http://orcid.org/0000-0002-5393-0733
Figures
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.
Citation: Almand EA, Moore MD, Outlaw J, Jaykus L-A (2017) Human norovirus binding to select bacteria representative of the human gut microbiota. PLoS ONE 12(3): e0173124. https://doi.org/10.1371/journal.pone.0173124
Editor: Adriana Calderaro, Universita degli Studi di Parma, ITALY
Received: December 20, 2016; Accepted: February 15, 2017; Published: March 3, 2017
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This material is based upon work that is supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, under award number 1111-2011-0494 and 2011-68003-30395, under the NoroCORE project. Support was provided by an US Air Force Institute of Technology Civilian Institute Program Scholarship (EAA). The views expressed in this article are those of the author and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government, https://nifa.usda.gov/grants. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Human norovirus is the leading cause of acute viral gastroenteritis worldwide [1], and also the most common cause of foodborne disease, at least within the United States [2]. The public health and economic burden of norovirus infection is substantial in the Western world, and may be crippling in developing countries [3]. Increased awareness of this ubiquitous pathogen has placed considerable interest in developing diagnostics, antivirals and vaccines, as well as finding more effective ways to halt its transmission. Critical to these efforts is elucidating the details of the norovirus infection cycle. Until recently, the lack of a reliable propagation model for human norovirus (reviewed in [4,5]) had been a significant roadblock, and little is still known about the viral replication cycle.
About fifteen years ago, researchers identified human histo-blood group antigens (HBGAs) as a putative human norovirus cellular receptor, [6,7]. HBGAs are complex terminal carbohydrates present on key cellular surfaces (i.e. red blood cells and mucosal epithelium) or secreted into biological fluids (i.e. saliva, intestinal lumen) [8]. Most, but not all, human norovirus strains interact with these sugar moieties in specific binding patterns [9] linked to residues on the A/B or Lewis antigens [10].
Structures similar to human HBGAs are found ubiquitously in other animals, plants, and even in bacteria [8,11]. Recently, Miura et al. [12] demonstrated that Enterobacter cloacae spp. cloacae (ATCC 13047) binds to human norovirus GI.1 and GII.6, interactions that appears to be mediated by bacterial HBGA-like moieties. These findings and others [13,14] informed the development of a mammalian cell culture model for human norovirus propagation in which the target cell line (Burkitt’s Lymphoma (BJAB) B cells) could support modest replication of the virus only when in the presence of E. cloacae or after supplementation with synthetic HBGAs [15]. This system has recently been replicated and used to study a viral polymerase inhibitor [16]. Recently, another report of successful in vitro human norovirus replication has been reported in human intestinal enteroids. Interestingly, bacteria was not required for productive infection in this model. However, the dependency upon enteroids derived from secretors (who express HBGAs) for GII.4 Sydney 2012 suggested that HBGAs are involved to some degree in pathogenicity for some strains [17].
Despite this impressive achievement, there is still a poor understanding of the dynamics and importance of bacteria-norovirus interactions. Certainly, HBGAs are not the only compounds implicated in human norovirus binding, and E. cloacae may not be the only strain to which the virus binds. The studies to date have focused on only a limited number of bacterial strains when, in reality, microbial populations within the human gastrointestinal tract are quite complex. This finding begs the question as to whether human norovirus-bacteria interactions are an isolated occurrence or a more widespread phenomenon. The purpose of this research was to characterize the specificity and intensity of these interactions using a select group of bacterial species representative of human gut microbiota. Both Gram-negative and Gram-positive bacterial strains (reference and fecal isolates) were screened for binding with genotypes GI.6 and GII.4 (New Orleans 2009 and Sydney 2012) human norovirus strains, as well as the cultivable Tulane virus surrogate. Tulane virus is a recently discovered Calicivirus that infects the gastrointestinal tracts of rhesus monkeys, and has been shown to selectively bind type A and B HBGAs and sialic acid [18–21]. Additionally Turnip Crinkle Virus, a plant virus whose capsid organization and structure (T = 3, icosahedral symmetry, one major coat protein) closely resembles noroviruses [22] was chosen as a negative control as it does not target HBGAs as receptors. Information was collected on binding efficiency, the impact of growth conditions on bacteria-virus binding, and the location on the bacteria to which the virus bound.
