Previous studies have identified clinical or inflammatory phenotypes of asthma on the basis of clinical and demographic parameters or sputum cell counts; however, few studies have defined transcriptional phenotypes of asthma.

To investigate asthma phenotypes at a transcriptional level by using gene expression profiling of induced sputum.

Induced sputum samples were collected from 59 people with asthma with a mean age of 58 years and an FEV₁% predicted of 76%, and 69% were taking inhaled corticosteroids. Thirteen healthy controls without asthma were also assessed. Inflammatory cell counts were performed, and RNA was extracted from selected sputum. Transcriptional profiles were generated (Illumina Humanref-8 V2) and analyzed by using GeneSpring GX11.

Unsupervised hierarchical clustering of gene expression profiles in asthma revealed 3 distinct groups. The first transcriptional phenotype (n = 21) had lower FEV₁% predicted and higher asthma control questionnaire scores, exhaled nitric oxide, and sputum eosinophils. The second transcriptional phenotype (n = 14) had lower FEV₁% predicted and forced vital capacity % predicted and higher sputum neutrophils compared with a third transcriptional phenotype (n = 24) that had higher sputum macrophages and resembled healthy controls. Differentially expressed genes between transcriptional asthma phenotypes were related to inflammatory and immune responses. Genes in the IL-1 and TNF-⍺/nuclear factor-κB pathways were overexpressed and correlated with clinical parameters and neutrophilic airway inflammation.

Gene expression profiling provides a novel means to investigate the molecular mechanisms and classifications of asthma phenotypes. There are 3 distinct transcriptional phenotypes of asthma that relate to both clinical and inflammatory parameters.