Secretory Phospholipase A₂ in Airway Pathophysiology). The study population consisted of persons aged 18–59 with a physician diagnosis of asthma recruited at the University of Washington who were treated with only a short-acting β₂-agonist as needed for the duration of the study, and had a forced expiratory volume in 1 s (FEV₁) >65% predicted and a positive methacholine challenge test (PC₂₀ ≤ 8 mg/mL). Subject groups were further separated based on the presence of EIB, defined by a decrease of 15% or greater in the FEV₁ from the pre-exercise level. Spirometry (25), methacholine challenge, and dry air exercise challenges (26) were conducted in accordance with American Thoracic Society Standards. A total of 39 subjects were evaluated in the laboratory for this study. A total of 18 subjects completed both induced sputum visits, and 11 subjects completed both research bronchoscopy visits (four total visits).

Study participants had a total of four visits following the initial laboratory evaluation for inclusion in the study: two for obtaining induced sputum and two for research bronchoscopy with bronchoalveolar lavage. Induced sputum was obtained on two visits in random order, either as a baseline value or 30 min following exercise challenge. Of the 18 patients to complete both induced sputum visits, 10 were EIB positive (EIB⁺) and 8 were EIB negative (EIB⁻). Of the 11 patients to complete both research bronchoscopy visits (two additional visits, four total visits as all subjects who completed research bronchoscopies also completed induced sputum testing), 7 were EIB⁺ and 4 were EIB⁻.

Induced Sputum and BAL Analysis

The BAL and induced sputum supernatant were centrifuged at 45,000 g for 60 min. The resulting supernatant, representing the small aggregate (SA) fraction, was removed and the pellet, representing the large aggregate (LA) fraction, was collected and both fractions were stored at −80°C. The phospholipid content of LA and SA fractions was measured by the method of Rouser (27). The phospholipid composition of the LA fraction was determined using HPLC with electronic light scatter detection (28). Phospholipase activity in BAL supernatants was measured as hydrolysis of PG using a model surfactant as previously described (28). Surfactant function was assessed with a pulsating bubble surfactometer (22). The results are reported as γ_min, the minimum stable surface tension achieved over 10 min. The levels of the eicosanoids CysLTs, PGD₂, and PGE₂ were measured via ELISA. Levels of sPLA₂-X protein in induced sputum were determined by time-resolved fluorescence immunoassay (TRFIA) (16) with recombinant human sPLA₂-X used to generate a linear standard curve (24).

Primary Airway Epithelial Cell Model

Primary airway epithelial cells (AECs) were isolated from epithelial brushings obtained during research bronchoscopy and placed in primary culture as previously described (15, 29, 30). Cells were cryopreserved and/or subcultured at 90%–95% confluence. Cryopreserved primary epithelial cells at passage 1 or 2 were used for differentiated air-liquid interface (ALI) organotypic cultures.

Osmotic Stress Model

Fully differentiated primary airway epithelial cell organotypic cultures from brushings obtained from 3 separate EIB⁺ individuals with asthma were used for the osmotic stress model (30). To apply osmotic stress, 200 µL of buffer was added to the apical surface with or without sorbitol at the concentrations ranging from 250 to 750 mOsm. The volume of the basolateral media was adjusted to 1 mL. After applying osmotic stress to the apical surface for 24 h, the apical and basolateral fluid volumes were collected and measured using a pipette. The epithelial cells were also collected using a Tris lysis buffer and gentle sonication. The levels of sPLA₂-X were assayed using the TRFIA (see above).

Statistical Analysis

For comparisons among groups, P values were calculated as appropriate for the data: a two-tailed Student’s t test for two-group comparisons for normally distributed continuous data and a Wilcoxon matched-pairs signed rank test for two-group comparisons for non-normally distributed continuous data. A two-way ANOVA with correction for multiple comparisons was used to assess the effect of EIB status and exercise condition on LA/SA ratio and eicosanoid levels in the airways. A one-way ANOVA with correction for multiple comparisons using the two-stage step-up method of Benjamini, Krieger, and Yekutieli was used for multiple group comparisons of the effects of osmotic stress on the secretion of fluid and sPLA₂-X to the apical surface in the epithelial organotypic ALI cell culture model. Associations between continuous variables were assessed by linear regression. A P value less than 0.05 was considered significant.

