SEARCH

SEARCH BY CITATION

Keywords:

  • adenovirus;
  • coliphage;
  • enterovirus;
  • groundwater;
  • norovirus;
  • stability

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Aim:  To investigate the potential health hazard from infectious viruses where coliphages, or viruses by polymerase chain reaction (PCR), have been detected in groundwater. Two aspects were investigated: the relationship between infectivity and detection by PCR and the stability of coliphage compared to human viruses.

Methods and Results:  Virus decay (1 year) and detection (2 years) studies were undertaken on groundwater at 12°C. The order of virus stability from most to least stable in groundwater, based on first-order inactivation, was: coliphage ΦX174 (0·5 d−1) > adenovirus 2 > coliphage PRD1 > poliovirus 3 > coxsackie virus B1 (0·13 d−1). The order for PCR results was: norovirus genotype II > adenovirus > norovirus genotype I > enterovirus.

Conclusions:  Enterovirus and adenovirus detection by PCR and the duration of infectivity in groundwater followed similar trends over the time period studied. Adenovirus might be a better method for assessing groundwater contamination than using enterovirus; norovirus detection would provide information on a significant human health hazard. Bacteriophage is a good alternative indicator.

Significance and Impact of the Study:  PCR is a useful tool for identifying the health hazard from faecal contamination in groundwater where conditions are conducive to the survival of viruses and their nucleic acid.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

The Polymerase Chain Reaction (PCR) is a well established and rapid method, which can be used to identify the presence of a specific virus based on the detection of its genetic material. It is a commonly used and valuable tool for assessing the microbial quality of water resources (Abbaszadegan et al. 1999; Powell et al. 2003). Virus detection by PCR may provide a more sensitive method of detection than cell culture assay because of the amplification process used, with low copy numbers of nucleic acid able to be detected. However, there is concern regarding the use of PCR results as a tool for assessing health hazard, as the presence of viral RNA or DNA detected by PCR does not indicate directly that there is an infectious level of contamination, but it does indicate that there is a source of viral contamination (e.g. sewage).

It has traditionally been assumed that the use of PCR as a water quality monitoring tool will overestimate the level of microbial hazard. PCR should result in more virus being detected than cell culture assays as viral nucleic acid can be present when there are no infectious viruses present (Abbaszadegan et al. 1999). However, recent studies have reported correlations between detection of virus genomes and infectivity under some specific conditions. From studies of the duration of detection by PCR in seawater, enterovirus was suggested as a useful indicator of recent contamination events (Wetz et al. 2004); however, the results were dependent on the water type, with greater stability in sterile waters resulting in detection by PCR after no infectious viruses were detectable. For astrovirus and rotavirus, correlations between the concentrations of infectious viruses and of the viral genomes were reported in surface water exposed to light; whereas, in groundwater (not exposed to light), viral genomes were more stable than infectious viruses (Espinosa et al. 2008). In surface and groundwaters, Bae and Schwab (2008) quantified the reduction rate of nucleic acid of norovirus surrogates and reported the rate to be significantly slower than the reduction rate of infectivity at 25°C.

Groundwater has traditionally been perceived as being of high microbiological quality because of the natural filtration in aquifers and the often long time periods involved for water to travel from the point of contamination (e.g. sewage leak) to the point of extraction (e.g. spring or well) (Howard et al. 2006). In recent years in the USA, viruses have been detected by PCR in groundwater, including in drinking water production wells and in confined and unconfined aquifers (Abbaszadegan et al. 1999; Borchardt et al. 2003, 2007). In the UK, norovirus and enterovirus were detected by PCR at depths of up to 47 m in unconfined Permo-Triassic sandstone aquifers (Powell et al. 2003). This has significant implications for groundwater quality. In England and Wales, groundwater accounts for 33% of public water supplies (Downing 1998), rising to 80% in some areas of southern England, a quarter of which is produced from Permo-Triassic sandstone aquifers (Bouchier 1998).

The survey of microbiological quality of groundwater in multi-level wells undertaken by Powell et al. (2003) also reported detection of viable faecal indicator bacteria (thermotolerant coliforms, faecal streptococci and sulphite-reducing clostridia) and culturable viruses (enteroviruses and coliphage) at depths of up to 60 m in unconfined sandstone and 90 m in confined sandstone aquifers. Powell et al. identified what appeared to be rapid transport of viruses in sandstone matrices, as the detection of viruses, particularly enteroviruses, corresponded with the seasonal infection peaks in the community (Moore et al. 1984) and the anticipated short survival times of enteroviruses reported in groundwater (West et al. 1998). While it was clear that there was virus contamination in these aquifers, the relationship between detection of the viruses by PCR and health hazard was not known. It was also not known what the duration of survival of infectivity of these viruses was in this type of aquifer.

To understand better, the implications of detection of viruses by PCR in groundwater and the stability of these organisms, we have studied the duration of detection by PCR of viruses in groundwater from a Permo-Triassic sandstone aquifer and the duration of virus and phage infectivity, for up to 2 years. The duration of detection by PCR was studied for enteroviruses and noroviruses. Infectivity studies were undertaken on enteroviruses (poliovirus and coxsackie B virus) and coliphages (PRD1 and ΦX174). These viruses were the same as those investigated by Powell et al. (2003). Soil was not used in the study as the rapid transport in the field suggested that attachment was not a significant mechanism. In addition, adenoviruses, a useful marker of faecal pollution in water (pers. comm. Jane Sellwood), were included in PCR and infectivity studies. First order and biphasic decay models were applied to the infectivity study results to enable statistical comparisons of the decay rate coefficients of different viruses and in different water types.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Viruses

A range of human viruses and coliphage was studied. Enteroviruses (EV) are single stranded (ss-)RNA, nonenveloped viruses. Enteroviruses poliovirus 3 (PV3) and Coxsackie virus B1 (CB1) were isolated from sewage, then grown in cell culture with infectious viruses enumerated by suspended cell plaque assay and the presence of nucleic acids by reverse transcriptase (RT-)PCR.

