key: cord-0800610-6jyz5msy
authors: Gallardo-Escárate, Cristian; Valenzuela-Muñoz, Valentina; Núñez-Acuña, Gustavo; Valenzuela-Miranda, Diego; Benaventel, Bárbara P.; Sáez-Vera, Constanza; Urrutia, Homero; Novoa, Beatriz; Figueras, Antonio; Roberts, Steven; Assmann, Paulina; Bravo, Marta
title: The wastewater microbiome: a novel insight for COVID-19 surveillance
date: 2020-10-09
journal: Sci Total Environ
DOI: 10.1016/j.scitotenv.2020.142867
sha: 8602c661c456ad8efd993ac7290f70d7c9ce4035
doc_id: 800610
cord_uid: 6jyz5msy

Wastewater-Based Epidemiology is a tool to face and mitigate COVID-19 outbreaks by evaluating conditions in a specific community. This study aimed to analyze the microbiome profiles using nanopore technology for full-length 16S rRNA sequencing in wastewater samples collected from a penitentiary (P), a residential care home (RCH), and a quarantine or health care facilities (HCF). During the study, the wastewater samples from the RCH and the P were negative for SARS-CoV-2 based on qPCRs, except during the fourth week when was detected. Unexpectedly, the wastewater microbiome from RCH and P prior to week four was correlated with the samples collected from the HCF, suggesting a core bacterial community is expelled from the digest tract of individuals infected with SARS-CoV-2. The microbiota of wastewater sample positives for SARS-CoV-2 was strongly associated with enteric bacteria previously reported in patients with risk factors for COVID-19. We provide novel evidence that the wastewater microbiome associated with gastrointestinal manifestations appears to precede the SARS-CoV-2 detection in sewage. This finding suggests that the wastewaters microbiome can be applied as an indicator of community-wide SARS-CoV-2 surveillance.

The SARS-CoV-2 is a new member of the Coronaviridae family and the etiologic of COVID-19, officially declared a pandemic by the World Organization of the Health in January 2020. Patients can present a wide variety of symptoms, and the prognosis varies from mild or moderate illness to severe and death (Guan et al., 2020; He et al., 2020) . A significant percentage of infected are symptomatic, and between 18 and 40% of cases have evidenced to be asymptomatic, a condition that helps the silent spread of the disease (Hadi et al., 2020; Jiang et al., 2020) .

The S protein gives the virus its distinctive crown of spikes and is responsible for binding to angiotensin-converting enzyme receptor 2 (ACE2), which allows the virus to enter the host cell Zhou et al., 2020b) . These receptors are present in various human cell types, with particular abundance in respiratory and gastrointestinal epithelial cells .

Indeed, an analysis of the ACE2 receptor distribution in human tissues found the highest levels of expression in intestine (Vuille-dit-Bille et al.; Zhang et al., 2020) . Although respiratory J o u r n a l P r e -p r o o f symptoms are the most frequently described in patients with , several studies have shown that the gastrointestinal tract can also be affected by SARS-CoV-2. A meta-analysis found that 15% of patients had gastrointestinal symptoms and that about 10% of patients had gastrointestinal symptoms but no respiratory symptoms Liang et al., 2020) .

In contrast, SARS-CoV-2 RNA has been found in the feces of people without gastrointestinal symptoms (Xu et al., 2020) . It has been found that more than half (55%) of those tested for fecal viral RNAs were positive and noted that the virus is excreted in the feces for long periods, and in some cases, beyond negative testing or with respiratory symptoms (Effenberger et al., 2020) . Wastewater-Based Epidemiology (WBE) is an epidemiological tool improve predictions and assist in mitigating COVID-19 outbreaks by evaluating biomarkers in a specific community (Daughton, 2020) . The primary advantage is related to minimizing domino effects such as unnecessarily long stay-at-home policies that stress humans and economies alike (de Lima et al., 2020) . While researchers have evaluated biomarkers for SARS-CoV-2 there is no information of how the wastewater microbiome is altered in communities infected by coronavirus or if the enteric bacteria can display species-specific signatures detected in sewage either from symptomatic or asymptomatic patients (Dhar and Mohanty, 2020) . Notably, cumulative evidence has been associated with dysbiosis of the gastrointestinal microbiota, and the relationship with pathologies such as obesity, diabetes, immunosuppression states, hypertension among others (Cani, 2018; Cryan et al., 2019; Lee et al., 2020) . With this in mind, according to the Center for Disease Control and Prevention (CDC), these pathologies increase the risk of severe in the human population. Surprisingly, the WBE tool has not yet been integrated with metagenomic analysis based on sequencing DNA technology and risk factors for viral diseases.

