Reduced infectivity of waterborne viable but nonculturable Helicobacter pylori strain SS1 in mice

Abstract Background Helicobacter pylori infection has been consistently associated with lack of access to clean water and proper sanitation, but no studies have demonstrated that the transmission of viable but nonculturable (VBNC) H. pylori can occur from drinking contaminated water. In this study, we used a laboratory mouse model to test whether waterborne VBNC H. pylori could cause gastric infection. Materials and Methods We performed five mouse experiments to assess the infectivity of VBNC H. pylori in various exposure scenarios. VBNC viability was examined using Live/Dead staining and Biolog phenotype metabolism arrays. High doses of VBNC H. pylori in water were chosen to test the “worst‐case” scenario for different periods of time. One experiment also investigated the infectious capabilities of VBNC SS1 using gavage. Further, immunocompromised mice were exposed to examine infectivity among potentially vulnerable groups. After exposure, mice were euthanized and their stomachs were examined for H. pylori infection using culture and PCR methodology. Results VBNC cells were membrane intact and retained metabolic activity. Mice exposed to VBNC H. pylori via drinking water and gavage were not infected, despite the various exposure scenarios (immunocompromised, high doses) that might have permitted infection with VBNC H. pylori. The positive controls exposed to viable, culturable H. pylori did become infected. Conclusions While other studies that have used viable, culturable SS1 via gavage or drinking water exposures to successfully infect mice, in our study, waterborne VBNC SS1 failed to colonize mice under all test conditions. Future studies could examine different H. pylori strains in similar exposure scenarios to compare the relative infectivity of the VBNC vs the viable, culturable state, which would help inform future risk assessments of H. pylori in water.


| INTRODUCTION
Helicobacter pylori (H. pylori) is a gastrointestinal bacterium that causes gastritis, peptic ulcers and, over time, gastric adenocarcinoma. 1,2 Helicobacter pylori infection is hypothesized to be transmitted through multiple routes, including vertically from mother to child and through contaminated reservoirs like food and water. 3,4 A body of evidence suggests that contaminated water may be a source of H. pylori infection, with epidemiological studies consistently associating H. pylori infection with lack of access to potable drinking water and proper sanitation. 3,[5][6][7][8][9] Furthermore, H. pylori has been detected in water using various molecular biology techniques, such as quantitative polymerase chain reaction (qPCR) and microscopy methods, 5,[10][11][12][13] and there are reports that it has been cultured from water. [14][15][16][17] Helicobacter pylori enters a viable but not culturable (VBNC) state within a few days after inoculation into water. [18][19][20] This change is often accompanied by a morphological change from a spiral bacillus to a U-shaped or coccoid form, and H. pylori has been found in the VBNC state in all these morphologies in the natural environment. 18,21 However, although H. pylori has been cultured from wastewater and drinking water, it is unclear whether this was due to the culturable form being present in the water or investigators being able to revert the VBNC form back to a culturable form using appropriate media.
The fact that H. pylori is present in both a culturable and VBNC state has not been accounted for when assessing risk associated with waterborne H. pylori. For example, a risk model of waterborne H. pylori infection using a quantitative microbial risk assessment methodology 22

| Transmission and exposure groups
Our studies were carried out sequentially following our initial dosing experiments that examined the infectious dose of viable, culturable H. pylori in water. 20  We increased the number of days of exposure (six instead of one), and also exposed severe combined immunodeficient mice to a single day of waterborne H. pylori, hypothesizing that more doses and immunocompromised hosts would be more likely to increase infection based on the results of our previous experiments. 20 When these also failed to induce infection, we increased the exposure length again and increased the number of mice to 100 to increase the likelihood of seeing infection. In these experiments, we used a similar experimental design to our original dosing studies, 20 exposing the mice to 56 days of contaminated water (experiment 4), and further decreasing the time until euthanasia. When this also failed to induce infection, we did a final follow-up study in which we gavaged mice with four doses of ~2*10 8 cells of VBNC SS1 over 2 weeks. This, too, failed to induce infection. The mice were exposed to water contaminated with ~10 9 cells/L VBNC H. pylori (See Table 1). In experiments 1-3, contaminated water was removed after 24 hours and replaced with either a bottle of freshly contaminated water or (when appropriate) sterilized, filtered tap water.
Each exposure group had 20 cages, with two mice per cage per the Animal Care and Use Committee regulations. In experiment 4, water was changed twice per week, every 3-4 days. As a negative control, 10 mice (five cages) were given sterile, filtered tap water for 9 weeks. As a positive control, 10 mice (five cages) were given sterile, filtered tap water inoculated with viable, culturable H. pylori for 9 weeks. All mice were housed at University Laboratory Animal Medicine facilities at the University of Michigan Medical School, and all experiments were approved by the Animal Care and Use Committee.

