Health Risk Assessment in Connection with the Use of Microbiologically Contaminated Source Waters for Irrigation
Report No.1226/1/04
March 2004

EXECUTIVE SUMMARY

1. Part A: Presence, antibiotic susceptibility patterns and chlorination resistance of faecal pollution from a dense settlement and the possible impact of polluted water on downstream water users

1.1 Background and motivation

The presence of dense settlements on the river banks in the Western Cape give rise to water pollution of nearby rivers and severely affect the water quality downstream. Most of the water pollution can be attributed to inadequate sanitation in these settlements, severe overcrowding, as well as failing sewerage systems. The resultant burden of disease, loss of productivity and health costs should make the reduction of pollution and improvement of sanitation in Kayamandi a priority for the local authorities. The study site chosen was the Plankenbrug River running through Stellenbosch and the dense settlement of Kayamandi on its banks.

1.2 Objectives

The objectives of this section of the investigation were:

1. Description of the exposed population in the dense settlement of Kayamandi;
2. Quantification of faecal pollution in the Plankenbrug River using Escherichia coli as an indicator organism;
3. Identification of the most likely pathogens involved in the Plankenbrug River by additional analyses of water samples;
4. Preliminary tests of antibiotic resistance of organisms in the river water; and
5. Preliminary tests on chlorination resistance of organisms on the river water

1.3 Summary of Results

There are many more persons living in the settlement than previously acknowledged. At least 25 000 persons live in the area, 80% of them in shack housing. Throughout the year there is a high prevalence of diarrhoea in the area.

1.3.2 Bacteriological analysis of water samples and biofilms

The faecal pollution in the river over the past four years reached a high count of 12 million E. coli per 100 ml water on one occasion, but frequently reached the million mark. It was above the recommended limit of 2000 E. coli per 100 ml water for 97% of the sampling occasions over the four years. Many pathogens carrying considerable health risks were identified in the water. Some, like ß haemolytic streptococcus Group A, were an unusual find in free-flowing water. A number of the organisms in the water and in the biofilms on stones in the river, exhibited signs of antibiotic resistance to some commonly used antibiotics and also resistance to chlorination. A further complication was that some of the organisms surviving the chlorination experiments showed enhanced antibiotic resistance.

1.3.3 Virological analysis of water samples and associated biofilms

Three sets, i.e. before and after the source of pollution, of water samples (2 litres) from the Plankenbrug River were referred for virological analysis between April and June 2002. In addition biofilms from recovered stones within the river were included with one of the sample sets. Viruses were recovered from the water samples, inoculated onto a variety of cell cultures, and then the concentrated sample and associated cell cultures analysed for enteroviruses, hepatitis A virus (HAV), human astrovirus (HAstV), human caliciviruses (HuCV) and human rotaviruses (HRV) by reverse transcriptase-polymerase reaction (RT-PCR).

Enteroviruses, namely polioviruses and an untypable enterovirus, were isolated from and HAstVs detected in water samples drawn from the Plankenbrug River below Kayamandi in April 2002. Enteroviruses were also detected in water drawn from a site further downstream on the same date. Water samples drawn from the Plankenbrug River below Kayamandi in May 2002 yielded adenoviruses and enteroviruses were detected by RT-PCR. HRVs, adenoviruses, enteroviruses and HAstVs were detected in samples collected in June 2002. Although the virological analysis was qualitative these data suggest that the Plankenbrug River below Kayamandi is heavily polluted with human faecal material.

1.4 Conclusions

The risks to health, environmental damage and the problems foreseen with economic activities downstream from the Plankenbrug River make it imperative that attention be given to the sanitation situation in Kayamandi, the state of the sanitation system in the whole of the town of Stellenbosch and assistance rendered to the local authority to start remedial action without delay.

