Monitoring levels of viable Helicobacter pylori in surface water by qPCR in Northeast Spain

Helicobacter pylori is a microaerophilic Gram-negative bacterium that has been estimated to infect the gastric epithelium of half of the human population (Brown 2000). Furthermore, H. pylori is considered the causative agent of chronic active gastritis and an important risk factor in the multifactorial aetiology of peptic ulcers, gastric carcinoma and MALT (mucosa-associated lymphoid tissue) lymphoma (Gasparetto et al. 2012). Although transmission presumably occurs through faecal–oral and oral–oral routes, the main route of H. pylori transmission remains unknown. Evidence favours a faecal–oral route (Rolle-Kampczyk et al. 2004), where water plays an important role, acting as a natural reservoir and a source of infection (Hulten et al. 1998; Azevedo et al. 2008). Although some reports indicate the presence of H. pylori DNA in water samples, the viability of the cells in the environment remains unknown.

In laboratory conditions, H. pylori's ability to survive in an infectious state is hindered because its cultivability is rapidly lost (Baker et al. 2002; Moreno et al. 2007). In detrimental environmental circumstances, such as in bodies of water, some authors have suggested that H. pylori may transform from the bacillary to the coccoid form, remaining viable and virulent but non-cultivable (Giao et al. 2008; Dube et al. 2009). However, results have been reported to challenge such a hypothesis. Interestingly, some studies have reported that H. pylori is capable of spreading and remaining viable for several weeks in the presence of amoebae and biofilms (Winiecka-Krusnell et al. 2002; García et al. 2014). Therefore, new methods should be developed to detect and study the viability of these bacteria in surface water.

Given the cultivability limitations under in vitro conditions, H. pylori detection in water samples has been conducted mainly using molecular techniques including polymerase chain reaction (PCR), quantitative PCR (qPCR) and fluorescent in situ hybridization (FISH) (Bragança et al. 2007). Although qPCR is a fast and sensitive method to detect microbial DNA, it does not differentiate between viable and non-viable cells. Therefore, sample pretreatment with viability dyes, such as propidium monoazide (PMA), has been used in combination with qPCR to distinguish between viable and non-viable cells (Cawthorn & Witthuhn 2008; Slimani et al. 2012). PMA is a DNA-intercalating agent that selectively penetrates cells with compromised membranes, traditionally considered to be dead. Once inside the cell, PMA is covalently cross-linked to DNA through light photoactivation, resulting in irreversible DNA modification and subsequent inhibition of its amplification (Nocker et al. 2009; Fittipaldi et al. 2012). The effectiveness of this technique has been tested on different types of bacteria, including H. pylori in drinking water samples and by in vitro studies (Agustí et al. 2010; Santiago et al. 2015).

The aim of this study was to determine the presence of viable H. pylori in urban surface waters using viability qPCR (PMA-qPCR) to understand the role of surface water in the pathogen's transmission and improve our understanding of infection pathways.

A total of 23 sampling points were randomly selected along a stream of water in the Terrassa's Vallparadís public park, in Northeast Spain. During the summer of 2014, four samples were taken each day, two in the morning and two in the afternoon, in different arbitrarily chosen parts of the stream. Each sample was collected aseptically in a 50 mL sterile vial and transported to the laboratory at 4 °C for subsequent analysis. In the laboratory the two samples collected in the morning were divided in half; two portions were treated with PMA and two served as controls; the same procedure was then performed for the afternoon samples, for a total of eight subsamples analysed each day. Subsamples were concentrated by centrifugation twice. A first step at 4,000 × g for 30 min was performed to reach a 2 mL volume of concentrate which was placed into a microcentrifuge tube. Afterwards, concentrates were centrifuged at 10,000 × g for 15 min (Minispin Plus-Eppendorf, Hamburg, Germany). The supernatant was discarded, and pellets were resuspended in 500 μL of phosphate buffered saline (PBS) at pH 7.4 for PMA treatment and DNA extraction.

Four concentrates from the daily samples were treated with PMA using a modified method (Agustí et al. 2010). To produce a final dye concentration of 50 μM, 12.5 μL PMA (GenIUL, Terrassa, Spain) was added to each concentrate. Samples were incubated for 10 min in the Dark Box System (GenIUL) with occasional mixing and photoactivated for 15 min using the PhAST^®-Blue system (GenIUL). The other four concentrates did not receive PMA treatment and were used as controls. Subsequently, both sets of concentrates were centrifuged at 10,000 × g for 5 min. The supernatants were discarded and the pellets were resuspended in 200 μL of sterile water.

