United States
Environmental Protection
Agency
Office of Science
and Technology
Washington, D.C.
EP A 822 R 06 002
March 5, 2006
EPA Office of Water
National Field Study for Coliphage Detection
in Groundwater: Method 1601 and 1602
evaluation in regional aquifers
FINAL REPORT
-------
EPA/822/R/06/002
National Field Study for Coliphage Detection in Groundwater: Method 1601 and
1602 Evaluation in Regional Aquifers
ACKNOWLEDGMENT
The contributions of the following are gratefully acknowledged:
Nena Nwachuku, Ph.D.
EPA Lead Scientist for the study
Office of Water, Office of Science and Technology
Mark Sobsey, Ph.D.
University of North Carolina
Chief Investigator for the study and
Principal Investigator for the South East US Region
Sagar Goyal, Ph.D.
University of Minnesota
Principal Investigator for the Midwest US Region
Aaron Margolin, Ph.D.
University of New Hampshire
Principal Investigator, for the New England Regions
Suresh Filial, Ph.D.
Texas A& M University,
Principal Investigator, South West US Region
Research Investigation assistants
Gregory Lovelace, Douglas Wait, Dorothy Thompson, Nicola Ballester, Baldev Gulati, Sigrun
Haugerud, Sunil Maherchandani, Yasphal Malik, James Totten and Robin Whitley.
n
-------
External Peer Reviewers
Dr. Pierre Payment
Dr. Juan Jofre
Dr. Charles Gerba
Dr. Marylynn Yates
Dr. Roger Fujioka
Dr. Sharon Long
Dr. Morteza Abbaszedegan
Dr. Dean Cliver
Dr. Gary Toranzos
Dr. Howard Kantor
Dr. Mohammad Karim
Dr. Bruce Keswick
Mr. Steve Via
University of Quebec, Canada
University of Barcelona, Spain
University of Arizona, Tucson
University of California, Riverside
University of Hawaii
University of Massachussetts, Amherst
Arizona State University
University of California
University of Puerto Rico
College of William and Mary, Virginia
American Water Works System
Proctor and Gamble, Mason, OH.
American Water Works Association
Internal Peer Reviewers
Dr. Nena Nwachuku
Dr. Stephen Schaub
Dr. Al Dufour
Dr. Paul Berger
Dr. James Sinclair
Dr. Ann Grime
Mr Stig Regli
Ms. Crystal Rodgers
Mr. Mark Messner
Dr Shay Fout
Dr. Phil Berger
USEPA
USEPA
USEPA
USEPA
USEPA
USEPA
USEPA
USEPA
USEPA
USEPA
USEPA
-OST, Washington, DC
-OST, Washington, DC
-ORD, Cincinnati- OH
- OGWDW, Washington, DC
- OGWDW, Cincinnati OH.
- ORD, Cincinnati, OH.
- OGWDW Washington DC
-OGWDW, Washington, DC
-OGWDW, Washington, DC
-ORD, Cincinnati, OH
-OGWDW, Washington, DC
Disclaimer
The mention of a product name or company does not constitute official USEPA endorsement of
the product or company.
111
-------
Executive Summary
The United States Environmental Protection Agency (EPA) office of water in compliance with
the Safe Drinking Water Act (SDWA) is responsible for developing regulations to protect the
nation's drinking water supply from drinking water contaminants.
EPA has proposed a groundwater rule which will require states to determine groundwater
systems that are vulnerable to fecal contamination. Studies were conducted by EPA on virus
fecal indicator occurrence across the U.S. A round robin testing for proposed coliphage indicator
has also been conducted.
A three year field study was conducted by Office of Science and Technology under the overall
supervision of Dr. mark Sobsey to determine the performance of method 1601 and 1602 in
detecting somatic and male-specific Coliphages in groundwater. In addition, Method 1601 and
1602 were tested using the confirmation procedure, as proposed by EPA, for all methods to
detect microbes in groundwater.
The investigation was conducted in four different regional aquifers across the united States.
These aquifers were in the Southeast region, the Northeast region, the south west region, and the
upper Midwest region.
Results obtained, show that coliphages can be used as a tool for screening groundwater samples
for the presence of fecal contamination. However the results show that there was no direct
correlation of the presence of human enteric viruses and the presence of viral indicators.