Results
Human norovirus binds to bacteria
Thirteen bacterial species were originally isolated from the GI.6-positive stool sample after growth on Tryptic soy broth (TSB), TSB+5% defibrinated sheep blood (blood), Brucella agar (BA), BA+blood, Brain Heart Infusion (BHI) and/or de Man, Rogosa and Sharpe (MRS) agars under anaerobic conditions. Primary microbial analysis putatively determined identity, and after deletion of duplicates, there were eight unique species. These eight strains were sent for 16S rDNA sequencing. Of the eight isolated, three appeared different physiologically but were duplicates on the genetic level. The five resulting species were evaluated for their ability to grow on different media. Additional reference bacteria were included, i.e., L. gasseri, L. plantarum and B. thetaiotaomicron based on their beneficial role within the human gut and overall prevalence in the gastrointestinal tract [23]. While most of the bacterial isolates were able to grow in all listed media types (Table 1), the lactobacillus and bacteroides species were more fastidious (Table 1), and were only able to be compared across different viruses, not different growth conditions (Fig 1).
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.
Suspension assays in PBS were used to evaluate the efficiency of binding of select bacteria to a group of viruses. Collectively, all bacterial strains bound all human norovirus strains with high efficiency, as in all cases <1 log10 of input virus was lost to supernatant, or >90% bound (Fig 1). There were no statistically significant differences in binding efficiency when percentages were compared across the three norovirus genotypes, although there was a general trend for better binding for the GII.4 strains vs. the GI.6 strain. There were differences between bacteria binding observed for Tulane virus, for which individual bacterial strains either bound to the virus at a level statistically equivalent to the other norovirus strains, or the bacteria did not bind the virus at all. None of the bacterial strains bound the negative control, Turnip Crinkle Virus (TCV), to any appreciable degree. Taken together, these data suggest that human norovirus-bacteria binding occurs with high efficiency but with relatively low inter-species specificity.
Human norovirus-bacteria binding efficiency is impacted by culture media
To further elucidate factors impacting the dynamics of bacteria-virus binding, experiments were undertaken in which the bacterial strains were grown in enriched media (chopped meat and TSB+blood), nutrient dense media (TSB and BHI), or minimal media and then equal amounts of bacterial cells assayed for binding in PBS with GII.4 Sydney 2012, the most recent epidemic human norovirus strain. For minimal media or nutrient dense media (TSB or BHI), there were no statistically significant differences between the input virus concentrations and virus concentrations estimated as bound to bacteria for any of the bacterial species, suggesting high capture efficiency (Fig 2). On the other hand, the bacteria-virus binding efficiency notably decreased in enriched media, a phenomenon that was statistically significant (p<0.05) in most cases, and most pronounced for Bacillus spp., E. faecium, Citrobacter spp., and H. alvei. In general, TSB+5% sheep blood had the greatest negative impact on human norovirus binding to bacteria. A minimal media or reduced strength media produced the most consistent, highest degree of binding.
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.
Human norovirus targets bacterial pili and cell membranes
Three Gram-negative and three Gram-positive bacteria (two ATCC strains and four fecal isolates) were viewed using transmission electron microscopy (TEM) after exposure to either GII.4 Sydney 2012 or GII.4 Farmington Hills 2002 virus-like particles (VLPs) that contain just the purified assembled capsid protein (Fig 3A–3C and 3D–3F, respectively). Because of the need for high assay resolution, TEM experiments were done using VLPs in place of fecally-derived virus. Of the bacteria imaged, three (S. aureus (Fig 3B), E. faecium (Fig 3D), and Citrobacter spp. (Fig 3E)) showed VLPs bound to the outer cell membrane; two (E. cloacae (Fig 3A) and Bacillus spp. (Fig 3C)) had VLPs bound to pili; and one (H. alvei (Fig 3F)) showed evidence of VLP binding to both structures. For the latter two instances, the binding was scattered around the cell membrane rather than localized to specific structures.
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.
Discussion
The purpose of this research was to characterize the specificity and intensity of human norovirus interactions with a select group of bacterial species representative of the human gut microbiota. Based on the results of a suspension-RT-qPCR binding assay, we found that naturally occurring human fecally-derived bacterial isolates, as well as repeatedly cultured reference bacteria, were all capable of binding to human norovirus with relatively high efficiency (Figs 1 and 2). These interactions were limited to human norovirus strains as the genotypes tested bound well to all bacteria screened. There were limited interactions between these same bacteria and the Tulane virus surrogate, and no binding was observed between any of the bacteria and the unrelated TCV. In short, virus-bacteria binding was specific at the norovirus genus level, but relatively promiscuous when considering the bacterial species. In addition, the presence of bacteria in the human stool matrix did not appear have a considerable effect on the virus-bacterial binding interaction (data not shown), which would not be surprising because the viral inoculum was diluted 1/500 (thus 0.2% stool) prior to use in the binding assay.