RESULTS

Study Population Characteristics

Study participants were confirmed to have asthma with a positive direct AHR test (methacholine challenge) and then underwent exercise challenge testing to define the severity of endogenous AHR. Subjects with a greater than 15% decline in their baseline FEV₁ values were defined as EIB positive (EIB⁺) and those with a less than 15% change in their baseline FEV₁ values were defined as EIB negative (EIB⁻) (Fig. 1, A and B). Both groups had similar baseline characteristics, including lung function testing (Table 1). In contrast, participants with a higher degree of EIB also had lower methacholine PC₂₀ levels indicating that there is a significant correlation between the severity of endogenous AHR with the degree of direct AHR in this population (r² = 0.46, P = 0.002; Fig. 1C).

Figure 1.

Characterization of the phenotype of airway hyperresponsiveness in asthma. All subjects had a positive methacholine challenge test with PC₂₀ < 8 mg/dL to confirm a diagnosis of asthma (A) and underwent exercise challenge testing to define the severity of exercise-induced bronchoconstriction (EIB) (B). A subject was determined to be EIB⁺ if there was a >15% decline in baseline forced expiratory volume in 1 s (FEV₁) following exercise challenge. There was a significant correlation between the severity of EIB (defined by the area under the curve of the percent change in FEV₁ over the 30 min postexercise challenge test or AUC30) and the provocative concentration to reduce the FEV₁ by 20% with methacholine challenge or PC₂₀ (C). A linear regression line is shown with the 95% confidence interval as dotted lines. AUC30, area under the FEV₁ curve; EIB, exercise-induced bronchoconstriction; FEV₁, forced expiratory volume in 1 s; PC₂₀, provocative concentration of methacholine to reduce the FEV₁ by 20% of the baseline.

Table 1.

Study population characteristics

	Asthma
	EIB+ (n = 10)	EIB− (n = 8)	P Value
Age
Mean (Range)	26.2 (18–34)	23.3 (19–36)	0.33
Sex
% Female (Number)	20 (2/10)	37.5 (3/8)	0.44
Asthma clinical features
Month prior to study
Seattle Asthma Severity and Control Questionnaire Score (Range 0–25)	3.9 (0–10)	4.5 (0–10)	0.73
SABA use (uses/month) (Range)	3.6 (0–12)	6.7 (0–16)	0.28
Days free of asthma symptoms (Range)	18 (2–31)	24 (15–31)	0.18
12 months prior to study
ED visit or hospitalization	0	0	N/A
Baseline spirometry
FVC (%) (Range)	110.9 (79–128)	109.1 (95–141)	0.79
FEV1 (%) (Range)	92.5 (75–123)	91 (83–104)	0.82
FEV1/FVC (Range)	0.72 (0.59–0.82)	0.76 (0.69–0.81)	0.23
FEF25-75 (%) (Range)	65.9 (50–101)	70.5 (47–82)	0.51
Direct AHR
Methacholine PC20 (mg/mL) (Range)	0.24 (0.05–0.91)	2.04 (0.12–4.63)	<0.001
After exercise
Maximum decrease in FEV1 (%) (Range)	29.5 (19.1–58.7)	7.4 (1.52–14.0)	<0.001
Area under the FEV1 curve (AUC30) (Range)	644.6 (133-1472)	130.9 (26-301)	0.002

Mean values and ranges are listed. P values represent the result of unpaired t tests. The Seattle Asthma Severity and Control Questionnaire scores range from 0 to 25 based on responses to frequency of nocturnal symptoms, daytime symptoms, frequency of asthma symptoms limiting participation, and frequency of sustained episodes of poor asthma symptom control with lower scores representing better disease control (31). FVC, forced vital capacity; FEV₁, forced expiratory volume in 1 s.

