15LO1 dictates glutathione redox changes in asthmatic airway epithelium to worsen type 2 inflammation

Abstract

Altered redox biology challenges all cells, with compensatory responses often determining a cell's fate. When 15 lipoxygenase 1 (15LO1), a lipid-peroxidizing enzyme abundant in asthmatic human airway epithelial cells (HAECs), binds phosphatidylethanolamine-binding protein 1 (PEBP1), hydroperoxy-phospholipids, which drive ferroptotic cell death, are generated. Peroxidases, including glutathione peroxidase 4 (GPX4), metabolize hydroperoxy-phospholipids to hydroxy derivatives to prevent ferroptotic death, but consume reduced glutathione (GSH). The cystine transporter SLC7A11 critically restores/maintains intracellular GSH. We hypothesized that high 15LO1, PEBP1, and GPX4 activity drives abnormal asthmatic redox biology, evidenced by lower bronchoalveolar lavage (BAL) fluid and intraepithelial cell GSH:oxidized GSH (GSSG) ratios, to enhance type 2 (T2) inflammatory responses. GSH, GSSG (enzymatic assays), 15LO1, GPX4, SLC7A11, and T2 biomarkers (Western blot and RNA-Seq) were measured in asthmatic and healthy control (HC) cells and fluids, with siRNA knockdown as appropriate. GSSG was higher and GSH:GSSG lower in asthmatic compared with HC BAL fluid, while intracellular GSH was lower in asthma. In vitro, a T2 cytokine (IL-13) induced 15LO1 generation of hydroperoxy-phospholipids, which lowered intracellular GSH and increased extracellular GSSG. Lowering GSH further by inhibiting SLC7A11 enhanced T2 inflammatory protein expression and ferroptosis. Ex vivo, redox imbalances corresponded to 15LO1 and SLC7A11 expression, T2 biomarkers, and worsened clinical outcomes. Thus, 15LO1 pathway-induced redox biology perturbations worsen T2 inflammation and asthma control, supporting 15LO1 as a therapeutic target.

Keywords: Asthma; Metabolism; Pulmonology; Radicals; Th2 response.

Figure 1. Abnormal redox balance, as measured by higher BAL fluid oxidized glutathione (GSSG) and lower intraepithelial cell GSH, is observed in asthmatic BAL fluid, as compared with healthy controls.

GSH, GSSG, and the GSH:GSSG ratio were measured by enzymatic assay in bronchoalveolar lavage (BAL) fluid and in fresh epithelial cells from healthy controls (HCs), mild/moderate (M/M), and severe asthma (SA) patients. (A) Higher levels of GSSG and lower GSH:GSSG ratios were observed in asthmatic BAL fluid. (B) Intracellular GSH levels and GSH:GSSG were lower in fresh asthmatic epithelial cells compared with HCs. (C) BAL fluid (n = 98) and intracellular GSH:GSSG (n = 26) positively correlated with FeNO. ANOVA with intergroup comparisons by t test was used for group comparisons. Pearson’s correlations were used for GSH:GSSG versus FeNO. Bonferroni-corrected significant P values for 3 groups was set at 0.0166. All data were log transformed prior to analysis and were converted back to the linear scale for presentation. Bars represent mean and error bars represent SEM. *P < 0.01, **P < 0.001, ***P < 0.0001.

Figure 2. IL-13 lowers alters the intra- and extracellular oxidative potential by decreasing intracellular GSH and GSH:GSSG and increasing extracellular GSSG.

(A) IL-13 decreases intracellular GSH levels and lowers the GSH:GSSG ratio (n = 6). (B) IL-13 also decreases extracellular GSH levels but increases extracellular GSSG, lowering GSH:GSSG (n = 6). (C) IL-13 (10 ng/mL) increases expression of 15LO1, GPX4, and SLC7A11 by Western blot, with GPX4 and SLC7A11 increases quantified by densitometry (n = 4–12). Matched-pair analysis of log-transformed data was used to compare conditions.

Figure 3. 15LO1 expression and activity regulate intracellular redox balance by lowering GSH and GSH:GSSG.

(A) Under IL-13–stimulated conditions, ALOX15 (15LO1) knockdown (KD) increases intracellular GSH and the GSH:GSSG ratio toward control conditions (n = 6). (B) Under IL-13 conditions, inhibition of 15LO1 by the specific 15LO1 inhibitor BLX2477 also restores intracellular GSH and GSH:GSSG (n = 6). (C) ALOX15 KD lowers extracellular GSSG and restores a normal GSH:GSSG ratio (n = 6). (D) Representative Western blots demonstrating that ALOX15 KD also decreases GPX4 protein expression (n = 4). (E) Representative Western blots show decreased expression of GPX4 under conditions of 15LO1 inhibition (n = 4). (F) Ex vivo 15LO1 protein expression correlates with intraepithelial cell GSH and GSH:GSSG (n = 15). (G) Indexing SLC7A11, as a critical regulator of intracellular GSH levels, to 15LO1 improves the relationship of 15LO1 to intracellular GSH (n = 12). Matched-pair analysis (A–C) of log-transformed data was used to compare conditions, while Pearson’s correlation was used for F and G.

Figure 4. Activity of the cystine-glutamate antiporter SLC7A11 modulates intracellular redox, cell death, and type 2 signature protein expression.

(A) Inhibition of SLC7A11 activity by erastin lowers intracellular GSH and the GSH:GSSG ratio (n = 6). (B) This decrease in GSH:GSSG is accompanied by increased expression of 15LO1 (n = 4). (C) The decreased GSH:GSSG increases LDH release, consistent with activation of oxidatively driven cell death (n = 6). (D) Representative Western blot of intracellular and secreted type 2 signature markers following SLC7A11 chemical inhibition with erastin for 24 hours (n = 3–4). (E) Densitometry results showing increases in intracellular periostin (POSTN), CCL26, and iNOS following treatment with erastin (n = 3–4). (F) Densitometry results showing parallel fold-change increases in basilar secretion of POSTN and CCL26, with increased apical secretion of MUC5AC protein following treatment with erastin (n = 4). Matched-pair analysis following log transformation was used to compare the conditions.

Figure 5. Ex vivo extra- and intracellular redox balance is associated with clinical outcomes relevant to asthma.

(A) Extracellular BAL fluid GSH:GSSG ratio correlates with lung function (n = 98), as measured by predicted FEV₁% (n = 75) and associates with asthma exacerbations. (B) Intracellular GSH:GSSG correlates with predicted FEV₁% (n = 26), while in small numbers the association with baseline exacerbations (n = 15) is not significant. Pearson’s correlation testing was performed for comparisons of predicted FEV₁% with GSH:GSSG, while t tests were used for group comparisons. Data were log transformed except for predicted FEV₁%, which was normally distributed. Bars represent mean and error bars represent SEM.