further studies on the incidence of helicobacter organisms in...

Further studies on the incidence of Helicobacter

organisms in water supplies.

A research report from the Health Protection Agency to the Drinking

Water Inspectorate for

Variation DWI/70/2/146

FINAL REPORT – December 2004

Geraldine Brick, Stephanie Chisholm, Sandra Lai, Bengü Said, Robert Owen

Gordon Nichols, John V. Lee, Susanne Surman-Lee

Final Report –DWI/70/2/146 extension December 2004

Contents

Further studies on the incidence of Helicobacter organisms in water supplies. ............................................................................................................ 1 1 Executive Summary ..................................................................................................... 5

1.1 Development of a protocol for culture of H. pylori from water and biofilm samples .......................5 1.2 Evaluation of solid culture media......................................................................................................6 1.3 Evaluation of enrichment broths .......................................................................................................6 1.4 Evaluation of methods for initial sample processing.........................................................................7 1.5 Evaluation of a preliminary culture protocol.....................................................................................8 1.6 Application of culture and molecular protocols to survey domestic properties ................................8 1.7 Risks to people consuming mains drinking water .............................................................................9

2 Introduction............................................................................................................... 11 2.1 Background .....................................................................................................................................12 2.2 Aims ................................................................................................................................................15 2.3 Objectives........................................................................................................................................15

3 Materials and methods .............................................................................................. 17 3.1 Quality control ................................................................................................................................17 3.2 Bacterial isolates .............................................................................................................................17 3.3 Storage and culture of Helicobacter pylori .....................................................................................17 3.4 Identification of H. pylori................................................................................................................18 3.5 H. pylori spiked experiments...........................................................................................................19 3.6 Acid treatment.................................................................................................................................19 3.7 Immunomagnetic separation (IMS).................................................................................................20 3.8 DNA extraction from water and biofilm samples ...........................................................................20 3.9 PCR-based detection of helicobacters .............................................................................................20 3.10 Recruitment of properties to the study........................................................................................21 3.11 First round of sampling ..............................................................................................................22 3.12 Second round of sampling ..........................................................................................................22 3.13 Third round of sampling.............................................................................................................22

4 Development of culture methods for H. pylori ......................................................... 23 4.1 Evaluation of solid culture media for recovery from Maximum Recovery Diluent (MRD) ...........23

4.1.1 Experimental design...................................................................................... 23 4.1.2 Results - evaluation of solid culture media for recovery from MRD ........... 23 4.1.3 Conclusion - evaluation of solid culture media for recovery from MRD..... 26

4.2 Evaluation of solid culture media for recovery from spiked sterilised tap water ............................26 4.2.1 Experimental design...................................................................................... 26 4.2.2 Results – evaluation of solid media for recovery from spiked sterilised tap water 26 4.2.3 Conclusion - evaluation of solid media for recovery from spiked sterilised tap water.................................................................................................................... 27

4.3 Evaluation of the Legionella antibiotic supplement ........................................................................28 4.3.1 Experimental design...................................................................................... 28 4.3.2 Results - Evaluation of the Legionella antibiotic supplement ...................... 28 4.3.3 Conclusion - Evaluation of the Legionella antibiotic supplement................ 28

5 Evaluation of broth media for enrichment of H. pylori ............................................ 29 5.1 Evaluation of blood-free broth media..............................................................................................29

5.1.1 Experimental design...................................................................................... 29 5.1.2 Results – evaluation of blood free broth media ............................................ 30

2


5.1.3 Conclusions - evaluation of blood-free broth media..................................... 31 5.2 Evaluation of blood-containing liquid media ..................................................................................31

5.2.1 Experimental design...................................................................................... 31 5.2.2 Results – evaluation of blood containing liquid media................................. 32 5.2.3 Conclusions – evaluation of blood containing liquid media......................... 33

5.3 Evaluation of charcoal as an alternative to blood as a broth supplement. .......................................34 5.3.1 Experimental design...................................................................................... 34 5.3.2 Results - Evaluation of charcoal as an alternative to blood as a broth supplement ................................................................................................................ 34 5.3.3 Conclusion - Evaluation of charcoal as an alternative to blood as a broth supplement. ............................................................................................................... 36

5.4 Effect of solid medium on growth after enrichment........................................................................36 5.4.1 Experimental design...................................................................................... 36 5.4.2 Results - Effect of solid medium on growth after enrichment...................... 37 5.4.3 Conclusion - Effect of solid medium on growth after enrichment ............... 38

6 Evaluation of methods for sample processing .......................................................... 39 6.1 Evaluation of acid treatment of samples .........................................................................................39

6.1.1 Experimental design...................................................................................... 39 6.1.2 Results - Evaluation of acid treatment of samples........................................ 40 6.1.3 Conclusion - Evaluation of acid treatment of samples ................................. 41

6.2 Evaluation of effect of the acid treatment duration on the levels of bacterial contamination .........41 6.2.1 Experimental design...................................................................................... 41 6.2.2 Results - Evaluation of effect of the acid treatment duration on the levels of bacterial contamination............................................................................................. 41 6.2.3 Conclusion - Evaluation of effect of the acid treatment duration on the levels of bacterial contamination......................................................................................... 42

6.3 Evaluation of the effect of different filtration procedures ...............................................................42 6.3.1 Experimental design...................................................................................... 43 6.3.2 Results - Evaluation of the effect of different filtration procedures ............. 43 6.3.3 Conclusion - Evaluation of the effect of different filtration procedures....... 43

6.4 Evaluation of the effect of dual filtration and acid treatment ..........................................................43 6.4.1 Experimental design...................................................................................... 43 6.4.2 Results - Evaluation of the effect of dual filtration and acid treatment ........ 44 6.4.3 Conclusion - Evaluation of the effect of dual filtration and acid treatment.. 46

6.5 Effect of dual filtration and differential centrifugation ...................................................................47 6.5.1 Experimental design...................................................................................... 47 6.5.2 Results - Effect of dual filtration and differential centrifugation ................. 47 6.5.3 Conclusion - Effect of dual filtration and differential centrifugation ........... 48

6.6 Examination of Immunomagnetic Separation (IMS) ......................................................................50 6.6.1 Experimental design...................................................................................... 50 6.6.2 Results - Examination of Immunomagnetic Separation (IMS)..................... 50 6.6.3 Conclusion - Examination of Immunomagnetic Separation (IMS) .............. 51

7 Evaluation of combined sample processing, pre-enrichment and culture methods.. 52 7.1 Evaluation of acid treatment of samples with different culture media ............................................52

7.1.1 Experimental design...................................................................................... 52 7.1.2 Results - Evaluation of acid treatment of samples with different culture media 52

3


7.1.3 Conclusion - Evaluation of acid treatment of samples with different culture media 56

7.2 Evaluation of enrichment of acid treated samples cultured on four different media.......................57 7.2.1 Experimental design...................................................................................... 57 7.2.2 Results - Evaluation of enrichment of acid treated samples cultured on four different media .......................................................................................................... 59 7.2.3 Conclusion - Evaluation of enrichment of acid treated samples cultured on four different media .................................................................................................. 61

7.3 Evaluation of the improved protocol...............................................................................................61 7.3.1 Experimental design...................................................................................... 61 7.3.2 Conclusion - To assess the application of the improved protocol ................ 66

8 Investigation of Domestic Systems........................................................................... 67 8.1 Phase one of survey of domestic properties ....................................................................................67

8.1.1 Results - Phase one of survey of domestic properties................................... 67 8.1.2 Conclusion - Phase one of survey of domestic properties ............................ 69

8.2 Phase Two of survey of domestic properties...................................................................................70 8.2.1 Results - Phase Two of survey of domestic properties ................................. 71 8.2.2 Conclusion - Phase Two of survey of domestic properties........................... 77

8.3 Phase Three of survey of domestic properties.................................................................................78 8.3 Phase Three of survey of domestic properties.................................................................................79

8.2.3 Results - Phase Three of survey of domestic properties ............................... 79 8.2.4 Conclusion - Phase Three of survey of domestic properties......................... 88

9 Effluent Samples....................................................................................................... 89 9.1 Conclusion - Effluent Samples........................................................................................................89

10 Overview of Results.............................................................................................. 90 10.1 Types of property .......................................................................................................................91 10.2 Cold water systems.....................................................................................................................92 10.3 WC cistern..................................................................................................................................95

11 Discussion ............................................................................................................. 96 12 Acknowledgements............................................................................................. 100 13 References........................................................................................................... 101 14 Appendices.......................................................................................................... 105

14.1 Appendix 1 Letter to volunteers ...............................................................................................105 14.2 Appendix 2 Protocol for sampling water systems ....................................................................106 14.3 Appendix 3 Property survey questionnaire...............................................................................109 14.4 Appendix 4 Property survey repeat sample questionnaire........................................................112 14.5 Appendix 5 Composition of agar media (amount L-1) used for culture of H. pylori ...............113 14.6 Appendix 6 Main components of broth media (gL-1) tested for enrichment of H. pylori........114

4


Executive Summary 1

This study has conducted developmental work to improve the methodology for

isolation of Helicobacter species, including H. pylori. The study developed

improved protocols for the collection, processing and culture of samples. The

survival and recovery of H. pylori subjected to stress conditions, including limited

nutrients and low temperature, was investigated.

The study involved a sampling programme of three phases. Phase one was a

survey of domestic properties. Samples were collected from a total of 62

properties recruited from 13 different areas within Greater London; 41 from the

Northwest London area, 16 from North London, one from each of East, West and

South London and no information was obtained from two properties. They

consisted of 56 biofilm and 62 water samples. Phase two involved the

recruitment of samples from properties that were Polymerase Chain Reaction

(PCR)-positive for Helicobacter by at least one of the three PCR assays

(HGS16S targeting the 16S rRNA Helicobacter genus-specific gene, HPVacA

and 16SHP targeting the VacA and 16SrRNA H. pylori species-specific genes

respectively) in sampling phase one. Samples were collected from nine

properties; they consisted of both biofilm and water samples. A modified protocol

was tested against samples from domestic properties collected in phase three.

Both biofilm and water samples were collected from a total of 27 properties; eight

of which were properties that had provided PCR positive samples in the first

round. Seven properties were included in each sampling phase. In addition, two

water effluent samples; primary and final effluent, were supplied by the sewage

treatment works.

1.1 Development of a protocol for culture of H. pylori from water and

biofilm samples A range of parameters was evaluated in order to improve the detection, isolation

and recovery of H. pylori from domestic water and biofilm samples. This included

5


assessment of media containing different levels of nutrients and selective agents

and the development of enrichment broth for the primary isolation of H. pylori.

1.2 Evaluation of solid culture media Four solid culture media (Columbia Blood Agar, Columbia Blood Agar containing

twice the concentration of Dent’s antibiotic supplement, Degnan agar and Half

strength H. pylori special peptone agar) were evaluated for their ability to recover

helicobacter from Maximum recovery diluent (MRD) and H. pylori that had

undergone nutrient and temperature shock by suspension in sterile tap water.

Overall, CBA medium allowed the best recovery of H. pylori after suspension in

MRD for up to 24 hours, both in terms of speed of growth and colony counts.

The Degnan agar medium gave the least satisfactory results. There were

considerable variations in growth characteristics between strains. Nevertheless,

all strains of H. pylori showed poor recovery after suspension in filtered tap water.

The ability to recover each strain decreased significantly after 24h incubation in

tap water. For all strains, the highest colony counts from filter-sterilized tap water

were observed on CBA, although a longer incubation of 72h was required to

visualise colonies.

1.3 Evaluation of enrichment broths An enrichment step is beneficial for increasing the numbers of H. pylori but the

growth of background contamination needs to be controlled. Enrichment with

four blood-free broth media (MRD, Full Strength special peptone (FSPB), Half

strength special peptone broth (ASPB), Brucella broth with 0.1% β cyclodextrin

(BBβC)) were tested. None of the broths enriched the growth of H. pylori.

However, colonies were visible for longer on the plates when they were

incubated in BBβC. The reduction in colony forming units on CBA plates from 24

and 48h broth cultures suggests that there is cell death or conversion of cells to a

non-culturable (VNC) state. Exclusion of antibiotics from the broth did not

significantly improve enrichment.

6


The ability to enrich H. pylori growth with two blood-containing liquid media

(Brucella broth with 1% yeast extract (BBYE), Brucella broth with 1% yeast

extract and 7% defibrinated horse blood (BBYEB), Mueller-Hinton broth with 1%

yeast extract (MHYE), Mueller-Hinton broth with 1% yeast extract and 7%

defibrinated horse blood (MHYEB)) was also evaluated. The results

demonstrated the potential of blood-containing broths to enrich H. pylori growth.

The period of incubation required and the level of enrichment are strain-

dependent. Additionally prolonged incubation of broths lead to overgrowth of

bacterial contaminants.

1.4 Evaluation of methods for initial sample processing Both physical and chemical methods of sample processing were explored to

reduce levels of contaminating microflora and to concentrate H. pylori. These

included double filtration, differential centrifugation, acid treatment, use of

antibiotic supplements and the application of Immunomagnetic Separation (IMS).

Findings from experiments, which explored different filter pore sizes (0.65µm,

0.45 µm and 0.2 µm) for their capacity to trap H. pylori cells indicated that H.

pylori cells could not penetrate membranes with pore sizes of 0.45 µm or 0.2 µm.

Different strains of H. pylori show different levels of acid tolerance. However,

acid treatment reduced the amount and mixture of bacterial contamination.

Meanwhile, higher numbers of H. pylori colonies were recorded when levels of

bacterial contaminants were reduced.

A combination of differential centrifugation and double filtration revealed that

despite a reduction in bacterial contamination from water samples, the proportion

of the H. pylori cells lost during the first lower speed centrifugation outweighed

any benefit of its application.

The acid treatment step showed promise for reducing the level of background

contamination, without affecting the recovery of H. pylori. The highest rate of H.

7


pylori recovery was from the acid treated samples plated on Dent’s medium,

where the lowest levels of bacterial contamination were observed.

Concentrating H. pylori cells from a sample by magnetic beads coated with

specific (anti-H. pylori) antibody to capture H. pylori cells was explored. This

approach reduced the recovery of H. pylori and the method was not pursued

further. .

1.5 Evaluation of a preliminary culture protocol Testing the initial protocol developed during this study with samples collected at

phase two showed that contamination was a consistent problem on the majority

of the culture plates. This was more apparent on plates derived from water

samples than those from biofilm samples. Moreover, for the biofilm samples,

more than 55% of the Dent’s, CBA2D and Degnan plates inoculated with broth

and antibiotics, did not yield any bacterial growth, regardless of the incubation

time. This is surprising and may be explained by the fact that biofilm was not

recovered on the swab when sampling the toilet cistern.

1.6 Application of culture and molecular protocols to survey domestic

properties Contamination levels were too high in some samples collected in phase two, in

particular the water samples. Therefore, the possibility that H. pylori are present

should not be excluded. However, H. pylori were not isolated on any of the

plates where contamination was absent or only moderate. These results may

indicate that H. pylori are not present, not culturable from these samples, or that

the cells have entered a viable but non-cultivable state. Alternatively, H. pylori

may be present at a level that is below the threshold of sensitivity for the methods

developed.

A modified protocol for isolating H. pylori from domestic water samples was

applied to samples collected in phase three. This protocol included two extra

broths, one with antibiotics and one without, for samples that had not undergone

8


acid treatment. No H. pylori have been cultured in 54 samples from 27 properties

from this round of sampling. Observations indicate that contaminating growth

continues to be a problem. This study has highlighted the limitations of a wider

application of this protocol developed during the course of the study.

PCR results indicated that 15/118 (12.7%) of samples collected from phase one,

were weakly positive by at least one PCR assay. Two of the 15 were positive for

all three assays (Logan et al. 2000, Ho et al. 1991 and Chisholm et al. 2001)

although the Ho et al. 1991 and the Chisholm et al. 2001 assays were very faint

positives. One of the 15 was positive by the Logan et al. 2000 and the Ho et al.

1991 assay and the remaining 12 samples were positive by the Logan et al. 2000

assay alone. All 18 samples collected during phas two were PCR-negative, while

11/54 samples from phase three were positive by at least one PCR assay.

No H. pylori were recovered in culture from samples collected from phase 2. All

samples were PCR negative. This was unexpected as the samples were from

previously PCR positive properties, including one property that had a biofilm

sample positive in all three PCR assays. This may suggest that the presence of

H. pylori in domestic properties is a transient occurrence. It is possible that the

survival of helicobacter is poor during the summer months and / or there may be

a higher disinfectant residual chlorine/ monochloramine in the mains water during

the summer.