The method used in these studies provided results supporting norovirus-bacteria binding, but did not provide information on how or why that binding occurred. Perhaps the most logical explanation for this binding is the expression of HBGA-like moieties by bacterial cells [11]. There is recent evidence of a naturally-occurring E. cloacae strain that readily bound to human norovirus, and HBGA (specifically, A type)-like moieties in the bacterial EPS were associated with that binding [22]. Li et al. [23] screened multiple bacterial strains for norovirus VLP binding, finding that virus binding of the bacteria correlated well with their HBGA expression profiles. In an effort to identify HBGA-type molecules in the bacteria we studied, the same assays used by these previous investigators were applied here, in addition to a previously-established ELISA assay [24]. We found all three of the assays difficult to reproduce and/or interpret (data not shown). Whether this challenge was a function of the quality and/or source of the reagents, minor differences in assay design, or simply a lack of HBGA-like activity of the bacteria is unknown. We therefore have no concrete evidence on the definitive mechanism(s) responsible for the observed virus-bacteria binding.
There is, however, some evidence supporting a potential role for bacterial HBGA-like molecules in human norovirus binding. For example, Tulane virus bound to some but not all bacterial strains tested. Tulane is genetically related to human norovirus and binds HBGAs in a highly selective manner (i.e., B and H antigens as well as sialoglycoconjugates), overlapping receptors with some human norovirus strains [25]. The failure of TCV—a genetically unrelated plant virus of similar size, shape, and charge to human norovirus [26]—to bind to any of the bacteria suggests that human norovirus-bacteria binding is more likely to be associated with a receptor-ligand (HBGA) interaction rather than charge-based or indirect interactions. Differences in bacteria-virus binding efficiency as a consequence of media formulation is also supportive of this, as differential expression of many bacterial genes occurs as a consequence of growth conditions [27]. It is possible that poor virus-bacteria binding efficiency in the presence of high nutrient composition media could be a consequence of binding site occlusion, however bacterial cells were washed and diluted prior to binding assays. Taken together, we believe the data presented here provides indirect evidence that a common ligand, like HBGA-like molecules, mediates virus-bacteria interactions.
In addition to HBGA-like molecules, bacteria possess sialylated gangliosides commonly implicated in virus binding [25,28]. These sialylated lipopolysaccharides are a major component of the outer surface of Gram-negative bacteria, on molecules which are also associated with histo-blood group activity (i.e. galactose) [29]. In the same binding pocket as HBGAs and sialic acid, human norovirus has been shown to bind citrate. This molecule is comprised of a pyranoside ring, which closely resembles, and in this case may mimic, a terminal fucose characteristic of HBGAs [19]. As citric acid is an intermediate in the tricarboxylic acid cycle, it is possible a surplus of this molecule is also aiding in norovirus binding to different bacteria [29]. While the exact molecule responsible for norovirus binding to bacteria has yet to be conclusively identified, the evidence we and others present suggests specific binding motifs rather than nonspecific, indiscriminate binding.
Unlike the findings of Miura et al. [12], who reported norovirus binding occurred to the bacterial exopolysaccharide matrix, we were unable to observe consistent VLP binding in the EPS. It should be noted that the negative staining used to prepare samples for electron microscopy made it difficult to visualize the external structures of the bacteria without compromising the ability to see the VLPs. Thus, it is possible that additional localized interactions are occurring between the virus and the bacteria in the EPS that were not visible due to methodological limitations. However, we did visualize binding to different bacterial structures, specifically cell membranes and pili, consistently. These structures on the outside of the cell may help mediate adhesion, and/or may also camouflage the bacteria from the host to prevent interactions that could be detrimental to the bacteria or the virus. From the virus’ perspective, adhering to the pili might place it closer to host cell binding regions, facilitating infection.