For DNA extraction, PMA-treated and non-treated concentrates were processed using the E.Z.N.A.^® DNA tissue kit (Omega Bio-Tek, Doraville, USA), following the manufacturer's recommendations. The DNA isolated was eluted in 200 μL buffer solution and stored at −20 °C until PCR amplification.

qPCR was performed to simultaneously amplify s/m alleles of vacA gene H. pylori using previously reported specific primers according to Atherton et al. (1995). Additionally, qPCR specificity was tested using DNA of two laboratory H. pylori strains (Tx30a and 60190), which were donated by Dr Xavier Calvet (Department of Gastroenterology, Hospital de Sabadell, Barcelona); DNA from Pseudomonas aeruginosa (CECT 110), Escherichia coli (CECT 101) and Bifidobacterium longum (CECT 4551) was also used as amplification controls. A conventional PCR protocol previously standardized by Dr Calvert's laboratory was run on a LightCycler 1.5 system (Roche, Mannheim, Germany). The PCR mixture contained 5 μL of genomic DNA, 0.25 μL of each primer (0.5 mM) and 4 μL of HOT FIREPol^® EvaGreen^® qPCR Mix Plus (1×) dye solution (Solis BioDyne, Tartu, Estonia) prepared according to the manufacturer's recommendations. The mix was forced down the PCR tube by a short spin in a microcentrifuge. The protocol consisted of one step of 15 min at 95 °C for polymerase activation, and 45 cycles: 95 °C for 20 s, 50 °C for 20 s, and 72 °C for 20 s with slopes of 20 °C s⁻¹, 20 °C s⁻¹ and 20 °C s⁻¹, respectively. Finally, melting point analysis was performed by raising the temperature slowly (0.1 °C s⁻¹) from 65 °C to 97 °C. A standard curve with a linear range across six logarithms of H. pylori DNA concentration was obtained from a series of 10-fold dilutions of a DNA Tx30a solution. Some of these dilutions were used as positive amplification controls, and PCR-grade water was used as negative control in all qPCR assays. Test detection limit was calculated according to the standard approach described in ISO-12869 (ISO 2012). Mean values of the cycles threshold (Ct) obtained ranged from 17 to 33.48, corresponding to concentrations from 1 to 10 × 10⁷ colony-forming units (CFU) 100 mL⁻¹; thus, the detection limit was 150 CFU 100 mL⁻¹. Samples were considered positive when their melting temperature (Tm) values were similar to the Tm of the standard curve (89.33 °C); qPCR amplification of the non-Helicobacter bacteria showed negative signals in the melting curve analyses, demonstrating the assay's specificity. Bacteria concentration was expressed as logarithmic units of cells per 100 mL of sample.

The statistical analysis was conducted using SPSS for Windows version 20 (SPSS Inc., Chicago, IL, USA). Bacterial concentrations were expressed by the mean and the standard error of the mean. The coefficient of variation was used to compare the degree of variation across groups while the nonparametric, median-based Levene's test was used to compare sample variances between groups (Total cell and Live). A probability (P) value of less than 0.05 was used as criterion of significance.

As shown in Figure 1, 21 (91%) out of 23 surface water samples were positive for the presence of H. pylori using PMA-qPCR. The total average concentration of H. pylori cells was 4.2 ± 1.52 log cell 100 mL⁻¹, while the total average concentration of viable cells was 3.46 ± 1.06 log cell 100 mL⁻¹. However, the results showed that when total concentrations were lower than 3 log cell 100 mL⁻¹, the concentration of viable cells coincided with the total cells detected. Additionally, it was observed that the samples collected on rainy days contained increased amounts of total H. pylori cells, but when compared with viable cells there was no significant statistical difference.

In order to estimate the degree of variance, the coefficient of variation was estimated for each sampling and is shown in Table 1.

Although some assays indicated variation, the analysis showed that these differences were not statistically significant (P = 0.372) (Table 2).

Evidence supporting the H. pylori water transmission hypothesis comes largely from two areas of research: epidemiological studies showing an association between the prevalence of H. pylori and water-related sources, and laboratory studies that have detected or isolated H. pylori in water sources (Aziz et al. 2015). The presence of H. pylori DNA in river waters has been previously reported by using PCR techniques. However, as previously mentioned, PCR is not suitable for determining the infective potential of a given sample. In this study, qPCR was combined with PMA to study H. pylori viability in environmental samples, and thus better understanding the potential role of surface water as a H. pylori transmission vehicle. A high frequency of viable H. pylori cells (91.3%) was found in the analysed samples. The presence of H. pylori in surface water can be explained by its ability to adapt to harsh conditions. Therefore, the adherence of H. pylori to biofilms fosters a microaerophilic environment for the survival of the bacteria, which seems to be independent of temperature (Azevedo et al. 2008; Giao et al. 2008). In addition, it has been suggested that H. pylori can propagate and remain viable for several weeks in the presence of Acanthamoeba spp., which suggests that amoebae could be a natural reservoir for the bacterium (Winiecka-Krusnell et al. 2002; Greub & Raoult 2004). The PMA-PCR technique was used to demonstrate that H. pylori is able to survive chlorination treatment in association with A. castellani (Moreno et al. 2007).