Noroviruses (NoV) are also ss-RNA, nonenveloped viruses. NoV nucleic acids of both genogroups was detected by RT-PCR . No infectivity assay is available. NoV GI and GII were each detected in a faecal sample. The endpoint of detection was determined by testing serial dilutions by RT-PCR; both samples were found to have detectable virus at a dilution of 10−4, but no detectable virus at 10−5.

Adenoviruses are double stranded (ds-)DNA viruses. Adenovirus serotype 2 (AdV2) was sourced from a clinical sample grown in Caco-2 cells for research purposes. Infectious AdV2 was quantified by monolayer plaque assay and the presence of nucleic acid was determined by PCR.

PRD1 is a ds-DNA coliphage. PRD1 (ATCC-BAA-769-B1) and host (Escherichia coli ATCC-BAA-769) were sourced from the American Type Culture Collection and were grown/assayed using tryptone soya agar and broth. ΦX174 is an ss-DNA coliphage. ΦX174 (NCIMB# 10382) and host (E. coli NCIMB# 14067) were sourced from the National Collections of Industrial, Marine and Food Bacteria and grown and assayed using Luria agar and broth. Phage stock suspensions were centrifuged at 3300 g for 30 min to reduce bacterial debris. Both the phages were assayed using the double agar layer technique (Adams 1959).

Water types

Three water types were studied: raw groundwater (GW), synthetic groundwater (SGW) and 1/40 dilution of Ringer’s solution (RS). GW was taken from boreholes in a Permo-Triassic sandstone aquifer used for experimental purposes on the campus of Birmingham University, England. The quality of the GW has been described by Joyce et al. (2007). The formula for SGW was based on the quality of chemical analyses of the natural GW used (per l of ultrapure (Milli-Q) water: 0·284 MgSO4.7H2O, 0·032 g NaCl, 0·266 g CaCl2.2H2O, pH neutralized). SGW was studied as a standardized option suitable for laboratory experiments with phage and as a sterile water for comparison with GW. RS is an isotonic salt solution. RS was used to provide a control for human viruses and represented a sterile, low ionic strength water. SGW and RS were pH adjusted to match the natural groundwater (pH 7·5).

Human virus infectivity and nucleic acid stability studies

Experiments were set-up in microcosms (Elkay UK plastic bijou, 7 ml). Experiments were set-up for each of the viruses in GW and RS. For each water type, 5 ml of water was inoculated with virus to achieve a titre of 104 PFU ml−1. The microcosms were stored at 12 ± 2°C to simulate the conditions in the Permo-Triassic aquifer. The initial titre was determined by assaying three microcosms approximately 2 h after inoculation. Microcosms were sampled at monthly intervals for up to 2 years. Triplicate samples were taken, except for adenovirus for which five samples were taken. Infectivity studies were undertaken with enteroviruses PV3, CB1 and AdV2 by cell plaque assay. One millilitre volumes were inoculated into each of three Petri dishes in the plaque assay and 1 ml was used for molecular tests. The limit of detection in the plaque assay was 1 PFU ml−1.

Studies to monitor the duration for which viruses can be detected in groundwater by PCR were undertaken with EV, NoV GI, NoV GII and AdV2. AdV2 DNA was extracted using the Qiagen QIAamp DNA Mini Kit. Ten microlitres was used for the PCR. Extraction and PCR was undertaken in duplicate. DNA was amplified using primers from the hexon region of the adenovirus genome (Chapron et al. 2000). Amplified products were analysed by agarose gel electrophoresis and visualized under UV illumination after being stained with ethidium bromide. NoV was detected by RT-PCR. RNA was extracted using the Qiagen QIAamp Viral RNA Mini Kit. Twenty microlitres was used for the PCR. Extraction and PCR was undertaken in duplicate. RNA was reverse-transcribed to cDNA using random primers. DNA was amplified using primers from the polymerase region of the norovirus genome (Green et al. 1998). For EV, as for NoV, RNA was extracted using the Qiagen viral RNA mini kit, with RNA reverse-transcribed to cDNA using random primers. Twenty microlitres was used for the PCR. Extraction and PCR was undertaken in duplicate. DNA was amplified using primers from the 5′ noncoding region of the enterovirus genome (Sellwood et al. 1998). Amplified products were analysed as for AdV2.

Phage infectivity studies

Phage experiments were undertaken using GW and SGW in microcosms (Fisherbrand cryovials, polypropylene 4·0 ml external thread). Each type of water was inoculated with stock suspensions of PRD1 or ΦX174 to a titre of 106 PFU ml−1 as determined in stock suspension the day prior to inoculation. Approximately 5 ml of inoculum was added to the microcosms. The cryovials and caps were both filled with sample to ensure that no air bubbles were trapped within the vials. The microcosms were stored at 12 ± 2°C to simulate the conditions in the Permo-Triassic aquifer. The initial titre was determined by assaying three microcosms approximately 2 h after inoculation. Three microcosms were removed per time point with triplicate samples assayed from each. Sampling was undertaken four times in the first week, then weekly for 6 weeks and then monthly up to 12 months. The limit of detection was 1 PFU ml−1.