1.7 Risks to people consuming mains drinking water

The study has demonstrated that although H. pylori can be detected by molecular

methods in drinking water the organism has not been isolated, despite extensive

cultural investigation. Even if viable organisms were present in such low

numbers, that their successful culture using the presently available culture

protocols remains unresolved because of heavy overgrowths of other

microorganisms in the water and biofilms. It is most likely that these organisms

9


are dead or in a viable but non-culturable state. This is because there is

considerable evidence for the poor survival of H. pylori in water, evidence that

disinfection used in treatment and as a residual should prevent its survival and

lack of clear evidence that these organisms are viable but non-cultivable. While it

remains possible that occasional contamination of mains water with H. pylori

could occur the surveillance systems would not be able to detect an outbreak of

H. pylori gastritis. It remains likely that the predominant mode of transmission of

H. pylori in the UK is by person-to-person transmission in childhood.

10


2 Introduction

Helicobacter pylori is a Gram-negative bacterium associated with various chronic

infections including gastritis, duodenal or gastric ulcers, and gastric cancer in

humans (Sanders et al., 2002). H. pylori may infect around half the worlds

population and lifelong carriage is common (Axon, 1997). Although infections are

more prevalent in developing countries, around 30% of the UK population is

infected (Vyse et al., 2002). The principal reservoir of H. pylori appears to be the

human stomach, and transmission is likely to be person-to person via the faecal-

oral route (Brown, 2000; Go, 2002). However it has been suggested that water

and the associated biofilm may be an environmental transmission route for

spreading the disease. Several studies have reported evidence of Helicobacter

species in water by PCR based methods (Hegarty et al., 1999; Krumbiegel et al.,

2001; Park et al., 2001; Mazari-Hiriart et al., 2001; Sasaki et al., 1999) including

our initial study on the presence of Helicobacter in public mains water in England

(Watson et al., 2004; Whale et al., 2003). Drinking water, biofilm and deposit

samples were taken from domestic properties, schools, hydrants, reservoirs and

water meters fed by three different distribution systems. The samples were

tested for the presence of Helicobacter spp. by culture and by molecular

methods. Although Helicobacter were not cultured there was evidence of

Helicobacter spp. in 26% (39 of 151) of samples and six positives were identified

as H. pylori.

The absence of culturable H. pylori in samples implies that the methodologies for

isolating Helicobacter and specifically H. pylori from environmental samples are

inadequate. The growing body of data from this and other investigations suggest

that Helicobacter are present in water distribution systems, particularly within

buildings and domestic premises. It remains unclear to what extent there is a

health risk from this source and further development work is required for an

assured position on the significance of these results. Therefore the current

extension study focussed on improving the methodology for the isolation of

11


Helicobacter spp. using fresh water and biofilm samples taken from domestic

properties.

2.1 Background Helicobacter pylori was first isolated in 1982 by Warren and Marshall and was

first cultured in 1984 (Stevenson et al. 2000). It was originally called

Campylobacter pylori but in 1989 it was transferred to a new genus. It is a Gram-

negative organism that grows microaerophically and is catalase, oxidase and

urease positive (Velázquez and Feirtag 1999). It has been implicated as a cause

of gastric ulcers in humans. The World Health Organization has classified it as a

type 1 carcinogen due to its role in stomach cancer (Jiang and Doyle, 2002).

A paper by Stevenson et al. 2000 compared the growth of H. pylori on different

broths and on different agar. The broths used were Mueller-Hinton broth, brain

heart infusion broth and H. pylori special peptone broth (HPSPB). All the broths

were supplemented with calf serum containing iron. They found that the H. pylori

grew equally well in all of their broths. The following agar media were also

evaluated, Columbia agar, Mueller-Hinton agar, modified Glupczynski’s Brussels

campylobacter charcoal agar, Johnson-Murano agar and H. pylori special

peptone agar (HPSPA). They found that on Glupczynski’s agar there was poorer

growth of H. pylori than on the other agars. The rate of recovery was similar for

the other four agars; however they noted that the colony size was larger on the

HPSPA than on the others.

A study undertaken by Degnan, Sonzogni and Standridge, 2003, involved the

development of a selective medium for the isolation of H. pylori from water. They

first looked at several standard media (brain heart infusion, brucella agar,

Columbia blood agar base, campylobacter agar kit Skirrow and HPSPA medium),

on which they grew strains from common waterborne bacteria. They found that

these bacteria grew more quickly than the H. pylori and the antibiotic supplement

that they had added to their standard media was not able to inhibit all of the

12


growth. They took the HPSPA base and added the antibiotic supplement, and

two antifungals, polymyxin B and amphotericin B and a phenol red indicator.

After inoculating this with a spiked non-sterile well water sample they found that

during seven days of incubation only the H. pylori colonies grew.

Azevedo et al. 2004, looked at culturing H. pylori from water. The media used

consisted of Columbia agar (CA), Wilkins-Chalgren agar (WCA) and H. pylori

special peptone agar (HPSPA); supplemented with 5% defibrinated horse blood.

They used different strengths; full strength, half strength and quarter strength, of

these media to see if nutrient shock was a factor in recovering the organism from

water. The quarter strength media gave no results. Their results showed that

better rates of recovery were achieved with half strength media than full strength

media and half strength CA and half strength WCA were better than half strength

HPSPA.

Different studies have been undertaken on the culturing of H. pylori. Taneera et

al. 2002 studied the influence of different growth supplements on Helicobacter

species. A comparison was made between activated charcoal, β-cyclodextrin

and porcine gastric mucin. Gonococcal broth (GB) and GAB-Camp agar (GCA)

were used for culturing H. pylori. The study revealed that the highest colony

forming unit numbers and the largest colony diameter after four days of growth

was obtained on the GCA medium supplemented with only 0.1% (w/v) activated

charcoal (diameter 1.2 mm). In addition, a complete spiral to coccoid conversion

on GCA supplemented with activated charcoal occurred on day 13 and with β-

cyclodextrin on day 14. In their study, H. pylori cultured in GB supplemented only

with 0.1% (w/v) activated charcoal showed the highest numbers of cfu/ml.

Complete conversion to a spiral form occurred on day 8 for GB supplemented

with activated charcoal and on day 9 for GB supplemented with β-cyclodextrin.

Walsh and Moran, 1997, compared several different supplements in four liquid

based and solid based media to see the effect of their growth on H. pylori. These

13


were brain heart infusion broth, Brucella broth, Mueller-Hinton broth and tryptone

soya broth. Culture media were supplemented with combinations of 5% (v/v)

horse serum, 5% (v/v) fetal calf serum, 7% (v/v) defibrinated horse blood, 1%

(w/v) yeast extract, 2%(w/v) proteose peptone no. 3, 0.1% (w/v) β-cyclodextrin

and 0.1% (w/v) corn starch. Two strains of H. pylori NCTC 11637 and NCTC

11638 were employed to evaluate the media which supported the best growth.

Growth of NCTC 11637 was best on tryptone soya agar containing blood and for

NCTC 11638 strain, Mueller-Hinton agar supplemented with blood was preferred.

Assessment of liquid based media showed that Mueller-Hinton broth

supplemented with 1% (w/v) yeast extract and 7% defibrinated horse blood gave

the highest growth rate for both of the strains. For blood free liquid media, either

the Mueller-Hinton broth or the tryptone soya agar supplemented with horse

serum and either yeast extract or proteose peptone no. 3 supported the highest

growth.

14


2.2 Aims • To continue developmental work to improve the methodology for the

isolation of Helicobacter species including H. pylori.

• To develop and test improved protocols for the collection, processing and

culture of samples to improve the isolation of helicobacter.

• To carry out a further programme of sampling and to report on the

prevalence and significance of any helicobacter isolated from water, biofilm

and deposit samples in domestic properties

2.3 Objectives • To assess H. pylori growth on a range of media containing different levels

of nutrients and selective agents.

• To examine survival and recovery of H. pylori that had been subjected to

stress (limited nutrients and low temperatures).

• To determine the impact of culture medium on recovery of H. pylori

through spiking experiments.

• To develop and assess enrichment broth media for the primary isolation of

H. pylori.

• To optimise sample processing to reduce levels of contaminating

microflora and to improve H. pylori recovery.

• To develop and evaluate a protocol for detection and isolation of H. pylori

from domestic water supply water and biofilm samples

• To identify a PCR-positive source of H. pylori from a water or biofilm

sample for evaluating culture work described in 2.3.6.

15


• To undertake further sampling of domestic properties for validation of

culture and PCR-based methods

16


3 Materials and methods

3.1 Quality control The laboratories operate under UKAS (FSML accreditation No. 1595) and CPA

(LEP accreditation No. 1683) quality systems. Throughout the study known

negative controls and positive controls of H. pylori were included in each set of

analyses.

3.2 Bacterial isolates Six cultures of Helicobacter pylori were used in this study, all of which originated

from human gastric biopsies, five from England and one from Australia.

The two reference strains included were NCTC 12445 (H899=genome

sequenced strain 26695) and NCTC 11637 (H509; the type strain), both of which

have been frequently subcultured on artificial media (laboratory adapted strains)

since isolation in 1983 (England) and in 1982 (Australia) respectively. The four

clinical isolates (H4780, H5027, H5029 and H5425) were from patients in

England (mid-Essex), and had been subjected to minimal (< four) subcultures

since primary isolation.

3.3 Storage and culture of Helicobacter pylori Stock cultures were stored in cryopreservative medium on glass beads at -800C

(Microbank system, Pro-Lab Diagnostics, Wirral, UK).

The H. pylori were cultured on Columbia agar base (Oxoid) containing 10% (v/v)

defibrinated horse blood (CBA), and incubated for 2-3d at 37°C under

microaerobic conditions (4% oxygen, 5% hydrogen, 5% carbon dioxide and 86%

nitrogen) in a modular atmosphere-controlled system (MACS-VA500) workstation

(Don Whitley Scientific Ltd., Yorkshire, United Kingdom).

The following other agar-based culture media were used (see Appendix 5 for

composition details):

• Columbia Blood Agar (CBA),

17


• Columbia Blood Agar containing twice the concentration of Dent’s

antibiotic supplement (CBA2D),

• H. pylori agar (DPA) as described by Degnan et al. (2003)

• H. pylori special peptone agar (HPSPA) as described by Stevenson et al.

(2000) but modified for use at half strength as described by Azevedo et al.

(2004) – the latter authors reported that H. pylori grew better in “low

nutrient media”.

The following broth media were used:

• Maximum recovery diluent (MRD)

• Full Strength special peptone broth (FSPB) (Stevenson)

• Half strength special peptone broth (ASPB) (Azevedo)

• Brucella broth with 0.1% β cyclodextrin (BBβC), as a recent study showing

that β cyclodextrin delayed the spiral to coccoid conversion (Taneera et al

2000)

• Brucella broth + 1% Yeast Extract (BBYE)

• Brucella broth + 1% Yeast Extract + 7% Defibrinated Horse Blood

(BBYEB)

• Mueller-Hinton broth + 1% Yeast Extract (MHYE)

• Mueller-Hinton broth + 1% Yeast Extract + 7% Defibrinated Horse Blood

(MHYEB)

• Brucella broth + 1% Yeast Extract + 0.1% Activated Charcoal (BBYEC)

• Brucella broth + 0.2% Agar + 0.1% Activated Charcoal (BBCA)

3.4 Identification of H. pylori Cultures were confirmed as H. pylori by Gram stain, and by positive reactions for

catalase, cytochrome oxidase and urease tests.

18


3.5 H. pylori spiked experiments The selected strains were obtained from beads that were frozen at -80°C and

streaked onto CBA plates. These plates were incubated in the MAC at 37°C for

72 hours. Ten colonies were selected from each plate and were re-streaked onto

fresh CBA plates and incubated in the MAC at 37°C for 48h. Bacterial

suspensions were prepared in either MRD or filter sterilised or autoclaved tap

water, to give an OD600 reading of 0.2 – 0.25 by spectroscopy. Tap water were

collected into sterile bottles containing sodium thiosulphate 5-hydrate (20mgl-1) to

neutralise any chlorine or other oxidising biocides. Aliquots (20µl) of serial

decimal dilutions of this neat medium (10-1 – 10-5) were dropped in triplicate onto

the dried surface of an agar plate (CBA or other test medium). Following

microaerophilic incubation for 48h (or longer in some experiments), the growth at

each dilution was recorded and where possible, individual colonies counted. The

number of colony forming units (cfu)/ml in the original suspension was calculated

by multiplying the mean number of colonies in 20µl by the dilution factor they

were recorded at, and multiplying this figure to give the number of colonies in 1

ml:

Colony count = mean no. of x dilution x 1000 (cfu/ml) colonies in 20µl factor 20

3.6 Acid treatment The acid buffer described by Bopp et al. (1981) consists of a HCl-KCl buffer with

a pH of approximately 2.2.

200µl of sample was added to 200µl of Bopp acid with 20µl of urea (Marshall et

al. 1990). This was left for 5 mins and 80µl of acid-treated bacterial suspension

was inoculated into the enrichment broth or 100µl was added directly to solid

media and incubated microaerophilically at 37°C for six and ten days

respectively.

19


3.7 Immunomagnetic separation (IMS) 1ml of sample was added to 40µl pre-coated anti-H. pylori magnetic beads and

agitated at 4°C for 1h. Tubes were placed in a magnetic rack for magnetic

particle collection (MPC) and the supernatant removed. The beads were re-

suspended in 1ml PBS/Tween-20 buffer, agitated gently at room temperature for

30mins and the supernatant removed following MPC. These wash steps were

repeated 4-5 times and the beads re-suspended in 200µl PBS/BSA.

3.8 DNA extraction from water and biofilm samples A 1ml aliquot of water/biofilm sample was concentrated by centrifugation. A

volume of supernatant was discarded, leaving 200 µl of concentrate. DNA was

extracted from the concentrate by the guanidinium-silica based method (Boom et

al. 1990).

3.9 PCR-based detection of helicobacters All DNA extracts were tested by three PCR assays: HGS16S targeting the 16S

rRNA Helicobacter genus-specific gene, HPvacA and 16SHP targeting the vacA

and 16SrRNA H. pylori species-specific genes respectively (Table 1). Samples

were considered positive for Helicobacter species if one or more assays showed

a band of equal molecular weight to the positive control. Amplification conditions

were as described in the literature. PCR amplicons were electrophoresed on a

1% (w/v) agarose gel at 120 volts and stained with ethidium bromide.

20


Table 1 Details of Helicobacter genus and species specific PCR assays applied

Target Gene

Primer Name

Sequence

(5′ → 3′)

Ta

(°C)

Product size (bp)

Reference

Helicobacter genus-specific

16S rrn H297F

H1026R

GGC TAT GAC GGG TAT CCG GC

GCC GTG CAG CAC CTG TTT TC 58 749

(Logan et

al. 2000)

H. pylori specific

16Srrn Hp1

Hp2

CTG GAG AGA CTA AGC CCT CC

ATT ACT GAC GCT GAT TGT GC 60 109

(Ho et al.

1991)

vacA vac3624F

vac3853R

GAG CGA GCT ATG GTT ATG AC

ACT CCA GCA TTC ATA TAG A 53 230

(Chisholm

et al. 2001)

3.10 Recruitment of properties to the study For this study fresh water and biofilm samples that could be brought directly to

the laboratory and processed within a few hours of sampling were used in order

to give the optimum chance of detecting and isolating Helicobacter spp. if they

were present. Participants were recruited through a global internal e-mail across

the HPA-Colindale site (Appendix 1). Volunteers were then invited to a meeting

to explain the study, to sign up to sampling dates and to have a brief training on

sampling procedures. Volunteers were provided with a sampling kit with

instructions (Appendix 2) and requested to provide a water sample and a biofilm

sample. The water samples were taken from; the shower, the bath mixer taps,

the bath tap or the bathroom hand basin taps and the biofilm samples were taken

from inside the WC cistern, provided it did not contain a disinfectant tablet, using

a swab and a 100cm2 template. Volunteers were also requested to fill in a

Property Survey Questionnaire (PSQ – Appendix 3), which was modified from the

questionnaire used in the previous DWI study. For subsequent rounds of

sampling a shortened version of the PSQ was used (Appendix 4).

21


3.11 First round of sampling The initial round of sampling took place in February 2004 over a period of a week

between Monday 2nd and Monday 9th. Participating properties were requested to

provide a 2L water sample, a biofilm sample and a completed PSQ.

3.12 Second round of sampling The second round of sampling took place in August 2004, on Tuesday 3rd,

Tuesday 10th, Thursday 12th and Tuesday 17th. Only properties positive for PCR

in the first round were requested to provide fresh samples for re-testing using the

newly developed protocol incorporating both PCR and culture of the organism.

On this occasion larger volumes of water (10L) and a repeat biofilm sample were

requested.

3.13 Third round of sampling The third round of sampling took place between September and November 2004.

The majority of participants had already provided samples in the previous rounds

of sampling. The participants were asked to provide a 2L water sample (based

on the experience of the previous sampling), a biofilm sample and to complete

the short PSQ.

A total of 64 properties were sampled: 98 water samples and 92 biofilm samples

(190 samples in total). In addition a sample of primary effluent and a sample of

final effluent were taken from the sewage treatment works in the Thames area.

22


Development of culture methods for H. pylori 4

4.1 Evaluation of solid culture media for recovery from Maximum Recovery Diluent (MRD)

4.1.1 Experimental design The four culture media below were compared.

• Columbia Blood Agar (CBA),

• Columbia Blood Agar containing twice the concentration of Dent’s

antibiotic supplement (CBA2D)

• HP agar as described by Degnan et al. 2003 (DPA)

• Half strength H. pylori special peptone agar (ASPA)

In experiment 1, serial decimal dilutions of each of the six strains of H. pylori

were prepared in MRD, as described in Methods section 3.4, and these were

subcultured on the four media. Colony counts were then performed after

microaerophilic incubation at 48, 72, 96, 120, 144 and 168h. To ascertain the

best media for recovery of H. pylori after suspension in MRD, the above

procedure was repeated in experiment 2; diluted strain suspensions were in air at

room temperature for 30 min (experiment 2A) and 24h (experiment 2B) before

subculture.