This novel approach can help epidemiologists and public health officials improve SARS-CoV-2 surveillance outbreak predictions which could lead to better mitigation strategies (La Rosa et al., J o u r n a l P r e -p r o o f 2020).

In this study, we analyze the microbiome signals using nanopore technology for fulllength 16S rRNA sequencing in samples collected from sewage produced in a residential care home (RCH), a penitentiary (P), and a healthcare facility (HCF). The HCF corresponds to quarantine accommodations, where COVID-19 positive patients stay-short periods in isolation for health recovery. For each location and sampling time-point, wastewaters were analyzed to detect SARS-CoV-2 by qPCR. The study was conducted from May to August 2020 in Chillan City, southern Chile. This research aimed to evaluate if changes in the sewage microbiota can be associated with the prevalence of COVID-19 in a specific human community. More importantly, if particular bacteria are associated with chronic diseases or risk factors for illness severity in patients with COVID-19. Overall, we provide novel evidence that the wastewater microbiome shows a specific-pattern related to gastrointestinal disorders and appears to precede the SARS-CoV-2 detection in sewage. This finding suggests that integrating metagenomic analyses into WBE provided an early indicator for COVID-19 outbreaks.

Untreated wastewater (sewage) samples were obtained from May to July 2020 in Chillan (200,000 people), a city in southern Chile. Wastewater operators collected a 24h flowdependent composite sample of 10 L using a submersible in-situ high-frequency autosampler.

Samples were transported on ice to the laboratory and stored at 4 °C until further analysis. For this study, three sewage sampled points were selected according to their social isolation level, risk factor, and the confirmation for COVID-19 by qPCR. Herein, for high-isolation level and size. The filtrates were concentrated in Amicon®-50 kDa (Millipore) by centrifugation at 5000xg for 15 min. Finally, 200 µL of concentrate wastewater was used for RNA isolation using the SV Total RNA Isolation System (Promega, USA). The concentration and purity of the RNA were determined by Nanodrop™ One/One C (Thermo Fisher Scientific, USA). The integrity of total RNA was evaluated by TapeStation 

Wastewater samples from the three sample points and times were centrifugated at 4500xg for 30 min, and then the supernatant was removed, and the organic pellet was used for DNA isolation using the QIAamp Fast DNA Stool Mini Kit (Qiagen, USA) following the manufacturer"s instructions. DNA integrity was verified by electrophoresis as above mentioned, and concentration measured by fluorescence in a Qubit 4 (ThermoScientific, USA) using the Qubit J o u r n a l P r e -p r o o f dsDNA BR Assay Kit (ThermoScientific, USA) following the manufacturer's instructions.

Genomic DNA samples were diluted to 10 ng/µL, and amplification of full-16S rRNA gene was performed by PCR in 25 µL with the 27 F 5′-AGAGTTTGATCCTGGCTCAG-3′ y 1492 R 5′ GGTTACCTTGTTACGACTT-3′ primers and Taq DNA polymerase LongAmp (NewEngland Biolabs, USA). PCR conditions were the followings: initial denaturing step at 95 ºC for 1 min, followed by 25 cycles of 95 °C for 20 sec, 56 °C for 30 sec and 65 °C for 2 min, and an extension at 65 °C for 5 min. PCR products were evaluated by electrophoresis.

After confirming the presence of full-16S rRNA amplicon, PCR products were purified with Agencourt AMPure XP beads (Beckman Coulter, USA) in a 1:2 sample-beads ratio. The mixture was incubated for 5 min at room temperature, and the sample placed in the magnetic stand for ethanol washing (freshly prepared at 70%). The washed mixture was resuspended in ultrapure sterilized water, and the purified product was separated from beads by placing it in a magnetic stand. The purified amplicons were used as a template for library synthesis using the 16S Barcoding Kit (SQK-RAB204, Oxford Nanopore Technologies, UK) following the manufacturer"s instructions for 1D sequencing strategy. Purified 16S amplicon was mixed with barcodes accordingly, and using the LongAmp Taq Polymerase (New England Biolabs, USA).

