Identification of novel biomarkers in seasonal allergic rhinitis by combining proteomic, multivariate and pathway analysis

Abstract

Background: Glucocorticoids (GCs) play a key role in the treatment of seasonal allergic rhinitis (SAR). However, some patients show a low response to GC treatment. We hypothesized that proteins that correlated to discrimination between symptomatic high and low responders (HR and LR) to GC treatment might be regulated by GCs and therefore suitable as biomarkers for GC treatment.

Methodology/principal findings: We identified 953 nasal fluid proteins in symptomatic HR and LR with a LC MS/MS based-quantitative proteomics analysis and performed multivariate analysis to identify a combination of proteins that best separated symptomatic HR and LR. Pathway analysis showed that those proteins were most enriched in the acute phase response pathway. We prioritized candidate biomarkers for GC treatment based on the multivariate and pathway analysis. Next, we tested if those candidate biomarkers differed before and after GC treatment in nasal fluids from 40 patients with SAR using ELISA. Several proteins including ORM (P<0.0001), APOH (P<0.0001), FGA (P<0.01), CTSD (P<0.05) and SERPINB3 (P<0.05) differed significantly before and after GC treatment. Particularly, ORM (P<0.01), FGA (P<0.05) and APOH (P<0.01) that belonged to the acute phase response pathway decreased significantly in HR but not LR before and after GC treatment.

Conclusions/significance: We identified several novel biomarkers for GC treatment response in SAR with combined proteomics, multivariate and pathway analysis. The analytical principles may be generally applicable to identify biomarkers in clinical studies of complex diseases.

Figure 1. Pathways enriched with nasal fluid proteins identified in both HR and LR to treatment with Glucocorticoids (GCs).

A total of 953 unique proteins were identified in nasal fluids from both HR and LR and were mapped onto canonical pathways using the IPA software. The yellow threshold indicates 95% confidence.

Figure 2. Differences in nasal fluid protein profiles between symptomatic HR and LR.

OPLS-DA modelling was performed with 161 nasal fluid proteins identified in no less than 50% HR and LR. A) OPLS-DA score plot showed partial separation between HR and LR, where t represents the predictive component. All samples were within a ±2 standard deviation (SD) limit (according to Hotelling's T²). B) OPLS-DA loading plot with confidence intervals (according to the cross validation procedure). The top 40 proteins that contributed to separation between HR and LR were highlighted in red. The black line represents error bar. C) Pathway analysis with the top 40 proteins using the IPA software. The yellow threshold indicates 95% confidence.

Figure 3. Identification of candidate biomarkers with ELISA.

A) Nasal fluids from 40 patients with SAR before and after GC treatment were analyzed. B) Nasal fluids from 10 HR. Pre, patients before treatment with GCs; Post, patients after treatment with GCs.