Surfactant Degradation Is Increased in Individuals with Exercise-Induced Bronchoconstriction and is Further Triggered by Dry Air Exercise Challenge in Vivo

Both induced sputum and BAL aliquots were analyzed for the amount of LA and SA of pulmonary surfactant. Subjects with more severe EIB tended to have a lower baseline LA/SA ratio in BAL fluid suggesting there is increased baseline surfactant degradation or turnover that could be attributable to sPLA₂ hydrolysis in subjects with more severe EIB (r² = 0.35, P = 0.07; Fig. 2A). There was a significant decline in the LA/SA ratio in induced sputum samples following exercise challenge in all subjects indicating that exercise challenge triggers further surfactant degradation (P = 0.04; Fig. 2B). A similar but nonsignificant trend was seen in BAL samples (P = 0.12; Fig. 2C). There was no significant effect of the EIB phenotype on the change in LA/SA ratio with exercise challenge in either induced sputum (P = 0.85) or BAL samples (P = 0.22) further indicating that the primary difference in phenotype is the baseline difference and that exercise triggers surfactant degradation in all asthmatics.

Figure 2.

Phospholipid hydrolysis in relation to the severity of EIB and effects of dry air exercise challenge on measures of phospholipid hydrolysis and phospholipase activity. A: subjects with more severe EIB tended to have a lower LA/SA ratio (linear regression line shown with 95% confidence interval as dotted lines). The LA/SA ratio in induced sputum (B; EIB⁺ n = 9, EIB⁻ n = 5) and BAL fluid (C; EIB⁺ n = 7, EIB⁻ n = 3) at baseline and following exercise challenge (mean values are shown; P values represent the effect of exercise challenge on LA/SA ratio based on a two-way ANOVA modeling EIB phenotype and exercise condition). D: effects of exercise challenge on measures of ex vivo hydrolysis of surfactant-associated PG in BAL fluid following exercise challenge (EIB⁺ n = 4, EIB⁻ n = 2; solid bars represent median values; P value represents the result of a paired t test). BAL, bronchoalveolar lavage; EIB, exercise-induced bronchoconstriction, LA, large aggregate; SA, small aggregate.

Exercise Challenge Induces the Hydrolysis of a Major Phospholipid Component of Surfactant and Surfactant Dysfunction

Due to limited sample volume in some subjects, only a subset of BAL samples could be analyzed for surfactant phospholipid composition of the LA fraction (n = 8) and phospholipase activity as determined by ex vivo phosphatidylglycerol (PG) hydrolysis (n = 6). There was no significant change in the percentages of phosphatidylcholine, phosphatidylglycerol, phosphatidylethanolamine, phosphatidylinositol, or sphingomyelin in the large aggregate fractions of BAL samples following exercise challenge. Similarly, there was no difference between baseline and postexercise percentages of the above individual phospholipids between EIB⁺ and EIB⁻ subjects (data not shown). However, there was a significant increase in PG hydrolysis in postexercise challenge BAL samples (mean values pre-exercise 4.86% versus 10.19% postexercise, P = 0.002; Fig. 2D) indicating an increase in phospholipase activity with exercise. Sufficient paired samples were available to assess surfactant function utilizing a pulsating bubble surfactometer. Two of these samples demonstrated an increase in the minimum surface tension following exercise, and the third sample demonstrated elevated minimum surface tension at baseline that persisted with exercise challenge (Supplemental Fig. S1).

Exercise Challenge Induces the Generation of Eicosanoids in the Airways That Is Greater in Subjects with More Severe Exercise-Induced Bronchoconstriction

We measured the levels of CysLTs, PGD₂, and PGE₂ in induced sputum at baseline and following exercise challenge (Fig. 3). We found that EIB⁺ subjects tended to have higher baseline levels of CysLTs in comparison with EIB⁻ subjects (mean values of 392.4 vs. 271.4 pg/mL, P = 0.09, based on unpaired t test, data not shown) and that CysLTs tended to increase with exercise challenge in all subjects (median values pre-exercise 270.2 vs. 323.1 pg/mL postexercise, P = 0.06; Fig. 3A). There were no significant differences in the baseline levels of either PGD₂ or PGE₂ between the two groups. Exercise challenge resulted in a significant increase in the levels of PGD₂ (median values pre-exercise 103.4 vs. 207.5 pg/mL postexercise, P < 0.001; Fig. 3B) and PGE₂ (median values pre-exercise 116.3 vs. 188.5 pg/mL postexercise, P = 0.009; Fig. 3C) in all subjects. A two-way ANOVA model with EIB status and exercise condition demonstrated an effect of EIB phenotype on the levels of CysLTs (P = 0.07; Fig. 3D) and PGD₂ (P = 0.02; Fig. 3E) but not on the levels of PGE₂ (P = 0.40; Fig. 3F).