PCR was performed with a final reaction volume of 50 µL. The PCR product was purified again with Agencourt AMPure XP beads and incubated in HulaMixer for 5 min. After the magnetic beads washing step, the purified product was eluted in 10 µL of elution buffer (10 mM Tris-HCl pH8.0 with 50 mM NaCl).

The final concentration of the library was quantified by fluorometry using Qubit 4 (Thermo Fisher, USA) and evaluated by TapeStation Bioanalyzer 2200 system (Agilent Technologies, USA) using DNA ScreenTape (Agilent Technologies, USA) following the manufacturer's instructions. A single library was synthesized from wastewaters DNAs for three J o u r n a l P r e -p r o o f sample points at three different times. Then libraries were pooled in multiplex mode following the protocol of Oxford Nanopore Technologies, and the flowcell MK1 Spot-ON FLO-MIN107-R9 was accordingly loaded. For each sample point, three replicates were sequenced. A total of 36 libraries were run in three flowcells for the MinION platform (Oxford Nanopore Technologies, UK). Sequencing efficiency was monitored through the software MinKNOW 2.0 (Oxford Nanopore Technologies, UK). Finally, as internal control, the DNA of a microbial mock community (ZymoBiomics Microbial Community Standard, USA) was extracted and sequenced following the same described procedures to evaluate the sequencing run accuracy through the observed/expected taxa abundances.

After the sequencing runs, the yielded fast5 files were base-called using Guppy (version3.2.2, Oxford Nanopore Technologies, UK), and a filter step was applied to retain only sequences with Q-score ≥7 (quality filter). Sequences were then evaluated using the Fastq 16S v3.2.1 (Oxford Nanopore Technologies, UK) pipeline, which included demultiplexing, primers, and barcode trimming. Sequences with ambiguous identification were excluded from subsequent analysis.

BLASTN aligned cleaned sequences against NCBI RefSeq for 16S ribosomal RNA database for bacteria and archaea within the EPI2ME software package (Oxford Nanopore Technologies), and only the sequences with identity ≥75% were retained (confidence level filter) as mentioned by Edwards, et al.(2019) . The occurrence of each NCBI taxonomy identifier was evaluated for each sample using R, and the relative abundance of each NCBI taxa id was calculated per sample to normalize. Taxonomical identifiers were ordered, and to minimize singletons, doubletons product of sequencing errors taxon with <0.01% of relative abundance in at least one sample were eliminated. For each sample, an equivalent to OTU (Operative Taxonomic Unit) frequency table J o u r n a l P r e -p r o o f was built using R, and rarefaction analysis was performed to assess sequencing depth performance. The alpha diversity of microbiota communities associated with wastewater was compared among sample points by calculating the Shannon and Simpson 1-D diversity indices, the Chao-1 richness index and community evenness using "Vegan" package in R. The difference among sample points was evaluated with One-Way ANOVA followed by the posthoc multiple comparison Tukey-HSD test when applicable, considering a P-value < 0.05. Microbial community structure was analyzed based on taxon relative abundance data, and principal coordinate analysis (PCoA) were performed as a multivariate unsupervised data exploration.

Variations of the microbial community along sample time points were explored by plotting relative bacteria abundance, which corresponded to normalized reads of each taxon by the total of reads of the corresponding sample. Also, multivariate exploratory analyses were performed through PCoA from the microbiome collected in positive and negative sewage samples for SARS-Cov2. Distance between communities of different sample and time points was evaluated at the species level by hierarchical clustering using the UPGMA method. A group of selected species and genera were also evaluated due to previous findings related to COVID-19 outbreaks.

Pearson"s correlations between the relative abundance of these bacteria and the SARS-CoV-2 viral load in the samples were calculated using the R software by the Corrplot package (Wei et al., 2017 ).