4.1.2 Results - evaluation of solid culture media for recovery from MRD Summary results are shown in Figure 1.

The mean values for CBA2D were calculated from data on four strains, as two

other strains failed to grow.

23


020406080

100120140160

0h (Experiment 1)

0.5h (Experiment 2)

24h (Experiment 2)

Incubation time in MRD

Mea

n (x

106 ) c

fu/m

l

0

1

2

3

4

5

6

7

Mea

n re

cove

ry ti

me

(d)

CBA ASPA CBA2D DPACBA ASPA CBA2D DPA

Hp recovery Recovery time

Figure 1: A comparison of the ability of four solid media to recover H. pylori (as

assessed by bacterial levels and recovery time) from MRD after 0, 0.5 and 24h incubation in air at room temperature

CBA was the optimal medium as the highest number of colonies were recovered

on CBA. In addition, colonies were visible on CBA at the lowest mean recovery

time. All six strains grew well on CBA and most were visible after 48h in all the

above experiments. The exception to this was H5027, which required 72h

incubation.

For four of the strains, growth on CBA2D was similar to that of CBA. However,

growth of two strains (H4780 and H5425) was completely inhibited. While the

higher antibiotic concentrations in CBA2D could decrease the growth of

contaminating micro-organisms, our results indicate this medium may also

prevent the growth of some H. pylori strains.

24


The inhibitory effect of CBA2D on H. pylori was tested further using twenty other

clinical isolates. Interstrain differences were observed as sixteen, (80 %) were

able to grow while the remainder were inhibited.

Isolates grew more slowly on the ASPA. However the number of visible colonies

on ASPA was comparable to those observed on the CBA media after an

extended period of incubation in the MAC at 37°C.

Growth of isolates on DPA was comparatively poor. Not only were bacteria

counts lower but prolonged incubation in the MAC of up to one week was

necessary for colonies to become sufficiently large for accurate counting. Some

strains of H. pylori may be sensitive to the polymyxin B component of the DPA

medium. Nevertheless, the DPA medium was evaluated further in subsequent

experiments to determine if it might be of value in decreasing the levels and

range of contaminating bacteria present in domestic water samples.

The lab-adapted strains (H509 and H899) grew consistently well on all media,

whereas the clinical isolates varied considerably in their ability to grow on the

different media under investigation. One of the clinical isolates (H5029)

appeared from these experiments to be a robust strain while the other three grew

poorly and/or slowly. This finding highlighted the importance of using a number

of clinical isolates, as examination of laboratory adapted strains alone does not

provide an accurate representation of the utility of the media.

The colony counts of the six strains decreased after prolonged incubation (24h)

in air at room temperature, regardless of the isolation medium used. Additionally,

for all culture media, H. pylori required a longer incubation to form visible colonies

for counting. This demonstrated that survival of H. pylori was poor under such

conditions, and that none of the test media were able to revive the strains.

25


4.1.3 Conclusion - evaluation of solid culture media for recovery from MRD

Overall, CBA medium allowed the best recovery of H. pylori after suspension in

MRD for up to 24h, both in terms of speed of growth and colony counts. The

DPA medium gave the least satisfactory results. Strain variations in growth

characteristics highlight the need to include a representative number of strains in

evaluation experiments.

4.2 Evaluation of solid culture media for recovery from spiked sterilised tap water

4.2.1 Experimental design As the overall aim of this part of the study was to develop methods to recover H.

pylori from domestic water systems, the following experiments aimed to evaluate

the ability of the four media (examined in section 4.1) to recover H. pylori that had

undergone nutrient and temperature shock by suspension in sterile tap water.

Six test strains of H. pylori were each suspended in tap water that had been filter-

sterilised using a 0.2µm pore size filter and serial decimal dilutions prepared (10-

1–10-5). Following incubation in air at 4°C, three 20µl aliquots were plated onto

the four test media at time intervals of 0 hours, 4h, and 24h. Colony counts were

recorded after microaerophilic incubation at 48, 72, 96, 120, 144 and 168h and

up to 264h for the 24h sample.

4.2.2 Results – evaluation of solid media for recovery from spiked sterilised tap water

For all strains, the highest colony counts from filter-sterilised tap water were

observed on CBA (Figure 2), although prolonged incubation (72h) at 37°C was

required to visualise colonies.

For each strain the number of visible colonies on each media decreased

significantly after 24h incubation in tap water (Figure 2). Bacterial growth on CBA

decreased by approximately 36%, while recovery on ASPA and CBA2D was

26


even lower (50%). Only the two reference strains could be recovered on the

DPA, where colony counts were reduced by 60%.

As in section 4.1, strains H4780 and H5425 failed to grow on CBA2D, although

scant growth of a small, possible subpopulation of H4780 was observed on the

4h sample plate. CBA with antibiotic supplement (Dent’s plates) were also

included in these experiments for analysis of the 0h and 24h time-points. The

same two strains also took longer to grow on Dent’s plates (data not shown).

Growth of contaminating bacteria occurred on some of the ASPA plates.

4.2.3 Conclusion - evaluation of solid media for recovery from spiked

sterilised tap water Overall, strains of H. pylori showed poor recovery on all culture media after

suspension in filtered tap water. After 4h incubation under stress conditions,

recovery was reduced on all media. Again, two strains (H4780 and H5425) failed

to grow on CBA2D.

0

2000

4000

6000

8000

10000

12000

14000

16000

0h 4h 24h

Incuba tion tim e in fi lte r-ste ril ise d ta p w a te r

Mea

n (x

104 ) c

fu/m

l

0

1

2

3

4

5

6

Mea

n re

cove

ry ti

me

(d)

CB A A S P A CB A 2D DP ACB A A S P A CB A 2D DP A

Hp recovery Recovery time

Figure 2: A comparison of the ability of four solid media to recover H. pylori that have

undergone nutrient shock in water after 0, 0.5 and 24h incubation in air at room temperature. *Mean values for CBA2D were calculated from four strains, as two failed to grow.

27


4.3 Evaluation of the Legionella antibiotic supplement 4.3.1 Experimental design. Contamination is a potential problem for the isolation of H. pylori directly from

drinking water samples.

As an alternative to the antibiotic combinations examined in sections 4.1 and 4.2,

the Legionella (GVPN) Selective Supplement (1.5g of glycine (ammonia free),

0.5mg of vancomycin hydrochloride, 40,000IU of polymyxin B sulphate and 20mg

natamycin per vial) (Oxoid) was examined. Culture plates were prepared by the

addition of one vial of supplement to 500 ml CBA. Plates were inoculated directly

with cultures of H. pylori and with cultures that had been pre-enriched for 48 h in

Brucella broth with yeast extract and horse blood (BBYEB, see section 5.1 for

details)

4.3.2 Results - Evaluation of the Legionella antibiotic supplement Only clinical isolate H5027 was able to grow on this medium, but with very scant

growth. The growth of all other strains of H. pylori was inhibited by the antibiotic

supplement.

4.3.3 Conclusion - Evaluation of the Legionella antibiotic supplement The Legionella selective supplement was not a suitable alternative to Dent’s

selective supplement as the antibiotics present were too inhibitory for H. pylori.

28


Evaluation of broth media for enrichment of H. pylori 5

5.1 Evaluation of blood-free broth media The numbers of H. pylori present in water systems may be extremely low. A

pre-enrichment step would increase H. pylori numbers and so increase the

sensitivity of the culture method.

5.1.1 Experimental design Four blood-free broth media ;(Maximum recovery diluent (MRD), Full Strength

special peptone (FSPB), Half strength special peptone broth (ASPB), Brucella

broth with 0.1% β cyclodextrin (BBβC), (Appendix 6) were tested to determine

their potential to enrich growth of the six H. pylori strains. All broths were pre-

incubated microaerophilically for 24h with lids loose to allow gases to reach

equilibrium.

For the first experiment, Dent’s selective supplement was added to each broth

medium and those were inoculated with 20µl of serially diluted (neat to 10-5)

suspensions of H. pylori in water. Broths were incubated microaerophilically with

lids loose, and three 20µl aliquots plated on CBA after 24 and 48h incubation.

For each strain, colony counts were recorded daily from 2d to 18d and the first

day of visible colony was recorded.

The second experiment was a repeat of experiment 1. The exception was the

omission of antibiotic supplement from the enrichment broth to assess the impact

of the antibiotic supplement. Additionally, the volume of media in the bijou was

increased to 8ml to reduce the headspace and therefore improve the

microaerophilic atmosphere.

29


5.1.2 Results – evaluation of blood free broth media Results obtained are summarised in Figure 3.

1

100

10000

1000000

24h 48h

Incubation time

cfu

log1

0BB BB + D MRD MRD + DASPB ASPB + D FSPB FSPB + D

Figure 3 Evaluation of four blood-free liquid media (with and without supplementary Dent’s antibiotics (D), experiments 1 and 2, respectively) for potential enrichment of H. pylori growth

Prolonged incubation in BBβC, MRD, ASPB and FSPB, both with and without

Dent’s antibiotic supplement, decreased bacterial counts. In all cases colony

counts were higher after 24h than 48h incubation in each respective broth.

Overall, the highest bacterial counts at both time intervals were observed for the

BB supplemented with 0.1% β cyclodextrin, although one clinical strain (H5425)

consistently failed to grow after 48h incubation.

Increased colony counts were observed for most strains after 48h incubation in

FSPB, but this was only observed in a single experiment, in the absence of

supplementary antibiotics.

Growth of strain H509 (NCTC 11637) from the MRD suspension sampled at 48 h

was recorded after the CBA had been incubated for 10 days, at the most

concentrated inoculum (data not shown). This highlights the need for prolonged

incubation periods and daily checking for growth.

30


5.1.3 Conclusions - evaluation of blood-free broth media While none of the broths enriched the growth of H. pylori, isolates generally

remained culturable for longer in the Brucella broth with 0.1% β cyclodextrin

(BBβC). The reduction in colony counts for all enrichment broth cultures between

24 and 48h suggests cell death or conversion to a viable but non-culturable

(VNC) state. Exclusion of antibiotics from the broth did not significantly improve

enrichment.

5.2 Evaluation of blood-containing liquid media 5.2.1 Experimental design Four broths originally described by Walsh and Moran (1997) were evaluated for

ability to enrich H. pylori growth:

• Brucella broth with 1% yeast extract (BBYE)

• Brucella broth with 1% yeast extract and 7% defibrinated horse blood

(BBYEB)

• Mueller-Hinton broth with 1% yeast extract (MHYE)

• Mueller-Hinton broth with 1% yeast extract and 7% defibrinated horse

blood (MHYEB)

Serial dilutions (neat to 10-5) of three strains - laboratory adapted strain H899

(=NCTC 12445) and two clinical isolates (H5027 and H5425) - were prepared in

autoclaved tap water. Enrichment broths were aliquoted in 8-ml volumes into

sterile bijoux bottles. An inoculum, which consists of 80µl of each spiked water

sample, was introduced into the enrichment broth and incubated in the MAC.

Subcultures for colony counts were undertaken after 24, 48 and 120 hour

incubation. The plates were read from day 2 to day 7 on CBA.

31


5.2.2 Results – evaluation of blood containing liquid media

Results for this experiment are summarised in Figure 4

0.011

10010000

1000000100000000

10000000000

24h 48h 120h

Incubation time

BBYE BBYEB MHYE MHYEB

Figure 4: Evaluation of the potential of blood free broths (BBYE and MHYE) and

blood-containing broths (BHYEB and MHYEB) to enrich growth of three strains of H. pylori.

While all strains survived in the blood-free broths BBYE and MHYE, even after

48h, there was no obvious increase in the number of visible colonies recovered

(Figure 4).

In contrast, the two clinical isolates in particular, incubated in broths containing

blood (BBYEB and MHYEB), displayed as much as 1000-fold enrichment

between 24 h and 48 h incubation (Figure 4). All BBYEB and MHYEB therefore

were sampled further after a total of 120 h incubation.

Enrichment of the laboratory adapted strain NCTC 12445 was not observed until

the broths had been incubated for a prolonged period (120h).

This experiment was repeated for BBYE and BBYEB to include two additional

strains: lab adapted strain H509 (NCTC 11637), and clinical isolate H4780. The

32


broth cultures were plated onto CBA at intervals of 0, 24, 48 and 120 hours and

plates read from day 2 to day 7.

Results for this second experiment are summarised in Figure 5.

As observed in other experiments, the number of visible colonies declined with

time when incubated in BBYE.

Growth of clinical isolates H5027 and H5425 was enriched 1x104 and 1x105 fold

respectively after prolonged incubation (120h) in BBYEB. Less enrichment

occurred for strain H899 however, with only a tenfold increase in colony count.

Growth characteristics for strains NCTC 11637 and H4780 could not be

determined in this experiment due to overgrowth of bacterial contaminants after

24h incubation.

1

1000

1000000

1000000000

0h 24h 48h 120h

Incubation time

BBYE BBYEB

Figure 5 Evaluation of the potential of blood free broths (BBYE and MHYE) and

blood-containing broths (BHYEB and MHYEB) to enrich growth of five strains of H. pylori.

5.2.3 Conclusions – evaluation of blood containing liquid media It can be concluded that H. pylori growth can be enriched by incubation in blood-

containing broths. The period of incubation required and the level of enrichment

would appear to be strain-dependent. Additionally prolonged incubation of broths

could lead to overgrowth of bacterial contaminants.

33


5.3 Evaluation of charcoal as an alternative to blood as a broth supplement.

5.3.1 Experimental design Broth containing charcoal instead of blood was evaluated as suggested by

Taneera et al. 2002. The advantages of charcoal are that it is less expensive

and has a longer shelf life than blood.

All six test strains were suspended in autoclaved tap water and serial decimal

dilutions were inoculated into BBYEB and Brucella broth + 1% yeast extract +

0.1% activated charcoal (BBYEC) (0.1% - Taneera et al. 2002).

Results for this experiment are presented in Figure 6.

1

100

10000

1000000

100000000

0h 24h 48h 120h

Time

BBYEBBBYEC

Figure 6 Comparison of the enrichment effects of BBYE supplements blood (B) and

charcoal (C) 5.3.2 Results - Evaluation of charcoal as an alternative to blood as a broth

supplement Greater enrichment of both clinical and laboratory isolates was observed for

BBYEB after 48h, compared with the BBYEC (Figure 6).

Higher levels of enrichment (1x104) were observed in BBYEB for the clinical

strains than for the laboratory adapted strains. Some enrichment was evident for

H899 (1x102) but not for H509.

34


As the charcoal precipitated from suspension in the first experiment, the BBYEC

broth was amended to contain 0.2 % agar (BBCA) and the experiment above

repeated.

Broths were inoculated with spiked water samples and left in the MAC for 4 hours

at 37°C. An aliquot of 50µl Dent’s supplement was added to each tube and the

broths were reincubated.

The results for this second experiment are summarised in Figure 7.

1

100

10000

1000000

0h 24h 48h 120h

Time

BBYEB BBCA

Figure 7 Comparison of the enrichment effects of BBYE supplemented with blood or

BB supplemented with charcoal in semi-solid agar (BBCA)

The charcoal remained in suspension in the BBCA broth, but as this was semi-

solid, it was more difficult to visualise bacterial growth on plates.

The number of colonies recovered from all strains incubated in BBCA, except for

NCTC 11637, decreased after 24h, with no colonies visible after 48h.

The addition of the Dent’s supplement to BBYEB and BBCA seemed to inhibit the

growth of the bacteria at first, in particular H4780 and H5425 (these were

inhibited by the CBA 2xA plates). H4780 had too numerous to count colonies at

10-4 after 120 hours in BBYEB while H5425 had too numerous to count colonies

35


10-2 dilution in BBYEB. The laboratory adapted strains NCTC 11637 and H899

did not survive in the broth. 5.3.3 Conclusion - Evaluation of charcoal as an alternative to blood as a

broth supplement. It was concluded that the best broths for enrichment were BBYEB and MHYEB.

Substitution of blood with charcoal appeared to decrease the enrichment effect.

5.4 Effect of solid medium on growth after enrichment 5.4.1 Experimental design All of the previous enrichment experiments used CBA to enumerate H. pylori.

This experiment evaluated growth of the H. pylori strains on CBA, ASPA, DPA

and CBA2D following enrichment.

The enrichment broth used was 28mls of BBYEB, which was inoculated with

280µl of spiked water (diluted neat to 10-5) and after 4 hours incubation 110µl of

Dent’s was added. A broth with greater depth was used here as it was suggested

that the microaerophilic atmosphere would increase with depth and so

enrichment of the H. pylori strains would also be expected to improve. The broth

was plated onto four solid media (listed above) at time intervals of 0, 24, 48, 72

and 144 hours and these were incubated in the MAC at 37°C. All plates were

read from day 2 to day 13.

Results are summarised in Figures 8a and 8b.