Statistical analyses and modelling

A first order and a biphasic decay model were considered. First order decay was modelled as

  • image( eqn (1))

where C is the concentration of virus, C0 is the concentration of virus at time = 0, t is time and μ is the decay rate coefficient. Biphasic decay was modelled as

  • image( eqn (2))

where α is the fraction of less stable viruses with a higher decay rate coefficient μ1 and (1−α) is the fraction of more stable viruses with a lower decay rate coefficient μ2 (Cerf 1977; Petterson and Ashbolt 2001).

A log-likelihood method was used to fit the decay models to the data and compare the first order and biphasic models. For first order decay and assuming normally distributed errors, a log-likelihood function, L, which includes parameters μ, ln C0 and s (standard error) was formulated for each set of n measured concentrations (Ci; i = 1–n) at time ti from each experiment (Hogg and Craig 1995):

  • image( eqn (3))

Mathematica (ver. 5.1, Wolfram Research, Champaign, IL) was used to obtain values for μ, ln C0 and s for the different experiments by maximizing the value of the log-likelihood function (equivalent to least squares solution) using numerical optimization. For biphasic decay, the following log-likelihood function was used:

  • image( eqn (4))

Using the likelihood ratios test (Cox and Hinkley 1974), the first order and biphasic decay models were compared for each dataset. The first order decay model was used to compare differences in decay rate coefficients under the various experimental conditions. The log-likelihood functions were applied to individual and combined datasets. For first order decay, common values of μ were assumed for combined datasets. The sum of the log-likelihoods of the individual datasets were compared to that of the combined data. The differences were interpreted as a χ2-deviate with the number of degrees of freedom equal to the difference in the number of parameters in the combined dataset and the total number of parameters of all separate data sets (Teunis et al. 1996). Where the log-likelihood of the combined data was found to be statistically significantly (compared to χ2df, 0·95) greater than that of the individual datasets, significant differences existed between the datasets.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Infectivity reductions in human viruses

Human viruses remained detectable in 1 ml in the cell culture plaque assay in GW for at least 364 days. AdV2 was the most stable of the virus genomes tested by PCR (4·1 log10 reduction in 364 days), followed by PV3 (4·7 log10 reduction in 140 days) and CB1 (4·2 log10 reduction in 70 days). The infectivity reduction curves were modelled with the first order decay model and biphasic decay model, with decay rate coefficients provided in Table 1. In groundwater, the decay rate coefficients (μ) for the first order decay model were 0·01 d−1 for AdV2 and 0·131 d−1 for CB1. The biphasic decay model provided a statistically better fit based on the chi-squared test (likelihood ratio >χ2(4–2), 0·95 = 5·99) in all three studies in groundwater. The secondary decay rate coefficient in the biphasic model (μ2), which may represent the rate of decay for a more stable subpopulation of viruses or aggregated viruses, were on average 0·4 of the first order model rate coefficient for enteroviruses in groundwater. This highlights the apparent increase in stability with time, which is not addressed with the first order model. The secondary rate coefficient for AdV2 in groundwater was 0·2 of the first order model rate coefficient.

Table 1.   Summary of human virus infectivity studies and results from application of the first order and biphasic decay models for decay rate coefficients (μ, μ1 and μ2), modelled initial concentrations (C0) and the proportion of viruses in the biphasic model (α) which were found to decrease at the primary decay rate (μ1)
VirusWater‡Measured titre at time = 0 (PFU ml−1)Duration detected (limit of study)*, dFirst order modelBiphasic modelLikelihood ratio†
μ (d−1)C0 (PFU ml−1)αμ1 (d−1)μ2 (d−1)C0 (PFU ml−1)
  1. Detection limit = 1 PFU ml−1.

  2. †A statistically significant likelihood ratio using the chi-squared test (χ24-2, 0·95 = 5·99) indicates that the biphasic model provided a better fit for the data.

  3. §Measured titre after 1 day.

  4. ‡GW, groundwater; RS, Ringer’s solution.

PV3GW1·6 × 104140*0·0732·0 × 1030·999 0·1430·019 9·7 × 0320·8†
 RS1·6 × 104140*0·0231·3 × 1040·232 37·5  0·023 1·6 × 1041·0
CB1GW5·6 × 10370 0·1311·9 × 1030·993 0·2970·069 1·0 × 048·5†
 RS5·8 × 103112*0·0223·7 × 1030·463 2·5310·02  5·8 × 1031·6
AdV2GW4·8 × 104364*0·0104·1 × 1010·99980·4890·002 2·4 × 10434·2†
 RS2·3 × 104§728*0·0025·0 × 1030·902 0·0350·00031·8 × 1045·8

Duration of detection of viruses by PCR

Of the viruses studied, enteroviruses had the least stable nucleic acid in groundwater, with the virus genome only consistently detected in 20 μl by PCR over 140 days and periodically detected for up to 1 year. The period of consistent detection by PCR was comparable to the duration that PV3 remained detectable in the infectivity assay (Fig. 1). AdV2 was detectable in 10 μl by PCR for 672 days of the 728 day study (Fig. 2). This corresponded with the stability of infectious AdV2. However, there was a rapid decrease in the infectivity of AdV2 (4·2 log10 over the initial 21 days; Fig. 2). No corresponding qualitative reduction in the strength of the nucleic acid signal was recorded over this time. The stability of both EV and AdV2, in terms of infectivity and detection by PCR, were greater in RS than in GW (Supporting information Figs S1 and S2).

image

Figure 1.  Infectivity and detection of enteroviruses in groundwater (a) survival of infectious PV3 with the first order (solid line) and biphasic (dashed) models; (b) survival of infectious CB1 with the first order (solid line) and biphasic (dashed) models; and (c) detection of virus nucleic acid by PCR. Samples were taken weekly for first 4 weeks, then fortnightly for 4 weeks, then four-weekly. Black indicates the period up until a positive sample, white indicates the period up until a negative.

