key: cord-0867685-lzo0dz7z authors: Gu, Silan; Chen, Yanfei; Wu, Zhengjie; Chen, Yunbo; Gao, Hainv; Lv, Longxian; Guo, Feifei; Zhang, Xuewu; Luo, Rui; Huang, Chenjie; Lu, Haifeng; Zheng, Beiwen; Zhang, Jiaying; Yan, Ren; Zhang, Hua; Jiang, Huiyong; Xu, Qiaomai; Guo, Jing; Gong, Yiwen; Tang, Lingling; Li, Lanjuan title: Alterations of the Gut Microbiota in Patients with COVID-19 or H1N1 Influenza date: 2020-06-04 journal: Clin Infect Dis DOI: 10.1093/cid/ciaa709 sha: be47a5668d6c9db9bbdc4daa2ae43862264a36f9 doc_id: 867685 cord_uid: lzo0dz7z BACKGROUND: Coronavirus disease 2019 (COVID-19) is an emerging serious global health problem. Gastrointestinal symptoms are common in COVID-19 patients, and SARS-CoV-2 RNA has been detected in stool specimens. However, the relationship between the gut microbiome and disease remains to be established. METHODS: We conducted a cross-sectional study of 30 COVID-19 patients, 24 influenza A (H1N1) patients, and 30 matched healthy controls (HC) to identify differences in the gut microbiota by 16S ribosomal RNA (rRNA) gene V3-V4 region sequencing. RESULTS: Compared with HC, COVID-19 patients had significantly reduced bacterial diversity, a significantly higher relative abundance of opportunistic pathogens, such as Streptococcus, Rothia, Veillonella and Actinomyces, and a lower relative abundance of beneficial symbionts. Five biomarkers showed high accuracy for distinguishing COVID-19 patients from HC with an area under the curve (AUC) up to 0.89. Patients with H1N1 displayed lower diversity and different overall microbial composition compared with COVID-19 patients. Seven biomarkers were selected to distinguish the two cohorts with an AUC of 0.94. CONCLUSION: The gut microbial signature of patients with COVID-19 was different from that of H1N1 patients and HC. Our study suggests the potential value of the gut microbiota as a diagnostic biomarker and therapeutic target for COVID-19, but further validation is needed. Coronavirus disease 2019 (COVID-19) is an emerging respiratory infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and rapidly spread throughout the world [1] . The clinical manifestations and transmission routes of seasonal influenza A (H1N1) are similar to those of COVID-19 [2, 3] . WHO reported that the prevalence of influenza and influenza-like illness were high in China and the United States in the 2019-2020 season [4] . Distinguishing COVID-19 and H1N1 at the early stage of outbreaks is essential because occupational safety measures, treatment, and prognosis are different between these entities. SARS-CoV-2 binds to angiotensin-converting enzyme 2 (ACE2) receptors to invade human cells, and these receptors are highly expressed in the intestinal epithelium [5] . ACE2 might be associated with cardiopulmonary disease via alterations in the gut and/or lung microbiomes [6] . Previous studies have found that 3.34-11.4% of COVID-19 patients had gastrointestinal symptoms, such as vomiting and diarrhea, especially in critically ill patients [7, 8] . Moreover, viral RNA and live viruses were detected in fecal samples, suggesting that the digestive tract might be a site of viral replication and activity [9] . However, the interaction between COVID-19 and intestinal microorganisms is not fully understood. The intestinal flora is involved in host nutrient absorption and metabolism and has a profound impact on human health and disease [10] [11] [12] . Previous studies indicated that the intestinal flora was closely related to respiratory virus infection and could affect the occurrence and development of diseases through the gut-lung axis [13] . Moreover, influenza infection can affect the composition of the intestinal microbiota [14] , and intestinal microflora disorders reduce host antiviral immune response, thereby aggravating lung damage caused by these infections [15] . This cross-sectional study analyzed the gut microbiome of COVID-19 patients, H1N1 patients, and healthy controls (HC) by high-throughput sequencing of the 16S rRNA gene to provide a theoretical basis for differential diagnosis and intestinal microbial intervention. The results showed that biomarkers could be used to identify changes in the structure, composition, and function of the microbiome between these A c c e p t e d M a n u s c r i p t patient groups. The study