36


1

100

10000

1000000

0h 24h 48h 72h 144h

Incubation time in BBYEB

CBA ASPA CBA2D DPA

a)

0

50000

100000

150000

0h 24h 48h 72h 144h

Incubation time in BBYEB

CBA ASPA CBA2D DPA

b) Figure 8 Evaluation of recovery of H. pylori after a pre-enrichment step in broth

(BBYEB) a) Logarithmic scale b) Linear scale

5.4.2 Results - Effect of solid medium on growth after enrichment As has been demonstrated in previous experiments, growth of all strains (with the

exception of H509) was enriched by incubation in BBYEB (Figure 8a). However

37


the recovery of these enriched levels varied according to the media used (Figure

8b).

Highest bacterial counts were observed on CBA. The ASPA gave the next best

results (Figure 8b).

Growth of the strains H4780 and H5425 was inhibited on CBA2D. However with

the increased incubation in the broth, colonies were seen for the most

concentrated inocula, suggesting that a resistant sub population is present.

As was demonstrated previously, bacterial growth was poor on DPA, even after a

pre-enrichment step.

The increased depth of the broth did not show any advantage for enrichment of

the H. pylori colonies.

5.4.3 Conclusion - Effect of solid medium on growth after enrichment

It can be concluded that while enrichment increases H. pylori levels, relative

recovery of each strain on each different medium was similar to that observed in

section 4.1.3 (see above).

38


6 Evaluation of methods for sample processing

Experimental data in sections 4 and 5 indicate that prolonged incubation in

enrichment broths and on solid media would be required to optimise recovery of

H. pylori. Under such conditions, overgrowth of bacterial contaminants present in

water samples could be a potential problem.

A series of experiments were conducted to see if levels of contaminants could be

reduced in samples whilst also concentrating H. pylori at the initial sample

processing stage.

Both physical and chemical methods have been employed successfully to reduce

levels of contamination. For this study, the use of dual filtration system,

differential centrifugation, immunomagnetic separation, acid treatment and the

use of antibiotics were investigated.

6.1 Evaluation of acid treatment of samples

To survive the acid environment of the gastric niche, H. pylori has adapted

mechanisms to tolerate low pH, such as production of a potent urease to

hydrolyse urea to ammonia. Few other organisms are known to survive

exposure to low pH and this differential acid tolerance can be exploited, to reduce

levels of contaminating bacteria while maintaining H. pylori levels in a sample.

This approach has been applied in the culture of H. pylori from gastric juice

aspirates, and similarly has been exploited in the isolation of Legionella

pneumophila directly from water samples.

6.1.1 Experimental design The acid tolerance of all six strains was examined in the first experiment to

assess the impact of acid treatment on H. pylori survival and the success of

reducing background microflora.

39


Bacterial suspensions of each strain were prepared in 5ml distilled water at initial

concentrations that would facilitate accurate enumeration of colonies (OD600 =

0.1 for H509, H4780 and H5027 and OD600 = 0.07 for H899, H5029 and H5425).

Aliquots (200µl) of each suspension were acid treated as described in section

3.5. 100µl of each acid treated suspension was spread-plated onto CBA in

duplicate after acid contact intervals of 0, 1, 2.5, 5 and 10 min.

The results are presented in Figure 9.

02468

10

Untreated 0 min 1 min 2.5 min 5 min 10 min

Duration of acid treatment

Gro

wth

sco

re*

H509 H899 H4780 H5029 H5027 H5425

Figure 9 Evaluation of the effect of acid contact time on six different strains of H.

pylori *Growth score determined as accurate enumeration was not possible on plates

with semi-confluent and confluent growth. Each plate was assigned an arbitrary

number based on colony counts as follows: 0 = no growth, 1 = <10 cfu; 2 =

>10<50 cfu; 3 = >50<150 cfu; 4 = >150<300 cfu; 5 = >300<500cfu; 6 =

>500<1000cfu; 7 = >1000 cfu (semi-confluent growth); 8 = >>1000 (confluent

growth).

6.1.2 Results - Evaluation of acid treatment of samples Two strains of H. pylori (H509, H4780) were tolerant to acid treatment, regardless

of the duration of the contact time (Figure 9). One strain (H5425) was acid

40


tolerant, but the number of colonies recovered was lower when the duration of

the acid treatment was prolonged (10min) (Figure 9).

6.1.3 Conclusion - Evaluation of acid treatment of samples The recovery of three strains of H. pylori (H899, H5027, H5029) was markedly

reduced, even if the acid treatment was of minimal duration (Figure 9). However,

different strains of H. pylori show different levels of acid tolerance.

6.2 Evaluation of effect of the acid treatment duration on the levels of

bacterial contamination Water (1l) from a domestic water supply was filtered through a 0.2µm pore sized

membrane (Pall Corporation). The cells on the filter membrane was eluted into

10ml MRD held in a sterile stomacher bag by rubbing for 30s. The cell

suspension was centrifuged (5000g for 30min) and a volume of supernatant was

removed so as to leave 1.5ml as the sample concentrate.

6.2.1 Experimental design A 1-ml aliquot of the sample concentrate was acid treated as described in section

3.5. Following acid contact times of 0, 1, 2.5, 5 and 10 min. 100µl was spread-

plated onto CBA, Dent’s, CBA2D and DPA. Plates were incubated at 37oC for 48

hours under microaerophilic conditions.

6.2.2 Results - Evaluation of effect of the acid treatment duration on the levels of bacterial contamination

The results of this experiment are presented in Figure 10. Semi-confluent and

confluent growth precluded accurate enumeration of H. pylori colonies.

Therefore, a growth score was used as a measure to access the level of

contamination. Each plate was scored an arbitrary number based on colony

counts as follows: 0 = no growth, 1 = <10 cfu; 2 = >10<50 cfu; 3 = >50<150 cfu; 4

= >150<200 cfu; 5 = >200.

41


0123456

Untreated 0 min 1 min 2.5 min 5 min 10 min

Duration of acid treatment

CBA Dents CBA2D DPA

Figure 10 Evaluation of the effect of acid treatment duration on growth of background

flora on four different solid media.

Untreated sample concentrate showed bacterial contamination on all four types

of solid media. However acid treatment was effective in reducing contamination

on all media. This was apparent after minimal acid contact time of (0 min)

(Figure 10).

Contamination levels remained highest on CBA and Dent’s media, but these

were markedly reduced after 5 min compared with 2.5 min acid treatment (Figure

10)

6.2.3 Conclusion - Evaluation of effect of the acid treatment duration on the levels of bacterial contamination

Acid treatment decreases the level of bacterial contamination. If samples are to

be plated on CBA or Dent’s, the optimal acid treatment would be of 5 min

duration.

6.3 Evaluation of the effect of different filtration procedures We hypothesised that pre-filtration of water samples with a larger pore size

membrane, followed by filtration using a 0.2 µm pore size membrane would

reduce contaminating bacteria; larger cells would be trapped on the prefilter and

H. pylori cells would be retained on the smaller pore size filter.

42


6.3.1 Experimental design The first experiment examined different filter pore sizes (0.65µm, 0.45µm and

0.2µm) for their capacity to trap H. pylori cells.

Suspensions (OD600 = 0.6) of all the six test strains were prepared in distilled

water. Each filter was placed on CBA and 500µl of the suspension was added to

the filter. After 40 min, the filter was aseptically removed.

6.3.2 Results - Evaluation of the effect of different filtration procedures Following 5 days microaerophilic incubation, growth of all H. pylori strains (except

H899) was observed on the plates where the 0.65µm filter had been applied.

No growth was visible on the plates where either the 0.45µm or the 0.2µm filters

had been applied.

6.3.3 Conclusion - Evaluation of the effect of different filtration procedures It can be concluded that with the exception of H899, all H. pylori cells penetrated

the 0.65 µm membrane. Meanwhile, the 0.45 µm and 0.2 µm filter membranes

retained H. pylori cells examined in this study so the 0.65 µm membrane would

be necessary for dual filtration experiments.

6.4 Evaluation of the effect of dual filtration and acid treatment 6.4.1 Experimental design In the second experiment, suspensions (OD600 = 0.5) of H899 (NCTC 12445) and

H4780 (clinical isolate) were prepared in sterile distilled water and diluted by a

ratio of 1:10. Aliquots of each dilution were added to 500ml of filter-sterilized

water.

Inoculated water samples were serially filtered through a 0.65µm pore size filter

and then a 0.2µm filter. The filter papers were placed in 10ml of MRD and

rubbed vigorously for 30s. The MRD was then centrifuged (3500rpm for 30min)

and the supernatant removed to leave a 1ml concentrate. Aliquots (100µl) were

43


spread-plated onto CBA (in duplicate). Of the remaining concentrate, 200µl was

acid-treated, as described in section 3.5. After 5 min acid treatment 100µl was

spread plated (in duplicate) onto CBA.

After 48h microaerophilic incubation, plates were examined for growth of H. pylori

and of bacterial contaminants.

6.4.2 Results - Evaluation of the effect of dual filtration and acid treatment Growth of bacterial contaminants was observed on all 32 plates. As growth was

confluent on some plates, and a wide range of different colony types were

observed, it was not possible to enumerate these accurately. For this reason, an

arbitrary qualitative growth score was devised as follows:

0 = no growth

1 = scant growth

2 = poor growth

3 = moderate growth

4 = semi-confluent growth

5 = confluent growth

Photos of examples of each of these is provided in Figures 11-14

Figure 11 Examples of poor growth of microbial contaminants

44


Figure 12 Example of moderate growth of microbial contaminants

Figure 13 Examples of semi-confluent growth of microbial contaminants

Figure 14 Examples of confluent growth of microbial contaminants

A summary of the contamination levels for each sample treatment is

provided in Figure 15.

45


02468

0.65um 0.65um 0.2um 0.2um

UT AT UT ATSample processing (filter pore size, + or - acid treatment)

H899 N H899 -1 H4780 N H4780 -1

Figure 15 Evaluation of the efficacy of dual filtration combined with acid treatment in

reducing levels of bacterial contaminants in water samples *Growth score was defined, as accurate enumeration of plates was not possible.

Plates were assigned an arbitrary number as follows: 0 = no growth, 1 = scant

growth; 2 = poor growth; 3 = moderate growth; 4 = semi-confluent growth; 5 =

confluent growth, AT; acid treated, UT; untreated.

Acid treatment reduced the level of bacterial contamination for both sample sets

(0.65µm and 0.2µm filters) (Figure 11).

Contamination levels were only marginally lower for the dual filtered (0.2µm)

samples compared with the single filtration (0.65µm) (Figure 15).

Colonies from the sample inoculated with the 10-1 dilution of H4780 that had

been filtered by the 0.65µm pore size filter alone and acid treated were confirmed

to be H. pylori (positive by the urease, catalase, oxidase, and gram stain tests).

Both duplicate plates for this sample were positive but H. pylori were not

recovered from any of the other plates.

6.4.3 Conclusion - Evaluation of the effect of dual filtration and acid treatment

It was concluded that the acid treatment step reduced the amount and mixture of

contamination present on the plates. Lowering the levels of bacterial

contaminants increases the chance of recovering H. pylori.

46


6.5 Effect of dual filtration and differential centrifugation Centrifugation of samples at low speed to pellet larger contaminating bacteria

followed by high speed centrifugation of the supernatant has been proposed as a

potential method for decreasing contaminants, while concentrating H. pylori

(Andersen, et al. 1997).

Differential centrifugation was evaluated along with dual filtration to see if a

combination of these methods would be optimal for recovery of H. pylori from

water.

6.5.1 Experimental design Suspensions of H899 and H5425 (OD600 = 0.69) were prepared in 10ml of

distilled water. This total volume was added to 240ml of filter-sterilised water

(1/25 dilution) and 25ml was then added to a second 225ml sample of filter-

sterilised water (1/10 dilution).

Each sample was filtered through the 0.65µm pore size filter and then through

the 0.2µm pore size filter. Filter papers were then removed and rubbed

vigorously in 10ml of MRD for 30s.

Each sample of MRD was centrifuged at 1500g for 15min. The supernatant was

transferred to a sterile universal and the pellet was plated onto CBA plates. The

remaining supernatant was centrifuged at 5000g for 15mins (Andersen, et al.

1997). The supernatant was removed to concentrate the sample to 1ml and

100µl of this was spread plated onto CBA (in duplicate). After 48h

microaerophilic incubation, all plates were examined for growth of H. pylori and of

bacterial contaminants.

6.5.2 Results - Effect of dual filtration and differential centrifugation The level of contamination in this experiment was low overall (Figure 16),

although there appeared to be marginally more contamination in the samples that

47


were centrifuged at lower speed, suggesting that the larger contaminating

bacteria are being concentrated by this preliminary step.

H. pylori was recovered both in samples that had undergone low speed

centrifugation and in those that had undergone an additional higher speed

centrifugation (Figure 17). This suggests that the differential centrifugation is not

concentrating the H. pylori cells and that this approach would lead to loss of H.

pylori from samples (as only the sample that had been centrifuged at two speeds

would be processed).

Filter pore size did not appear to affect contamination levels (Figure 16). While

the clinical isolate (H5425) was recovered from all filters, the laboratory adapted

strain (H899) was recovered from the larger pore size filter only, suggesting that

either the bacterial cell is larger, or that this strain may form aggregates.

6.5.3 Conclusion - Effect of dual filtration and differential centrifugation It can be concluded that differential centrifugation reduced some bacterial

contamination but H. pylori cells may be lost during the first lower speed

centrifugation. Therefore, differential centrifugation is not a suitable approach for

processing water samples,

48


01234

0.65um 0.65um 0.2um 0.2um

1500 g +5000 g 1500 g +5000 g

Sample processing - filter size and centrifugation speed

H899 N H899 1/10 H5425 N H5425 1/10

Figure 16 Evaluation of the efficacy of dual filtration combined with differential

centrifugation in reducing levels of bacterial contaminants in H. pylori-spiked water samples *Growth score was defined, as accurate enumeration of plates was not possible.



confluent growth

01234

0.65um1500 g

0.65um +5000 g

0.2um1500 g

0.2um +5000 g

Sample processing - filter size and centrifugation speed

H899 N H899 1/10 H5425 N H5425 1/10

Figure 17 Evaluation of the efficacy of dual filtration combined with differential centrifugation in recovering H. pylori from spiked water samples *Growth score was defined, as accurate enumeration of plates was not possible.



confluent growth

49


6.6 Examination of Immunomagnetic Separation (IMS) IMS is used widely in food and water microbiology as a means of concentrating

specific organisms while decreasing levels of contaminating microbes. Magnetic

beads are coated with specific (e.g. anti-H. pylori) antibody that captures

bacterial cells (e.g. H. pylori) in a sample. Application of a magnet allows the

bead-organism complex to be retained in the reaction vessel while contaminating

(unbound) organisms are removed by a series of washing steps.

6.6.1 Experimental design Three strains of H. pylori (H899, H5027, H4780) were suspended in 5ml sterile

water (OD600 = 0.2) and decimal dilutions to 10-5 prepared. This was repeated for

strains H5027 and H4780. Colony counts of all dilutions were undertaken on

CBA as described in section 3.4.

All dilutions were processed by IMS as described in section 3.6 and colony

counts from all bead suspensions were determined. Additionally, bead

suspensions were boiled for 10min and H. pylori-specific PCR (Ho et al. 1991

assay) was performed on all resultant DNA preparations.

6.6.2 Results - Examination of Immunomagnetic Separation (IMS) After IMS, only 5 colonies were visible at the most concentrated (neat)

suspension for strain H899. No growth was observed for the other two strains. A

majority of the preparations (neat and diluted samples) could not be cultured after

IMS. However, H. pylori DNA was detected by the Ho et al. 1991 PCR assay

(Figure 18). H899 and H4780 could be detected down to the 10-4 dilution while a

faint band was visible for H5027 at the lowest dilution (10-5) (Figure 18).

50


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Figure 18 PCR based detection of H. pylori serially diluted and bound to IMS beads. Lanes 1-6 H899; Lanes 7-12; H5027, Lanes 13-18, H4780; Lane 19 positive

control (H. pylori DNA); Lane 20 negative control (distilled water); Lane 21 123bp

MW marker. All strains are in the dilution order neat, 10-1, 10-2, 10-3, 10-4, and 10-

5.

PCR demonstrated that the beads are binding some H. pylori cells although the

lower detection limit compared with primary culture suggests that cells are being

lost during the IMS procedure. The failure to obtain cultures for these cells

suggests that H. pylori may enter a non-cultivable or inviable state as a result of

the IMS procedure. Alternatively the levels of H. pylori that complex with the

beads are so low that they are detectable by PCR, but may be below the

sensitivity threshold of culture.

6.6.3 Conclusion - Examination of Immunomagnetic Separation (IMS) It can be concluded that the IMS protocol led to a decrease in H. pylori recovery

by culture. As H. pylori would possibly be present in water at very low levels, this

method will not be used for processing samples.

51


7 Evaluation of combined sample processing, pre-enrichment

and culture methods.

7.1 Evaluation of acid treatment of samples with different culture media

7.1.1 Experimental design Suspensions of all six test strains were prepared in 10 ml distilled water (OD600 =

0.5). Each bacterial suspension was added to 240ml of filter sterilised domestic

property water (1/25 dilution). The spiked water sample was processed through

a 0.45 µm filter. All filters were rubbed in MRD for 30s, and the supernatant

centrifuged (5000g for 30min) to concentrate the sample to a final volume of 1ml.