Figure 3.

Effects of airway hyperresponsiveness phenotype and exercise challenge on the levels of induced sputum eicosanoids in asthma (total n = 16; EIB⁺ n = 8, EIB⁻ n = 8). A–C: levels of PGD₂, CysLTs, and PGE₂ increased following exercise challenge in individuals with asthma (solid bars represent median values; P values result from a Wilcoxon matched-pairs signed rank test). D–F: eicosanoid levels grouped by EIB status identify differential effect of the levels of CysLTs and PGD₂, but no difference in PGE₂ (P values in the upper left-hand corner represent the effect of EIB status on eicosanoid levels based on a two-way ANOVA with multiple comparisons modeling EIB phenotype and exercise condition; P values comparing postexercise eicosanoid levels result from the post hoc multiple comparison analysis). CysLT, cysteinyl leukotriene; EIB, exercise-induced bronchoconstriction; PGD₂, prostaglandin D₂; PGE2, prostaglandin E₂.

We also found a significant correlation between the baseline levels of CysLTs and PGD₂ and the degree of direct AHR. Higher levels of CysLTs (r² = 0.30, P = 0.03; Fig. 4A) and PGD₂ (r² = 0.40, P = 0.009; Fig. 4B) were associated with lower methacholine PC₂₀ values and thus more severe direct AHR. There was a weak association between higher baseline levels of CysLTs and severity of endogenous AHR (r² = 0.19, P = 0.09; data not shown) but otherwise no correlation between baseline levels of PGD₂ or PGE₂ and endogenous AHR. However, both the postexercise values of CysLTs and PGD₂ as well as the change in CysLTs and PGD₂ levels with exercise correlated with endogenous AHR. Higher postexercise levels of CysLTs (r² = 0.28, P = 0.03; Fig. 4D) and PGD₂ (r² = 0.33, P = 0.02; Fig. 4E) were associated with more severe EIB, and similarly, greater increases in CysLTs (r² = 0.22, P = 0.07; Fig. 4G) and PGD₂ (r² = 0.27, P = 0.04; Fig. 4H) following exercise challenge were associated with more severe EIB. There was no significant correlation between pre-exercise PGE₂ levels and direct AHR or postexercise PGE₂ levels or the change in PGE₂ levels with exercise and endogenous AHR (Fig. 4, C, F, and I).

Figure 4.

Levels of eicosanoids in induced sputum correlated with different aspects of airway hyperresponsiveness (AHR) (total n = 16; EIB⁺ n = 8, EIB⁻ n = 8). A–C: pre-exercise levels of CysLTs and PGD₂ inversely correlated with the methacholine PC₂₀ (i.e., direct AHR), but there was no correlation with pre-exercise PGE₂ levels. D–-F: postexercise levels of CysLTs, PGD₂, and to a lesser degree PGE₂ positively correlated with EIB severity (determined by AUC30). G–I: the absolute difference between pre- and postexercise levels of CysLTs and PGD₂ positively correlates with EIB severity (determined by AUC30), whereas there was less of an effect for the change in PGE₂ levels. A linear regression line is shown with the 95% confidence interval as dotted lines. CysLT, cysteinyl leukotriene; AHR, airway hyperresponsiveness; AUC30, area under the curve of the percent change in FEV₁ over the 30 minutes post-exercise challenge test; EIB, exercise-induced bronchoconstriction; PGD₂, prostaglandin D₂; PGE2, prostaglandin E₂.

Overall, these findings indicate that basal levels of PGD₂ and CysLTs are associated with the severity of direct AHR, whereas the increased eicosanoid production with exercise challenge and higher levels of PGD₂ and CysLTs in the airways following exercise correlates with more severe EIB.