Download figure to PowerPoint

image

Figure 2.  Infectivity and detection of adenovirus in groundwater (GW) (a) survival of infectious viruses with the first order (solid line) and biphasic (dashed) models and (b) detection of virus nucleic acid by PCR. Samples were taken weekly for first 4 weeks, then fortnightly for 4 weeks, then four-weekly. Black indicates the period up until a positive sample, white indicates the period up until a negative, cross hatch indicates no sample.

Download figure to PowerPoint

Similar to AdV2, NoV was very stable in groundwater. Qualitative differences were evident between the two genotypes (Figs 3 and 4). In groundwater, NoV GII nucleic acid was detectable for 728 days, whereas NoV GI was detectable only in 50% of samples after 140 days.

image

Figure 3.  Detection of norovirus genotype I in (a) groundwater and (b) 1/40 Ringer’s solution by PCR. Samples were taken weekly for first 4 weeks, then fortnightly for 4 weeks, then four-weekly. Black indicates the period up until a positive sample, white indicates the period up until a negative.

Download figure to PowerPoint

image

Figure 4.  Detection of norovirus genotype II in (a) groundwater and (b) 1/40 Ringer’s solution by PCR (black indicates positive, white indicates negative).

Download figure to PowerPoint

Infectivity reduction rates in coliphage

PRD1 remained detectable in 0·5 ml by plaque assay for up to 238 days. First order decay rate coefficients in microcosms for PRD1 (Table 2) were 0·017 d−1 in SGW and 0·050 d−1 in natural groundwater. ΦX174 remained culturable and detectable for up to 106 days in groundwater (2·3 log10 reduction). First order decay rate coefficients in microcosms for ΦX174 (Table 2) were 0·022 d−1 in SGW and 0·050 d−1 in natural groundwater. For PRD1, the biphasic model provided a statistically significant better fit in GW and SGW. Secondary decay rate coefficients, 0·009 d−1 in SGW and 0·029 d−1 in natural groundwater, were half that of the rate coefficients for the first order model. For ΦX174, the biphasic model provided a statistically significant better fit in SGW, with the secondary decay rate coefficient of 0·016 d−1 in SGW approximately 70% of the first-order model rate coefficient.

Table 2.   Summary of modelling results for coliphage infectivity studies using the first order and biphasic decay models. Detection limit = 1 PFU ml−1
VirusWater‡Measured titre at time = 0 (PFU ml−1)Duration (d)First orderBiphasicLikelihood ratio†
μ (d−1)C0 (PFU ml−1)αμ1 (d−1)μ2 (d−1)C0 (PFU ml−1)
  1. †A statistically significant likelihood ratio using the chi-squared test (χ24-2, 0·95 = 5·99) indicates that the biphasic model provided a better fit for the data.

  2. ‡GW, groundwater; SGW, synthetic groundwater.

PRD1GW5·1 × 1052380·0503·4 × 1050·8720·12 0·0294·8 × 10514·8†
SGW1·5 × 1062380·0175·9 × 1050·9140·2350·0092·5 × 10612·2†
ΦX174GW1·1 × 1061060·0501·2 × 1060·2260·0860·0471·2 × 1060·6 
SGW4·0 × 105570·0221·3 × 1060·4280·5770·0161·8 × 1067·7†

The biphasic model provided a statistically significantly better fit than the first order model in six of the 10 human virus and phage studies. This two-phase behaviour (e.g. Fig. 2) was more prevalent in groundwater. Evidence of two-phase behaviour was more common for phages than human viruses and more common for PRD1 than ΦX174. From these experiments, the biphasic model provided a statistically significantly better fit of the data in some cases for all virus types and for GW and SGW.

Comparison of infectivity reduction rates

A comparison of the infectivity decay rate coefficient for the first order model (μ) with varying virus and water type was undertaken. In virus infectivity studies, AdV2 was the most stable, but with rapid loss of 3 to 4 log10 over the first month. PV3 was less stable in GW than the phages. Very little scatter in the results for PRD1 and ΦX174 in GW (Fig. 5) meant that a statistically significant difference between the decay rate coefficients could be determined using the chi-squared test (likelihood ratio >χ26–4, 0·95 = 5·99), despite the decay rate coefficient for PRD1 being only 0·01 greater than that for ΦX174 (Table 2). In SGW, where there was considerable scatter in the PRD1 data (Fig. 5), the decay rate coefficient for ΦX174 was significantly higher than that for PRD1. In RS, AdV2 had a significantly (likelihood ratio >χ26–4, 0·95 = 5·99) lower decay rate coefficient than with CB1 or PV3.

image

Figure 5.  Results of coliphage infectivity studies with the first order (solid line) and biphasic (dashed) models (a) PRD1 in natural groundwater; (b) ΦX174 in natural groundwater; (c) PRD1 in synthetic groundwater and (d) ΦX174 in synthetic groundwater (error bars indicate standard deviation).