was approved by the Research Ethics Committee of the First Affiliated All subjects who received antibiotics, probiotics, or both within 4 weeks before enrollment were excluded. Viral infections were confirmed by real-time reverse-transcription polymerase chain reaction. Only laboratory-confirmed cases with clinical symptoms were included in the study. Patient data, including laboratory test results, clinical manifestations, and disease course, were obtained from medical records and laboratory information systems. Fecal samples from COVID-19 and H1N1 patients were collected at admission, A c c e p t e d M a n u s c r i p t and fresh stools from healthy subjects were collected during physical examination. Fecal samples were processed in the laboratory within 4 hours after collection and stored at -80°C until analysis. Blood samples were taken from all study subjects for analyzing hematological variables, liver functions, kidney functions and serum cytokines, using routine clinical laboratory methods as described in the Supplementary Methods. Given the potential presence of live virus in feces, all fecal samples were inactivated For biomarker identification, a two-step approach was adopted. First, a random forest model [20] was constructed for distinguishing between the two groups, and ten most predominant genera as candidate biomarkers were selected on the basis of A c c e p t e d M a n u s c r i p t importance values (using the R package "randomForest", ntree=500) (Supplementary Table S1 ). Second, the differences in taxonomic composition taxa between two cohort groups were identified using the linear discriminant analysis (LDA) effect size (LEfSe) analysis (http://huttenhower.sph.harvard.edu/galaxy/). Candidate biomarkers with a LDA>3.5 were selected as final biomarkers. The discriminatory ability of the biomarkers was evaluated by plotting receiver-operating characteristic (ROC) curves and calculating the area under the ROC curve (AUC) using R software's pROC package. Continuous variables were reported as means ± standard deviations, and statistical comparisons were made using the independent t-test. Non-normally distributed variables were expressed as interquartile range (IQR), and comparisons were conducted using the Mann-Whitney U test. For correlation analysis, Spearman's rank test was performed. Statistical analysis was performed using SPSS version 20.0 (SPSS Inc., Chicago, IL). P values of less than 0.05 after multiple-comparison correction using the false discovery rate method were considered significant. The study population included 30 hospitalized patients with confirmed SARS-CoV-2 There were no significance differences in age, gender, and body mass index (BMI) between the groups. However, there were significant differences in platelet A c c e p t e d M a n u s c r i p t count, aspartate aminotransferase (AST), IL-4, and TNF-α between H1N1 and COVID-19 patients (P<0.05). The rate of hypertension did not differ significantly between these two groups (P=0.445). With regard to inflammatory markers, there was a significant difference in procalcitonin (P=0.016) but no significant difference in C-reactive protein (CRP) (P=0.832) between H1N1 and COVID-19 patients. There were significant differences in lymphocyte count, alanine aminotransferase (ALT), IL-2, IL-4, IL-6, IL-10 and TNF-α between HC and COVID-19 patients. According to clinical guidelines, COVID-19 severity on admission was categorized as general in 15 patients and severe in 15 patients. The clinical data of these subjects were summarized and compared (Supplementary Table S2 ). There was a significant difference in white blood cell (WBC) count, neutrophil count, lymphocyte count, and lactate dehydrogenase (LDH) between these two groups (P<0.05). The characteristics of the gut microbiome in patients with respiratory virus infections were analyzed by 16S rDNA gene sequencing of 84 fecal samples (one sample per patient). After merging and filtering, 4,105,869 high-quality sequence reads were generated, with an average of 48,879 sequences per sample for subsequent data analysis. The mean community richness and microbial diversity were significantly lower in COVID-19 and H1N1 patients than in HC, according to Shannon diversity index and Chao diversity index ( Figures 1A, B) . The number of OTUs in the