Baseline Levels of sPLA₂-X in Induced Sputum Are Associated with Severity of EIB

The levels of sPLA₂-X were measured in induced sputum samples at baseline and postexercise using a time-resolved fluoroimmunoassay. EIB⁺ subjects tended to have higher baseline levels of sPLA₂-X in induced sputum compared to EIB⁻ subjects (P = 0.19; Fig. 5A). There was no significant difference between pre- and postexercise levels of sPLA₂-X (median values 0.62 vs. 1.06 ng/mL, P = 0.67; Fig. 5B); however, there was an association between baseline levels of sPLA₂-X and endogenous AHR with higher baseline levels of sPLA₂-X correlating with more severe EIB (r² = 0.38, P = 0.009; Fig. 5C). Similarly, the absolute change in sPLA₂-X levels with exercise challenge was positively correlated with EIB severity (r² = 0.36, P = 0.01; Fig. 5D). Although we did not observe significant overall changes in measured levels of sPLA₂-X in the airways with exercise in our study population, these findings suggest that subjects with higher levels of sPLA₂-X at baseline and those who develop increases in airway sPLA₂-X levels with exercise challenge experience more severe EIB.

Figure 5.

Levels of sPLA₂-X in induced sputum in relation to severity of EIB and in response to exercise challenge (total n = 17; EIB⁺ n = 9, EIB⁻ n = 8). A: there was a trend toward increased levels of sPLA₂-X in the induced sputum of EIB⁺ subjects as measured by time-resolved fluorescence immunoassay (TRFIA) (bar graphs represent mean values, and error bars represent the standard error of the mean; P value represents the result of an unpaired t test). B: there was no significant change in the level of sPLA₂-X in induced sputum with exercise challenge (solid bars indicate median values of pre- and postexercise sPLA₂-X levels as measured by TRFIA). Baseline induced sputum levels of sPLA₂-X (C) and the absolute change in sPLA₂-X with exercise challenge (D) positively correlated with EIB severity (determined by AUC30). A linear regression line is shown with the 95% confidence interval as dotted lines. AUC30, area under the forced expiratory volume in 1 s curve; EIB, exercise-induced bronchoconstriction; TRFIA, time-resolved fluorescence immunoassay.

Osmotic Stress Causes the Apical Secretion of sPLA₂-X Protein in an Organotypic Epithelial Cell Culture Model

We examined the levels of the sPLA₂-X protein in primary cultures of epithelial cells collected during research bronchoscopy from three EIB⁺ individuals with asthma. Epithelial cells were differentiated in organotypic culture, and the accumulation of sPLA₂-X was assessed over 24 h. Since organotypic epithelial cell cultures have a minimal amount of fluid at the apical surface but have a large volume of media in the basolateral compartment, we added 200 µL of isotonic fluid to the apical compartment in order to measure protein secretion. Although we reduced the basolateral fluid volume, we found that the level of sPLA₂-X was below the level of detection in the basolateral media. However, we have previously demonstrated significant accumulation of sPLA₂-X protein in the apical fluid in comparison with the cell lysates (15). To simulate the osmotic stress that accompanies water loss from the epithelial surface during a period of hyperventilation, we stimulated epithelial cells in organotypic culture with varying levels of sorbitol to increase the osmolarity of the apical fluid over a period of 24 h. We found that the addition of sorbitol caused the rapid movement of water from the basolateral to the apical surface such that the volume of fluid in the apical compartment significantly increased from 150 µL to a maximum of 355 µL at an apical concentration of sorbitol of 500 mOsm (P = 0.0004), whereas there was a decrease in the volume of the basolateral fluid from 960 µL to 780 µL (P < 0.0001). The application of osmotic stress to the apical fluid resulted in a significant increase in the total amount of sPLA₂-X in the apical compartment taking into account the dilution factor (P = 0.003, Fig. 6A). The figure shows the response to osmotic stress of epithelial cells from three different subjects within the cohort to demonstrate the change in sPLA₂-X levels with greater clarity given the baseline differences in secretion from these different donors. The increasing levels of osmotic stress also generally corresponded to a reduction in the amount of sPLA₂-X protein in the epithelial cells further indicating secretion in response to osmotic stress (Fig. 6B). The largest individual change in sPLA₂-X was noted at the sorbitol concentration of 250 mOsm (P = 0.05), and a test for linear trend suggested a reduction in sPLA₂-X with increasing concentrations of sorbitol (P = 0.06). The results indicate that sPLA₂-X can be secreted by polarized epithelial cells in the apical direction in response to osmotic stress.

Figure 6.

