Transcriptomic Profiles of Well-Differentiated Airway Epithelial Cells in Response to Environmental Triggers of Asthma ExacerbationAntonella Marrocco1, Jennifer A. Mitchel1, Margaret Parker2, Maureen McGill1, Robert P. Chase2, Scott T. Weiss2, Diane R. Gold1,2, Peter J. Castaldi2, Jin-Ah Park1, Joanne Sordillo34Department of Environmental Health, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA2Channing Division of Network Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA.3Division of Chronic Disease Research Across the Lifecourse (CoRAL), Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, Boston, MA, USA.Corresponding Author: Joanne E. SordilloEmail : [email protected]:Landmark Center West 401 Park Dr., 4th Floor Boston, MA 02115The authors have no conflicts of interest to disclose.Funding: R01HL148152, P30ES000002, T32HL007118, Francis Family FoundationKeywords: Bronchial epithelial cells, gene expression, RNAseq, asthma

Direct cleavage and activation of gasdermin B by asthma trigger allergensTo the Editor:Recent fine-mapping studies have pointed to gasdermn B (GSDMB ) as a potential asthma susceptibility gene in 17q21 locus, the strongest and most highly replicated signal in genome-wide association studies1. The GSDMB protein is a member of the gasdermin family that, when cleaved, triggers an inflammatory cell death known as pyroptosis2. Caspase-1 and granzyme A have been shown to cut GSDMB at specific sites to release the N-terminal fragment of the protein (GSDMB-NT) that has the ability to induce pyroptosis in cells, including airway epithelial cells3,4. These findings suggest that the role of GSDMB in asthma lies in its ability to be activated through cleavage to induce pyroptosis; however, it remains unclear whether GSDMB cleavage and activation occur in the context of asthma.Common asthma trigger allergens often possess protease activities that cause airway epithelial injury and inflammation5,6. We thus tested whether the allergens directly cleave GSDMB. Incubation of extracts from house dust mite (HDM), a common asthma trigger, with lysates from human bronchial epithelial cells, which express endogenous GSDMB3, resulted in GSDMB cleavage as evidenced by the appearance of a smaller protein around 17kD (Figure 1A). Since the GSDMB antibody used in the Western blotting targets the C-terminus of the protein, the 17kD protein band likely represents the C-terminal GSDMB fragment. Such GSDMB cleavage was also observed when lysates from cells expressing C-terminal-FLAG-tagged GSDMB were mixed with HDM extract (Figure 1B). Furthermore, mold or cockroach extract also cleaved tagged GSDMB (Figure 1C). The cleavage of GSDMB protein by all allergen extracts resulted in a single product of similar size (about 17 kD), suggesting a specific cutting site.To identify the cleavage site, we incubated recombinant full-length GSDMB with HDM extract and resolved the cleaved protein products on SDS-PAGE (Figure 1D). We excised the putative 17 kD C-terminal fragment (GSDMB-CT, Figure 1D) and determined the N-terminal amino acid sequence of the fragment via Edman sequencing (Supplemental Figure S1, Figure 1E). Despite some ambiguities, the first ten amino acid residues of the 17 kD GSDMB-CT largely map to position 245 to 254 (SLGSEDSRNM) of the full length GSDMB protein (Figure 1E). This result indicates that GSDMB was cleaved immediately after the lysine residue at position 244 (K244). Interestingly, granzyme A also cuts GSDMB at the same K244 site4. To confirm K244 as the site of cleavage, we mutated lysine 244 to alanine (K244A) in GSDMB and tested whether the mutant protein can be cleaved by HDM. As shown by Western blotting, HDM was able to cleave wild type (WT) GSDMB but failed to cleave K244A GSDMB as evidenced by the absence of the 17 kD fragment (Figure 1F).The cleavage of GSDMB by HDM is expected to release an N-terminal fragment of 244 amino acids (GSDMB-NT-K244) (Figure 2A). We next tested whether GSDMB-NT-K244 triggers pyroptosis. Transfection of GSDMB-NT-K244 induced cell morphological changes characteristic of pyroptosis, including rounding up and detachment (Figure 2B). LDH release assay confirmed increased toxicity in these cells (~3.4 fold) as compared to cells transfected with the full-length GSDMB (Figure 2C). Consistent with our previous finding on GSDMB-NT shortened by a functional asthma-associated splice variant3, transfection of a truncated GSDMB-NT from the variant (NT-K231var) did not induce pyroptosis (Figure 2B,C).While future studies are needed to identify the specific proteases within the allergen extracts that cleave GSDMB, our current study demonstrates that asthma triggers such as HDM can directly cleave and activate GSDMB, thus providing biochemical evidence linking GSDMB-mediated pyroptosis to asthma.

