Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Medical University of Warsaw and São Paulo Research Foundation (FAPESP). Introduction Coronavirus disease 2019 (COVID-19) is associated with an increased risk of mortality and adverse cardiovascular (CV) occurrences, especially in a group of patients with CV diseases (CVD). SARS-CoV-2 virus enters cells through angiotensin converting enzyme 2 (ACE2) cell membrane enzymes, leading to the downregulation in the ACE2-related pathways. Based on our previous in silico [1] results we hypothesized that alteration in the ACE2 interaction network caused by virus infection can affect the expression of miRNAs and influence severity of the disease. Purpose In this study we aimed to analyse the diagnostic and predictive utility of ACE2-related miRNAs, which were identified in our previous in silico analysis (miR-10b-5p, miR-124-3p, miR-200b-3p, miR-26b-5p, miR-302c-5p) in patients with COVID-19. We also aimed to unravel the functions of analyzed miRNAs, by using machine learning based tools, including SHAP for clinical data analysis, and bioinformatic tools for data mining, target predictions and enrichment analyses. Methods Blood samples and clinical data of 79 COVID-19 patients were collected at three different time points, including the day of admission, 7-days and 21-days after admission, as well as one time-point of 32 healthy volunteers. The composite endpoint was hospitalisation length of stay (>21 days) and/or death in follow-up. Results Delta low miR-200b-3p expression (after 7-days of admission) presents predictive utility in assessment of the hospital length of stay and/or death in ROC curve analysis (AUC:0.730, p=0.002) (Fig. 1). Delta low miR-200b-3p expression, together with diabetes mellitus (DM) are independent predictors of increased hospital length of stay and/or death (OR= 5.775; CI= 1.572- 21.214; p= 0.008, OR= 4.888; CI= 1.001- 23.858; p= 0.050, respectively). The expression levels of miR-26b-5p and miR-10b-5p in COVID-19 patients were found lower at the baseline, 7 and 21-days after admission compared to the healthy controls (p<0.0001 for all time points). SHAP analysis indicated levels of miR 200b-3p (day 7th), miR-302c-5p (day 7th), CRP (day 7th), neutrophils (day 0), and D-Dimer (day 0) as the most promising predictors of long hospitalisation for the COVID-19 patients. Pathway enrichment analysis showed that among top shared pathways between targets of analyzed miRNAs were IL-2 signaling pathway, and Pathways in cancer. miR-200b-3p showed regulation of COVID-19-related targets associated with T cell protein tyrosine phosphatase and HIF-1 transcriptional activity in hypoxia, key pathways in COVID-19. Bioinformatics analysis pointed out the role of those miRNAs in multiple CVD phenotypes associated with COVID-19 (Fig. 2). Conclusions In this study we validated and characterized miRNAs which could serve as novel, predictive biomarkers of the COVID-19 long term hospitalisation and can be used for early stratification of patients and prediction of severity of infection development in an individual.
Read full abstract