Effects of osmotic stress on the secretion of fluid and sPLA₂-X to the apical surface in an epithelial organotypic air-liquid interface (ALI) cell culture model. A and B: there was an increase in the total amount of sPLA₂-X in the apical compartment and a decrease in the amount of sPLA₂-X in epithelial cell lysates with application of osmotic stress. Connected data points represent the response to osmotic stress of epithelial cells from 3 different EIB⁺ asthmatic individuals (P value represents the results of a one-way ANOVA model). ALI, air-liquid interface; EIB, exercise-induced bronchoconstriction.

DISCUSSION

The mechanisms responsible for driving endogenous AHR in the form of EIB are not fully understood, and we aimed to better characterize the role of phospholipid hydrolysis, surfactant composition and function, and dysregulated eicosanoid synthesis in EIB by collecting airway samples before and after exercise challenge in a group of individuals with asthma and varying levels of direct and endogenous AHR. We discovered that exercise challenge resulted in a decline in the LA/SA ratio of pulmonary surfactant in sputum and BAL as well as increased ex vivo PG hydrolysis in BAL samples indicating that exercise challenge promotes phospholipid hydrolysis and surfactant degradation as well as increased airway phospholipase activity. We also identified that EIB⁺ subjects had a reduced LA/SA ratio at baseline in comparison with EIB⁻ subjects suggesting increased basal phospholipid hydrolysis and surfactant degradation in this population. Surfactant has a large phospholipid component that contributes to its function in maintaining airway homeostasis, and previous studies have demonstrated that bronchoconstriction following dry air or exercise challenge in the conducting airways results in increased peripheral airway resistance due to heterogeneous airway narrowing and regional airway closure (32–34), which are consistent with the potential effects on airway patency directly attributable to surfactant function (11–13). These baseline differences in surfactant degradation appear to be critical for the development of bronchoconstriction in the face of further phospholipid hydrolysis induced in response to exercise challenge.

We also examined levels of specific key eicosanoids in induced sputum samples and demonstrated that exercise challenge increased levels of PGD₂, PGE₂, and CysLTs. Additionally, we discovered associations between higher postexercise levels of PGD₂ and CysLTs and the propensity to develop EIB. This is consistent with prior studies which have shown that CysLTs and PGD₂ are released following exercise challenge in subjects with EIB (9, 10). We extend these results here by analyzing the release of these mediators in subjects with a definite diagnosis of asthma but without EIB, demonstrating that CysLTs and PGD₂ levels at baseline are associated with the severity of AHR to an exogenous bronchoconstrictor (methacholine), while both the postexercise values of CysLTs and PGD₂ as well as the absolute change in CysLTs and PGD₂ with exercise are associated with the propensity to develop EIB. These results provide new insights into the underlying immunopathology leading to AHR in general and the specific mediators that contribute to endogenous AHR. In particular, mast cells are the key source of PGD₂ in the airways (35–37). Given that endogenous AHR is specific for asthma, these results also further refine the basis of mediator release in asthma and how this system is activated in response to exercise challenge.

Lastly, we explored the role of sPLA₂-X in driving both surfactant degradation and dysregulated eicosanoid synthesis. Consistent with prior work, we found that the levels of sPLA₂-X were associated with the level of endogenous AHR, implicating this enzyme as a potential etiology for the elevated basal levels of phospholipid turnover and eicosanoid synthesis present in subjects with EIB (15, 30). In fact, elevated basal levels of this enzyme are associated with the severity of both direct and endogenous AHR (38). Although there was a minimal overall increase in sPLA₂-X after exercise challenge in the present study, this lack of a significant change was due in part to higher basal levels of the enzyme in those with EIB, and the change with exercise challenge was associated with the severity of EIB. We acknowledge that our findings could reflect the absence of sPLA₂-X release in response to exercise challenge, but also wanted to explore the role of additional factors such as changes in the dilution factor or clearance and/or degradation of the enzyme. Models of the effects of ventilation indicate that, in the absence of water movement to the apical surface, the majority of the water in the airway surface liquid would evaporate into the exhaled air (39). Mucociliary clearance is also initially decreased and then enhanced after exercise, indicating dynamic changes in the airway surface liquid (40). These considerations motivated the development of an osmotic stress model of a high ventilation stimulus in primary epithelial cells from EIB⁺ asthmatics, revealing that there were both a marked increase in the water movement to the apical surface and a total increase in the amount of sPLA₂-X with osmotic stress. It is also possible that the elevated levels of both sPLA₂-X and phospholipid turnover at baseline are sufficient for the generation of peak eicosanoid production in the setting of a second signal that further facilitates eicosanoid synthesis, such as the activation of leukocytes that lead to the terminal synthesis of eicosanoids relevant to EIB and endogenous AHR. We have previously shown that osmotic stress leads to the generation of IL-33, which can serve as a trigger for eicosanoid synthesis by innate cells (15, 41), and others have demonstrated that osmotic stress triggers release of adenosine triphosphate that can promote bronchoconstriction by triggering CysLT release by mast cells (42).