To the Editor,Folliculin, a protein expressed in various types of cells including airway epithelial cells, encoded by the FLCN gene, is associated with the 5′ AMP-activated protein kinase (AMPK) and mammalian target of rapamycin complex 1 (mTORC1) signaling pathways and, it is thought to alter cell-to-cell adhesion and contribute to the pathogenesis of cystic lung disease in Birt-Hogg-Dubé syndrome (1-5). In addition, the geneFLCN regulates the E-cadherin-LKB1-AMPK axis, which controls lung epithelial cell survival and alveoli size (2). In a recent study, serum folliculin levels were found to be higher in patients with asthma than in healthy control groups and high folliculin levels were associated with increased airway hyperresponsiveness in patients with asthma. In vitro data demonstrated the eosinophil-induced release of folliculin from epithelial cells. These clinical and in vitro observations suggest that folliculin may play some role in the interaction between the eosinophils and airway epithelium (6).To investigate the relationship between clinical characteristics and the level of folliculin in asthmatics, the data of a total of 404 patients with asthma and 94 of controls were enrolled and retrospectively reviewed. To correct for the heavily skewed distributions of the serum folliculin levels, values were log-transformed. Study methods, design and definitions used can be found in the online supplement (Study S1).The proportions of males and smokers were significantly higher among the patients with asthma than in the controls, and the mean serum folliculin level in asthmatics was significantly higher than that in controls (4.80 pg/mL versus 4.13 ng/mL; P < 0.001) (Table S1). As the control group was comprised only of males and a significant difference in smoking history was noted, adjusting for sex and smoking history was performed and significantly higher serum folliculin levels were still observed (P < 0.001). We compared the serum folliculin levels between asthmatics and controls subdivided by sex and smoking status. In these subgroups, the serum folliculin levels were still significantly higher in asthmatics than in the control group (Table S2). ROC curve analysis revealed a significant difference in serum folliculin levels between asthmatics and controls (area under the curve = 0.846, confidence interval [CI] 0.80–0.89, P < 0.001); the optimal cut-off value of serum folliculin level that distinguished asthma patients from controls was 4.31 pg/mL after log-transformation, correlating with 83.91% sensitivity and 77.66% specificity (Figure 1). When we perform ROC curve analysis with only the males, the optimal cut-off value of serum folliculin level was 4.33 pg/mL (Figure S1).We compared folliculin levels among the four groups divided by pre- pre-bronchodilator (BD) predicted FEV1 (%) and found a significant difference in serum folliculin level (P < 0.001, Figure S2). Simple and multiple linear regression analysis was performed to determine the correlation between serum folliculin level and lung function in patients with asthma. In simple linear regression analysis, serum folliculin level were significantly correlated with pre-BD FEV1% predicted (β-coefficient = −4.848, P = 0.013), however significance was only marginal after adjusting for age and sex (β-coefficient = −3.199,P = 0.096) in multiple linear regression analysis. This is because, firstly, there was collinearity of folliculin level and age in our data, and secondly, the rate of smokers (85.22%) among males was higher than among females (14.91%), so it seems that lung function in females is higher.Patients with asthma were divided into two groups using the mean value of the logarithmic serum folliculin levels (4.80 pg/mL). Patients in the high-folliculin group were older at the onset of symptoms, heavier smokers and had a significantly lower lung function. The number of acute exacerbations occurring per year was more frequent in the high folliculin group than in the low folliculin group, but no statistical significance was noted (Table 1). When patients with asthma were divided into the upper quartile of folliculin levels and the lower three quartiles combined, those from the high folliculin group in the upper 25 percentile were found to be older and had lower atopy and lung function than the lower folliculin group with the lower 75 percentile combined. (Table S3). Likewise, we also divided the patients into 4 quantile groups according to serum folliculin levels and identified differences in each group in lung function and age (Table S4).A previous in vitro study showed that human airway epithelial cells (HAECs) exposed to leukotriene E4 and peripheral blood eosinophils released folliculin and interleukin (IL)-8, which resulted in the destruction of the integrity of the epithelial cells. The knockdown of folliculin expression resulted in a decrease in IL-8 release and suppression of epithelial cell activation, which restored the epithelial integrity in HAECs. In their study, folliculin was suggested to be associated with a higher serum transforming growth factor-β1 level, which was associated with worsening of airway inflammation and remodeling (6,7). Consistent with theses result, a higher serum folliculin level in patients with asthma than in healthy controls was also observed in our study, in addition, we showed that an increase in serum folliculin level was associated with a decrease in basal lung function. As folliculin is released from bronchial epithelial cells in response to compressive stress that mimics a bronchospasm (8, 9), we postulate that chronic airway inflammation produces mechanical stress on the airway epithelium, thereby inducing oxidative damage and release of folliculin with changes in the epithelial cell structure. Therefore, we assume that folliculin is associated with the airway inflammation and remodeling pathway in patients with asthma. In our study, serum folliculin level showed no association with serum laboratory variables, suggesting that the increase in folliculin level following mechanical stress is independent of other serum inflammatory markers.In conclusion, our study demonstrates for the first time that serum folliculin concentration is higher in patients with asthma, and it is associated with worse lung function independent to other serum inflammatory markers. Thus, folliculin may represent a novel biomarker related to lower pulmonary function in patients with asthma and further studies are warranted to evaluate the mechanism and test our hypothesis.Keywords: asthma; folliculin; biomarker