1.5 Recommendations and Future Research

Several aspects of the present study should be investigated on a larger scale:

1. There is an urgent need for research into the feasibility and practical applications of monitoring water quality in natural watercourses by means of microbiological determinations.
2. The antibiotic resistance of organisms in natural water sources and the possible impact of such resistance on health in the area need to be assessed;
3. Additional research regarding chlorination resistance of organisms reaching natural waters, either in a treated or untreated state is required, especially the link between chlorination resistance and antibiotic resistance ought to receive urgent attention.
4. Research into the links between community behaviour, availability of sanitation services and receptivity to educational programmes should be expanded, especially for the local populations of impoverished people in dense settlements.

2. Part B: Microbiological analysis of irrigation waters and food crops

2.1 Background and Motivation

Despite major advances in preventative health foodborne illnesses remain a widespread and growing public health problem in the developed and developing world (Satcher, 2000), and the burden of infection is grossly underestimated (Marx, 1997; Koopmans et al., 2002). Many factors, including changing lifestyles and demographics, faster and more frequent travel, decreasing water supplies in certain countries and enhance importation of foods have contributed to the increase in food- and waterborne infections (Cuthbert, 2001; Koopmans et al., 2002). The threat of foodborne viral diseases has triggered worldwide interest in the fate of viruses on fresh produce irrigated with sludge or sewage effluents (Petterson and Ashbolt, 2001). Minimally processed foods (MPFs) such as salads, vegetables, fruits and other fresh produce that require only minimal processing before consumption are usually contaminated through human contact during harvesting or processing but contamination via wastewater and sludge, used for crop irrigation and fertilization, has also been documented (Petterson and Ashbolt, 2001). The use of wastewater for irrigation purposes has been responsible for many disease outbreaks caused by bacteria, protozoa, parasitic helminths and viruses (Bitton, 1980). The World Health Organization (WHO) estimates that 70% of diarrhoeal episodes are caused by biologically contaminated food (Satcher, 2000).

A large number of viruses are found in the human intestinal tract, with three disease categories being associated with food- and waterborne viruses, namely:

1. Gastroenteritis: caused by HRV, HuCV which include the "Norwalk-like viruses" (NLVs) or noroviruses, and the "Sapporo-like viruses" (SLVs) or sapoviruses, HAstV and the enteric adenoviruses;
2. Hepatitis: caused by the faecally transmitted hepatitis viruses, namely HAV and hepatitis E virus (HEV); and
3. Other severe illnesses such as myocarditis: caused by enteroviruses which include polioviruses, coxsackie A and B viruses, echoviruses and enteroviruses 68-71 (Koopmans et al., 2002).

HAV is a major cause of morbidity associated with faecally contaminated food and water (Mead et al., 1999). HuCVs (Parashar and Monroe, 2001; Koopmans et al., 2002) and HAstVs (Glass et al., 1996; Ferrari and Torres, 1998; Walter and Mitchell, 2000) are increasingly being identified as important foodborne viruses. Viruses are strict intracellular parasites and cannot replicate in food and water. Viral infection via contaminated food therefore depends on viral stability, amounts of virus shed by an infected person, method of processing of the food or water, infective dose and susceptibility of the host (Koopmans et al., 2002). Most of the food- and waterborne viruses are non-enveloped and are resistant to heat, disinfection and pH changes. Many of these viruses cannot be isolated or detected by conventional routine laboratory techniques and molecular methods for the detection of HAV, HuCV and HAstV have, until recently, had limited applicability in food and environmental virology (Richards, 1999).

There is very little data on the presence of HAV, HuCV and HAstV in food sources and domestic, agricultural and recreational water sources in SA. This is partly due to the lack of infrastructure for the detection and recording of such infections (Grabow, 1996). There is no reason to believe that risks of food- and waterborne disease in SA are any different from those in the rest of the world. Escalating demands and pollution of already limited water sources, particularly in rural and developing communities, may even elevate risks (DAWN, 2001). Common-source viral foodborne outbreaks, such as SRSV-associated gastroenteritis, have been described in SA (Taylor et al., 1993), but to date, no food- and waterborne outbreaks of HA or HAstV have been documented due to the relative underreporting of these diseases. HAV and HAstV have been detected in raw and treated water sources (Marx et al., 1998; Grabow et al., 2001; Taylor et al., 2001) confirming the risk of waterborne transmission of these viruses. In view of these facts better approaches to prevent the contamination of foods with potentially pathogenic viruses via irrigation or processing water, food handlers and sewage are needed. In addition, more effective techniques for the recovery and detection of these viruses in food and water will be beneficial (Keddy, 1998).