The strengths of this study are in the careful measurements that were made both before and after dry air exercise challenge in a group of subjects that were carefully characterized for features of AHR and were not using controller therapies. Using this unique design, our study revealed new knowledge about the environment that is induced in the airways following dry air exercise challenge and its relationship to the severity of AHR. Previous studies of exercise challenge in subjects with EIB have relied on indirect measurements of lipid mediators within the airway such as exhaled breath condensates or urinary metabolites or more direct assessment via induced sputum. To our knowledge, this is the first report of research bronchoscopy in subjects with EIB immediately following exercise challenge and first description of phospholipid composition and phospholipase activity from BAL fluid in this patient population. The primary weakness of this study is in the modest sample size, resulting in marginal statistical power to achieve statistical significance for some of the secondary measurements obtained in this study given the inherent variability of human samples. Although we show clearly that exercise challenge leads to phospholipid hydrolysis and that this leads to features of surfactant degradation and eicosanoid production, the identity of the enzymes responsible for this activity is not entirely clear from the present study. The results suggest that other sPLA₂s and/or mechanisms leading to phospholipid hydrolysis may also be implicated in addition to the enzyme sPLA₂-X that we focused on in these experiments. Ultimately, these results move the field forward with a better understanding of the events that occur following exercise challenge, but also raise additional questions about the factors that serve as intermediates in the activation of this system that leads to phospholipid cleavage and lipid mediator generation.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author upon reasonable request.

SUPPLEMENTAL DATA

Supplemental Fig. S1 and Methods: https://doi.org/10.6084/m9.figshare.13428419.v1.

GRANTS

This work was supported by NIH National Heart, Lung, and Blood Institute Grant R01HL089215 and National Institute of Allergy and Infectious Diseases Grant K24AI130263 to T.S.H.

DISCLOSURES

A. Chakrabarti reports his current employment at Genentech. M. C. Seeds reports current funding by Biomedical Advanced Research and Development Authority (BARDA) and by Wake Forest University School of Medicine but no conflicts of interest regarding this work. W. A. Altemeier reports grants from NIH/National Institute of Allergy and Infectious Diseases/National Heart, Lung, and Blood Institute (NIAID/NHLBI) and personal fees from NIH. M. H. Gelb reports consulting fees for PerkinElmer Genetics. R. D. Hite reports grants from NIH/NHLBI, ALung Technologies, and Lungpacer Medical, and personal fees from NIH. T. S. Hallstrand reports grants from NIH/NIAID/NHLBI and personal fees from NIH. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

M.C.S., W.A.A, M.H.G., R.D.H., and T.S.H. conceived and designed research; Y.L., J.D.N., R.A.A.P., A.C., M.V.N., M.C.S., R.D.H., and T.S.H. performed experiments; R.C.M., Y.L., J.D.N., R.A.A.P., A.C., M.V.N., M.C.S., W.A., M.H.G., R.D.H., and T.S.H. analyzed data; R.C.M., Y.L., J.D.N., R.A.A.P., A.C., M.V.N., M.C.S., W.A.A., M.H.G., R.D.H., and T.S.H. interpreted results of experiments; R.C.M. prepared figures; R.C.M. and T.S.H. drafted manuscript; R.C.M., M.C.S., W.A.A., R.D.H., and T.S.H. edited and revised manuscript; R.C.M., Y.L., J.D.N., R.A.A.P., A.C., M.V.N., M.C.S., W.A.A., M.H.G., R.D.H., and T.S.H. approved final version of manuscript.