Download figure to PowerPoint

Overall, in natural groundwater, the phages were more stable than the enteroviruses studied. Overall, the order of stability for infectivity in viruses in natural groundwater, based on first-order decay rates, was: ΦX174 > AdV2 > PRD1 > PV3 > CB1. The qualitative order for stability of nucleic acid, based on PCR results, was: NoV GII > AdV2 > NoV GI > PV3. There were statistically significant differences between the decay rate coefficient in different water types, with decay greatest in GW > SGW and RS.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

PCR provides a sensitive method for assessing the presence of viral RNA or DNA in groundwater and other environments; however, it does not quantify the health hazard (Abbaszadegan et al. 1999). We have studied the stability of viruses in groundwater from a UK sandstone aquifer by PCR and infectivity assay to determine the relationship between infectious viruses and the presence of virus genomes or coliphage. EV (PV3 and CB1) remained infectious for up to 140 days and were able to be detected by molecular techniques for 140 days. AdV2 was stable for longer periods. AdV2 remained infectious for the duration of the study (364 days), with detection by molecular techniques possible for 672 days. NoV GI and GII remained detectable by RT-PCR for 588 (albeit sporadically) and 728 days respectively. Coliphage ΦX174 and PRD1 were infectious for approximately 100 days, but were not at the limit of detection at that time, with 215 days required to achieve the same reduction as experienced with the poliovirus based on the first order decay model or 250 days based on the biphasic model for PRD1.

Of the water types studied (GW, SGW and RS), virus stability was greatest in the sterile, low ionic strength conditions of RS. In particular, enterovirus stability was greatly increased in RS compared to GW. SGW, a sterile, standardized, laboratory substitute for groundwater, was formulated to reflect the chemical characteristics of the natural groundwater. Decay rate coefficients in SGW were also significantly lower than in GW.

The rate coefficients for the decay or reduction of infectivity, reported here were generally lower than reported by others in groundwaters. Yates et al. (1985) highlighted the variability of decay rate coefficients of poliovirus in different groundwaters. The authors reported a range of decay rate coefficients for poliovirus in nine groundwaters at 12–13°C from 0·08–0·3 d−1 (all values converted to ln), the lower end of which is comparable to those reported here. Matthess et al. (1988) reported lower decay rate coefficients for poliovirus 1 at 10°C in groundwater (0·013 d−1); however, higher decay rate coefficients for coxsackie B1 in the same groundwater with coefficients of 0·19 d−1 in groundwater. Similarly, reported decay rate coefficients for PRD1 (Schijven and Hassanizadeh 2000) are generally higher than reported here, although comparable with those collated by John and Rose (2005) at 0–10°C. Limited data are available on the survival of ΦX174 in groundwater, although Schijven et al.(2002) reported a decay rate coefficient of 0·012 d−1 at 5 ± 3°C. Schijven et al. (2006) used a value of 0·02 d−1 as a mean value for virus inactivation to calculate setback distance for protecting groundwater wells.

Virus detection by PCR provides a more sensitive method than cell culture because of the amplification process used, but without assessment of infectivity. NoV and AdV2 were the most stable viruses in PCR studies independent of water type, with AdV2 the most stable in the infectivity study as well. Previous studies have found similar stability of the norovirus genome (Bae and Schwab 2008). The stability of AdV2 in the infectivity study indicates a strong relationship between detection by PCR and infectivity and detection by PCR followed similar trends over the time period of this study.

The present study and previous studies have been limited in their findings by the long durations over which viruses are stable in groundwater. While Espinosa et al. (2008) reported that the correlation of the presence of viral genomes with infectivity was limited in groundwater because of the stability of the viral genome, there was a good qualitative relationship between the two in groundwater as over the duration of the study, 210 days for rotavirus and 120 days for astrovirus, both viruses remained detectable by PCR and retained infectivity. In the study reported here, AdV2 and NoV genomes have proved to be stable in groundwater over the duration of the study (2 years), with AdV2 also retaining infectivity over the duration of the 1 year infectivity study.

In the case presented by Powell et al. (2003), the detection of coxsackie B virus in groundwater was reported to coincide with seasonal infection in the community, with the conclusion being that rapid transport of viruses had occurred. Our results support that conclusion. Powell et al. (2003) reported detection in sandstone aquifers of enteroviruses of 5–10 PFU l−1 and of coliphage of 30–1000 PFU l−1. Assuming that the detection of viruses in the aquifer represents contamination with sewage and concentrations in sewage for enteroviruses of 3500 PFU l−1 (HPA, 2006) and for coliphage of 3 × 106 PFU l−1 (Zhang and Farahbakhsh 2007), this detection represent reductions of approximately 3 log10 and 3 to 5 log10 respectively. Based on the rates of reduction of infectivity in the laboratory studies, the maximum period that the viruses might have been in groundwater was 52 to 94 days for enterovirus and 148 days for coliphage. This result supports the conclusions of Powell et al. (2003) that rapid transport of viruses occurred in the sandstone aquifer. It is possible that the duration was considerably shorter as dilution, filtration and attachment will also reduce virus concentrations in groundwater. However, in a Permo-Triassic sandstone aquifers, the mobility of virus has been reported to be low (Joyce et al. 2007) and the vertical flow rates have been reported to be less than a metre per year (Taylor et al. 2003). This indicates that virus transport may not have been via matrix flow, but through fissures in the sandstone where the mechanisms of filtration and attachment are reduced.