2.2 Objectives

The objectives of this investigation were:

1. To establish recovery and molecular detection and characterisation techniques for HuCV, HAstV, and HAV in raw and partially treated irrigation waters and selected food crops.
2. To monitor selected irrigation waters and associated fruit and vegetable crops for the following pathogens which can be waterborne and associated with foodborne disease, namely HuCV, HAstV and HAV, as well as selected indicator organisms.

2.3 Summary of Results

Irrigation water: A glass wool adsorption-elution method described by Grabow and Taylor (1993) was optimised and used for the recovery of viruses from water samples >2L. For sample volumes <2L, and for the eluate from the glass wool adsorption-elution method, viruses were recovered in a final volume of 6 ml phosphate buffered saline (PBS) containing 1% bovine serum albumin (BSA) by precipitation using the NaCl/polyethylene-glycol method described by Minor (1985).

Heavily polluted water: Three different methods, namely glass wool adsorption-elution, NaCl/PEG precipitation and a SiO2 method described by Baggi and Peduzzi (2000) for the concentration of rotaviruses from surface water and raw sewage were compared for the recovery of viruses from heavily polluted water. Of these three methods, the NaCl /PEG precipitation was shown to be the most cost-effective and efficient for the recovery of viruses from the heavily polluted water samples.

Minimally processed foods: For the purposes of this investigation the methods described by Bidawid et al. (2000) for the recovery of HAV from lettuce and strawberries, and Schwab et al. (2000) for the recovery of NLVs and HAV from delicatessen foods were adapted and optimised for the recovery of viruses and bacteria from MPFs. Three buffer systems, namely PBS pH 7.4, Tris-EDTA pH 8.0 and minimal essential medium (MEM) pH 7.2 were compared for the recovery of viruses from MPFs with rough (e.g. strawberries) and smooth (grapes, cherry tomatoes) surfaces. PBS was found to be the most efficient recovery buffer. For field analyses, ± 100 gm MPF was washed overnight at 4ºC in 50 ml PBS. Bacteriological analysis was done directly on the PBS washing while the viruses were recovered from the PBS washing in a final volume of 6 ml PBS containing 1% BSA by NaCl/polyethylene-glycol precipitation.

RNA extraction: Two extraction procedures, namely the QIAGEN QIAamp viral RNA extraction kit and TRIZOL® reagent, were compared to ascertain which method was the most suitable for the extraction of viral RNA from concentrates from fruit washings, water and sewage samples and cell culture extracts. The procedure using the TRIZOL® reagent proved to be the most suitable for the type of samples used. Subsequently a newer more effective RNA extraction kit, the QIAGEN QIAamp UltraSens Virus Kit, which uses larger volumes of sample, was used.

Hepatitis A virus: The RT-PCR oligonucleotide probe assay (Taylor et al., 2001) was used for the detection of HAV.

Human astroviruses: The RT-PCR oligonucleotide probe assay as described by Marx et al. (1998) and modified by Taylor et al. (2001) was used for the detection of HAstVs.

Human caliciviruses: A sensitive and specific RT-PCR was developed and optimised for the detection of HuCVs using the primer pair described by Jiang et al. (1999) for the detection of both noro- and sapoviruses.
Multiplex RT-PCR for the simultaneous detection of HAV and HuCVs: A multiplex RT-PCR, described by Schwab et al. (2000), was assessed for the detection of HAV and HuCVs. This multiplex RT-PCR was found to be less sensitive than the single RT-PCRs, i.e. using the multiplex RT-PCR no HuCVs could be detected and HAVs were only detected up to a dilution of 1:8 compared to detection at dilutions of 1:2 and 1:32 respectively for the individual RT-PCRs.