From May to July 2020 in Chillan, Chile, we monitored wastewater SARS-CoV-2 load from a residential care home (RCH), penitentiary (P), and health care facilities (HCF) selected according to their social isolation level and risk factor. COVID-19 was confirmed by qPCR analysis of J o u r n a l P r e -p r o o f ORFab1 and N protein genes. The quantity of virus genome units per liter of sewage samples was calculated by a 10-log standard curve using 10-fold dilutions of a Positive control TaqMan TM 2019-nCoV Control Kit v1 (Thermo Fisher Scientific, USA) (Fig. S1 ). The positive control group, HCF, scored positive for SARS-CoV-2 genes with average C T values of 32 and 33 for ORFab1 and N protein genes. These C T values corresponded to an estimated maximum concentration of 1x10 4 Genome Unit/L (Fig. 1) . Samples obtained from RCH, with middleisolation level and high-risk factor, registered an outbreak of COVID-19 from July 10 th to July 21 st (corresponding to week 4 of the study) with C T values between 32 and 33 for both genes to an average virus load of 7x10 3 Genome Unit/L. In addition, penitentiary samples evidenced an outbreak of COVID-19 during July 10 th with a 4x10 3 Genome Unit/L, observing C T values of 35.6 and 33.4 for ORFab1 and N protein genes, respectively. The demography data obtained for each studied location and group of people showed substantial differences in sex proportions, age, and comorbidity. This last characteristic associated with risk factors for COVID-19 indicated less prevalence in youngest individuals and under confinement in prison.

Metagenome analyses were conducted in sewage samples collected from the three different sites and four sampling points (Fig. 1) based on Oxford nanopore sequencing. Reads obtained from the DNA sequencing from the different samples collected were annotated through BLASTn against the NCBI RefSeq for 16S ribosomal RNA database for bacteria and archaea, and an equivalent to OTU (Operative Taxonomic Unit) frequency table was built. Rarefaction analysis indicated that all sequenced libraries reached a plateau, indicating that the sequencing effort was enough to capture most of the microbial variability in the samples (Fig. S2 ). In parallel, NCBI Taxonomy J o u r n a l P r e -p r o o f tree from the wastewater microbiome was produced, indicating high bacterial diversity represented by the most abundant (30) taxa (Fig. 2) .

Diversity indexes were calculated based on the generated OTU table. These indexes included Shannon, Simpson's, Evenness, and Chao-1 indexes (Fig. 3) . Even though no significant variations were perceived on Shannon, Simpson's, and Evenness, a marked variation in the Chao-1 index was evidenced in the different sampled sites and points. Thus, an increase in the diversity value from samples collected during the COVID-19 outbreak was observed using one index.

Notably, the highest index for the residence care home and the penitentiary was observed at timepoint S3 (Fig. 3d) . Concerning the health care facility (HCF), no differences in taxa diversity during the study were found. Interestingly, a variation in the relative abundance of identified phylum was observed. Specifically, there was a marked reduction in the relative abundance of Proteobacteria and corresponding increase in other phylum at the residential care home and the penitentiary during the outbreak (Fig. 3e ).

There were substantial differences in microbiota communities in wastewater collected from all locations (Fig. 4) . At the level of species, the penitentiary showed the highest number of exclusive taxa (304); meanwhile, the residence care home and the health care facility samples contained 135 and 45 species. respectively. Notably, a core microbiome was observed with 333 bacterial species (Fig. 4A and Table 1S ). We analyzed samples from all the localities according to the viral quantification with the aim of evaluating differences between exclusives and shared species for wastewaters tested for SARS-CoV-2. Samples collected from locations testing positive or negative for SARS-CoV-2 contained 521 and 198 exclusive taxa, respectively. There were 409 taxa present in both SARS-CoV-2 positive and SARS-CoV-2 negative samples. (Fig.   4B and Table 1S ). Furthermore, the analysis evidenced that exclusives species were most represented in positive samples for coronavirus collected from RCH (25.7%) and P (28,6%) J o u r n a l P r e -p r o o f locations. Unexpectedly, negative samples for SARS-CoV-2 were associated to exclusives bacterial species from wastewaters collected from the Penitentiary (76.3%). With respect to exclusives bacteria found in HCF, the analysis showed a low representativity either for positive or negative SARS-CoV-2 samples. This analysis also shows that overall high SARS-CoV-2 loads of sewage are associated with infected individuals or communities. (Fig. 5B) . The top species associated with positive samples (particularly samples from HCF in all the sampling times, P and RCH on July 17th) were Arcobacter suis, A. venerupis, A. cloacae, Prevotella copri, A. aquamarines, and Bacteroides vulgatus. Notably, these species were also more abundant in P and RCH sampling points in the sampling time before the first SARS-CoV-2 detection (July 7th) even when the virus was not detected.

