Human norovirus binding to select bacteria representative of the human gut microbiota

Abstract

Recent reports describe the ability of select bacterial strains to bind human norovirus, although the specificity of such interactions is unknown. The purpose of this work was to determine if a select group of bacterial species representative of human gut microbiota bind to human norovirus, and if so, to characterize the intensity and location of that binding. The bacteria screened included naturally occurring strains isolated from human stool (Klebsiella spp., Citrobacter spp., Bacillus spp., Enterococcus faecium and Hafnia alvei) and select reference strains (Staphylococcus aureus and Enterobacter cloacae). Binding in PBS was evaluated to three human norovirus strains (GII.4 New Orleans 2009 and Sydney 2012, GI.6) and two surrogate viruses (Tulane virus and Turnip Crinkle Virus (TCV)) using a suspension assay format linked to RT-qPCR for quantification. The impact of different overnight culture media prior to washing on binding efficiency in PBS was also evaluated, and binding was visualized using transmission electron microscopy. All bacteria tested bound the representative human norovirus strains with high efficiency (<1 log10 of input virus remained unbound or <10% unbound and >90% binding efficiency) (p>0.05); there was selective binding for Tulane virus and no binding observed for TCV. Binding efficiency was highest when bacteria were cultured in minimal media (<1 log10 of input virus remained unbound, so >90% bound), but notably decreased when cultured in enriched media (1-3 log10 unbound or 0.01 -<90% bound)) (p<0.05). The norovirus-bacteria binding occurred around the outer cell surfaces and pili structures, without apparent localization. The findings reported here further elucidate and inform the dynamics between human noroviruses and enteric bacteria with implications for norovirus pathogenesis.

Fig 1. Binding efficiency of human norovirus and representative surrogate viruses to select bacteria.

The line indicates the total virus input. Data are expressed as log₁₀ mean concentration ± the standard deviation of bacteria bound (in RT-qPCRU) (bars) and percent binding efficiency as determined by loss-to-supernatant ([total input virus-supernatant virus]/total input virus) (numerical). The black bars correspond to ATCC or control strains. The gray bars correspond to bacteria isolated in this study. All bacteria grew anaerobically in TSB with the following exceptions: B. thetaiotaomicron (chopped meat medium), L. gasseri (MRS) and L. plantarum (MRS). Asterisks (*) represent values for which there was a statistically significant difference (p<0.05) between the viral input load and the bacterial capture amount based on log₁₀ RT-qPCRU. Data represents averages and standard deviations of the assays performed in triplicate.

Fig 2. Binding efficiency of the GII.4 Sydney 2012 when bacterial strains were grown in different media.

The line indicates the total virus input. Data are expressed as mean log₁₀ concentration of bacteria bound ± the standard deviation (in RT-qPCRU) (bars) and percent binding efficiency as determined by loss-to-supernatant ((total input virus-supernatant virus)/total input virus) (numerical). Letters indicate statistically significant differences (p<0.05) between the amount of virus bound for each bacterial strain cultured using different growth media. Different letters within the same bacteria indicate statistical difference. Statistical differences in binding between different bacteria was not tested. Data represent averages and standard deviations of the assays performed in triplicate.

Fig 3

Transmission electron microscopy (50,000x) photos of select bacteria to which GII.4 Sydney 2012 (A-C) and GII.4 Farmington Hills 2002 (D-F) VLPs are bound. Bacteria-VLP interactions are shown as follows: (A) E. cloacae, (B) S. aureus, (C) Bacillus spp., (D) E. faecium, (E) Citrobacter spp., and (F) H. alvei. Representative VLPs are pointed out with the arrows, although additional VLPs are frequently also present in the image. Images shown are representative of multiple fields of view.