Genetic analysis in mice reveals polygenic basis and candidate transcriptional regulators of airway remodeling features

AUTHORS: Lauren Donoghue1,2, Kathryn McFadden1, Greg Keele1, and Samir Kelada1,2,3

1Department of Genetics, University of North Carolina at Chapel Hill
2Curriculum in Genetics and Molecular Biology, University of North Carolina at Chapel Hill
3Marisco Lung Institute, University of North Carolina at Chapel Hill

BACKGROUND: Airway wall remodeling in individuals with asthma is thought to contribute to hallmark disease features and is associated with disease severity. Identifying the biological drivers of remodeling features and potential therapeutic targets remains challenging given the invasive nature of characterizing remodeling in patients. Numerous studies have shown that genetic variation plays a significant role in asthma risk, but the extent to which genetics may regulate airway remodeling has not been thoroughly investigated. We used systems genetics approaches in a population of genetically diverse mice chronically exposed to allergen to identify the genetic and transcriptomic underpinnings of variation in remodeling features.

METHODS: The Collaborative Cross (CC) is a panel of recombinant inbred mouse strains whose genomes vary by over 45 million genetic variants. We chronically exposed BALB/cJ mice and 31 CC strains to saline or house dust mite allergen (25ug intra-nasally, 3x/week for 5 weeks) to induce airway inflammation and remodeling. We performed morphometric analysis of mucous cell metaplasia, subepithelial fibrosis, and smooth muscle thickening in addition to profiling inflammatory cells and mucin proteins in lavage fluid. For each phenotype, we estimated the contribution of genetic variation to overall phenotypic variation (heritability). Select strains were assessed for airway hyperresponsiveness. We performed RNAseq on airway tissue and integrative gene expression analyses including co-expression, eQTL mapping, and differential expression to identify phenotype-associated gene networks and candidate transcriptional regulators.

RESULTS: The majority of CC strains exhibited remodeling and inflammation either more or less extreme than the oft-used BALB/cJ strain. Across CC strains, phenotypic outcomes were continuously distributed, indicative of polygenic architectures for each trait. Pairwise correlations between remodeling and inflammatory features varied from non-significant to r=0.37, suggesting remodeling is driven by both trait-shared and trait-specific molecular mechanisms. Through gene expression analyses we identified gene co-expression modules significantly associated with one or more remodeling traits and enriched for disease-associated pathways such as endoplasmic reticulum stress and TGFb-1 signaling. Genes highly-interconnected in modules included those differentially expressed in human asthma, near GWAS loci for asthma or lung function, and/or with unknown roles in disease. We identified candidate transcriptional regulators for subepithelial collagen deposition from co-expression modules enriched for expression in epithelial cells, fibroblasts, and plasma cells.

CONCLUSIONS: Our results highlight that genetic variation significantly influences features of airway remodeling in the context of chronic allergen exposure in mice. Furthermore, remodeling traits have shared and distinct mechanisms of regulation that can be associated with gene expression programs containing genes with unknown functional roles. CC strains identified as having more extreme responses than BALB/cJ mice represent new models for the investigation of airway remodeling or severe asthma.

Lauren Donoghue BS

University of North Carolina

Appearances