Host Biomarkers of Respiratory Infection

Abstract

Background: Respiratory tract infections are responsible for millions of deaths annually. Interferon-stimulated genes (ISGs) play a significant role in fighting off viral respiratory tract infections in the antiviral defence system. Measuring extracellular protein products of ISGs could be potential biomarkers of viral infection. Although, the feasibility and performance of ISGs as functional and robust clinical biomarkers from a non-invasive sample format remains unknown.

Methods: Three ISGs, CXCL10, CXCL11, and TNFSF10, were examined in in-vivo and in-vitro gene expression datasets (RNA-sequencing and microarray) infected with common respiratory tract infections (Rhinovirus, Respiratory syncytial virus, influenza A and SARS-CoV-2) samples and compared to negative controls. Using qualitative selection criteria of 1) elevated presence in at least one dataset with viral infection, 2) secreted protein product, and 3) commercially available antibodies for detection, CXCL10, CXCL11 and TNFSF10 gene expression levels were assessed. A correlation analysis was performed with SARS-CoV-2 infection severity and gene expression kinetics. CXCL10 was subsequently validated at the protein level in saliva as a prerequisite for developing a host-response LFA.

Results: CXCL10 and CXCL11 upregulation were positively correlated with RSV compared to control (p < 0.05). CXCL10/CXCL11/TNFSF10 were not different between samples collected from RV infected subjects relative to controls (p > 0.05). No significant association was found with influenza A for all three genes. CXCL10/CXCL11/TNFSF10 upregulation was positively correlated with SARS-CoV-2 infection compared to control (p < 0.001). CXCL10 expression correlated with COVID-19 viral load. CXCL10 was chosen as a lead biomarker candidate based on these analyses that included different virus infections, time-courses, and measures of severity. CXCL10 was not detected at the protein level in healthy saliva but was elevated in saliva from COVID-19 patients. A CXCL10 LFA was developed with a sensitivity of 2 ng/ml in a buffer and artificial saliva.

Conclusion: We establish and validate the potential of developing rapid test techniques to examine host immune response from a bioinformatic approach to developing a prototype rapid test with capabilities to be used in point-of-care settings.