2.3.3 Molecular characterisation of viruses

A dideoxynucleotide chain termination sequencing method was established for the direct sequencing of PCR amplicons. For the characterisation of the HAV, HuCV and HAstV sequences of field isolates were then compared to sequences from well-described reference strains using the CLUSTAL X program. HRVs and human adenoviruses were typed by specific RT-PCRs and restriction enzyme analysis respectively.

2.3.4 Analysis of irrigation water and minimally processed food samples.

Commercially available MPFs: Six samples of packaged strawberries, from two different sources, and two of cherry tomatoes were analysed for HAV, HAstV and HuCVs. No viruses were detected on the surfaces of any of these MPFs.

Irrigation water and associated MPFs: Samples of MPFs and associated irrigation water were obtained from two rural areas, namely Letaba and Venda. The MPF samples from Venda were washed on site, where possible, in 25 ml PBS/50 g sample. The MPF-PBS was analysed at the University of Venda for indicator organisms and pathogenic bacteria and the remaining MPF-PBS wash was forwarded to the University of Pretoria for virological analysis. Two litres samples of irrigation water, i.e. either borehole or river, were forwarded to the University of Pretoria for virological analysis while at the University of Venda bacteriological analysis was performed on simultaneously water drawn samples.
Letaba: A single river sample and associated MPF (cabbage) was analysed for HAV and HAstV, but no viruses were detected.
Venda: A total of nine sets of samples, i.e. irrigation and one or more associated MPFs, were received from Venda, five in 2001 and four in 2002. Analysis for bacteria, indicator organisms and selected pathogens was performed on the samples. Viruses were recovered from both irrigation water and associated MPFs, inoculated onto a variety of cell cultures and then the sample and associated cell cultures analysed for HAV, HAstV, HuCV and HRV by RT-PCR.

Faecal coliform counts of >1000 per 100 ml were evident in a number of the irrigation water samples. This water should therefore not be used for any produce in contact with humans (Department of Water Affairs and Forestry, 1996). In addition pathogenic bacteria such as Salmonella, Shigella and E. coli were detected in a number of irrigation water and MPF samples. HRVs were detected in one irrigation water sample and from one MPF, namely tomatoes.

2.4 Conclusions

The objectives of the project have been met and the following conclusions drawn:

1. Techniques for the recovery, detection and characterisation of potentially pathogenic food- and waterborne viruses in raw and partially treated irrigation water and associated MPFs have been established.
2. Although a multiplex RT-PCR may be an attractive cost-effective alternate to individual RT-PCRs for the detection of selective viruses, the multiplex RT-PCR was found not be as sensitive as individual RT-PCRs for the viruses assessed.
3. No HAVs, HAstVs or HuCVs were detected in any of the irrigation water samples or associated MPFs from Venda. However, HRVs were detected in one of the irrigation water samples, namely Tshikuwi River sample, drawn on 2002/07/09. This finding is highly significant as the peak incidence of HRV infection is in the winter months. The occurrence of HRV in this water sample correlates with the high thermotolerant or faecal coliform count indicating faecal pollution of the water. However no HRVs were detected on any of the associated MPFs for that particular sampling date. HRVs were however detected on a MPF (tomato) on a previous sampling date where no HRVs were detected in the associated irrigation water. Characterisation data on the HRV isolates are still outstanding
4. The isolation and detection of enteroviruses, human adenoviruses, HRVs and HAstVs from a river used for domestic purposes and as irrigation water suggests that this water could pose a potential health risk, but more data are required to quantify the risk.

2.5 Future research

Future research should focus on the following:

1. The use of larger volumes of irrigation water for the recovery and detection of
2. potentially pathogenic viruses
3. Analysis of irrigation water and associated MPFs for pathogenic protozoan parasites
4. The development of quantitative real-time RT-PCRs to quantify the viruses present in field samples
5. Determine whether there are seasonal fluctuations in the type of micro-organisms present in irrigation water and associated MPFs.