In the laboratory, NoV was detected in groundwater by PCR over 2 years, which suggests that, unlike with EV, detection of NoV in the field was not necessarily from recent contamination. We suggest that detection of NoV in groundwater by PCR does represent a health hazard. Groundwater in the UK represents a dark and cool environment (e.g. 12°C in the Permo-Triassic aquifer). It is also an environment which, when not polluted, typically has low concentrations of micro-organisms, with concentrations decreasing with depth (Madsen et al. 1991). The stability of infectivity in viruses is known to be greater in these conditions: low temperatures, no light and limited predatory populations (Pedley et al. 2006). Furthermore, recent work supports that in these groundwater environments, the reduction in infectivity in human noroviruses may not be significantly different from reduction in nucleic acid. In groundwater at 4°C, the stability of MS2 and poliovirus genomes was reflected in the stability of the infectivity (Bae and Schwab 2008) and no significant difference was reported in the reduction in infectivity and in nucleic acid for murine norovirus suggesting that human noroviruses may have a similar relationship between infectivity and detection of nucleic acid in such conditions.

Different viruses lose infectivity and undergo break up of genomic material at different rates. Viruses with a double-stranded genome are generally considered to be more stable than those whose genome is single-stranded (Espinosa et al. 2008). In this case, the ds-DNA viruses AdV2 and PRD1 were more stable than the ss-RNA, nonenveloped enteroviruses. However, the most stable viruses were the ss-RNA, nonenveloped norovirus and ss-DNA coliphage ΦX174. Recent work has reported that murine norovirus may undergo a capsid maturation process, where the mature state may be more stable (Katpally et al. 2008), which may account for the stability experienced with NoV in this study.

There were qualitative differences between the stability of NoV GI and GII. It is notable that NoV GII was the more stable as NoV GII, in particular variant 4, has been the most prevalent form of norovirus in outbreaks in Europe in recent years (Kroneman et al. 2006).

Assessing microbial groundwater quality

The results of this study support that, while PCR might be a good measure of infectivity of enteroviruses in groundwater, enteroviruses themselves are not necessarily good indicators of wider faecal contamination in groundwater. In this case, it is the pathogen, norovirus, which is a better test for contamination than using an indicator organism because of the greater stability of NoV. The phage studied, ΦX174 and PRD1 and adenovirus have the greatest stability of infectivity. The stability of these viruses also make them useful indicators of faecal contamination of groundwater.

The contamination of groundwater with viruses, through sewage contamination, animal faecal contamination or other methods, is a threat to the quality of the drinking water supplies. Evidence from surveys of the presence of virus in groundwater (Powell et al. 2003; Pedley et al. 2006; Borchardt et al. 2007) highlights that aquifers are susceptible to human faecal contamination, in particular, with viruses. Despite quality controls and treatment of UK water supplies, outbreaks of disease from groundwater supplies have been recorded (Short 1988; HPA, 2006), typically where a contamination event has coincided with a failure in treatment. The evidence from outbreaks illustrates the vulnerability of groundwater to faecal contamination and the mobility of pathogens in groundwater drinking water supplies, which is a major concern to water quality managers. Even at very low concentrations, the presence of human enteric viruses are a concern because of the small number of infectious viruses required to cause illness and the potentially high concentrations of viruses in sewage. For example, norovirus infections can potentially be caused by as little as 10 viral particles (Sair et al. 2002), but concentrations have been reported in sewage of 107 norovirus particles (detected by PCR) per litre (Lodder et al. 1999).

In addition to the duration for which viruses are stable in groundwater, it is important to consider water the flow and therefore the potential for transport. Permo-Triassic sandstones generally have low flow rates of <200 m y−1 (Tellam and Barker 2006), which would limit the spread of any contamination. However, the rapid transport of viruses in a Permo-Triassic aquifer described by Powell et al. (2003) and discussed by Taylor et al. (2004) raises questions about the use of average flow rates for predicting virus transport in groundwater. Furthermore, the current groundwater protection zone in England and Wales is 50 days travel time for potential sources of bacterial and viral contamination (Environment Agency, 2007), which is considerably shorter than the duration viruses are stable in groundwater. This protection zone is based on bacterial survival and while it has been adapted to accommodate the greater stability of Cryptosporidium, it still needs to be adapted to accommodate the body of evidence for the stability of viruses in groundwater.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

The stability of viruses in groundwater from a UK Permo-Triassic sandstone aquifer was studied to investigate the relationship between infectivity and detection by PCR and the stability of coliphage compared to human viruses. The results of relatively short-term stability of enteroviruses, both in terms of infectivity and nucleic acid, supports the conclusions of Powell et al. (2003) that detection of enterovirus in a Permo-Triassic aquifer was because of rapid transport mechanisms. The low stability of enteroviruses suggests that enteroviruses are not good indicators of potential faecal contamination of groundwater, but that NoV as a pathogen can be used to measure the contamination and the health hazard directly. Adenovirus, which has been shown to have greater stability than enteroviruses, is present more consistently in the environment than norovirus and so may be the best marker of human faecal contamination. Alternatively, coliphage would potentially be better indicator than enteroviruses where PCR facilities are not available. In general, the results show that the survival of viruses in groundwater over many months is possible and indicate that the detection of viruses in groundwater conditions (dark, limited predatory populations, low temperature) by PCR has a good potential correlation to infectivity and therefore does indicate a health hazard. This study, shows for the first time, the differences in the environmental stability of NoV GI and GII.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