The impact of growth conditions on the norovirus-bacteria binding is intriguing. Mechanistically, one could make the argument that the small intestine is a “nutrient rich” environment, in which case lower degrees of bacteria-virus binding might be expected. On the other hand, the nutrient dense times in this organ are short (approximately 1–2 h [30]) and are separated by long periods devoid of nutrients. Since the bacteria would do the majority of their growth in periods with low nutrient density, it is possible that expression of binding ligands might be up-regulated during fasting. It is possible that the bacteria were competing for binding with residual HBGA-like substances in the richer blood and chopped meat media, however the cells were washed once with PBS and no difference in binding ability was seen with additional washes. Further, the washed bacteria were diluted 1/5 in a different tube for performing the suspension assay. However, there are many other factors involved (e.g., pH, temperature, bacterial interactions) that could influence bacteria-virus binding, and given the effect of growth medium we report, further elucidation of their effect on viral binding would be a logical future direction for investigation [27].
Improved understanding of norovirus-bacterial interactions has widespread implications ranging from understanding virus infectivity [13,14]; aiding in in vitro cultivation of human norovirus [15]; developing concentration and purification methods for detection [12]; and design of novel removal/inactivation methods [31,32]. Regardless of potential downstream applications, the phenomenon of bacteria and virus interaction resulting in advantages to one or both of the infectious entities is being increasingly recognized. The impact of this shifting paradigm will need to be considered in many different aspects of human norovirus study, providing fruitful avenues for further research and applications. The data we report here builds on and advances the increasingly important field of human norovirus-bacteria interactions.
Materials and methods
Virus strains and virus-like particles (VLPs)
Human fecal specimens derived from outbreaks and confirmed positive (by sequencing) for GI.6, GII.4 New Orleans 2009, and GII.4 Sydney 2012 norovirus were obtained courtesy of Dr. Shermalyn Greene, North Carolina State Laboratory of Public Health, Raleigh, NC. Stool samples were diluted to 20% (v/v) in 1X PBS, pH 7.2 (Life Technologies, Carlsbad, CA), aliquoted, and stored at -80°C until use. Tulane virus, (obtained courtesy of Dr. Jason Jiang, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH) was cultivated in the LLC-MK2 cell line (ATCC CCL7, American Type Culture Collection, Manassas VA) as previously reported [19], and harvested by three cycles of freeze-thaw at -80°C followed by centrifugation. Aliquots of the virus-rich supernatant were stored at -80°C. A purified stock of Turnip Crinkle Virus (TCV) was obtained from the laboratory of Dr. Steven Lommel (North Carolina State University, Raleigh, NC) and diluted in PBS. Virus-like particles (VLPs) for strains GII.4 Sydney 2012 (1.0mg/ml) and GII.4 Farmington Hills 2002 (1.3mg/ml) were provided courtesy of Dr. Robert Atmar (Baylor College of Medicine, Houston, TX).
Bacterial isolates
Bacterial strains used in this study are shown in Table 1. Stock cultures of Lactobacillus plantarum and Lactobacillus gasseri were obtained courtesy of Dr. Todd Klaenhammer (North Carolina State University, Raleigh, NC). Reference strains, provided by the American Type Culture Collection (ATCC; Manassass, VA) included Staphylococcus aureus (ATCC 25235), Enterobacter cloacae (ATCC 13047) and Bacteroides thetaiotaomicron (ATCC 29148). To obtain natural bacterial isolates, human stool from one patient positive for GI.6 norovirus was diluted and plated in 100μl aliquots onto tryptic soy agar (TSA; Thermo Fisher Scientific, Waltham, MA); TSA supplemented with 5% defibrinated sheep blood (Lampire Biological Products, Pipersville, PA); Brucella agar (BA; Thermo Fisher Scientific); and BA supplemented with 5% defibrinated sheep blood. For the lactobacillus species, de Man, Rogosa and Sharpe agar or broth (MRS; Thermo Fisher Scientific) was used for cultivation. These plates were incubated anaerobically (5% carbon dioxide, 10% hydrogen, 85% nitrogen) in an anaerobic glove box with CAM-12 oxygen monitoring system and heating unit (Coy Laboratories, Grass Lake, MI) for 24 h at 37°C. Unique colonies from each agar plate (i.e. size, color, shape, hemolysis) were isolated, purified, and grown anaerobically at 37°C overnight on both the agar and in the broth formulations from which they were initially isolated. Stocks were made in 30% glycerol and stored at -80°C. Each of the isolated fecally derived cultures underwent further characterization using Gram stain, motility tests and the Oxoid Microbact GNB 24E reagent system (Thermo Fisher Scientific). To identify the bacteria, whole agar plates were sent and a single colony of each bacterial isolate was sequenced via 16S rDNA by GENEWIZ (South Plainfield, NJ) (S1 Fig). Bacterial IDs were confirmed using NCBI Blast (S1 Table).