The inclusion of coliphages along with conventional bacterial indicator analysis increases the
likelihood for detection of fecally contaminated samples. The absence of detection of human
enteric viruses in the presence of viral indicator suggest that the presence of pathogens may not
routinely be detected unless under heavily contaminated conditions.
IV
-------
Table of Contents
Cover page i
Acknowledgment ii
Disclaimer iii
Executive Summary iv
INTRODUCTION 5
Background 5
Experimental Approach 6
Coliphages and their Detection Methods 6
EPA Methods for Coliphage Detection in Ground water 8
Method 1601 9
Method 1602 10
Confirmation of Positive Results by Methods 1601 and 1602 12
Simultaneous Detection of both Somatic and Male-specific
Coliphages on a Single Host 12
Survival of Coliphages in Groundwater 13
Field Application of Methods 1601 and 1602 to Detection of
Coliphages as Indicators of Fecal Contamination in
Vulnerable Groundwater 14
-------
PHASE I STUDIES 15
PURPOSES, GOALS, AND TASKS OF PHASE I STUDIES 15
PHASE I METHODS AND MATERIALS 18
Method 1601 18
Method 1602 19
PHASE I RESULTS 20
Coliphage Recovery by Method 1602 20
Confirmation of Plaques in Method 1602 25
Summary of Method 1602 Results 29
Coliphage Recovery by Method 1601 30
Summary of Method 1601 Results 36
PHASE I CONCLUSIONS 38
PHASE II 40
Statement of Work: Coliphage Method 1601 and 1602
Validation and Field Testing 40
Background 40
Purpose and Objectives of the Study 41
Specific Objectives 41
Schedule of Deliverables 45
PHASE II METHODS AND MATERIALS 48
Groundwater Samples and Wells 48
Coliphage Analysis of Groundwater 52
-------
Coliphage Isolate Characterization 53
Bacteriological Analysis of Groundwater 55
Analysis of Groundwater for Human Enteric Viruses 57
Primary virus concentration from groundwater 57
Virus isolation in cell cultures 58
Virus detection by nucleic acid amplification 61
Nucleic acid amplification by (RT-)PCR 62
PHASE II RESULTS AND DISCUSSION 75
Introduction 75
Results of Field Sample Analysis of Coliphage and
Bacterial Indicators in Groundwater 75
Comparative Detection of Two Indicators in Groundwater Samples 80
Statistical Comparisons of Fecal Indicators in Groundwater Samples 89
Analysis for Enteric Viruses in Groundwater 91
Survival of Coliphage in Seeded Groundwater 92
Comparison of Coliphage, Bacterial Indicator and
Enteric Virus Detection in This Study and in Previous
Studies in the USA 95
Responses to Questions and Comments of the
April 2004 Coliphage Workshop 98
SUMMARY AND CONCLUSIONS 102
REFERENCES 105
-------
APPENDIX I: Summary Report of the Northeast Region 108
APPENDIX H: Summary Report of the Southwest Region 122
APPENDIX IE: Summary Report of the Upper Midwest 135
-------
INTRODUCTION AND BACKGROUND
Background
This report summarizes the results of Phase I and phase studies on a project to evaluate and, if
necessary further improve, EPA Methods 1601 and 1602 to detect coliphages in groundwater. In
the first phase of the study samples of groundwater were seeded with known quantities of
naturally occurring coliphages from sewage and the recovery efficiency of the methods in
detecting these added coliphages was determined in a series of controlled experiments performed
concurrently by four participating laboratories located in different regions of the country. The
data from the seeded sample recovery experiments were used to further establish and quantify the
performance characteristics of the methods.
