COVID-19 continues to extensively impact every facet of our lives. Central to our efforts to slow the spread of SARS-CoV-2 is accurate and sensitive diagnostic tests. Unfortunately, news reports continually highlight examples of individuals initially deemed uninfected, only to test positive later, contributing to the spread throughout the community. Conventional diagnostic tests target the viral genome, which may not be readily detectable in the early, particularly pre-symptomatic, stages of infection. Host-encoded microRNAs (miRNA) have shown promise as biomarkers of infection; they respond rapidly and are easily detectable within biofluids. While other host responses, especially inflammatory cytokines, have been extensively studied, the host-encoded miRNA response to SARS-CoV-2 remains poorly defined.
Here, we used next-generation sequencing to profile the circulating miRNAs from 10 COVID-19 patients, sampled longitudinally, and age and gender matched controls. We found 55 differentially expressed miRNAs (FDR adjusted p-value <0.05) during early-stage disease, with several miRNAs having previously established roles in inflammation. We then employed a multivariate machine learning approach to identify a three-miRNA biomarker signature (miR-423-5p, miR23a-3p, miR-195-5p) that identifies COVID-19 cases with 99.9% accuracy, 99.8% sensitivity, 99.8% recall, and a receiver operator characteristic area under the curve (ROC AUC) of 1. This signature is associated with early-stage disease; as the patients recovered, their samples clustered with the healthy controls. Further validation was conducted in a ferret model, where these three miRNAs identified SARS-CoV-2 infection with 99.7% accuracy, and distinguished SARS-CoV-2 infection from influenza A (H1H1) infection and uninfected animals with 95% accuracy.
This study, recently published in PLoS Pathogens [1], gives insights into the host miRNA response to SARS-CoV-2 infection and highlights the use of these molecules as circulating biomarkers for the detection of COVID-19. This biomarkers signature has great potential to compliment existing diagnostic tests and help to identify infections that may otherwise be missed.