Bacteria-virus binding assays
The ability of bacteria to bind norovirus was modified from the plate-based assay of Miura et al [12] to a suspension test design. Since preliminary studies showed that bacteria-virus binding was highly dependent upon initial bacterial growth conditions (data not shown), several different media formulations were used for bacterial propagation. Specifically, for minimal media, two formulations were used. The Gram-negative medium consisted of M9 minimal salts (supplemented with 20% glucose, 0.2% MgSO4 and 0.01% CaCl2; Sigma-Aldrich, St. Louis, MO), while the Gram-positive bacteria were grown in one-half strength Tryptic Soy Broth (TSB). On the other end of the nutrient spectrum both Gram-negative and Gram-positive bacteria (E. cloacae, S. aureus and the fecal isolates) were grown in two rich media types typically used to cultivate fastidious microbes, i.e., chopped meat broth (Anaerobe Systems, Morgan Hill, CA) and TSB supplemented with defibrinated sheep blood (Lampire Biological Laboratories, Pipersville, PA). Prior to experimentation, starting concentrations (input) of bacteria were enumerated for each bacterial strain and media combination via growth curves to ensure that comparable amounts of bacteria were used in each binding assay.
Prior to binding assays, bacterial cultures were grown anaerobically for 24 h at 37°C in 10 ml of select medium. The cells were then pelleted, washed once, and resuspended in 10 ml PBS, pH 7.2. These stocks were diluted to concentrations of approximately 1x107cfu/ml. Human norovirus suspensions were diluted 100-fold in PBS to concentrations around 1x105 RT-qPCR amplifiable units (RT-qPCRU)/ml (GII.4 Sydney 2012), 1x103 RT-qPCRU /ml (GII.4 New Orleans 2009) and 1x104 RT-qPCRU /ml (GI.6). Tulane virus and TCV were likewise diluted in PBS to reach a final concentration of 1x107 RT-qPCRU /ml and 1x106 RT-qPCRU/ml, respectively.
For each suspension assay, 100μl of resuspended bacteria was diluted in 300μl PBS, and 100μl of diluted virus suspension was added. The mixture was incubated for 2 h at 37°C with rotation. The mixture was pelleted at 10,000 x g for 5 min at room temperature and the supernatant was removed for enumeration of remaining (unbound) viruses using RT-qPCR. The pellet was also retained for comparison. Positive controls consisted of input virus suspension without exposure to bacteria; negatives controls consisted of PBS alone.
Nucleic acid extraction and detection of viral RNA
The bacterial supernatants and in some instances, pellets, were processed for RNA extraction using the NucliSENS® easyMAG automated system (bioMérieux SA, Marcy l’Etoile, France) as per manufacturer instructions, with a final resuspension volume of 50μl.
Detection of viral RNA was carried out by Reverse Transcriptase quantitative PCR (RT-qPCR) using primers and probes outlined in Table 2. The 25μl reaction consisted of 12.5μl 2X reaction buffer (SuperScript® III One-Step qRT-PCR Kit, Invitrogen, Grand Island, NY), 0.5μl RT/Platinum® Taq Mix, 400nM forward and reverse primers (Table 2; Integrated DNA Technologies, Coralville, IA), 200nM fluorescently labeled TaqMan probe (Table 2; Integrated DNA Technologies) and 2.5μl of RNA. A CFX96 Touch™ Real Time PCR Detection System (Bio-Rad, Hercules, CA) thermocycler was used with the following amplification conditions: (1) reverse transcription for 30 min at 50°C; (2) denaturation for 15 min at 95°C; and (3) 45 cycles of 15 s at 95°C then 30 s at 60°C. All primers and probes utilized the same amplification protocol.