3. Collaboration

Part A: Invaluable collaboration was built up in this project with the following organisations:

1. Division of Agrimeteorology of the Agricultural Research Council for the present and future analyses of pollution patterns in relation to weather data.
2. Department of Microbiology, University of Stellenbosch for the exchange of knowledge and experience in the field of environmental pollution.
3. Department of Water Affairs and Forestry, Western Cape for the sharing of samples and sample data as well as co-operation in many spheres of knowledge of water sources in the province.
4. Prof EP Jacobs, Chief Researcher, Institute for Polymer Science, University of Stellenbosch for discussions and collaboration on membrane technology used to remove pathogens from water.
5. The Kayamandi Steering Committee and the greater community of Kayamandi itself with whom valuable bonds have been established and from most of whom good co-operation had been received.
6. The Wildlife and Environment Society of South Africa, and especially Mr Stephen Finnemore, who shared previous data and correspondence on the river.

Part B: Prof JE Walter and Prof DK Mitchell, Center for Pediatric Research, Eastern Virginia Medical School, Norfolk, Virginia gave advice with regard to the molecular detection and characterisation of HAstVs and supplied valuable references viruses

4 Technology transfer

Part A: Technology used in the investigations of river water has been shared with other laboratories as well as the Department of Water Affairs and Forestry and the Stellenbosch Municipality

Part B: Technology developed during the course of this project has been transferred to personnel from other institutions, namely University of Venda, Medical University of Southern Africa, Noguchi Memorial Institute for Medical Research, University of Ghana (Dr G Armah), and University of Botswana (Ms M Kasule).

CAPACITY BUILDING

Part A:

1. Ms Wesaal Khan, a PhD student of the Department of Microbiology, Faculty of Science, University of Stellenbosch used some of the organisms isolated during the course of this study for further investigation
2. Several community workshops involving various segments of the Kayamandi community, DWAF, the Stellenbosch Municipality and other stakeholders.

Part B:
 
The work was carried out by post-graduate students and the project therefore contributed to the training of manpower in advanced technology for the microbial analysis of water and minimally processed foods.

1. This project formed the basis for the practical training of Ms T Naus for the degree BSc (Hons) Medical Virology;
2. University of Pretoria students Ms S Nadan and Ms JME Venter working on HAstVs and HAV respectively participated in the project as part of their investigations for their MSc degrees;
3. Student assistants from the University of Venda gained experience in the bacteriological analysis of irrigation waters and associated minimally processed foods. Neither student chose to continue with further post-graduate studies.
4. Two PhD students from the Medical University of Southern Africa gained experience with regard to the typing of environmental viral isolates;
5. Post-doctoral fellow, WB van Zyl, gained experience with the virological
6. analysis of irrigation waters and associated minimally processed foods

RESEARCH OUTPUTS

Reports

Barnes JM, Taylor MB, Slabbert M, Huisamen M and Post-graduate students. Part A: Presence, antibiotic susceptibility patterns and chlorination resistance of faecal pollution from a dense settlement and the possible impact of polluted water on downstream water users. Water Research Commission

Taylor MB, Nadan S, Naus TM, Venter JME, van Zyl WB, Potgieter N. Health risk assessment in connection with the use of microbially contaminated source waters for irrigation. Part B: Microbiological analysis of irrigation waters and food crops. Water Research Commission

Barnes JM. Report on water quality analyses of the Plankenbrug River. 13 March 2002. Issued to Municipality of Stellenbosch, Dept of Water Affairs and Forestry, and Stellenbosch Hospital.

Theses

BSc (Hons) Medical Virology
Naus T. The recovery and detection of hepatitis A virus and human astrovirus from irrigation waters and associated minimally processed foods. Pretoria, South Africa: University of Pretoria; 2001.

MSc (Medical Virology)
Nadan S. Characterisation of astroviruses from selected clinical and environmental settings. Pretoria, South Africa: University of Pretoria; 2001.

PhD (Community Health)
Barnes JM. The impact of water pollution from formal and informal urban developments along the Plankenbrug River on water quality and health risk. Stellenbosch, South Africa: University of Stellenbosch (submitted)

Publications
Nadan S, Walter JE, Grabow WOK, Mitchell DK, Taylor MB. Molecular characterization of astroviruses by reverse transcriptase PCR and sequence analysis: Comparison of clinical and environmental isolates. Appl Environ Microbiol 2003;69:747-753.

