Identification of genes expressed by human airway eosinophils after an in vivo allergen challenge

Abstract

Background: The mechanism for the contribution of eosinophils (EOS) to asthma pathophysiology is not fully understood. Genome-wide expression analysis of airway EOS by microarrays has been limited by the ability to generate high quality RNA from sufficient numbers of airway EOS.

Objective: To identify, by genome-wide expression analyses, a compendium of expressed genes characteristic of airway EOS following an in vivo allergen challenge.

Methods: Atopic, mild asthmatic subjects were recruited for these studies. Induced sputum was obtained before and 48h after a whole lung allergen challenge (WLAC). Individuals also received a segmental bronchoprovocation with allergen (SBP-Ag) 1 month before and after administering a single dose of mepolizumab (anti-IL-5 monoclonal antibody) to reduce airway EOS. Bronchoalveolar lavage (BAL) was performed before and 48 h after SBP-Ag. Gene expression of sputum and BAL cells was analyzed by microarrays. The results were validated by qPCR in BAL cells and purified BAL EOS.

Results: A total of 299 transcripts were up-regulated by more than 2-fold in total BAL cells following SBP-Ag. Mepolizumab treatment resulted in a reduction of airway EOS by 54.5% and decreased expression of 99 of the 299 transcripts. 3 of 6 post-WLAC sputum samples showed increased expression of EOS-specific genes, along with the expression of 361 other genes. Finally, the intersection of the 3 groups of transcripts (increased in BAL post SBP-Ag (299), decreased after mepolizumab (99), and increased in sputum after WLAC (365)) was composed of 57 genes characterizing airway EOS gene expression.

Conclusion: We identified 57 genes that were highly expressed by BAL EOS compared to unseparated BAL cells after in vivo allergen challenge. 41 of these genes had not been previously described in EOS and are thus potential new candidates to elucidate EOS contribution to airway biology.

Figure 1. Bronchoalveolar lavage and sputum timelines.

Atopic mild asthmatics underwent bronchoalveolar lavage (BAL) followed by segmental challenge with an allergen (SBP-Ag) on visit 1 (BAL V1). 48 h later on visit 2 (BAL V2), BAL was performed at the site of the challenged segment. One month later, a single dose of mepolizumab was administered. One month after dosing, BALs were repeated both before and after SBP-Ag. BAL cells (BALC) were prepared on visits 1, 2, 3, and 4 and airway EOS (BALEOS) were purified on visit 2. For the induced sputum study, sputum collection was performed followed by a whole lung allergen challenge (WLAC) on visit 1 (sputum V1). Collection of a second induced sputum was done 48 h after WLAC on visit 2 (sputum V2).

Figure 2. Heat map generated using Cluster 3.0 and presenting the 299 genes up-regulated by more than 2-fold in BAL cells from both subjects #1 and #2, 48 h after SBP-Ag (BAL V2 as shown in Figure 1) compared to before SBP-Ag (V1).

Eosinophil-markers (IL5RA, RNASE2, RNASE3 and SIGLEC8) are highlighted. The full list of gene names is presented on Table S2.

Figure 3. Venn diagram identifying 57 genes highly associated with airway EOS after an allergen challenge.

57 genes defined at the intersection of genes up-regulated in BAL cells after SPB-Ag, down-regulated when SBP-Ag was performed following mepolizumab, and part of the EOS-associated genes in the sputum following WLAC.

Figure 4. Validation by qPCR of the 57 genes associated with airway EOS after airway allergen challenge.

qPCR was performed on BAL cells before SBP-Ag (BAL V1 as shown on Figure 1), 48 h after allergen with (V4) or without mepolizumab treatment (V2), and on purified BAL EOS (BALEOS) after allergen challenge (no mepolizumab, V2). Fold changes compared to expression in BAL cells before allergen challenge and mepolizumab (V1) are presented. Box plots depict the median and the interquartile range between the 25^th and the 75^th percentiles. 6 subjects are included in each group and means are presented in parentheses. ** indicates that mRNA level in BAL cells is significantly decreased by mepolizumab compared to BAL cells on V2 (p<0.01). # indicates that mRNA level in BAL EOS at V2 is significantly higher or lower compared to total BAL cells on V2 (^#, p<0.05; ^##, P<0.01: ^###, P<0.001).