A subset of bacteria was used to evaluate the association with COVID-19 outbreaks (Zuo et al., 2020) . These taxonomic units were selected according to their functions and previously reported association with SARS-CoV-2 infections, including commensal symbionts and opportunistic pathogens. Relative abundance was calculated, as previously described. Normalized reads of selected bacteria were correlated with viral load values through Pearson"s correlations using all of the data. The commensal symbionts Eubacterium ventriosum, and commensal species from the Roseburia genus and the Lachnospiraceae family were significantly correlated to the viral load in the samples (Fig. 6A) . However, the opportunistic pathogen species Bacteroides nordii, and species from the Rothia and Veilonella genera were positively correlated with SARS-CoV-2 levels in the sewage samples.

In contrast, pathogens from the Clostridium, Actinomyces, and Streptococcus genera were negatively correlated with the viral load, and the same was obtained for the commensal species from the Ruminoccoccaceae family. Evaluating how the abundances of this selected subset vary in the different locations, a heatmap of normalized reads was built (Fig. 6B ). In general, the most variable location was the Penitentiary (P), where bacteria from the Lachnospiraceae and Rumnococcaceae families were the most abundant after the COVID-19 outbreak. A similar trend was also obtained for Alistipes genus in this location. Most of the selected bacteria were identified in the sampling point before COVID-19 outbreak detection (July 7th), particularly in the P and RCH locations. The location with constant COVID-19 positive detection (HCF) had lower variations of this group.

Our data are consistent with the idea that the wastewater microbiome reveals species-specific profiling associated with human communities where SARS-CoV-2 is prevalent, independently J o u r n a l P r e -p r o o f the symptomatic/asymptomatic status of those individuals. This finding strongly suggests that the virus alters the microbiome of infected individuals similarly to risk factors of severe COVID-19 (Zhou et al., 2020a) . Several studies have implicated the microbiome in a range of physiologic processes that are vital to host health, including energy, metabolism, gastrointestinal health, immune response, and neurophysiology (Cani, 2018; Cryan et al., 2019) . The microbial community supports metabolic interactions connecting those with the host and its physiological state. Notably, the enteric microbiome has an active role in humans' well-being, involving complex molecular signaling between the gut-brain axis (Cryan et al., 2019; Rieder et al., 2017) .

Moreover, dysbiosis in the gastrointestinal microbiome is strongly associated with chronic diseases and psychological disorders in humans (Jie et al., 2017; Rieder et al., 2017; Taylor and Holscher, 2020; Villette et al., 2020) . Our study revealed substantial differences in the microbiota associated with wastewaters collected from three communities with different levels of social isolation and risk factors for COVID-19. However, a core microbiota community was significantly found in sewage laboratory-tested for SARS-Cov-2. This finding suggests that the putative microbiome dysbiosis of infected individuals is the primary factor for the wastewater microbiota variations. However, this study does not support how the risk factors for are involved in the magnitude of the microbiota dysbiosis, and consequently, in the wastewater microbiome associated with SARS-CoV-2 detection. Notably, the demographic data obtained for this study did not indicate a high prevalence of comorbidities related to the Penitentiary population. Unexpectedly, the analysis conducted in wastewaters collected from patients with positive SARS-CoV-2 testing indicated more exclusive bacterial species than samples associated with individuals negatives for coronavirus. Our results indicate that Simpliscira, Flavobacterium, Acidovorax, and Acinetobacter genera were negatively associated with SARS-CoV-2. aquamarines, and Bacteroides vulgatus appear to precede the SARS-CoV-2 detection in sewage.

Previous studies have suggested strong correlations among these taxonomic units with chronic diseases such as obesity (Crovesy et al.) , immune-mediated inflammatory diseases (Geva-Zatorsky et al., 2017) , and cardiovascular diseases (Tang et al., 2017) .

Studies connecting the microbiota with distal organs in regulatory functions such as gutlung axis have revealed a key role in the respiratory system (Groves et al., 2018) . It has been reported that respiratory viral infections alter the gut microbiota, modulating the antiviral immune response through TNF-and CD8 + cells (Groves et al., 2020) . For example, gastric disorders during respiratory syncytial virus (RSV) or influenza virus infection in mice resulted in significantly altered gut microbiota diversity, increased Bacteroidetes, and a concomitant decrease in Firmicutes phyla abundance. Viral infection also led to changes in the fecal gut metabolome, with a significant shift in lipid metabolism. Sphingolipids, polyunsaturated fatty acids, and the short-chain fatty acid valerate were all increased in abundance in the fecal metabolome following RSV infection (Hasegawa et al., 2017) . Our findings showed that Eubacterium ventriosum, and commensal species from the Roseburia genus and the Lachnospiraceae family were significantly correlated to the viral load in wastewater samples.

