Gut microbiota composition in COVID-19 hospitalized patients with mild or severe symptoms

Abstract

Background and aim: COVID-19, the infectious disease caused by SARS-CoV-2 virus that has been causing a severe pandemic worldwide for more than 2 years, is characterized by a high heterogeneity of clinical presentations and evolution and, particularly, by a varying severity of respiratory involvement. This study aimed to analyze the diversity and taxonomic composition of the gut microbiota at hospital admission, in order to evaluate its association with COVID-19 outcome. In particular, the association between gut microbiota and a combination of several clinical covariates was analyzed in order to characterize the bacterial signature associate to mild or severe symptoms during the SARS-CoV-2 infection.

Materials and methods: V3–V4 hypervariable region of 16S rRNA gene sequencing of 97 rectal swabs from a retrospective cohort of COVID-19 hospitalized patients was employed to study the gut microbiota composition. Patients were divided in two groups according to their outcome considering the respiratory supports they needed during hospital stay: (i) group “mild,” including 47 patients with a good prognosis and (ii) group “severe,” including 50 patients who experienced a more severe disease due to severe respiratory distress that required non-invasive or invasive ventilation. Identification of the clusters of bacterial population between patients with mild or severe outcome was assessed by PEnalized LOgistic Regression Analysis (PELORA).

Results: Although no changes for Chao1 and Shannon index were observed between the two groups a significant greater proportion of Campylobacterota and Actinobacteriota at phylum level was found in patients affected by SARS-CoV-2 infection who developed a more severe disease characterized by respiratory distress requiring invasive or non-invasive ventilation. Clusters have been identified with a useful early potential prognostic marker of the disease evolution.

Discussion: Microorganisms residing within the gut of the patients at hospital admission, were able to significantly discriminate the clinical evolution of COVID-19 patients, in particular who will develop mild or severe respiratory involvement. Our data show that patients affected by SARS-CoV-2 with mild or severe symptoms display different gut microbiota profiles which can be exploited as potential prognostic biomarkers paving also the way to new integrative therapeutic approaches.

Keywords: COVID-19; SARS-CoV-2; biomarkers; intensive and critical care; microbiota (16S rRNA).

Figure 1

Fecal microbiota composition (i.e., mean relative abundance %) at Phylum (A), Family (B), Genus (C), and Species (D) levels in patients with coronavirus disease (COVID-19) at their hospitalization, for mild and severe symptoms separately. “Others (<1%)” category includes bacteria with mean relative abundance less than 1%.

Figure 2

Scatter plots of the Z-scores computed within each cluster (i.e., centroid) detected by PEnalized LOgistic Regression Analysis at Genus (C), and Species (D). Each point represents the Z-scores pair computed at each individual and were filled with blue and red colors to designate COVID-19 patients with milder or more severe evolution. Moreover, a polygon connecting the outermost data points is shown for both group. As a single cluster of bacteria population was detected at both Phylum (A) and Family levels (B), box plots were presented (instead of scatter plots).

Figure 3

Heatmaps of relative abundance (%) of bacterial populations identified (into different clusters) by the PEnalized LOgistic Regression Analysis at different taxonomic hierarchy, grouped by COVID-19 patients with mild and severe symptoms, respectively. The order of the rows (patients) was determined by the main heatmap, which has been defined at the Phylum level. All other heatmaps are automatically adjusted according to the settings in the main heatmap.

Figure 4

Prediction of the functional capabilities of microbial communities based on 16S datasets.