In the present study, bacterial concentration levels ranged from 2.5 to 6 log units per 100 mL. It is important to note that higher bacterial levels were detected in samples collected on rainy days. This could be explained by surface water run-off, which can be caused by stormwater, a leading source of pollution to fresh incoming urban waters. Previous studies have indicated that surface water pollution by faecal bacteria and both inorganic and organic chemicals could be exacerbated following rainfall. Furthermore, precipitation causes a lowering in temperatures and a dilution effect that reduces salinity, and thus, favours the establishment of bacteria, including H. pylori (Gubler et al. 2001).

Although the minimum H. pylori infectious dose has not yet been established in humans, some studies suggest that 10⁵ CFUs might be close to the minimum infectious dose (Graham et al. 2004). In the present study, 21.7% (5/23) of the samples showed viability levels in concentrations over 4 log cell 100 mL⁻¹, while 8.7% (2/23) presented viable cell concentrations higher than 5 log cell 100 mL⁻¹. These levels might be considered an environmental hazard to human health, given that viable coccoid forms are able to maintain their urease activity and preserve their ability to adhere to epithelial cells from gastric mucosa (She et al. 2003).

In this study, H. pylori sample positivity was the same for both qPCR and PMA-qPCR methods alike. However, when bacterial concentrations were high (>4 log cell 100 mL⁻¹), mostly on rainy days, an average difference of 1.7 log cell 100 mL⁻¹ was observed between methods, being higher for qPCR, which might mainly be due to the fraction of dead cells also accounted for by this technique. Although PMA has been proposed as a more appropriate DNA-intercalating dye with a substantially higher specificity for live cells than ethidium bromide monoazide (EMA) (Fittipaldi et al. 2012), the possibility of PMA uptake in bacterial cells with intact membranes should also be considered. It has been shown that membrane permeability undergoes changes in live cells dependent on the physiological status (Gedalanga & Olson 2009). However, a previous permeability study has shown that membranes of live H. pylori cells effectively prevent penetration of PMA (50 μM, exposure for 5 min) but allow the passage of EMA (Nam et al. 2011).

Sample pre-treatment with viability stains has been used mainly in combination with PCR. However, some practical limitations still remain when applied to environmental samples. In some cases, treatment with PMA can lead to false positive or negative results (Fittipaldi et al. 2011). The efficiency of the PMA-qPCR technique depends on a complex set of parameters, which include: dye concentration, cell concentration, live and dead cells ratio, PCR amplicon length and potentially targeted DNA sequence, turbidity, pH and salt sample concentration, incubation temperature, light source and even the microbial species. It has been suggested that the activity of the PMA can be saturated if high numbers of cells are present in the sample (Varma et al. 2009). Thus, the presence of high concentrations of other microorganisms different from the target microorganisms may lead to false positives. In order to achieve an improved estimation of the number of viable cells, the performance of three independent PCR amplifications for each sample – one regular qPCR reaction, one v-qPCR reaction and one v-qPCR on an aliquot subjected to lethal conditions inflicting membrane damage – was previously proposed (Fittipaldi et al. 2012). These assays are intended to provide more reliable data regarding the number of live microbes by comparison and subtraction of results. The approach might help to reduce overestimation of bacterial viability in complex matrices like environmental samples. Another alternative is the use of PMA-qPCR combined with other specific techniques, such as direct viable count combined with fluorescence in situ hybridization (DVC-FISH) (Santiago et al. 2015).

Our study is the first to successfully identify viable H. pylori cells in small volumes of environmental water samples. It demonstrates the usefulness of PMA-based qPCR to detect H. pylori contamination and to estimate the concentration of live cells exposed to environmental stress conditions in a complex matrix such as urban surface water. In conclusion, our investigation provides a quick and simple procedure for evaluating H. pylori viable cells in environmental samples by quantitative viability PCR. Moreover, the results suggest that urban surface water can be contaminated with viable H. pylori, which strengthens the argument in favour of water as an environmental reservoir for H. pylori transmission. Nevertheless, and given the small number of samples analysed, it is not possible to state that H. pylori is common in aquatic habitats. Consequently, we recommend additional larger studies using the PMA-qPCR technique to establish its presence and viability in other surface water sources.

The authors wish to acknowledge Dr Xavier Calvet from the Department of Gastroenterology, Hospital de Sabadell, Barcelona, for providing the Helicobacter pylori strains 60190 and Tx30a. Additionally, the authors wish to thank Universidad del Cauca, Fundación InnovaGen, GenIUL R&D Laboratory, Cristina Carlos-Roca Font and the UNESCO Sustainability conference of the Catalonia Polytechnic University for their support in conducting this study.

This study was financed by the Programa Nacional de Ciencia y Tecnología de Salud de Colciencias from Colombia (Grant No. 110356934947) and the Ministry of Science and Innovation of Spain (Grants CTM2012-39858).

The authors declare no conflicts of interest.