Notably, opportunistic pathogens from the Bacteroides genera showed high modulation during the study, and some species positively correlated with the SARS-CoV-2 levels. This evidence can be connected with a recent study where the sphingosine gene expression was significantly correlated with increased Bacteroides in patients with acute respiratory viral infections (Groves et List of figures Figure 1 . Experimental design and sampling points/times for microbiome sequencing. (A) SARS-CoV-2 quantification. The full-16S rRNA gene was sequenced using MinION platform (Oxford Nanopore Technology). The sequencing runs are indicated as orange rectangles (S1-4). The virus load was estimated by qPCR in untreated wastewaters from three locations: (P) Penitentiary, (RCH) Residence Care Home and (HCF) Health Care Facilities. The study was conducted from May to July 2020. (B) Demography data for P, RCH and HCF. Figure 2 . NCBI Taxonomy tree from the wastewater microbiome showing the top 30 taxa.

Minimum abundance cutoff was 0.5%. 

Bacteroides-Derived Sphingolipids Are Critical for Maintaining Intestinal Homeostasis and Symbiosis

Human gut microbiome: hopes, threats and promises

Can the COVID-19 Epidemic Be Controlled on the Basis of Daily Test Reports?

Novel coronavirus: where we are and what we know

Gastrointestinal Manifestations of SARS-CoV-2 Infection and Virus Load in Fecal Samples From a Hong Kong Cohort: Systematic Review and Meta-analysis

Profile of the gut microbiota of adults with obesity: a systematic review

THE MICROBIOTA-GUT-BRAIN AXIS

Wastewater surveillance for population -wide Covid-19: The present and future

Effects of quarantine on mental health of populations affected by Covid-19

Gut microbiota and Covid-19-possible link and implications

In-field metagenome and 16S rRNA gene amplicon nanopore sequencing robustly characterize glacier microbiota

Faecal calprotectin indicates intestinal inflammation in COVID-19

Mining the Human Gut Microbiota for Immunomodulatory Organisms

Respiratory Disease following Viral Lung Infection Alters the Murine Gut Microbiota

Respiratory Viral Infection Alters the Gut Microbiota by Inducing Inappetence

Clinical Characteristics of Coronavirus Disease 2019 in China

A Review on COVID-19: Origin, Spread, Symptoms, Treatment, and Prevention

Sphingolipid metabolism potential in fecal microbiome and bronchiolitis in infants: a case-control study

Coronavirus disease 2019: What we know

Review of the Clinical Characteristics of Coronavirus Disease 2019 (COVID-19)

The gut microbiome in atherosclerotic cardiovascular disease

First detection of SARS-CoV-2 in untreated wastewaters in Italy

Gut microbiome and its role in obesity and insulin resistance

Physiological and pathological regulation of ACE2, the SARS-CoV-2 receptor

Diarrhoea may be underestimated: a missing link in 2019 novel coronavirus

Characterization of spike glycoprotein of SARS-CoV-2 on virus entry and its immune cross-reactivity with SARS-CoV

Detection of SARS-CoV-2 in Fecal Samples From Patients With Asymptomatic and Mild COVID-19 in Korea

Microbes and mental health: A review

The 2019-2020 Novel Coronavirus (Severe Acute Respiratory Syndrome Coronavirus 2) Pandemic: A Joint American College of Academic International Medicine-World Academic Council of Emergency Medicine Multidisciplinary COVID-19 Working Group Consensus Paper

Gut Microbiota in Cardiovascular Health and Disease

A review of dietary and microbial connections to depression, anxiety, and stress

Unraveling Host-Gut Microbiota Dialogue and Its Impact on Cholesterol Levels

SARS-CoV-2 receptor ACE2 gene expression in small intestine correlates with age

Characteristics of pediatric SARS-CoV-2 infection and potential evidence for persistent fecal viral shedding

Specific ACE2 expression in small intestinal enterocytes may cause gastrointestinal symptoms and injury after 2019-nCoV infection

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

A pneumonia outbreak associated with a new coronavirus of probable bat origin

Alterations in Gut Microbiota of Patients With COVID-19 During Time of Hospitalization

The authors declare no competing interest.