RT-qPCR standard curves for quantification were constructed (in triplicate) as previously described [26] using ten-fold serial dilutions of viral RNA in DEPC-treated water. The CT value corresponding to each serial dilution was plotted and the data analyzed using linear regression to determine a slope. The lower limit of detection for a given standard curve was considered the lowest dilution at which all three replicates were positive for viral RNA; this was designated as 1 RT-qPCRU. The virus input concentration (in RT-qPCRU) before exposure to the bacteria, and that in the supernatant after exposure to the bacteria, were estimated by comparison to the standard curve. Because the amplifications associated with the viruses bound to the pellet were so variable (potentially due to PCR interference associated with excessive amounts of background DNA, data not shown), virus binding efficiency was computed based on loss to supernatant, i.e., ((Total virus input-virus lost in supernatant)/total virus input), expressed as a percentage [36].
Transmission electron microscopy (TEM)
The bacteria-virus binding experiments were done as described above except for the use of 10μg of VLP (GII.4 Sydney 2012 or GII.4 Farmington Hills 2002). After the bacteria-virus mixture was pelleted, the precipitate was washed and resuspended in 500μl 20mM HEPES (Life Technologies, Carlsbad, CA). Fifteen μl of each sample was applied to 400-mesh carbon-nickel coated grids (Ladd Research, Williston, VT) for 10 min and the excess sample removed using Whatman filter paper (Grade 2, 8μm pore size, Sigma-Aldrich, St. Louis, MO). Negative staining was done with 10μl 2% aqueous uranyl acetate for 60 s. Negative controls consisted of bacteria without exposure to VLPs. Images were visualized using the JEOL 1210 transmission electron microscope (JEOL-USA, Inc., Peabody, MA) at 80 kV at the Center for Electron Microscopy (North Carolina State University, Raleigh, NC).
Statistical analysis
Three independent experiments were performed for each binding assay. Binding efficiency data were expressed as described above, and also as ± standard deviations across the experiments, depicted by error bars. Statistical comparisons between the log10 concentrations of unbound virus based on different bacterial strains (Fig 1) or media type (Fig 2) were done in JMP 11 (SAS Institute, Cary, NC) using the Tukey-Kramer honest significant difference test and considered significant if p <0.05.
Supporting information
S1 Fig. Raw 16S sequence information for bacteria isolated from human stool.
Data is provided for both forward and reverse reads and isolate names refer to isolates described in S1 Table. Raw forward and reverse reads of the isolates relevant to this paper are provided in FASTA format below.
https://doi.org/10.1371/journal.pone.0173124.s001
(PDF)
S1 Table. Top matches for 16S rRNA sequences of bacteria isolated from human stool samples.
Different samples of human stool were streak plated and grown in both aerobic and anaerobic conditions. Specific colonies were isolated, further cultured and their 16S rRNA region sequenced. Below is a summary table of the top sequence isolates selected for further analysis.
https://doi.org/10.1371/journal.pone.0173124.s002
(PDF)
Acknowledgments
This material is based upon work that is supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, under award number 1111-2011-0494 and 2011-68003-30395, under the NoroCORE project. Support was provided by an US Air Force Institute of Technology Civilian Institute Program Scholarship (EAA). The views expressed in this article are those of the author and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government. We thank V. Lapham for her endless patience and knowledge when it comes to all things TEM. We thank Capt C. Runch for his coordination and assistance. The views expressed in this article are those of the author and do not reflect the official policy or position of the United States Air Force, Department of Defense, or the U.S. Government.
Author Contributions
- Conceptualization: EAA LAJ.
- Data curation: EAA LAJ.
- Formal analysis: EAA.
- Funding acquisition: LAJ.
- Investigation: EAA MDM JO.
- Methodology: EAA MDM JO LAJ.
- Project administration: EAA MDM JO LAJ.
- Resources: LAJ.
- Supervision: LAJ.
- Validation: EAA MDM JO LAJ.
- Visualization: EAA.
- Writing – original draft: EAA.
- Writing – review & editing: EAA MDM LAJ.
References
- 1. Ahmed SM, Hall AJ, Robinson AE, Verhoef L, Premkumar P, Parashar UD, et al. Global prevalence of norovirus in cases of gastroenteritis: a systematic review and meta-analysis. The Lancet Infectious Diseases. 2014;14: 725–730. pmid:24981041
- 2. Scallan E, Hoekstra RM, Angulo FJ, Tauxe RV, Widdowson M-A, Roy SL, et al. Foodborne illness acquired in the United States—major pathogens. Emerg Infect Dis. 2011;17.
