Multi-kingdom gut microbiota analyses define COVID-19 severity and post-acute COVID-19 syndrome

• Published in: Nature communications Vol. 13; no. 1; pp. 6806 - 11
• Main Authors: Liu, Qin, Su, Qi, Zhang, Fen, Tun, Hein M., Mak, Joyce Wing Yan, Lui, Grace Chung-Yan, Ng, Susanna So Shan, Ching, Jessica Y. L., Li, Amy, Lu, Wenqi, Liu, Chenyu, Cheung, Chun Pan, Hui, David S. C., Chan, Paul K. S., Chan, Francis Ka Leung, Ng, Siew C.
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 10.11.2022; Nature Publishing Group; Nature Portfolio
• Subjects: 45/23; 631/326/107; 692/308; 692/4020; Clustering; Coronaviruses; COVID-19; Feces - microbiology; Gastrointestinal Microbiome - genetics; Humanities and Social Sciences; Humans; Intestinal microflora; Metabolomics; Metagenomics; Metagenomics - methods; Microbiomes; Microbiota; Microorganisms; multidisciplinary; Patients; Phenotypes; Post-Acute COVID-19 Syndrome; Respiratory tract; Science; Science (multidisciplinary); Viral diseases; Viruses
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-022-34535-8

Our knowledge of the role of the gut microbiome in acute coronavirus disease 2019 (COVID-19) and post-acute COVID-19 is rapidly increasing, whereas little is known regarding the contribution of multi-kingdom microbiota and host-microbial interactions to COVID-19 severity and consequences. Herein, we perform an integrated analysis using 296 fecal metagenomes, 79 fecal metabolomics, viral load in 1378 respiratory tract samples, and clinical features of 133 COVID-19 patients prospectively followed for up to 6 months. Metagenomic-based clustering identifies two robust ecological clusters (hereafter referred to as Clusters 1 and 2), of which Cluster 1 is significantly associated with severe COVID-19 and the development of post-acute COVID-19 syndrome. Significant differences between clusters could be explained by both multi-kingdom ecological drivers (bacteria, fungi, and viruses) and host factors with a good predictive value and an area under the curve (AUC) of 0.98. A model combining host and microbial factors could predict the duration of respiratory viral shedding with 82.1% accuracy (error ± 3 days). These results highlight the potential utility of host phenotype and multi-kingdom microbiota profiling as a prognostic tool for patients with COVID-19. Here, by performing an integrative multi-omics analysis coupled to clinical features of COVID-19 patients prospectively followed for up to 6 month, the authors identify specific gut microbiome patterns associated with disease severity and development of post-acute COVID-19 syndrome.