In the second phase of the study EPA Methods 1601 and 1602 were applied to geographically
representative samples of groundwater potentially vulnerable to fecal contamination in order to
compare the performance of the different coliphage methods and to compare their ability to
detect fecally contaminated groundwater relative to the detection of fecal indicator bacteria and
the detection of culturable enteric viruses. Each of the four geographically representative
laboratories (southeast, northeast, upper Midwest and southwest) was to analyze at least 16
groundwater samples for coliphages, indicator bacteria and enteric viruses, for a total target
number of 64 samples to analyzed for second and final phase of the project. Several other tasks
were linked to this effort to further validate and improve the coliphage methods and their ability
to detect and characterize coliphages in groundwater
-------
Experimental Approach
Coliphages and their detection methods. Coliphages are viruses infecting Escherichia coli
bacteria. Coliphages are present at high concentrations in sewage and other fecal wastes and they
are indicators of fecal contamination of groundwater, other waters and other environmental
media. There are two main groups of Coliphages: somatic and male-specific. The relationships
between these Coliphages and their host bacteria, showing specific bacterial strains as examples,
are summarized in Figure 1. The conventional method to detect coliphages is by their ability to
infect host cells in which they replicate (proliferate), producing large numbers of progeny viruses
and lysing (killing) the host cells in the process. It is this killing and lysis of host cells that forms
the basis of most coliphage infectivity assay methods, including those employed for coliphage
analysis by the EPA methods.
Figure 1. Somatic and Male-specific (F+) Coliphages and their Relationship to Host Bacteria
Somatic Coliphage
F+ Coliphage
F+ Coliphage
£co//C3000
F-plaamid
Combined Host
F+ Coliphage
Somatic
Coliphage
Somatic Coliphage
-------
Somatic coliphages infect host bacteria by attaching directly to the outer cell wall (outer cell
membrane). The male-specific coliphages infect only male F+ strains of bacteria by attaching to
the hair-like appendages projecting from the cell surface, called F-pili or fimbrae, that are the
characteristic male trait. Somatic coliphage hosts lack the F-pili and cannot be infected by F+
coliphages. F+ coliphage hosts differ in their ability to be infected by somatic coliphages. Some
F+ coliphage hosts are very resistant to somatic coliphage infection because they have an outer
cell membrane that differs from those of E. coli (such as the Salmonella typhimurium strain
WG49) and E. coli Famp (which was experimentally selected as a somatic-coliphage resistant
mutant). Other F+ coliphage hosts such as E. coli C3000 have not been subjected to selection for
resistance to somatic coliphages and are susceptible to F+ coliphage infection as well as somatic
coliphage infection. Therefore, some host bacteria are infected only by somatic coliphages (E.
coli C and CN13), others only by male-specific coliphages (E. coli Famp and Salmonella
typhimurium WG49) and yet others by both groups of coliphages (E. coli C3000).
There are still questions about which groups of coliphages, somatic, male-specific or both groups
together, are the appropriate and preferred indicators of fecal contamination. There is evidence
in support of both somatic and male-specific coliphages as being effective and useful virus
indicators of fecal contamination. Some have suggested that both somatic and male-specific
coliphages should be detected as fecal indicator viruses of contamination of groundwater and
other waters. It is the understanding of the authors that EPA has so far not made any final
decisions about which of the coliphage groups to target for detection in future guidelines or
regulations. It also been suggested that both groups of coliphages, somatic and male-specific,
-------
could be simultaneously detected on a single coliphage host, thereby giving the greatest
probability and highest sensitivity in detecting any coliphage indicative of fecal contamination.
EPA Methods for Coliphage Detection in Ground Water. EPA Methods 1601 and 1602 were
developed to detect somatic and male specific coliphages in large volumes of groundwater, with
target sample volumes of up to 1000 mL in Method 1601 (an enrichment method) and 100 mL in
method 1602 (a Single Agar Layer plaque assay method), respectively. The methods are based
upon the ability of the coliphages to infect host bacteria, which results in the lysis of the host
bacteria. This a widely used approach to detect coliphages. In plaque assays or other assays on
solid media, such as those containing agar, the lysis of the host bacteria is visualized as zones of
lysis or clearing of the bacteria as discrete, circular areas (called a lysis zones or plaques) in a
confluent layer (or "lawn") of host bacteria in a solid nutrient medium. In liquid enrichment
cultures in broth media, the lysis of host bacteria can in principle be observed as the clearing of
turbidity from the culture as the bacteria are lysed and their resulting cell debris settles out of
suspension. Because such clearing of broth cultures as evidence of host cell lysis can be hard to
observe due to interference from other bacteria that may grow in the broth culture, other ways to
confirm the presence of phages are often used. One of the most common ways is to take some of
the enrichment culture containing phages, apply it to a lawn of bacteria in an agar medium, and
allow the phages to infect and lyse the host cells in the lawn to produce a clear zone of lysis that
can be readily observed.
