Distinct Transcriptome Networks Define Early Life Longitudinal Wheezing Phenotypes
Recommended Citation
Phelan K, Roskin KM, Burkle J, Chang W, Martin LJ, Satish L, Haslam D, Spagna D, Parmar E, Bacharier LB, Gebretsadik T, Gold DR, Jackson DJ, Cole Johnson C, Lynch S, Mccauley K, McKennan C., Miller RL, Ober C, Ownby DR, Ryan P, Schoettler N, Singh S, Visness C, Altman MC, Gern JE, Hershey GK. Distinct Transcriptome Networks Define Early Life Longitudinal Wheezing Phenotypes. Am J Respir Crit Care Med 2023; 207(1).
Document Type
Conference Proceeding
Publication Date
5-22-2023
Publication Title
Am J Respir Crit Care Med
Abstract
RATIONALE: Asthma is a complex and heterogenous disease most commonly with childhood onset. Previous studies from birth cohorts have identified temporal patterns of wheezing which have differential associations with asthma development. While multiple studies have replicated these phenotypes, the transcriptional patterns which regulate them have not yet been examined. In this study, we derived longitudinal wheezing phenotypes for participants in 7 pediatric cohorts from the ECHO-CREW consortium and leveraged matched nasal lavage specimens to investigate the underlying patterns of transcription and the biologic pathways defining each phenotype. METHODS: Participants (n=961) in 7 CREW birth cohorts had nasal lavages collected and RNA sequenced. Latent class analysis was used to derive 4 longitudinal wheezing phenotypes (infrequent, transient, late-onset, persistent) for each participant. After requiring samples to have ≥33% of read-pairs map to annotated regions of the genome, ≥600k aligned read-pairs, and cell-count differentials, 709 samples (1 sample per participant) also had a derived wheezing phenotype. Stepwise regression analysis determined white blood cell count, sequencing run, collection site, and epithelial cell count contributed significantly to the normalized data and were adjusted for in the models. We used Limma and weighted gene correlation network analysis to derive gene co-expression modules and assess average expression differences between wheezing phenotypes. STRING and Cytoscape were used to generate networks of known protein interactions in each module. RESULTS: We identified 11 co-expression modules which significantly differed in expression between the wheezing phenotypes. Those with persistent wheeze displayed an increase in a co-expression module enriched for mast cell activation and T2 immunity, consistent with pathways enriched in those with asthma. Those with transient wheeze exhibited the largest increase in a module enriched for infection response pathways, suggesting infection is an important cause of early childhood wheeze. Those with late-onset wheeze had the largest increase in a module related to epithelial development, indicating a regenerative response to barrier injury. CONCLUSIONS: Our data reveal distinct early life longitudinal wheeze phenotypes with different pathologic origins. Specific patterns of nasal epithelial gene expression demonstrate the unique biologic pathways regulating these phenotypes and identify novel therapeutic targets for the prevention and treatment of wheezing in early life.
Volume
207
Issue
1