Aliquots (100µl) of the sample concentrate were immediately spread-plated onto

CBA, CBA2D, DPA and Dent’s media. Aliquots (200µl) of the remaining

concentrates were acid treated as described in section 3.5. After 5min, 100µl

aliquots were spread-plated onto CBA, CBA2D, DPA and Dent’s media.

All plates were examined for bacterial contamination and for H. pylori colonies

after 48h microaerophilic incubation at 37°C and then daily up to 7 days.

7.1.2 Results - Evaluation of acid treatment of samples with different

culture media Accurate enumeration of contaminant and H. pylori colonies was not possible in

many cases as growth was semi-confluent and confluent. As data recorded was

therefore not always quantitative, all plates were scored according to a growth

category as follows:

52


• No growth (0 cfu) = 0 points

• Scant growth (<10 cfu) = 1 point

• Poor growth (>10 cfu <50) = 2 points

• Moderate growth (>50 cfu <200) = 3 points

• Semi-confluent growth (>200 cfu) = 4 points

• Confluent growth (>>>200 cfu) = 5 points

In earlier experiments, semi-and fully confluent growth plates were not

distinguished and so all plates were assigned to the >200 cfu category (4 points).

To summarise data for the different strains for each experimental condition, a

growth index was calculated.

Growth Index = ∑ (Number of strains in each category x category points) The example below illustrates the conversion of raw data (Table 2) into a growth

index (Table 3)

While this is a qualitative system of analysis that we have developed, based on

the arbitrary categories defined, it nevertheless summarises the findings of each

experiment accurately. In the example provided above, it is evident from both the

raw data and the growth index that the highest rate of H. pylori recovery was from

the acid treated samples plated on Dent’s medium, while the lowest rates were

on DPA.

53


Table 2 : Raw data of H. pylori colony counts from the experiment described in section 7.1

Strain number

DENT’S DPA

Untreated Acid Treated Untreated Acid Treated

H899 TNC* TNC 0 0

H509 TNC TNC TNC TNC

H5027 0 78 0 0

H4780 0 TNC 0 0

H5029 38 TNC 0 0

Strain number

CBA CBA2D

Untreated Acid Treated Untreated Acid Treated



H5027 0 143 0 46

H4780 0 0 0 0


*TNC = “too numerous to count” (= semi-confluent/confluent growth)

54


Table 3: Conversion of raw data in Table 2, to determine the Growth Index

Number of strains in each category Solid medium

(+ or – acid treatment)

No growth

(0 points)

Scant Growth (1 point)

Poor growth

(2 points)

Moderate growth

(3 points)

S-Confluent growth

(4 points)

Confluent growth

(5 points)

Growth Index*

Dent’s UT 2

(0 points)

1

(2 points)

2

(8 points)

NA* 10

(0+2+8)

Dent’s AT 1

(3 points)

4

(16 points)

NA 19

(0+3+16)

DPA UT 4

(0 points)

1

(4 points)

NA 4

(0+4)

DPA AT 4

(0 points)

1

(4 points)

NA 4

(0+4)

CBA UT 2

(0 points)

3

(12 points)

NA 12

(0+12)

CBA AT 2

(2 points)

1

(2 points)

2

(8 points)

NA 12

(0+2+2+8)

CBA2D UT 2

(0 points)

3

(12 points)

NA 12

(0+2+8)

CBA2D AT 1

(0 points)

1

(2 points)

3

(12 points)

NA 14

(0+2+12)

*Not applicable as semi-confluent and confluent growth were not distinguished in this experiment

The results of this experiment are summarised in figures of the growth indices of

bacterial contamination (Figure 19) and of H. pylori (Figure 20).

55


05

101520

Dents DPA CBA CBA2D

Medium

Untreated Acid treated

Figure 19: Comparison of growth of bacterial contaminants before and after acid

treatment on different solid media. *Calculation of the growth index is explained in 7.1.6

Growth of contaminating bacteria was observed on 37/40 (93%) plates in this

experiment, with highest growth indices observed on CBA and CBA2D (Figure

19). With the exception of DPA, acid treatment lowered the level of

contamination on each medium (Figure 19). The lowest levels of bacterial

contamination were observed for acid treated samples that were plated on Dent’s

medium (Figure 19). As has been shown in previous experiments, the recovery

of H. pylori was poorest on DPA medium (Figure 20).

7.1.3 Conclusion - Evaluation of acid treatment of samples with different culture media

The highest H. pylori growth index was for acid treated samples plated on Dent’s.

This is likely to relate to the fact that contamination was lowest on these plates,

enabling H. pylori to grow with less competition from other bacterial species.

Similarly the higher recovery rate of H. pylori for the acid treated samples on

CBA2D may relate to the lower contaminant levels (Figure 20).

56


05

101520

Dents DPA CBA CBA2D

Medium

Untreated Acid treated

Figure 20: Comparison of growth of H. pylori before and after acid treatment on

different solid media. *Calculation of the growth index is explained in section 7.1.6

7.2 Evaluation of enrichment of acid treated samples cultured on four different media

7.2.1 Experimental design For the initial experiment (experiment A), suspensions of all six test strains were

prepared (OD600 = 0.25) in 10ml and added to 240ml of filter sterilised domestic

property water (1/25 dilution). This was filtered through a 0.45µm pore-sized

filter that was agitated for 30s in 10ml MRD. The MRD was then centrifuged

(5000g for 30min) and the supernatant discarded to obtain a final 1ml

concentrate. The concentrate (100µl) was spread plated onto CBA, CBA2D,

DPA and Dent’s.

An aliquot (200µl) of concentrate was acid treated for 5min as described in

section 3.5 and 100µl spread plated onto CBA, CBA2D, DPA and Dent’s media.

The acid treatment was repeated on a second aliquot of concentrate and 100µl

was added to MHYEB, in duplicate for each strain. After microaerophilic

incubation (2h), Dent’s antibiotic supplement was added to one set of broths.

57


Aliquots (100µl) of all broths were spread-plated onto CBA, CBA2D, DPA and

Dent’s after microaerophilic incubation at 37°C at time intervals of 24h (Day 1),

48h (Day 2) and 120h (Day 5).

After 48h microaerophilic incubation, all plates were examined for H. pylori

growth and for growth of bacterial contaminants. The growth indices to

summarise results for each set of growth conditions were calculated as described

in section 7.1.6, for both bacterial contamination and for H. pylori, the results of

which are presented in figures 21 and 22, respectively.

05

1015202530

Dents

Dents+

ADPA

DPA+A CBA

CBA+A

CBA2D

CBA2D+A

MHYEB (+ or - antibiotics "A") on solid media

Bac

teria

l Con

tam

inat

ion

Gro

wth

Inde

x*

Day 1 Day 2 Day 5

Figure 21 The level of bacterial contamination from a domestic water sample

(experiment A – see text below) that was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media. *Growth indices were calculated as described in section 7.1.6

Note that in this experiment semi-confluent was distinguished from confluent

growth.

58


7.2.2 Results - Evaluation of enrichment of acid treated samples cultured on four different media

Bacterial contamination was observed on 106/164 plates (65 %) and overall for

all media, this increased with incubation time in enrichment broth (Figure 21).

The highest levels of contamination were found on CBA plates, although

inclusion of antibiotics in the enrichment broth lowered these (Figure 21).

Similarly, all other samples showed lower contamination levels in the broths that

contained antibiotics.

0

5

10

15


Day 1 Day 2 Day 5

Figure 22 The level of H. pylori recovery from a spiked domestic water sample

(experiment A) that was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media. *Growth indices were calculated as described in section 7.1.6

Note that in this experiment semi-confluent and confluent growth was

distinguished.

For three strains of H. pylori (H509, H5425 and H4780) growth was not observed

under any of the combined conditions. This may be due to the observed

overgrowth of high levels of contaminating bacteria. Growth of the remaining

three strains (H899, H5027 and H5029) was observed on all media (Figure 22).

However the highest sustained levels of growth were observed from samples

enriched in MHYEB containing antibiotics and plated on CBA.

59


As water samples from different domestic properties are likely to vary in terms of

the numbers and range of different contaminating organisms: the above

experiment (experiment A) was repeated twice (experiments B and C) using

shower water from two other properties. Experiment B varied slightly from A and

C in that the initial bacterial suspension was marginally lighter (OD600 = 0.2 vs

0.25). Experiment C differed from A and B in that antibiotics were added to the

enrichment broths 4h instead of 2h after inoculation.

As for experiment A, growth indices were calculated for bacterial contamination

and H. pylori growth as described in section 7.1.6. Results for bacterial

contamination are presented in Figures 23 and 24.

In both experiments, confluent growth of bacterial contaminants was observed on

all plates after two days incubation in enrichment broth.

05

101520253035

NHYEB (+ or - antibiotics "A") on solid media

DAY 1 DAY 2


(experiment B) that was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media.

*Growth indices were calculated as described in section 7.1.6


growth.

60


010203040


DAY 1 DAY 2


(experiment C) that was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media.



growth.

7.2.3 Conclusion - Evaluation of enrichment of acid treated samples

cultured on four different media It can be concluded that the level of bacterial contamination in shower water

varies between different domestic properties. The protocol that has been

developed will overcome problems of contamination for some, but not all, water

samples. H. pylori can only be recovered if the background contamination is

sufficiently low.

7.3 Evaluation of the improved protocol

7.3.1 Experimental design This experiment was repeated to examine water from other sites in a domestic

property.

The design of the first experiment, that examined water from a kitchen tap, was

similar to that described in section 7.3.1-7.3.6 except that the initial H. pylori

61


inoculum was higher (OD600 = 0.35) and antibiotics were added to the broths 5h

after initial inoculation.

The growth indices of bacterial contamination and H. pylori recovery are

presented in Figures 25 and 26, respectively.

05

10152025

Dents

Dents+

ADPA

DPA+A CBA

CBA+A

CBA2D

CBA2D+A


Bac

teria

l con

tam

inat

ion

Gro

wth

Inde

x*

DAY 1 DAY 2 DAY 5

Figure 25 The level of bacterial contamination from a spiked domestic kitchen tap

water sample that was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media.



growth.

As would be expected for potable water, contamination levels were lower (Figure

25) than had been observed for any of the three shower water samples tested

(Figures 21, 23 and 24). The highest level of contamination was observed when

broths were pre-enriched without antibiotics and plated on the CBA plates.

The addition of antibiotics to the pre-enrichment broths decreased contamination

on all plates. The lowest levels of contamination were observed on CBA2D

plates.

62


05

1015202530

Dents

Dents+

ADPA

DPA+A CBA

CBA+A

CBA2D

CBA2D+A


H. p

ylor

i Gro

wth

In

dex*

DAY 1 DAY 2 DAY 5

Figure 26 The recovery of H. pylori from a spiked domestic kitchen tap water sample

that was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media.



growth.

Recovery of H. pylori was overall higher from spiked kitchen water (Figure 26)

compared with shower water (Figure 22). This is likely to relate partly to the

slightly higher initial inoculum and also to the lower levels of bacterial

contamination enabling H. pylori to grow without competition for space and

nutrients.

All six strains of H. pylori were recovered on all plates, although bacterial counts

for H5425 and H4780 were considerably lower than for the other strains. The

recovery of H. pylori decreased as incubation time in the pre-enrichment broths

increased. This is likely to be due to increased levels of contaminating bacteria

and may also be due to depletion of nutrients in the broth and H. pylori cells

entering a stationary or even autolytic death phase.

The second experiment, that examined water from a bath tap, was similar in

design to that described in section 7.3.1-7.3.6 except that the initial H. pylori

63


inoculum was higher (OD600 = 0.5) and antibiotics were added to the broths 5h

after initial inoculation.

The growth indices of bacterial contamination and H. pylori recovery are

presented in Figures 27 and 28, respectively. Note that in this experiment semi-

confluent was distinguished from confluent growth.

05

1015202530


DAY 1 DAY2 DAY 6

Figure 27 The level of bacterial contamination from a domestic bath water sample

that was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media.


As would be expected for non-potable water, the level of bacterial contamination

was higher in this bath tap water sample (Figure 27) than that seen for the

kitchen tap sample (Figure 25). However, overall levels of contamination were

lower than those seen for the three shower water samples (Figures 21, 23 and

24).

Again the highest level of contamination was observed for samples that had been

pre-enriched without antibiotics and plated onto CBA.

64


For all plates, the level of contamination increased with time incubated in pre-

enrichment broths.

0102030


DAY 1 DAY 2 DAY 6

Figure 28 The recovery of H. pylori from a spiked domestic bath water sample that

was acid treated, enriched in MHYEB with and without antibiotics and plated onto four solid media.



growth.

Similarly to the results obtained for the kitchen tap water, the recovery of H. pylori

from the spiked bath water was higher (Figure 28) than was observed from the

spiked shower water (Figure 22). This may be partially due to the higher initial

inoculum in addition to the lower levels of background contaminating bacteria.

The highest recovery of H. pylori was observed in samples that had been

incubated in pre-enrichment broths without antibiotics for 0 – 24h and plated onto

CBA or Dent’s media.

As was observed for the spiked kitchen tap water sample, while all strains grew,

lower colony counts were recorded for H5425 and H4780. Thus, recovery of H.

pylori is to some extent strain dependent.

65


The recovery of H. pylori decreased as the incubation time in the pre-enrichment

broth increased (Figure 28). By day 6, only H899 and H5029 were observed on

all plates.

7.3.2 Conclusion - To assess the application of the improved protocol It can be concluded that the efficiency of pre-enrichment followed by antibiotic

treatment is dependent on the source of water sample being processed. For

non-potable waters where background is likely to be high, antibiotic treatment

followed by culture on CBA is effective in reducing the level of background.

Successful background reduction encourages H. pylori growth. For potable

waters where background microflora is low, prolonged pre-enrichment is

detrimental to the recovery of H. pylori as this facilitates the growth of

background microflora.

The acid treatment step showed promise for reducing the level of background

contamination, while still allowing the H. pylori to be isolated.

The enrichment step is beneficial for increasing the numbers of H. pylori but the

growth of background contamination is also increased. These experiments have

revealed that sampling the enrichment broth at 24h provided the best results.

The Dent’s media plates appeared to be optimal in terms of recovery of H. pylori

and in reducing the background contamination. The highest numbers of H. pylori

recovered were on CBA, but overgrowth of microbial contaminants was a

frequent problem.

66


8 Investigation of Domestic Systems

In this study water and biofilm samples for culture of H. pylori were taken from

domestic properties mostly from the Greater London area. The study was

conducted in three phases:

• Phase 1 02/02/04 – 09/02/04 Analysis by PCR

• Phase 2 03/08/04 – 17/08/04 Analysis by PCR and culture protocol A

• Phase 3 07/09/04 – 17/11/04 Analysis by PCR and culture protocol B 8.1 Phase one of survey of domestic properties The aim of this phase of sampling was to identify a PCR-positive source of H.

pylori from a water or biofilm sample that could be used for culture work in

subsequent phases of the survey.

Volunteers from the HPA Colindale site were recruited. They were supplied with

a protocol, a questionnaire and a pack (see Appendices 2 & 3) containing the

materials they would need for collecting water and biofilm samples. They

provided these samples from the 2nd to the 9th of February 2004.

8.1.1 Results - Phase one of survey of domestic properties Altogether samples were received from 62 domestic properties in phase one;

they consisted of 56 biofilm (WC cistern) and 62 x 2L water (bath cold tap and/or

shower) samples.

67


The following flowchart illustrates the processing method for biofilm and

shower water samples.

1ml to be used by CHRU for PCR 1ml to be kept by WEMRU for storageSo 1ml equivalent to 1L original

Remove supernatant to concentrate down to 2mls

Centrifuge the sample at 3500 rpm for 30 mins

Resuspend cells in 10mls PBS

Biofilm sample from WC

Remove supernatant to concentrate down to 2mls

Centrifuge the sample at3500 rpm for 30 mins

2L water sample

Elute cells into 5mls PBS

Filtration using Supor 200 47mm; pore size 0.2µm, hydrophilic polyethersulfone membrane

DNA was extracted from all 118 water and biofilm samples by the method of

Boom et al. 1990. All were PCR- tested by a Helicobacter specific assay

targeting 16S rRNA (Logan et al. 2000), and two H. pylori specific assays,

targeting 16S rRNA (Ho et al. 1991) and vacA (Chisholm et al. 2001) - as

described in the Report for the previous DWI study.

68


Fifteen of 118 samples (12.7%) were weakly positive by at least one PCR assay.

Two of the 15 were positive for all three assays (Logan et al. 2000, Ho et al. 1991

and Chisholm et al. 2001) although the Ho et al. 1991 and the Chisholm et al.

2001 assays were very faint positives. One of the 15 was positive by the Logan

et al. 2000 and the Ho et al. 1991 assay and the remaining 12 samples were

positive by the Logan et al. 2000 (genus specific) assay alone (Table 4).

In some cases the bands representing positive results on the agarose gel were

very faint (Figure 29). Reactions were too weak to provide sufficient template for

confirmatory sequencing.