- 3. Scharff RL. Economic Burden from Health Losses Due to Foodborne Illness in the United States. j food prot. International Association for Food Protection; 2012;75: 123–131. pmid:22221364
- 4. Thorne LG, Goodfellow IG. Norovirus gene expression and replication. J Gen Virol. 2014;95: 278–291. pmid:24243731
- 5. Karst SM, Wobus CE, Goodfellow IG, Green KY, Virgin HW. Advances in Norovirus Biology. Cell Host and Microbe. Elsevier Inc; 2014;15: 668–680. pmid:24922570
- 6. Harrington PR, Vinjé J, Moe CL, Baric RS. Norovirus capture with histo-blood group antigens reveals novel virus-ligand interactions. Journal of Virology. 2004;78: 3035–3045. pmid:14990722
- 7. Marionneau S, Ruvoën N, Le Moullac Vaidye B, Clement M, Cailleau Thomas A, Ruiz Palacois G, et al. Norwalk virus binds to histo-blood group antigens present on gastroduodenal epithelial cells of secretor individuals. Gastroenterology. 2002;122: 1967–1977. pmid:12055602
- 8. Marionneau S, Cailleau Thomas A, Rocher J, Le Moullac Vaidye B, Ruvoën N, Clement M, et al. ABH and Lewis histo-blood group antigens, a model for the meaning of oligosaccharide diversity in the face of a changing world. Biochimie. 2001;83: 565–573. pmid:11522384
- 9. Cao S, Lou Z, Tan M, Chen Y, Liu Y, Zhang Z, et al. Structural basis for the recognition of blood group trisaccharides by norovirus. Journal of Virology. American Society for Microbiology; 2007;81: 5949–5957. pmid:17392366
- 10. Huang P, Farkas T, Zhong W, Tan M, Thornton S, Morrow AL, et al. Norovirus and histo-blood group antigens: demonstration of a wide spectrum of strain specificities and classification of two major binding groups among multiple binding patterns. Journal of Virology. American Society for Microbiology; 2005;79: 6714–6722. pmid:15890909
- 11. Springer GF, Williamson P, Brandes WC. BLOOD GROUP ACTIVITY OF GRAM-NEGATIVE BACTERIA. The Journal of Experimental Medicine. The Rockefeller University Press; 1961;113: 1077–1093. pmid:19867191
- 12. Miura T, Sano D, Suenaga A, Yoshimura T, Fuzawa M, Nakagomi T, et al. Histo-blood group antigen-like substances of human enteric bacteria as specific adsorbents for human noroviruses. Journal of Virology. American Society for Microbiology; 2013;87: 9441–9451. pmid:23804639
- 13. Kuss SK, Best GT, Etheredge CA, Pruijssers AJ, Frierson JM, Hooper LV, et al. Intestinal microbiota promote enteric virus replication and systemic pathogenesis. Science. American Association for the Advancement of Science; 2011;334: 249–252. pmid:21998395
- 14. Robinson CM, Jesudhasan PR, Pfeiffer JK. Bacterial Lipopolysaccharide Binding Enhances Virion Stability and Promotes Environmental Fitness of an Enteric Virus. Cell Host and Microbe. 2014;15: 36–46. pmid:24439896
- 15. Jones MK, Watanabe M, Zhu S, Graves CL, Keyes LR, Grau KR, et al. Enteric bacteria promote human and mouse norovirus infection of B cells. Science. American Association for the Advancement of Science; 2014;346: 755–759. pmid:25378626
- 16. Kolawole AO, Rocha-Pereira J, Elftman MD, Neyts J, Wobus CE. Inhibition of human norovirus by a viral polymerase inhibitor in the B cell culture system and in the mouse model. Antiviral Res. 2016;132: 46–49. pmid:27210811
- 17. Ettayebi K, Crawford SE, Murakami K, Broughman JR, Karandikar U, Tenge VR, et al. Replication of human noroviruses in stem cell-derived human enteroids. Science. 2016;353: 1387–1393. pmid:27562956
- 18. Farkas T, Cross RW, Hargitt E III, Lerche NW, Morrow AL, Sestak K. Genetic Diversity and Histo-Blood Group Antigen Interactions of Rhesus Enteric Caliciviruses. Journal of Virology. American Society for Microbiology; 2010;84: 8617–8625. pmid:20554772
