Abstract
Asthma is a complex disease comprising various phenotypes and endotypes, all of which still need solid biomarkers for accurate classification. In a previous study, we defined specific genes related to asthma and respiratory allergy by studying the expression of 94 genes in a population composed of 4 groups of subjects: healthy control, nonallergic asthmatic, asthmatic allergic, and nonasthmatic allergic patients. An analysis of differential gene expression between controls and patients revealed a set of statistically relevant genes mainly associated with disease severity, i.e., CHI3L1, IL-8, IL-10, MSR1, PHLDA1, PI3, and SERPINB2. Here, we analyzed whether these genes and their proteins could be potential asthma biomarkers to distinguish between nonallergic asthmatic and asthmatic allergic subjects. Protein quantification was determined by ELISA (in serum) or Western blot (in protein extracted from peripheral blood mononuclear cells or PBMCs). Statistical analyses were performed by unpaired t-test using the Graph-Pad program. The sensitivity and specificity of the gene and protein expression of several candidate biomarkers in differentiating the two groups (and the severity subgroups) was performed by receiver operating characteristic (ROC) curve analysis using the R program. The ROC curve analysis determined single genes with good sensitivity and specificity for discriminating some of the phenotypes. However, interesting combinations of two or three protein biomarkers were found to distinguish the asthma disease and disease severity between the different phenotypes of this pathology using reproducible techniques in easy-to-obtain samples. Gene and protein panels formed by single biomarkers and biomarker combinations have been defined in easily obtainable samples and by standardized techniques. These panels could be useful for characterizing phenotypes of asthma, specifically when differentiating asthma severity.