Newspaper articles

Cloete L. Kayamandi on the road to cleanliness. Eikestadnuus 2001 November 9; Vol 52: 2.

Thom A. The story of a sick river. Cape Argus. 2002 August 28:16.

Retief E. Besoedel - rivier vuil en gevaarlik. Eikestadnuus 2002 September 13;1.

Anon. Langtermynplan nodig vir besoedeling. Eikestadnuus 2002 September 13;2.

Schumann C. Speedy action needed for Plankenbrug [Letter] Eikestadnuus. 2002 September 27;6.

Retief E. Stappe nodig vir besoedelde rivier. Eikestadnuus. 2002 October 4;2.

Editorial. Don't kill the messenger. Eikestadnuus. 2002 October 4;8.

Retief. E. A little can make a big difference. Eikestadnuus. 2002 October 11;7.

Van Zyl J. Hoe water, hoe kwater. Die Burger. 2002 October 18;11.

Bonthuys J. Eikestad-rivier 'n gesondheidsramp. Cholera-vrees: Munisipaliteit kry ultimatum van staat. Die Burger. 2002 October 23;11.

Anon. Project will divert polluted stormwater from river. Eikestadnuus. 2002 October 25.

Retief E. A clean Kayamandi, a cleaner river. Eikestadnuus. 2002 November 22;9.

Radio talks

Barnes JM. SA FM: Interview on paper presented to Infectious Diseases Conference. 2001 December 7;18h15 (Repeated Sunday 2001 December 9).

Barnes JM. SA FM: Interview with Sue Valentine 2002 June 19. Broadcast in three parts on SA FM (dates not known)

Barnes JM. Matie FM Radio: Interview about the pollution in the Plankenbrug River. 2002 September 12; 18h30.

Conference presentations
International : None

Conference presentations
National/Regional

Barnes JM, De Villiers AS. Complex interaction between river pollution and a dense settlement along the Plankenbrug River. 44th Academic Day, Faculty of Medicine, University of Stellenbosch, 24 August 2000: Tygerberg.

Barnes JM. Complex interaction of factors causing serious environmental pollution and health effects in a peri-urban dense settlement receiving clean municipal water. Joint Congress of the Infectious Diseases & Sexually Transmitted Diseases Societies of Southern Africa. 2 - 7 December 2001: Music Conservatoire, Stellenbosch, South Africa.

Kloppers W, Barnes JM. Implementation of a national strategy to manage water quality effects from densely populated settlements: Kayamandi model project. WISA Biennial Conference and exhibition of The Water Institute of Southern Africa, 28 May - 1 June 2000: Sun City Conference Centre, North West Province, South Africa.

Nadan S, JE Walter, Grabow WOK, Taylor MB. The molecular detection and characterisation of astroviruses from human stool specimens and sewage. [Presentation]. "Microbial Diversity" 12th Biennial Congress of the South African Society for Microbiology, Faculty of Health Sciences, University of the Free State 2-5 April 2002: Bloemfontein.

Naus T, Grabow WOK, Taylor MB. The assessment of techniques for the recovery and detection of hepatitis A- and astroviruses from minimally processed foods [Poster]. Faculty Day, Faculty of Health Sciences, University of Pretoria 21-22 August 2001: Pretoria.

Van der Linde M, Taylor MB, Grabow WOK, van Zyl WB. Occurrence of rotaviruses in selected water samples [Poster]. Faculty Day, Faculty of Health Sciences, University of Pretoria, 21-22 August 2001: Pretoria

Venter JME, Grabow WOK, Taylor MB. Comparison of methods for the isolation and detection of hepatitis A virus from water samples [Poster]. Faculty Day, Faculty of Health Sciences, University of Pretoria, 21-22 August 2001: Pretoria.

Venter JME, Grabow WOK, Taylor MB. Comparison of methods for the isolation and detection of hepatitis A virus in water samples [Presentation]. "Microbial Diversity" 12th Biennial Congress of the South African Society for Microbiology, Faculty of Health Sciences, University of the Free State, 2-5 April 2002: Bloemfontein.