8.1.2 Conclusion - Phase one of survey of domestic properties Of the samples tested only a small proportion tested positive for the presence of Helicobacter spp. and / or H. pylori. Table 4 Positive PCR results from filtered water samples and biofilms collected

from domestic properties in Phase One

Sample No. Date SampleType

Property Logan Ho Chisholm

RF04000023 02/02/04 Water A + - -

RF04000029 02/02/04 Water B + - -

RF04000030 02/02/04 Biofilm B + - -

RF04000033 02/02/04 Water C + + +

RF04000036 02/02/04 Biofilm D + + +

RF04000037 02/02/04 Water E + + -

RF04000060 04/02/04 Water F + - -

RF04000064 04/02/04 Water G + - -

RF04000065 04/02/04 Biofilm G + - -

RF04000103 05/02/04 Biofilm H + - -

RF04000107 05/02/04 Water I + - -

RF04000109 06/02/04 Biofilm J + - -

RF04000128 06/02/04 Water K + - -

RF04000130 06/02/04 Water L + - -

RF04000136 06/02/04 Water M + - -

69


14 15 16 17 18 19 20 21 22 23 24 25 26 27

1 2 3 4 5 6 7 8 9 10 11 12 131 Figure 29 Helicobacter specific PCR assay (Logan et al. 2000 assay)

Samples giving faint positive results are in lanes 5, 11,12,16 and 19. Lanes 13

and 26, positive control (H. pylori DNA); lane 27, negative control.

8.2 Phase Two of survey of domestic properties The aim of this phase of sampling was to recruit samples from the properties that

were PCR-positive in sampling phase one, to see if the protocol developed could

be applied to culture H. pylori from these.

It was possible to contact nine of the 13 properties that had been positive by at

least one PCR assay in phase one of sampling. Fresh water samples (10L –

large volume to increase sensitivity) and biofilm samples were submitted from

these nine properties for re-testing. This second phase of sampling took place

over four weeks in August 2004, on Tuesday 3rd, Tuesday 10th, Thursday 12th

and Tuesday 17th.

The protocol (Figure 38) developed for isolating H. pylori from domestic water

samples was applied to the nine water and nine biofilm samples collected. The

samples were processed, then either plated directly onto the different media in an

untreated or acid treated form, or acid treated and grown for 24, 48, 72 and 144

hours in MHYEB broth with or without antibiotics, before plating onto the four

70


different media (CBA, Dent’s, CBA2D, and DPA). A total of 40 plates were used

per sample for this protocol. The processed samples were also tested using the

three PCR assays (Logan et al 2000, Ho et al. 1991 and Chisholm et al. 2001

assays).

8.2.1 Results - Phase Two of survey of domestic properties No H. pylori were recovered in culture and in addition all the samples were PCR-

negative, this was unexpected as the samples were from previously PCR-positive

properties, including one property that had had a biofilm sample positive in all

three PCR assays. This may suggest that the presence of H. pylori in domestic

properties is a transient occurrence. It is possible that the survival of helicobacter

is poor during the summer months and / or there may be a higher disinfectant

residual chlorine/ monochloramine in the mains water during the summer.

Contamination was present on the majority of the culture plates however only two

water samples and one biofilm sample had such heavy growth that it would not

have been possible to see any helicobacter colonies, even if H. pylori was

present in the sample.

Extending the incubation period of the broths increased the background

contamination observed on the plates. After six days incubation in pre-

enrichment broth, just over half of the plates were too contaminated to visualise

any discreet colonies (Figure 30).

The plates on which there was no bacterial growth after 24 h incubation in pre-

enrichment broth generally remained clear even after six days incubation in the

broth. Thus the percentage of plates that showed no growth remained constant

(Figure 31).

71


0

20

40

60

80

100

24 48 72 144

Incubation Time in MHYEB (h)

% p

late

s w

ith c

onflu

ent

grow

th o

f co

ntam

inat

ion

Figure 30 The percentage of plates from sampling phase two with heavy confluent

growth of microbial contamination after incubation in pre-enrichment broth

Confluent growth of contaminants was observed on many more plates from the

water samples (Figure 32), compared with from the biofilms. Six of the nine

biofilm samples had several plates with no growth of any kind present (Figure

33).

0

20

40

60

80

100

24 48 72 144

Incubation time in MHYEB (h)

Figure 31 The percentage of plates from sampling phase two with no growth of microbial contamination after incubation in pre-enrichment broth

72


020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144

CBA Dent CBA2D DPA


% p

late

s w

ith

conf

luen

t gro

wth

of

cont

amin

atio

n

Antibiotics No Antibiotics

Figure 32 The percentage of plates with confluent growth of microbial contamination

for each solid medium after water samples were incubated in pre-enrichment broth with and without antibiotics

The highest amount of confluent growth was on the CBA plate with the samples

from the broth with no antibiotics (Figure 32). There was no significant difference

in the amount of confluent contamination between the other media.

The same pattern is also evident for the biofilms (Figure 33). The samples from

the broth with no antibiotics plated onto CBA have the highest level of

contamination but for the other media the contamination level is much lower

when compared to the water. This is unexpected as a biofilm should comprise a

complex mix of several different microbial communities and so, theoretically, the

level of contamination should be high. It is possible that when sampling the toilet

cistern a biofilm was not recovered on the swab.

73


020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144

CBA Dent CBA2D DPA



Figure 33 The percentage of plates with confluent growth of microbial contamination for each solid medium after biofilm samples were incubated in pre-enrichment broth with and without antibiotics

Conversely, if the number of plates with no bacterial growth is considered, the

biofilms have a large number of plates with no growth (Figure 34), compared with

the water samples (Figure 35).

No growth was observed on over 55% of the Dent’s, CBA2D and DPA plates

inoculated with broth with antibiotics, regardless of the incubation time (Figure

34). This again suggests that actual biofilm samples may not have been

obtained.

74


020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144

CBA Dent CBA2D DPA



Figure 34 The percentage of plates with no microbial contamination for each solid medium after biofilm samples were incubated in pre-enrichment broth with and without antibiotics

020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144

CBA Dent CBA2D DPA


% p

late

s w

ith n

o ba

cter

ial g

row

th


Figure 35 The percentage of plates with no microbial contamination for each solid medium after water samples were incubated in pre-enrichment broth with and without antibiotics

Fewer plates had no bacterial growth recorded for the water samples (Figure 35)

compared with the biofilms (Figure 34).

None of the water samples incubated in broth without antibiotics and plated on

CBA were completely free of contamination. No significant differences between

75


the other media were observed, although in general, more plates with no

contamination were observed from the broth containing antibiotics (Figure 35).

A number of plates were only lightly or moderately contaminated (<200 cfu) and

so, if present, H. pylori could still be visualised on a plate. The percentage of

plates with this lower level of contamination was greater for the water samples

(Figure 37), compared with the biofilm samples (Figure 36).

Overall almost 65% of plates from the biofilm samples had no bacterial growth.

The exception to this was the CBA plates that had been inoculated with broths

containing no antibiotics (Figure 36).

020406080

100

CBA Dent CBA2D DPA



Figure 36 The percentage of plates with moderate contamination (<200 cfu) from biofilm samples enriched in MHYEB with and without antbiotics

76


020406080

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144

CBA Dent CBA2D DPA


% p

late

s w

ith s

cant

to

mod

erat

e

cont

amin

atio

n


Figure 37 The percentage of plates with moderate contamination (<200 Cfu) from water samples enriched in MHYEB with and without antibiotics

8.2.2 Conclusion - Phase Two of survey of domestic properties It can be concluded from phase two of sampling that contamination levels were

too high in some samples, particularly water samples, to exclude the possibility

that H. pylori was present.

Compared to the Dent’s media, the DPA and the CBA2D media provided no

extra benefit in reducing contamination levels and so will not be used in phase

three of sampling.

H. pylori was not isolated on any of the plates where contamination was absent

or only moderate. These results may indicate either that H. pylori is not present

or not culturable in these samples. Alternatively, H. pylori may be present at a

level that is below the threshold of sensitivity for the methods developed.

However it is unlikely that there were any falsely negative results as all samples

were PCR-negative also.

It was demonstrated in section 6.4 that acid treatment can reduce the levels of

some H. pylori strains. To improve sensitivity of the method for phase three of

sampling a second set of broths were inoculated with samples that had not been

acid treated.

77


Water (10L)

Resuspend (10ml MRD/PBS)

Acid trea

Filtered (0.2µm)

Pellet in100µl on

each plate

Pre-enric

Biofilm

CBA

CBA2D

Dents

DPA

CBA

MHYEB100µl on

each plate

at t = 24h

t = 48h

t = 72h

t = 144h

DPA CBA CBA2D Dents

DPA CBA CBA2D Dents

DPA CBA CBA2D Dents

DPA CBA CBA2D Dents

Figure 38 Protocol for isolating H. pylori fthe 2nd round of sampling

5000g/30min

tment

3ml 100µl on each plate

hment

CBA2D

Dents

DPA

MHYEB +Antibiotics 100µl on

each plate

at t = 24h

t = 48h

t = 72h

t = 144h

DPA CBA CBA2D Dents

DPA CBA CBA2D Dents

DPA CBA CBA2D Dents

DPA CBA CBA2D Dents

rom domestic water samples and biofilms in

78


8.3 Phase Three of survey of domestic properties The aim of this phase of sampling was to continue to examine samples from

domestic properties for H. pylori using a modification of the protocol developed in

phase two of sampling.

A total of 27 properties were included in this period, eight of which were from

properties that had provided PCR-positive samples in the first round. Seven of

these had also been included in the second round of sampling.

Fresh water samples (2 L) and biofilm samples were submitted from these

properties. This third round of sampling took place weekly from Tuesday 7th

September to Wednesday 17th of November.

The protocol that had been developed for isolating H. pylori from domestic water

samples in phase two was modified to include two extra broths, one with

antibiotics and one without, for each sample that had not undergone acid

treatment. The DPA and CBA2D media were no longer used as the results from

the second round of sampling showed no added benefit of these. Although the

CBA media had shown the highest levels of contamination, this medium was still

used as spiking experiments (sections 5 and 7) had demonstrated that it was

optimal for recovery of H. pylori if contamination was low. This protocol required

the use of 36 plates per sample.

8.2.3 Results - Phase Three of survey of domestic properties No H. pylori have been cultured in 54 samples from 27 properties during this

phase of sampling.

From these 27 properties, eight water and two biofilm samples had sufficiently

heavy contaminating growth from the beginning that it would have been expected

to interfere with the growth and detection of helicobacter colonies.

79


Extending the incubation period of the broths increased the background

contamination observed on the plates. By day two in the broths, just over half of

the plates were too contaminated to visualise any discrete colonies (Figure 39).

0

20

40

60

80

100

24 48 72 144


Figure 39 The percentage of plates in sampling phase three with confluent growth of microbial contaminants

While there was a consistent increase in the number of plates with confluent

growth of contamination there was some variation with plates containing scant to

moderate growth. After 72 h incubation in pre-enrichment broth, there was a

noticeable decline in the number of plates with scant to moderate contamination.

This may be due to a number of the moderate growth plates becoming confluent.

80


0

20

40

60

80

100

24 48 72 144


% o

f pla

tes

with

sca

nt to

mod

erat

e co

ntam

inat

ion

Figure 40 The percentage of plates in sampling phase three with moderate growth

(<200 cfu) of microbial contaminants

0

20

40

60

80

100

24 48 72 144


Figure 41 The percentage of plates in sampling phase three with no growth of microbial contaminants

As was observed in phase two of sampling, the level of bacterial contamination

(confluent growth) was higher for the water samples (Figure 42) than for the

biofilms (Figure 43).

81


020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144

CBA - AT CBA - UT DENT - AT DENT - UT



Figure 42 The percentage of plates from acid treated and untreated water samples incubated in MHYEB broth with and without antibiotics that had confluent growth of microbial contaminants

020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144




Figure 43 The percentage of plates from acid treated and untreated biofilm samples incubated in MHYEB broth with and without antibiotics that had confluent growth of microbial contaminants

The highest level of contamination was again observed on the CBA plates,

particularly when samples were first incubated in broth with no antibiotics.

Fewer plates were heavily contaminated when the water samples were treated

with acid, compared to the untreated samples (Figure 42).

82


No bacterial growth was observed in over 60% of the Dent’s plates that were

inoculated with acid treated biofilm samples incubated in broth with antibiotics,

regardless of the duration of pre-enrichment. The proportion of plates without

contamination was slightly lower (55%) if the samples were incubated without

antibiotics. Similarly for the untreated samples, no growth was recorded for over

45% on the Dent’s plates (Figure 43). These low levels of contamination would

again suggest that actual biofilm samples are not being obtained.

020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144

CBA - AT CBA - UT DENT - AT DENT - UTIncubation Time in MHYEB (h)

% p

late

s w

ith n

o ba

cter

ial g

row

th


Figure 44 The percentage of plates from acid treated and untreated biofilm samples incubated in MHYEB broth with and without

antibiotics that had no growth of microbial contaminants

Similarly to the observations of sampling phase two, fewer plates that were free

of contamination were observed for the water samples (Figure 45), compared

with the biofilm samples (Figure 44).

The Dent’s plates did reduce the contamination compared with the CBA, as a

greater proportion of these were free of contaminants. For both the acid treated

and untreated samples, the broth containing antibiotics reduced contamination,

with more plates with no growth recorded (Figure 45).

83


020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144




Figure 45 The percentage of plates from acid treated and untreated water samples incubated in MHYEBB broth with and without antibiotics that had no growth of microbial contaminants

The biofilm and water samples can also be compared with regards to the amount

of plates that contain scant to moderate contamination, which would not be

expected to interfere with the recovery of H. pylori.

There is a higher level of scant to moderate growth on the CBA plates for the

biofilm samples than on the Dent’s plates. However for the water samples, a

higher level of scant to moderate growth is evident on the Dent’s plates in

particular those from the broth without antibiotics, this is possibly due to the fact

that more of the CBA plates have confluent growth of contaminants for the water

samples.

84


020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144



% p

late

s w

ith s

cant

to

mod

erat

e co

ntam

inat

ion Antibiotics No Antibiotics

Figure 46 The percentage of plates from acid treated and untreated biofilm samples incubated in MHYEB broth with and without antibiotics that had moderate (<200cfu) growth of microbial contaminants

020406080

100

24 48 72 144 24 48 72 144 24 48 72 144 24 48 72 144




Figure 47 The percentage of plates from acid treated and untreated water samples incubated in MHYEB broth with and without antibiotics that had moderate (<200 cfu) growth of microbial contaminants

85

All plates Figure 48 Flo

Water (2L)

Filtered (0.2µm)

Biofilm

Resuspend (10ml MRD/PBS)

Acid treated

Dents

CBA

Dents

CBA

Pellet in 4ml

MHYEB + AntibioticsMHYEB 100µeach plateat t = 24h

t = 48h

t = 72h

t = 144h

l on MHYEB + Antibiotics MHYEB 100µl on

each plateat t = 24h

t = 48h

t = 72h

t = 144h CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

CBA Dents

Pre-enrichment

5000g/30min 100µl on

each plate

200µl on each plate

Untreated

Pre-enrichment

incubated microaerophilically at 37 °C and examined for H. pylori and microbial contamination at 48h and up to 10 days

w chart of the sampling phase 3 protocol

Of the 54 samples tested, 11 were positive by at least one PCR assay (Table 5).

Eight of the 27 properties examined were PCR-positive (Table 5). Of these,

three had been positive in sampling phase one. Property “I” was not sampled in

the first phase.

Table 5 The samples from sampling phase three that were positive by at least one

PCR assay

Property Date

Received

Sample MOLIS

Number

Logan Ho Chisholm

Water RF0400030003 + A 07/09/2004

Biofilm RF0400030103

Biofilm RF0400032103 + +/- H* 28/09/2004

Water RF0400032403 + +/- +/-

Biofilm RF0400032203 + I† 28/09/2004

Water RF0400032503 +/- +/-

Biofilm RF0400032903 J 05/10/2004

Water RF0400033003 +

Biofilm RF0400033103 +/- + K 05/10/2004

Water RF0400033203 + +/- +

Biofilm RF0400034103 N 12/10/2004

Water RF0400034203 +/-

Biofilm RF0400035603 R* 19/10/2004

Water RF0400035703 +/-

Water RF0400039303 + Z* 17/11/2004

Biofilm RF0400039403

*Properties that were PCR-positive by at least one assay in the first round sampling

†This property was not examined in sampling phase 1.

Of the 11 positive samples, many plates were overgrown with confluent lawns of

contaminating bacteria (Table 6). Thus if Helicobacter spp. or H. pylori was

present in the samples, as suggested by PCR, it would not be possible to


visualise the small colonies. Additionally, the rapid growth of microbial

contaminants may have inhibited growth of helicobacter.

8.2.4 Conclusion - Phase Three of survey of domestic properties Lower level contamination that would allow any H. pylori growing to be seen was

observed on comparatively few plates (Table 6). However no H. pylori were

cultured. This may reflect differences in the sensitivity of PCR compared with the

culture protocol developed. Alternatively, the Helicobacter in water samples may

have entered a viable but non-cultivable state or growth of competing microbes

may still have been sufficient to inhibit H. pylori growth.