This work was funded by the NERC (NER/A/S/2001/00656) and the England and Wales Environment Agency. The authors would like to thank the members of the project steering committee for their guidance: John Tellam, Richard Taylor, Richard Greswell, Michael Riley and Eadaoin Joyce. The authors would also like to thank the reviewers for their comments.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information
  • Abbaszadegan, M., Stewart, P. and LeChevallier, M. (1999) A strategy for detection of viruses in groundwater by PCR. Appl Environ Microbiol 65, 444449.
  • Adams, M.H. (1959) Bacteriophages. New York: Interscience Publishers Inc.
  • Bae, J. and Schwab, K.J. (2008) Evaluation of murine norovirus, feline calicivirus, poliovirus, and MS2 as surrogates for human norovirus in a model of viral persistence in surface water and groundwater. Appl Environ Microbiol 74, 477484.
  • Borchardt, M.A., Bertz, P.D., Spencer, S.K. and Battigelli, D.A. (2003) Incidence of enteric viruses in groundwater from household wells in Wisconsin. Appl Environ Microbiol 69, 11721180.
  • Borchardt, M.A., Bradbury, K.R., Gotkowitz, M.B., Cherry, J.A. and Parker, B.L. (2007) Human enteric viruses in groundwater from a confined bedrock aquifer. Environ Sci Technol 41, 66066612.
  • Bouchier, I. (1998) Cryptosporidium in Water Supplies, Third Report of the Group of Experts to: Department of the Environment, Transport and the Regions and Department of Health. London: Drinking Water Inspectorate. Available at: http://www.dwi.gov.uk/pubs/bouchier/index.htm
  • Cerf, O. (1977) Tailing of survival curves of bacterial spores. J Appl Bacteriol 42, 119.
  • Chapron, C.D., Ballester, N.A., Fontaine, J.H., Frades, C.N. and Margolin, A.B. (2000) Detection of astroviruses, enteroviruses and adenovirus types 40 and 41 in surface waters collected and evaluated by the information collection rule and an integrated cell culture-nested PCR procedure. Appl Environ Microbiol 66, 25202525.
  • Cox, D.C. and Hinkley, D.V. (1974) Theoretical Statistics. London: Chapman and Hall.
  • Downing, R.A. (1998) Groundwater–Our Hidden Asset. Keywoth, Nottingham: British Geological Survey.
  • Environment Agency. (2007) How the Zones Work. Available at: http://www.environment-agency.gov.uk/maps/info/groundwater/963948/?version=1&lang=_e (accessed 8 August 2007).
  • Espinosa, A.C., Mazari-Hiriart, M., Espinosa, R., Maruri-Avidal, L., Mendez, E. and Arias, C.F. (2008) Infectivity and genome persistence of rotavirus and astrovirus in groundwater and surface water. Water Res 42, 26182628.
  • Green, J., Henshilwood, K., Gallimore, C.I., Brown, D.W.G. and Lees, D.N. (1998) A nested reverse transcriptase PCR assay for detection of small round-structured viruses in environmentally contaminated molluscan shellfish. Appl Environ Microbiol 64, 858863.
  • Hogg, C.V. and Craig, A.T. (1995) Introduction to Mathematical Statistics. Englewood Cliffs, NJ: Prentice Hall.
  • Howard, G., Bartram, J., Pedley, S., Schmoll, O., Chorus, I. and Berger, P. (2006) Chapter 1. Groundwater and public health. In Protecting Groundwater for Health: Managing the Quality of Drinking-water Sources ed. Schmoll, O., Howard, G., Chilton, J. and Chorus, I. pp. 320. London: IWA Publishing.
  • HPA. (2006) General outbreaks of infectious intestinal disease in England and Wales: 2005. Commun Dis Rep CDR Rev 16, 5763.
  • John, D.E. and Rose, J.B. (2005) Review of factors affecting microbial survival in groundwater. Environ Sci Technol 39, 73457356.
  • Joyce, E., Rueedi, J., Cronin, A., Pedley, S., Tellam, J. and Greswell, R. (2007) Fate and transport of phage and viruses in UK Permo-Triassic Sandstone aquifers. Science Report – SC030217/SR. Bristol, UK: Environment Agency of England and Wales.
  • Katpally, U., Wobus, C.E., Dryden, K., Virgin, H.W. 4th and Smith, T.J. (2008) Structure of antibody-neutralized murine norovirus and unexpected differences from viruslike particles. J Virol 82, 20792088.
  • Kroneman, A., Vennema, H., Harris, J., Reuter, G., Von Bonsdorff, C.H., Hedlund, K.O., Vainio, K., Jackson, V. et al. (2006) Increase in norovirus activity reported in Europe. Euro Surveill 11, E0612141.
  • Lodder, W.J., Vinj, J., Van De Heide, R., De Roda Husman, A.M., Leenen, E.J.T.M. and Koopmans, M.P.G. (1999) Molecular detection of Norwalk-like caliciviruses in sewage. Appl Environ Microbiol 65, 56245627.
  • Madsen, E.L., Sinclair, J.L. and Ghiorse, W.C. (1991) In situ biodegradation: microbiological patterns in a contaminated aquifer. Science 252, 830833.
  • Matthess, G., Pekdeger, A. and Schroeter, J. (1988) Persistence and transport of bacteria and viruses in groundwater – a conceptual evaluation. J Contam Hydrol 2, 171.