COVID-19, H1N1, and HC groups was 911, 960, and 922, respectively. More than 50% of 1242 OTUs were shared by the three groups, and 62.3% of OTUs overlapped between the COVID-19 group and HC ( Figure 1C ). PCoA of Bray-Curtis distances indicated differences in the fecal microbiota between COVID-19 and H1N1 patients and between these groups and HC (ANOSIM, R=0.36, p=0.001) ( Figure 1D ). Ternary plot showed that the relative abundance of Streptococcus and Escherichia-Shigella was significantly higher in COVID-19 and H1N1 patients, respectively ( Figure 1E ). The analysis of group similarities indicated that differences in richness, diversity, A c c e p t e d M a n u s c r i p t and structure of the gut microbiota were not significantly different between general and severe COVID-19 patients (ANOSIM, p=0.426; Supplementary Figure S1 ), indicating that the experimental design was adequate. To investigate changes in the microbiota of COVID-19 patients, we assessed relative abundance in the three groups at the phylum, class, family, and genus levels ( Figure 2 LEfSe analysis was used to determine and distinguish the composition of the gut microbiome between the COVID-19 group and HC. The gut microbiome of the Figure S2) . Figure S2) . The gut microbiota signature of COVID-19 and H1N1 patients was analyzed to assess correlations between disease characteristics and the microbiome. The abundance of Prevotella, Ezakiella, Murdochiella, and Porphyromonas was higher in the H1N1 group than in COVID-19 patients ( Figure 4A ). In addition, seven final biomarkers (Streptococcus, Fusicatenibacter, Collinsella, Dorea, Agathobacter, Eubacterium hallii group, Ruminococcus torques group) were selected reference the two-step schema in the method to distinguish the two cohorts, with an AUC of 0.94 (95% CI, 0.87-1.00) ( Figure 4B ). Spearman analysis was conducted to evaluate the correlation between genera (abundance >0.1%) and clinical indexes, including WBC, CRP, PCT, D-dimer, IL-2, IL-4, IL-6, and TNF-α in COVID-19 and H1N1 patients, respectively. The significance thresholds were absolute correlation coefficients higher than 0.4 and P values lower than 0.05, as shown in Figure 5 . Intestinibacter, and Prevotella) ( Figure 5B ). Increasing evidence indicates the intimate relationship between the gastrointestinal and respiratory tract, which is known as the gut-lung axis [13] . Deriu stimulating the production of IFN-γ by these cells [14] . Studies have identified SARS-CoV-2 RNA in stool specimens of infected patients [9] , and RNA analysis demonstrated that the duration of viral shedding from stool was longer than that from respiratory samples [32] . In addition, host ACE2 receptors are highly expressed in the gastrointestinal epithelium [7] . SARS-CoV-2 may interfere with nutrient absorption by binding to ACE2 receptors, causing gastroenteritis-like symptoms, and disrupting intestinal homeostasis. We declare no competing interests. M a n u s c r i p t A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China Estimates of global seasonal influenza-associated respiratory mortality: a modelling study A pneumonia outbreak associated with a new coronavirus of probable bat origin Emerging Targets for Cardiopulmonary Disease Therapy Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019 (COVID-19) with gastrointestinal symptoms Multi-omics Evaluation of Gastrointestinal and Other Clinical Characteristics of SARS-CoV-2 and COVID-19 COVID-19: Gastrointestinal Manifestations and Potential Fecal-Oral Transmission Structure, function and diversity of the healthy human microbiome The gut flora as a forgotten organ Homeostatic Immunity and the Microbiota The Gut-Lung Axis in Respiratory Disease Respiratory influenza virus infection induces intestinal immune injury via microbiota-mediated Th17 cell-dependent inflammation Microbiota regulates immune defense against respiratory tract influenza A virus infection Abbreviations: NA, not available; IQR, interquartile range; BMI, body mass index; IL, interleukin; TNF-α, tumor necrosis factor-α. a P values indicate differences between H1N1 and COVID-19 05 was considered to indicate significant difference between healthy controls and H1N1 and COVID-19, respectively. Lactate dehydrogenase, U/L NA A c c e p t e d M a n u s c r i p t A c c e p t e d M a n u s c r i p t