Table 6 The levels of microbial contamination on culture plates from phase three

samples that were positive by at least one PCR assay

Property Sample Confluent growth Plates that had ≤ semi-confluent growth (incubation time)

A Water All plates at 72h AT on CBA; AT+A on Dent’s (24h)

UT+A on CBA (48h)

Biofilm All plates at 48h UT+A on CBA (24h) H

Water All plates at 48h AT+A on Dent’s (24h)

Biofilm All CBA at 144h No growth on Dent’s; AT and UT on CBA (24h)

AT on CBA (72h)

I

Water All plates at 24h

J Water All plates at 48h UT sample on CBA at 24h

Water All plates at 24h K

Biofilm No growth on any plates

N Water All plates at 72h Dent’s plate (48h)

R Water All plates at 72h AT on CBA; AT+A on Dent’s (24h)

UT on CBA, UT+A on Dent’s (48h)

Z Water All plates at 24h

AT = Acid treated, UT = no acid treatment, A = antibiotics,

88


9 Effluent Samples

To date the study by Lu et al. 2002, which took place in Mexico and sampled

wastewater is the only one to have cultured H. pylori from a water source.

Primary and final effluent samples from the Thames were therefore examined for

H. pylori by culture. Samples (500ml) were filtered following the protocol for

phase three. The final effluent sample required two 0.2µm pore size membranes

to filter, while the primary effluent required one 0.2µm and nine 0.45µm

membranes to filter the required volume. Both samples were diluted in MRD

down to 10-5 and each dilution was added to the broths.

9.1 Conclusion - Effluent Samples No H. pylori was recovered as contamination proved to be a major problem with

96% of the primary effluent plates showing confluent growth (Table 7) and 83%

for the final effluent sample (Table 8). The dilution factor decreased the level of

contamination present.

Lu et al. 2002 included an IMS step in their protocol and from the amount of

contamination obtained in this experiment extra steps would need to be added to

the protocol to reduce the level of contaminants.

Table 7 Primary Effluent Sample Primary Effluent AT/UT 24hrs 48hrs 72hrs 144hrs Confluent Growth 17 46 46 46 46 Some Contamination 7 0 0 0 0 No Growth 0 2 2 2 2 Table 8 Final Effluent Sample Final Effluent AT/UT 24hrs 48hrs 72hrs 144hrs Confluent Growth 12 40 40 41 41 Some Contamination 11 0 4 3 2 No Growth 1 8 4 4 5

89


10 Overview of Results

10.1 Location of Domestic Properties

A total of sixty-four domestic properties were recruited to the study (two were

only recruited to the third phase – properties in NW and N London). Based on the

first two letters of the postcode properties were from 13 different areas mostly

within the Greater London area, the majority (50%) were from North West

London and Harrow (see Table 9).

At least one of the samples from 13 (21%) of the 62 properties tested in Phase

One and 8 (29.6%) of the 27 properties tested in Phase Two were positive by at

least one of the specific PCR assays for Helicobacter. Three of the properties

were positive in both Phase One and Phase Three. However, the nine properties

sampled in Phase Two were all negative even though they had previously been

PCR positive in Phase One.

Table 9: Property Information

Postcode Area Number of properties

(64)

PCR+ properties (14/62) Phase 1

PCR+ properties (8/27)

Phase 3 AL St Albans 4 1* 1* HP Harpenden 1 1 0 HA Harrow 11 2 0 E East London 1 1 N/A EN Enfield 2 1 0 LU Luton 4 1 0 ME Gillingham 1 0 1 MK Milton Keynes 1 1* 1* N North London 5 0 1 NW North West London 21 3 1 SG Stevenage 4 0 0 W West London 1 1 0 WD Watford 6 2 3* No code given

Unknown 2 0 0

*same property as Phase 1 N/A = not available

90


Overall 28% (18/64) of properties were Helicobacter PCR positive. There were a

total of 26 positive samples from these 19 properties. The positivity rate for

samples was 13.7% (26/190). There were 15/118 (13%) positive samples in

Phase One and 11/54 (20%) positive samples in Phase Three. Both properties

and samples show a higher rate of helicobacter in the last Phase of sampling.

These results are similar to those obtained in the previous DWI study.

10.1 Types of property Fifty of the properties were houses (two of these, a semi-detached and a

terraced house, were only sampled in Phase 3) and twelve were flats. Another

two were unknown as PSQ’s were not filled out (Table 10). The results suggest

that houses, in particular terraced houses, are more likely to be PCR positive

than flats. Overall 36% (18/50) houses and 15.5% of samples from houses

(24/155) were PCR positive for helicobacter.

Table 10: Types of Property

Type of Property

Description Number properties (64)

PCR+ properties (18)

Number samples (190)

PCR+ samples (26)

Detached 15 3 (20%) 43 5 (12%) Semi-detached 21 7 (33%) 69 12 (17%)

House (78%)

Terraced 14 6 (42%) 43 7 (16%) Flat (19%) - 12 2 (17%) 31 2 (6%) Unknown - 2 0 4 0 The number of rooms varied from 4 -15 with a mean of 7 rooms (median 6). All

properties had at least one toilet / bathroom and more than two-thirds of

properties also had a separate shower (two properties had a shower and no

bath). For properties with seven rooms or less 17% of properties and 15% of

samples were PCR positive. For properties with more than seven rooms 25% of

properties and 11% of samples were PCR positive. Helicobacter positive PCR

did not seem to be associated with the size of the property (Table 11).

91


Table 11: Size of Properties

No. of rooms No. of properties PCR+ properties PCR+ samples

7 or less 41 7 (17%) 17/113 (15%)

>7 20 5 (25%) 7/65 (11%)

The age of the properties varied from 4 – 200 years (mean 63 years, median, 60

years) (Table 12). Comparing properties under 50 years to those over 50 years

there is little difference in positive rates. Twenty-nine percent (4/21) of properties

that were less than 50 years old were PCR positive. For properties over 50 years

29% (11/38) are PCR positive. However, the oldest properties (>100 years)

appear to have the highest rate of PCR positives both in terms of properties

(63%) and samples (27%), although there were fewer of these properties.

Table 12: Age of Properties

Age of

property

Number of

properties

PCR+

properties

PCR+

samples

<20 years 7 2 (29%) 2/15 (13%)

20-50 14 4 (29%) 7/41 (17%)

50-100 30 6 (20%) 8/93 (9%)

>100 8 5 (63%) 8/30 (27%)

10.2 Cold water systems The majority (76%) of properties sampled were supplied by Three Valleys Water

Utility (Table 13). Twelve (25%) of these properties were PCR positive. Of the 11

properties supplied by Thames water four (36%) were PCR positive, which

suggests that properties supplied by Thames water are more likely to be PCR

positive. However when we compare the % positive of the samples these are

identical (12%) whether the property is supplied by Three Valleys or Thames.

92


Table 13: Water Utilities

Table: Water Utility No. of properties

(%properties served by

water utility)

No. PCR

positive

properties

No. PCR+

samples

Anglian 2 (3%) 1 3/4

Southern Water 1 (2%) 1 2/4

Thames 12 (17%) 4 (36%) 5/43 (12%)

Three Valleys 49 (76%) 12 (25%) 16/135 (12%)

Unknown 2 (2%) 0 0

Most properties had a cold tap fed directly off the mains and this was usually the

kitchen tap as the kitchen was where mains water entered the majority (80%) of

properties. Only two (3%) properties, both flats in the Thames area indicated that

there was no cold tap directly fed by the mains supply. There was no apparent

association between where the water entered the property and PCR positive

samples or properties.

The pipe work in most (64%) properties was copper or partially copper. Of the 41

properties that had copper pipes seven (17%) properties had at least one sample

that was PCR positive by one of the Helicobacter specific assays. Of the 23

properties that did not have copper pipes six (26%) were PCR positive.

Only about third of properties had water heaters fed directly from the mains,

while more than two thirds had storage tanks for hot water. Of forty properties

with storage tanks nine (22.5%) were PCR positive compared to four (22%) of 18

properties without storage tanks. Therefore there was no association with

storage tanks and PCR positive properties.

More than 60% (39 properties) had header tanks and the majority of these

header tanks also supplied cold water (33 properties) and were located in the loft

(34 properties). Twenty-five of the tanks were fully covered and eleven were

93


partially insulated. There was a possible association with properties which had

header tanks: eight (20.5%) of 39 properties with header tanks were positive,

compared to 2 (14%) of 14 properties without a header tank.

The age of the plumbing systems varied from one to 50 years old (Table 14).

Although there appears to be no correlation with PCR positive properties and the

age of the plumbing system, the comparison with percentage of PCR positive

samples seems to indicate an association with plumbing systems older than ten

years (between 30 and 50 years). This may be because the helicobacter are

surviving within biofilms and the biofilms require time to establish.

Table 14: Age of Plumbing System

Age of plumbing

system

No. of

properties

PCR+

properties

No. of

samples

PCR+

samples

5 years or less 14 3 (21%) 37 3 (8%)

6-10 years 7 2 (28%) 26 2 (8%)

11-25 years 21 6 (28%) 60 10 (17%)

>25 years 7 2 (28%) 28 4 (14%)

The last time any work was done on the plumbing also seems to have an affect

both on whether the property or the samples from the property are PCR positive

with the highest percentage in properties where there has recently been

plumbing work or at the other extreme where the plumbing system has not been

worked on for some time (between12 and 20 years).

94


Table 15: Work on Plumbing System

Last time work

done on

plumbing system

No. of

properties

PCR+

properties

No. of

samples

PCR+

samples

2 years or less 24 8 (33%) 81 11 (14%)

3-5 years 12 2 (17%) 30 2 (7%)

6-10 years 5 1 (20%) 18 2 (11%)

>10 years 7 2 (29%) 21 4 (19%)

All but five properties had a shower, although 22 (35%) were mixers connected

to the bath tap. For 55 showers the connection to the shower head was a flexible

connection. Only two had fixed connections: a mixer connected to bath taps and

a separate pumped shower. The showerheads were composed of plastic (38),

metal (16), or metal and plastic (2). The PSQ also asked if there were any

obvious faults with the shower, the main fault was listed as “scaling” (23) or

“dripping” (6). There were no apparent associations with PCR positive samples

or properties.

There were 18 (18%) PCR positive water samples from a total of 98 water

samples. Ten (16%) of 62 water samples were positive in Phase 1, none of the

nine water samples taken in Phase 2 were positive while eight (30%) of 27

samples were positive in Phase 3.

10.3 WC cistern The biofilm sample when taken was from the water closet cistern. The PSQ

collected information on the material the toilet cistern was made from and scores

were assigned for the presence of corrosion, dirt, or biofilm.

The majority (74%) of cisterns were made of porcelain. The corrosion, dirt and

biofilm was scored as 0, 1, 2, 3 or 4: 0 = none present; 4 = highest score. There

95


was no apparent correlation between the corrosion, dirt or biofilm scores and

PCR positive samples.

There were eight (9%) PCR positives from a total of 93 biofilm samples. Five

(8%) of 63 biofilm samples were positive in Phase 1, none of the nine biofilm

samples taken in Phase 2 were positive while three (11%) of 27 samples were

positive in Phase 3.

11 Discussion

The study has examined a number of methodologies for isolating H. pylori from

water samples and these have been improved. The most appropriate approach is

to use acid treatment and pre-enrichment in MHYEB with antibiotics and culture

on CBA with Dent’s medium as adopted in the phase 3 protocol. However the

methods remain imperfect, as others have found, and decisions need to be made

about how to take forward the results from this study and make practical public

health use of them.

This study has found that water samples and biofilms are commonly positive

when using PCR to detect the presence of Helicobacter genes but all of these

samples are negative for H. pylori isolation using an array of different media.

PCR detections are therefore likely to represent DNA released from :

1. viable H. pylori that are either surviving within a biofilm or in free living

protozoa but are not able to be cultured by current methodologies

2. dead H. pylori.

3. other bacteria that produce positive PCR products from the primers used

to detect Helicobacter genes.

Although it was not possible to confirm by sequencing that the bands detected

during the analysis of water samples in the present study belong to H. pylori, the

previous work by both this group (Watson et al 2004) and by others (Lu et al

96


2002; Bunn et al 2002; Park et al 2001) does indicate that these PCR detections

do represent Helicobacter genes, albeit sometimes not H. pylori for some of the

products being detected. All attempts at isolating H. pylori from biofilm and water

samples have proved unsuccessful. It is most likely that the PCR detections from

within water distribution systems represent dead organisms that retain a

membrane bound bacterial cell that protects the DNA from being denatured by

the nucleases generated by other organisms naturally inhabiting these waters. It

remains possible that they are viable but non-cultivable.

Evidence of the relatively poor survival of H. pylori strains in water over a few

days in this study suggests that long-term survival within biofilms is unlikely. If the

PCR detections did represent viable but non-cultivable H. pylori then this would

represent evidence of a breakdown in both water disinfection and residual

disinfection if the organisms were free living or residing within the biofilm. H.

pylori is a relatively fastidious organism which does not appear to grow in the

conditions of low nutrients and low iron that are found in water systems.

Interestingly studies in the USA indicate that H. pylori is able to enter the viable

but nonculturable state (VBNC) as cells age in the laboratory or are exposed to a

natural freshwater environment (Adams et al 2003). In the natural stream

environment it was found that culturability decreased over time and that within

70h the helicobacters were non-culturable. Nevertheless a large number of cells

apparently remained viable as indicated by BacLight live/dead assay. Unlike the

laboratory experiments, there was no significant change in the number of rod and

coccal forms in the environment.

Some further work to check whether H. pylori can survive within protozoa may be

justified to rule out this as a route for human disease transmission through

drinking water. This is because intracellular growth within protozoa may give H.

pylori better access to essential nutrients and may partially protect it from the

chlorine-based disinfectants used in water treatment. However the predominating

evidence from this work is that viable/cultivable H. pylori are not commonly

97


present in routinely collected water samples and therefore mains drinking water

is unlikely to be a significant source of sporadic human Helicobacter infection

where normal treatment processes and residuals are used. It remains possible

that, as with Campylobacter, outbreaks of H. pylori could occur as a result of

occasional breakdowns in water treatment. As occasional outbreaks of

Campylobacter infection are predominantly associated with water systems that

are unchlorinated (mainly private water supplies) or have experienced a failure of

chlorine treatment it might be presumed that similar outbreaks of Helicobacter

might occur. However, there is currently no surveillance system that would be

able to detect an H. pylori outbreak. As H. pylori is widely regarded as an

organism that is unlikely to be zoonotic it is less likely to be associated with

private water supplies where animal faeces probably represents a majority of the

infection risk, especially for Campylobacter.

The epidemiology of H. pylori remains unclear. Although there is epidemiological

evidence for associations between human H. pylori infection and contaminated

drinking water in developing countries there is little evidence that human disease

within the UK derives from mains drinking water and better evidence for person-

to-person transmission within childhood being important. Evidence of human

excretion of H. pylori antigen in faeces is being used more commonly in

diagnosis, but the difficulty in isolating the organisms from faecal samples partly

reflects the technical difficulties experienced in this study in growing H. pylori in

the presence of faster growing organisms. It may also partly reflect the relatively

low excretion of viable H. pylori in faeces although this is also relatively poorly

understood. It is likely that excretion may be more common in people with

diarrhoea. Understanding the epidemiology is also limited by an absence of

information about the dose of H. pylori required to cause human disease.

Is additional water testing for H. pylori likely to be useful? Given the relatively

poor performance of the current best available methods for H. pylori isolation we

feel that further testing of random samples is not likely to be fruitful. The most

98


important technical reason for poor isolation of H. pylori was the high level of

contaminating faster growing and less fastidious microbial flora. We believe that

the most likely explanation of PCR positive water and biofilm samples is that

these are dead H. pylori cells. While it remains theoretically possible that

treatment failures may very occasionally allow viable H. pylori into mains drinking

water we feel that in the absence of any way of demonstrating an outbreak it will

be impossible to know when and where to test for contamination. Given that

about 20% of the UK population have evidence of active infection the ability to

detect the source of infection for individual cases is poor.

There is little data on the persistence of bacterial DNA in the aquatic environment

although this is important to our understanding of the interpretation of the

detection of organisms by PCR. Experiments studying the persistence of DNA

within cells killed by different means would be of value.

What further work might usefully be done to improve the methods by which the

epidemiology would be better understood? There is still a need for work on

improved decontamination so that isolation of H. pylori from faeces and

environmental samples is easier. In particular the reasons for the poor

performance of immunomagnetic separation in this study could usefully be

examined. In addition the factors influencing the relatively poor survival of H.

pylori in water are poorly understood. In this respect the sensitivity to

atmospheric oxygen may be important although this study has not looked at this

specifically. The study has demonstrated evidence of considerable strain

variability to the factors influencing growth including acid resistance and

sensitivity to antibiotics. Any follow up work should ensure that multiple strains

are tested.

99


12 Acknowledgements

This work was funded by DEFRA and managed by the DWI. This work would not

have been possible without the help and co-operation of colleagues at the HPA

Centre for Infections who were kind enough to take samples from their properties

and bring them into the laboratory on the same day for processing.

100


13 References

Adams, B. L., Bates, T.C. and Oliver, J.D. (2003), Survival of Helicobacter pylori

in a natural freshwater environment, Applied and Environmental Microbiology, vol

69, pp. 7462-7466.