  • Moore, M., Kaplan, M.H., McPhee, J., Bregman, D.J. and Klein, S.W. (1984) Epidemiologic, clinical, and laboratory features of Coxsackie B1-B5 infections in the United States, 1970–79. Public Health Rep 99, 515522.
  • Pedley, S., Yates, M., Schijven, J.F., West, J., Howard, G. and Barrett, M.H. (2006) Chapter 3. Pathogens: health relevance, transport and attentuation. In Protecting Groundwater for Health: Managing the Quality of Drinking-water Sources ed . Schmoll, O., Howard, G., Chilton, J. and Chorus, I. pp. 4980 London: IWA Publishing.
  • Petterson, S.R. and Ashbolt, N.J. (2001) Viral risks associated with wastewater reuse: modelling virus persistence on wastewater irrigated salad crops. Water Sci Technol 43, 23226.
  • Powell, K.L., Taylor, R.G., Cronin, A.A., Barrett, M.H., Pedley, S., Sellwood, J., Trowsdale, S.A. and Lerner, D.N. (2003) Microbial contamination of two urban sandstone aquifers in the UK. Water Res 37, 339352.
  • Sair, A.I., D’Souza, D.H. and Jaykus, L.A. (2002) Human enteric viruses as causes of foodborne disease. Compr Rev Food Sci Food Saf 1, 7389.
  • Schijven, J.F. and Hassanizadeh, S.M. (2000) Removal of viruses by soil passage: overview of modeling, processes, and parameters. Crit Rev Environ Sci Technol 30, 49127.
  • Schijven, J.F., Hassanizadeh, S.M. and De Bruin, R.H.A.M. (2002) Two-site kinetic modeling of bacteriophages transport through columns of saturated dune sand. J Contam Hydrol 57, 259.
  • Schijven, J.F., Mulschlegel, J.H., Hassanizadeh, S.M., Teunis, P.F. and De Roda Husman, A.M. (2006) “Determination of protection zones for Dutch groundwater wells against virus contamination – uncertainty and sensitivity analysis”. J Water Health 4, 297312.
  • Sellwood, J., Shore, J., Read, S. and Wyn-Jones, P. (1998) The use of reverse transcriptase polymerase chain reaction to investigate environmental samples for the presence of enteroviruses. Commun Dis Public Health 1, 5860.
  • Short, C.S. (1988) The Bramham incident, 1980 – an outbreak of water-borne infection. J Inst Water Environ Mngmt 2, 383390.
  • Taylor, R.G., Cronin, A.A., Trowsdale, S.A., Baines, O.P., Barrett, M.H. and Lerner, D.N. (2003) Vertical groundwater flow in Permo-Triassic sediments underlying two cities in the Trent River Basin (UK). J Hydrol 284, 92.
  • Taylor, R., Cronin, A., Pedley, S., Barker, J. and Atkinson, T. (2004) The implications of groundwater velocity variations on microbial transport and wellhead prtection – review of field evidence. FEMS Microbiol Ecol 49, 1726.
  • Tellam, J.H. and Barker, R.D. (2006) Towards prediction of saturated zone pollutant movement in groundwaters in fractured permeable-matrix aquifers: the case of the UK Permo-Triassic Sandstones. In Fluid Flow and Solute Movement in the Fluvial and Aeolian Sandstones of the UK Permo-Triassic Red Bed Sequence ed. Barker, R.D. and Tellam, J.H., Geol Soc Sp Pub 263, 148.
  • Teunis, P.F.M., Van Der Heijden, O.G., Van Der Giessen, J.W.B. and Havelaar, A.H. (1996) The Dose–Response Relation in Human Volunteers for Gastro-Intestinal Pathogens. RIVM Report 284550002. Bilthoven, Netherlands: RIVM.
  • West, J., Pedley, S., Baker, S., Barrott, L., Morris, B., Storey, A., Ward, R. and Barrett, M. (1998). A review of the impact of microbiological contaminants in groundwater. Bristol, UK: UK Environment Agency, R&D Technical Report no. P139.
  • Wetz, J.J., Lipp, E.K., Griffin, D.W., Lukasik, J., Wait, D., Sobsey, M.D., Scott, T.M. and Rose, J.B. (2004) Presence, infectivity, and stability of enteric viruses in seawater: relationship to marine water quality in the Florida Keys. Mar Pollut Bull 48, 698.
  • Yates, M.V., Gerba, C.P. and Kelley, L.M. (1985) Virus persistence in groundwater. Appl Environ Microbiol 49, 778781.
  • Zhang, K. and Farahbakhsh, K. (2007) Removal of native coliphages and coliform bacteria from municipal wastewater by various wastewater treatment processes: implications to water reuse. Water Res 41, 28162824.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Figure S1. Infectivity and detection of enteroviruses in 1/40 Ringer’s solution (a) survival of infectious PV3 (&bsl00066;) and CB1 (Δ); and (b) detection of virus nucleic acid by PCR. Samples were taken weekly for first four weeks, then fortnightly for four weeks, then four-weekly. Black indicates the period up until a positive sample, white indicates the period up until a negative.

Figure S2. Infectivity and detection of adenovirus in 1/40 Ringer’s solution (a) survival of infectious viruses and (b) detection of virus nucleic acid by PCR. Samples were taken weekly for first four weeks, then fortnightly for four weeks, then four-weekly. Black indicates the period up until a positive sample, white indicates the period up until a negative.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
JAM_4150_sm_FigsS1-S2.doc39KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.