- 19. Farkas T, Sestak K, Wei C, Jiang X. Characterization of a rhesus monkey calicivirus representing a new genus of Caliciviridae. Journal of Virology. American Society for Microbiology; 2008;82: 5408–5416. pmid:18385231
- 20. Tan M, Wei C, Huang P, Fan Q, Quigley C, Xia M, et al. Tulane virus recognizes sialic acids as cellular receptors. Sci Rep. 2015;5: 11784. pmid:26146020
- 21. Zhang D, Huang P, Zou L, Lowary TL, Tan M, Jiang X. Tulane Virus Recognizes the A Type 3 and B Histo-Blood Group Antigens. Journal of Virology. American Society for Microbiology; 2015;89: 1419–1427. pmid:25392226
- 22. Prasad BVV, Matson DO, Smith AW. Three-dimensional Structure of Calicivirus. Journal of Molecular Biology. 1994;240: 256–264. pmid:8028008
- 23. Kleerebezem M, Vaughan EE. Probiotic and gut lactobacilli and bifidobacteria: molecular approaches to study diversity and activity. Annu Rev Microbiol. 2009;63: 269–290. pmid:19575569
- 24. Manuel CS, Moore MD, Jaykus LA. Destruction of the Capsid and Genome of GII.4 Human Norovirus Occurs during Exposure to Metal Alloys Containing Copper. Appl Environ Microbiol. American Society for Microbiology; 2015;81: 4940–4946. pmid:25979897
- 25. Rydell GE, Nilsson J, Rodriguez-Diaz J, Ruvoën-Clouet N, Svensson L, Le Pendu J, et al. Human noroviruses recognize sialyl Lewis x neoglycoprotein. Glycobiology. Oxford University Press; 2009;19: 309–320. pmid:19054801
- 26. Gentry-Shields J, Jaykus L-A. Comparison of process control viruses for use in extraction and detection of human norovirus from food matrices. Food Research International. 2015.
- 27. Moody MR, Young VM, Faber JE. Effect of cultural environment on the blood group activity of microorganisms. Appl Microbiol. 1969;18: 262–267. pmid:4896884
- 28. Han L, Tan M, Xia M, Kitova EN, Jiang X, Klassen JS. Gangliosides are ligands for human noroviruses. J Am Chem Soc. 2014;136: 12631–12637. pmid:25105447
- 29. Hansman GS, Shahzad-Ul-Hussan S, McLellan JS, Chuang G-Y, Georgiev I, Shimoike T, et al. Structural basis for norovirus inhibition and fucose mimicry by citrate. Journal of Virology. American Society for Microbiology; 2012;86: 284–292. pmid:22031945
- 30. Macfarlane S. Microbial Biofilm Communities in the Gastrointestinal Tract. Journal of Clinical Gastroenterology. 2008;42: S142–S143. pmid:18806707
- 31. Li D, Breiman A, Le Pendu J, Uyttendaele M. Binding to histo-blood group antigen-expressing bacteria protects human norovirus from acute heat stress. Frontiers in Microbiology. Frontiers Media SA; 2015;6: 659. pmid:26191052
- 32. Amarasiri M, Hashiba S, Miura T, Nakagomi T. Bacterial histo-blood group antigens contributing to genotype-dependent removal of human noroviruses with a microfiltration membrane. Water research. 2016.
- 33. Kageyama T, Kojima S, Shinohara M, Uchida K, Fukushi S, Hoshino FB, et al. Broadly reactive and highly sensitive assay for Norwalk-like viruses based on real-time quantitative reverse transcription-PCR. Journal of Clinical Microbiology. 2003;41: 1548–1557. pmid:12682144
- 34. Jothikumar N, Lowther JA, Henshilwood K, Lees DN, Hill VR, Vinjé J. Rapid and sensitive detection of noroviruses by using TaqMan-based one-step reverse transcription-PCR assays and application to naturally contaminated shellfish samples. Appl Environ Microbiol. American Society for Microbiology; 2005;71: 1870–1875. pmid:15812014
- 35. Sestak K, Feely S, Fey B, Dufour J, Hargitt E, Alvarez X, et al. Experimental Inoculation of Juvenile Rhesus Macaques with Primate Enteric Caliciviruses. Cameron DW, editor. PLoS ONE. 2012;7: e37973. pmid:22666426
- 36. Stevens KA, Jaykus LA. Direct detection of bacterial pathogens in representative dairy products using a combined bacterial concentration-PCR approach. J Appl Microbiol. Blackwell Science Ltd; 2004;97: 1115–1122. pmid:15546401