| Bacterial strain
SS1 (Sydney Strain 1) was selected for this study for consistency with our previous studies, 20 and because it colonizes mice more successfully than other H. pylori strains. 26

| H. pylori cultivation, counting, and inoculation
Helicobacter pylori cultivation was carried out as previously described. 20 Briefly, SS1 was plated from stocks and grown at 37°C on 5% sheep blood tryptic soy agar II plates (BBL, Franklin Lakes, NJ, USA) in microaerobic conditions. After 3 days, colonies were collected and suspended

| Dose estimation
To estimate the doses consumed by the mice, water bottles were weighed before being placed in cages and immediately after their removal. As water drips out of water bottles when they are placed in the cage and when the cages are moved, "dummy" bottles were filled with water and treated in the exact same way as experimental bottles.
The amount of water lost from dummy bottles was averaged and that average was subtracted from the total water lost from each bottle. As mice were housed two per cage, the adjusted total per cage was then halved to provide the individual dose per mouse.

| Morphology, culturability, and metabolic activity of H. pylori in water
In all experiments, there was a complete loss of culturability of H. pylori 2-3 days after initial inoculation into water. Despite being nonculturable, cells were still found to be membrane intact via Live/Dead staining 8 days after inoculation into water (Figure 1).
The VBNC H. pylori cells also induced color changes in the Biolog PM1 panels at each time point, respectively (Figure 2). This suggests that the cells were metabolically active, as they were metabolizing the carbon sources in each well. The cells in the viable, culturable state utilized many more carbon sources than any of the cells in the VBNC state. No differences in metabolic activity were seen between VBNC cells on days 3, 4, 7, and 8 (data not shown).  Table 2. Further, the mice dosed with SS1 via gavage were also not infected.

| Exposure to waterborne H. pylori
The negative controls showed no signs of infection, and confirmed H. pylori cultures were recovered from 8 of 10 positive controls. None of the mice exposed to VBNC H. pylori showed any sign of infection, either via culture or PCR.
F I G U R E 1 60× magnification of Helicobacter pylori suspension in water after 8 d of incubation at room temperature. Green cells are membrane intact, and red cells have membrane damage. The predominant form was coccoid F I G U R E 2 PM1 plates of Day 0 (viable and culturable) and Day 8 Helicobacter pylori cells (VBNC). Each well contains a different carbon source, and wells with a purple color change indicate that the carbon source was being used. Viable culturable H. pylori utilized a much wider variety of carbon sources than the VBNC H. pylori

| DISCUSSION
We were unable to cause infection in mice with the VBNC form of SS1, either in drinking water or via gavage. Our inability to cause infection was surprising, given the known capacity of this strain to successfully infect mice, 26,32 our wide range of exposure scenarios, and our previously published study that showed that SS1 in water could infect mice in a dose-dependent manner. 20 In our previous study, 4 weeks of exposure to water spiked with 10 9 CFU/L, 10 8

| Limitations and public health implications
As with any animal study, we cannot be certain that our results accurately reflect what would occur with human exposure. As H. pylori is a human pathogen, it is possible that the VBNC form is more infectious in humans than in mice. Further, we only exposed mice to one strain of H. pylori, and it is possible that other strains would be more infectious in the VBNC state than SS1, as has been seen in other published papers in the literature. 19

| CONCLUSIONS
We found that mice exposed to VBNC SS1 H. pylori via drinking water were not infected, despite the various exposure scenarios