SARS-CoV-2 infectivity can be modulated through bacterial grooming of the glycocalyx

Abstract

The gastrointestinal (GI) tract is a site of replication of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and GI symptoms are often reported by patients. SARS-CoV-2 cell entry depends upon heparan sulfate (HS) proteoglycans, which commensal bacteria that bathe the human mucosa are known to modify. To explore human gut HS-modifying bacterial abundances and how their presence may impact SARS-CoV-2 infection, we developed a task-based analysis of proteoglycan degradation on large-scale shotgun metagenomic data. We observed that gut bacteria with high predicted catabolic capacity for HS differ by age and sex, factors associated with coronavirus disease 2019 (COVID-19) severity, and directly by disease severity during/after infection, but do not vary between subjects with COVID-19 comorbidities or by diet. Gut commensal bacterial HS-modifying enzymes reduce spike protein binding and infection of authentic SARS-CoV-2, suggesting that bacterial grooming of the GI mucosa may impact viral susceptibility.IMPORTANCESevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus responsible for coronavirus disease 2019, can infect the gastrointestinal (GI) tract, and individuals who exhibit GI symptoms often have more severe disease. The GI tract's glycocalyx, a component of the mucosa covering the large intestine, plays a key role in viral entry by binding SARS-CoV-2's spike protein via heparan sulfate (HS). Here, using metabolic task analysis of multiple large microbiome sequencing data sets of the human gut microbiome, we identify a key commensal human intestinal bacteria capable of grooming glycocalyx HS and modulating SARS-CoV-2 infectivity in vitro. Moreover, we engineered the common probiotic Escherichia coli Nissle 1917 (EcN) to effectively block SARS-CoV-2 binding and infection of human cell cultures. Understanding these microbial interactions could lead to better risk assessments and novel therapies targeting viral entry mechanisms.

Keywords: Covid; Heparan Sulfate; SARS-CoV-2; aging; human microbiome.

Fig 1

Metabolic task analysis of bacterial species capacity for heparan sulfate (HS) modification across the phylogenetic tree of life. SARS-CoV-2 relies on HS, a major component of the human gut mucosa, for host cell adhesion and entry. Human gut bacteria can alter mucosal surfaces potentially disrupting SARS-CoV-2 cell entry (A). Here, we examine the HS degrading genes in 7,490 bacterial strains measured in the FINRISK 2002 data set and quantify the HS catabolic capacity of each strain based on the set of tasks defined in degradation (B). Bacterial tree of life colored by superphylum groups and phyla containing predicted HS-modifying species. The bar chart represents the predicted capacity for HS modification of each species colored by phyla (C). Created in BioRender. Martino, C. (2025) https://BioRender.com/x09k421.

Fig 2

HS-modifying bacteria are inversely enriched according to host age, sex, but not diet. The log-ratio of predicted HS-modifying species relative to those with no predicted capacity for HS degradation (y-axis) in the FINRISK 2002 data set compared by host age (x-axis). The log-ratio of HS lyase genes relative to a set of housekeeping genes (y-axis) in the FINRISK 2002 fecal data compared by host age (x-axis) (A). The log-ratio of predicted HS-modifying species relative to those with no predicted capacity for HS degradation (y-axis) in the AGP fecal data set, compared over host age (x-axis) (B). Log-ratios are colored by participant sex (women, black; men, red). Log-ratio of predicted HS-modifying species relative to those with no predicted capacity (y-axis) in the FINRISK 2002 data set compared by dietary fiber and red meat intake above (high; blue) or below (low; green) the median consumption across ages (x-axis) and between sex (panel columns) (C and D). All log-ratio plots across age were annotated by the number of subjects at that time point. Error bars represent the standard error of the mean. Presented P values and test statistics are from unpaired two-tailed t tests evaluated on each host age group between host sex.

Fig 3

Commensal human gut bacteria block SARS-CoV-2 spike protein binding, through heparan sulfate degradation, in axenic culture supernatant. Growth of Bacteroides ovatus and Bacteroides thetaiotaomicron (A) measured by optical density (y-axis) across time from inoculation (x-axis) in minimal media (black; negative control NC), minimal media with 22 mM glucose and 1.4 mg/mL heparin (blue), or minimal media with 1.4 mg/mL heparin alone (red). Comparison of heparin concentration (y-axis; mg/mL) before inoculation (white; 0 h) and at stationary phase (gray; 84 h) for B. ovatus and B. thetaiotaomicron (B). Geometric mean of flow cytometry data (y-axis) of cultured human A549 bronchial epithelial cells stained with the anti-HS antibody 10E4 or incubated with biotinylated SARS-CoV-2 spike protein (D). Cells were incubated with culture media (Media), cell-free supernatant of B. ovatus (B. ova), or B. thetaiotaomicron (B. thet.) or purified heparin lyase from Pedobacter heparinus (F. hep.). Presented P values are from unpaired t-test statistics compared to the untreated control (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001).

Fig 4

Purified enzymes derived from human commensal gut bacteria block SARS-CoV-2 spike protein binding and viral infection through heparan sulfate modification. Schematic diagram of activity and putative specificity on typical highly and lowly sulfated HS. The bacterial species sourced for enzymes with X-ray crystallographic structures (A). Fold change of the geometric mean of flow cytometry data across two experiments relative to untreated control (y-axis) of cultured human A549 cells stained with the HS antibody 10E4 (red) or incubated with biotinylated SARS-CoV-2 spike protein (black) compared with and without 100 µg/mL heparin as a competitive substrate between purified Pedobacter heparinus (F. hep.) heparin lyase (Hsase), purified B. thetaiotaomicron heparin Lyase BT4652 (polysaccharide Lyase family 15; PL15), Escherichia coli strain Nissle 1917 (EcN) wild-type (WT) culture supernatant, EcN encoding B. thetaiotaomicron HS targeting genes BT4662 (polysaccharide Lyase family 12; PL12) and BT4675 (polysaccharide Lyase family 13; PL13) (B). SARS-CoV-2 infection of Vero cells performed in the absence and presence of P. heparinus Hsase, or with different concentrations of B. thetaiotaomicron purified BT4662 heparin lyse. The graph shows the percent infection from a composite of two separate experiments, each performed in triplicate (C). The same data but with the experimental data normalized to the mock infection for each respective experiment (D). Presented P values are from unpaired t-test statistics w.r.t. no treatment (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001).