Axon, A. T. R. 1997. “Transmission of Helicobacter pylori”. Yale J Biol &

Medicine, vol. 70, no. 1, pp. 1-6.

Azevedo, N. F., Pacheco, A. P., Keevil, C. W. and Vieira, M. J. (2004). “Nutrient

Shock and Incubation Atmosphere Influence Recovery of Culturable Helicobacter

pylori from Water,” Applied and Environmental Microbiology, vol.70, no. 1, pp.

490-493.

Boom, R., Sol, C., Salimans, M. M., Jansen, C. L.,Wertheim-van Dillen, P. M.

and van der, N. J. (1990), “Rapid and simple method for purification of nucleic

acids”, Journal of Clinical Microbiology, vol. 28, pp. 495-503.

Brown, L. M. 2000. “Helicobacter pylori: epidemiology and routes of

transmission.” Epidemiol. Rev. vol. 22, pp. 283-297.

Chisholm, S. A., Owen, R. J., Teare, E. L. and Saverymuttu, S. (2001), “PCR-

based diagnosis of Helicobacter pylori infection and real-time determination of

clarithromycin resistance directly from human gastric biopsy samples”, Journal of

Clinical Microbiology, vol.39, no. 4, pp. 1217-1220.

Degnan, A. J., Sonzogni, W. C. and Standridge, J. H. (2003). “Development of a

Plating Medium for Selection of Helicobacter pylori from Water Samples”, Applied

and Environmental Microbiology, vol. 69, no. 5, pp. 2914-2918.

101


Go, M. F. 2002. “Natural History and Epidemiology of Helicobacter pylori

infection.” Aliment. Pharmacol. Ther., vol. 16, suppl 1, pp. 3-15.

Hegarty, J. P., Ma, Y. L., Lebo, C. M., Konopka, T. L. and Baker K. H. 1999.

“Waterborne transmission of Helicobacter pylori : convergence of environmental

and clinical observations.” American Society for Microbiology 99th General

Meeting.

Ho, S. A., Hoyle, J. A., Lewis, F. A., Secker, A. D., Cross, D., Mapstone, N. P.,

Dixon, M. F., Wyatt, J. I., Tompkins, D. S. and Taylor, G. R. (1991), “Direct

polymerase chain reaction test for detection of Helicobacter pylori in humans and

animals”, Journal of Clinical Microbiology, vol. 29, pp. 2543-2549.

Jiang, X. and Doyle, M. P. (2002). “Optimizing Enrichment Culture Conditions for

Detecting Helicobacter pylori in Foods”. Journal of Food Protection, vol. 65, no.

12, pp. 1949-1954.

Krumbiegel, P. K., Lehmann, I., Herbarth et al. (2001) H. pylori in German non-

municipal drinking water: epidemiology and microbiology in evidence. Int J Med

Microbiol, 291 (Suppl no.31): A-03.

Logan, J. M., Burnens, A., Linton, D., Lawson, A. J. and Stanley, J. (2000),

“Campylobacter lanienae sp. nov., a new species isolated from workers in an

abattoir”, Internation Journal of Systematic and Evolutionary Microbiology, vol.

50, pp. 865-872.

Marshall, B. J. and Warren, J. R. 1984. “Unidentified curved bacilli in the

stomach of patients with gastritis and peptic ulceration.” Lancet, vol. 1, pp. 1311-

1315.

102


Mazari-Hiriart, M., Lopez-Vidal, Y. and Calva, J. J. 2001. “Helicobacter pylori in

water systems for human use in Mexico City.” Water Sci Technol, vol, 43, no.

12, pp. 93-98.

Park, S. R., Mackay, W. G.and Reid, D. C. 2001. “Helicobacter sp. recovered

from drinking water biofilm sampled from a water distribution system.” Water

Res, vol. 35, no. 6, pp. 1624-1626.

Poms, R. E. and Tatini, S. R. 2001, “Survival of Helicobacter pylori in ready-to-

eat foods at 4°C”, International Journal of Food Microbiology, vol. 63, pp. 281-

286.

Sanders, M. K. and Peura, D. A. 2002. “Helicobacter pylori-Associated

Diseases”. Cur Gastroenterol Rep, vol. 4, no. 6, pp. 448-454.

Sasaki, K., Tajiri, Y., Sata, M., Fujii, Y., Matsubara, F., Zhao, M., Shimizu, S.,

Toyonaga, A.and Tanikawa, K. 1999. “Helicobacter pylori in the natural

environment.” Scand J Infect Dis, vol. 31, no. 3, pp. 275-279.

Stevenson, T. H., Castillo, A., Lucia, L. M. and Acuff, G. R. (2000). “Growth of

Helicobacter pylori in various liquid and plating media”, Letters in Applied

Microbiology, vol. 30, pp. 192-196.

Taneera, J., Moran, A. P., O Hynes, S., Nilsson, H. O., abu Al-Soud, W. and

Wadstrom, T. (2002). “Influence of activated charcoal, porcine gastric mucin and

β-cyclodextrin on the morphology and growth of intestinal and gastric

Helicobacter spp.”, Microbiology, vol, 148, pp. 677-684.

Vyse, A. J., Gay, N. J., Hesketh, L. M., Andrews, N. J., Marshall, B., Thomas, H.

I., Morgan-Capner, P.and Miller, E. 2002. “The burden of Helicobacter pylori

103


infection in England and Wales”. Epidemiology and Infection, vol, 128, no. 3, pp.

411-417.

Walsh, E. J. and Moran, A. P. 1997, “Influence of medium composition on the

growth and antigen expression of Helicobacter pylori”, Journal of Applied

Microbiology, vol. 83, pp. 67-75.

Watson, C. L., Owen, R. J., Said, B., Lee, J. V., Surmann-Lee, S. and Nichols, G.

2004. “Detection of Helicobacter pylori by PCR but not culture in water and

biofilm samples from drinking water distribution systems in England.” Journal of

Applied Microbiology, vol. 97, pp.690-698.

Whale, A., Said, B., Owen, R., Watson, C. L., Drobniewski, F., Nichols, G., Lee,

J. V., Weldon, L., Edwards, R., Lai, S. and Surman, S. 2003. “Further studies on

the Incidence of Mycobacterium avium Complex (MAC) in drinking water supplies

(Including the detection of Helicobacter pylori in water and biofilm samples).” A

research report from the Health Protection Agency to the Drinking Water

Inspectorate DWI/70/2/146.

104


14 Appendices

14.1 Appendix 1 Letter to volunteers Sent: 15 January 2004 13:15 Subject: Volunteers for water and swab samples Dear all The Water & Environmental Microbiology Reference Unit, in collaboration with the Helicobacter Reference Unit and the Environmental Surveillance Unit, are currently involved in a project funded by the Drinking Water Inspectorate, to investigate the quality of water in domestic properties. To do this we would like two samples; a 2 litre water sample taken from the shower or bathroom and a swab sample from the toilet cistern. We will provide the sampling equipment. If you are willing to collect the two samples from your property, please contact either Dr J. V. Lee (ext 4012) e-mail: [email protected] or Dr S. Lai (ext 4247) [email protected] for more information. All volunteers appreciated. Many thanks, Dr J. V. Lee Dr S. Lai Unit Head Clinical Scientist Water & Environmental Microbiology Reference Unit (WEMRU) Food Safety Microbiology Laboratory Health Protection Agency Specialist & Reference Microbiology Division 61 Colindale Avenue London NW9 5HT Tel: +44 (0) 20 8200 4400 ext 4247 Fax: +44 (0) 20 8358 3112

105

mailto:[email protected]

mailto:[email protected]


14.2 Appendix 2 Protocol for sampling water systems

PROTOCOL FOR SAMPLING VOLUNTEERS

PLEASE COLLECT ONE 2 LITRE WATER SAMPLE AND ONE SWAB SAMPLE. NB: WEAR DISPOSABLE GLOVES AND PUT ON A FRESH PAIR FOR EACH SAMPLE.

1. Water sampling PLEASE COLLECT FIRST FLUSH SAMPLES i.e. DO NOT FLUSH THE TAP BEFORE COLLECTING THE SAMPLE.

a) For showers with flexible hoses 1. Wipe down the surface of the outlet with alcohol wipe provided.

2. Label bottle

3. Unscrew the showerhead from the hose

4. Adjust the temperature knob to cold (if the water comes out slightly warm it does

not matter but it should definitely not be hot)

5. Unscrew and remove bottle cap being careful not to touch the inside of the cap

6. Collect 2 litres of water

7. Screw cap back onto bottle and mix the contents.

OR b) Showers with fixed heads and no flexible hose

1. Wipe down the surface of the outlet with alcohol wipe provided.

2. Label a bottle

3. Unscrew and remove cap (store cap in a clean new plastic bag provided)

4. Cut the end or a corner off a new food grade plastic bag.

5. Insert the showerhead into the plastic bag and secure with elastic band provided

6. Insert the other end of the bag into the open sample container.

7. Adjust the temperature knob to cold (if the water comes out slightly warm it does

not matter but it should definitely not be hot).

8. Carefully turn on the water flow so that a gentle flow is created and allow water to

flow into the container. You may need to hold the bag to the showerhead

9. Collect 2 litres of water.


106


OR

11. c) Bathroom mixer tap If you do not have a shower but have a blender tap on the bath or bathroom wash hand

basin this can be used instead.


2. Label a bottle

3. Unscrew cap from the sampling bottle

4. Collect 2 litres of cold water. It is important to ensure that the water is definitely not hot. If it is not possible to get the 2 litre sample bottle under the tap use a

sterile 1 litre container and transfer the water to the 2 litre vessel. Discard the 1

litre container.


OR d) Bathroom with separate hot & cold taps


2. Label a bottle

3. Unscrew cap from the sampling bottle

4. Collect 2 litres of cold water.


107


2 Swab sampling Toilet cistern Select a toilet cistern for sampling. If there is a choice avoid using ones with disinfectant blocks immersed in the cistern.

1. Label the Phosphate Buffered Saline (PBS) bottle with your name and date.

2. Remove the lid of the cistern

3. Using a garden tie provided, tie off the ball valve.

4. Flush the toilet so the water level in the tank goes down.

5. Remove the sterile template from its package and hold in one hand.

6. Take a sterile swab provided and using the blue template as a guide, swab an

area of 100cm2.

7. This is your biofilm sample.

8. Place the swab into the Phosphate buffered saline (PBS). You may have to

break off part of the shaft of the swab so that you can screw the cap back onto

the PBS bottle.

9. Release the ball valve ensuring the water flows back in properly.

10. Recover the tank making sure not to allow dirt to enter it.

Storage and transport of samples Protect samples from direct sunlight and heat during transport. Transport samples to the laboratory as soon as possible so that they can be processed on the day of collection.

Please send samples to: Dr. S. Lai

Room 2B32 or 4A35 Water and Environmental Microbiology Research Unit

Food Safety Microbiology Laboratory, Central Public Health Laboratory,

61 Colindale Avenue, London NW9 5HT

108


14.3 Appendix 3 Property survey questionnaire

Property Survey Questionnaire

Property details 1. Contact name………………………………………………………………………….

2. Address……………………………..………………………………………………..

…

……………………………………………………….… Postcode....………...………..…

3. Telephone Number…………………………………………………………………...

4. Type of property (tick all that apply) Single occupancy Multi-occupancy

Domestic Detached Semi-detached Terraced Flat Purpose-built block

5. Age of property…………………………

6. Number of rooms………………………

7. Number of toilets………………………

8. Number of baths.………………………

9. Number of shower booths (as opposed to showers over baths)…….……….…..

10. Number of kitchen / utility room

sinks….………………………………………….…

11. When was the last time any work was done on the plumbing?……………………

12. How old is the system? (give different ages of tank, shower, etc separately)

…………………………………………………………………………………..…….

…

Cold system – Mains supply 13. Water Utility supplying

water…………………………………………………………...

14. Where does supply enter the property?….……….………………………….………..

15. What are the pipes made of? plastic copper stainless steel galvinised steel

16. Is there a cold tap directly off the main supply? kitchen bathroom outside

109


Header tank

17. Is there a header tank? Yes No

18. Where is the tank

located?…..….….……………………………………………..…….

19. Is the tank covered? completely partially not at all

20. Is the tank insulated? completely partially not at all

21. Does it supply cold water to any taps? Yes No

Hot water system 22. Is the water heater fed from the cold tank? Yes No

23. Is the water heater fed directly from the mains supply? Yes No

24. Is there a storage tank? Yes No

25. Is the storage tank insulated? completely partially not at all

26. What are the pipes made of? plastic copper stainless steel galvinised steel

Shower 27. The shower is… (tick all that apply)

a mixer connected to hot and cold tap on bath

a separate pumped shower

an instantaneous electrically heated shower

28. Type of connection to showerhead? Flexible hose Fixed head

29. Are there any faults? (tick all that apply) dripping scaling mould /other

growth deterioration of the hose other please

describe……………………………………

30. The showerhead is… metal plastic

Water Closet cistern 31. WC cistern (tick all that apply) plastic porcelain metal low level high

level

32. On a scale of 0-4 (0=none; 4=worst) score:

Corrosion (if

metal)…………..………………………………………………………..

Dirt in the bottom of the cistern..……………………………………………………..

Biofilm or slime on the side of the cistern near the air / water interface.…………

110


Samples 33. Samples collected at.……..…(am/pm) on (dd/mm/yy) ……./…..../.…...

34. Samples check-list (tick which were taken at this property):

2L water sample (mixed hot and cold) shower

bath mixer tap

bathroom wash handbasin

Swab sample (100cm2biofilm) from WC

111


14.4 Appendix 4 Property survey repeat sample questionnaire

Property Survey Questionnaire – repeat sample

Property details 1. Contact name………………………………………………………………………….

2. Address……………………………..…………………………………………………

……………………………………………………….… Postcode....………...………..…

3. Water Utility supplying water Thames

Three Valleys

Other please state………………………………

4. When was the last time any work was done on the plumbing?……………………

5. Please describe briefly, if any work was done since the last sample was taken

………………………………………………………………………………………

………………………………………………………………………………………

6. Is there a disinfectant block (eg”loo bloo”) in use in the WC? Yes No

Samples

7. Samples collected at.……..…(am/pm) on (dd/mm/yy) ……./…..../.…...

8. Samples check-list (tick which were taken at this property):

2L water sample cold water only

mixed hot and cold (eg: mixer tap)

shower

bath tap

bathroom wash handbasin

Swab sample (100cm2biofilm)

WC cistern

112


14.5 Appendix 5 Composition of agar media (amount L-1) used for culture of H. pylori

Ingredients Columbia

Blood Agar (CBA)

Columbia Blood Agar with 2x Dent’s (CBA2D)

Degnan (DPA)

Special peptone agar (1/2 strength) (ASPA)

Dent’s Agar

Agar 10g 10g 15g 15g 10g Special Peptone 23g 23g 5g 12g 23g Sodium chloride 5g 5g 5g 2.5g 5g Starch 1g 1g 1g Yeast Extract 5g 2.5g Beef Extract 5g 2.5g Defibrinated Horse Blood

11% 11% 2.5%

Laked Horse Blood 7% Calf Serum with Iron 70ml Pyruvic acid sodium salt

0.25g

Vancomycin 20mg 10mg 10mg Trimethoprim 10mg 5mg 5mg Cefsoludin 10mg 5mg 5mg Amphotericin B 10mg 7.5mg 5mg Polymyxin B 3500U Urea 600mg Phenol Red 100mg 1N HCl 0.8ml

113


114

14.6 Appendix 6 Main components of broth media (gL-1) tested for enrichment of H. pylori

Ingredients Broth medium* BBβC FSPB** MRD BBYE BBYEB MHYE MHYEB BBYEC BBAC Special peptone 10g Peptone 1g Yeast Extract 2g 5g 12g 12g 10g 10g 12g Beef Extract 5g Sodium Chloride 5g 5g 8.5g 5g 5g 5g 5g Pyruvic acid sodium salt

5g

Iron supplement calf serum

0.04%

Casein Enzymic Hydrolysate

10g 10g 10g 10g 10g

Dextrose 1g 1g 1g 1g 1g Peptic Digest of Animal Tissue

10g 10g 10g 10g 10g

Sodium bisulphate 0.1g 0.1g 0.1g 0.1g 0.1g Dehydrated infusion from beef

300g 300g

Casein hydrolysate 17.5g 17.5g Starch 1.5g 1.5g Defibrinated Horse Blood

7% 7%

β cyclodextrin 0.1% Activated charcoal 0.1% 0.1% Agar 0.2% * Abbreviations used for the media are as follows (in the order used in the table): BBβC = Brucella broth + 0.1% β cyclodextrin HPSPB = H. pylori special peptone broth BBYE = Brucella broth + 1% yeast extract BBYEB = Brucella broth + 1% yeast extract + 7% defibrinated horse blood MHYE = Mueller-Hinton broth + 1% yeast extract MHYEB = Mueller-Hinton broth + 1% yeast extract + 7% defibrinated horse blood BBYEC = Brucella broth + 1% yeast extract + 0.1% activated charcoal BBAC = Brucella broth + 0.2% agar + 0.1% activated charcoal ** For ½ Strength HPSPB (HPSPB*) all quantities were halved except for the iron supplemented calf serum

further studies on the incidence of helicobacter organisms in...

Documents