-------
Method 1601. Method 1601 is a so-called two-step "enrichment" method and the steps of the
method are outlines in Figure 2. In the first step of this method, liquid bacterial media,
magnesium chloride (to promote coliphage attachment to the host bacteria), and the E. coli host
are added to the water sample, making a liquid (broth) culture for coliphage infection of the E.
coli host bacteria. After allowing for coliphage infection and lysis of the host bacteria during
overnight incubation, a small volume (several micro liters) of the enrichment culture is placed on
the surface of a Petri dish of agar medium containing E. coli host bacteria (a spot). This is the
second step of the method. If the applied sample contains coliphages able to infect the host
bacteria, a circular zone of host cell lysis (clearing) develops after several hours of incubation in
the spot where the sample was applied. Such a lysis zone in the spot indicates coliphage
presence in the enrichment broth and is a positive result. If no such lysis zone develops in the
sample spot on the plate, the enrichment culture of the sample is considered negative for
coliphages.
-------
Figure 2.
Method 1601 - Two-Step Enrichment-Spot Plate Method for Coliphage Presence-Absence
GroundwaterSample:
A
Add MgCfe, E coli host and broth culture medium
.1
Incubate overnight at 37SC
I
Remove a small volume of enrichment culture and spot onto surface of Petri
plate containing agar medium and E. co//host
I
Appearance of lysis zone in the enrichment spot, indicating coliphage presence
When Method 1601 is applied to a single sample volume, the analysis provides a determination
of the presence or absence of coliphages in the sample volume analyzed. If the method is applied
to multiple sample volumes, each in separate enrichment cultures, the method is capable of
giving an estimation of the concentration of coliphages in the water sample, based on which
sample enrichment volumes become positive and negative for coliphages.
Method 1602. EPA Method 1602 is a so-called single agar layer method for the enumeration of
coliphage plaques (discrete clear zones of lysis of host bacteria) developing in a culture of host
bacteria in an agar medium in a Petri dish. As shown in Figure 3, a 100-mL sample of
10
-------
groundwater is supplemented with magnesium chloride, host bacteria and then combined with
molten agar medium. The mixture is then distributed into Petri plates, the agar medium is
allowed to solidify and the plates are incubated overnight for the development of coliphage
plaques, which are clear, circular zones of lysis, each produced by a separate or individual
coliphage. The plaques are then counted to determine the total number of number of coliphages
in the sample, assuming each plaque arose from an individual infectious coliphage.
Figure 3.
Method 1602 - Single Agar Layer (SAL) Plaque Assay Method for Coliphage Enumeration
Groundwater Sample
Add MgCI2, E. coli host and molten agar culture medium
•1
Mix briefly and then distribute contents into several Petri dishes
.1
Incubate at 3G:;C overnight
I
•I
Count coliphage plaques
(clear zones of lysis appearing in agar medium-host mixture)
11
-------
Confirmation of Positive Results by Methods 1601 and 1602. For both Methods 1601 and
1602, EPA has proposed a method to confirm positive results. For confirmation of positive
results, material is removed (picked or aspirated with a capillary pipette or a micropipettor) from
the lysis zones of enrichment spots on agar medium-host cell plates (Method 1601) or from the
plaques that develop in agar medium-host cell plates of Method 1602. The recovered material is
transferred to a small volume of buffered water, mixed briefly, and then a small volume (several
microliters) of the material is placed ("spotted") on the surface of a Petri dish of agar medium
containing E. coli host bacteria. If the applied material contains coliphages capable of infecting
the host bacteria, a circular zone of host cell lysis develops after incubation (for several hours or
overnight) in the spot where the sample was applied. Such a lysis zone in the spot is indicative
of coliphage presence in the material recovered from either a lysis zone on the spot plate of an
enrichment broth (Method 1601) or from the plaque of a Single Agar Layer plate (Method 1602).
If no such lysis zone develops in the sample spot on the confirmation plate, the sample
(presumptive lysis zone from an enrichment culture or presumptive plaque from an SAL plate) is
considered negative for coliphages.
Simultaneous Detection of both Somatic and Male-specific Coliphages on a Single Host.
EPA Methods 1601 and 1602 were originally developed to separately detect somatic and male-
specific coliphages using separate E. coli hosts able to support the growth of only one or the
other coliphage group (somatic or male-specific, respectively). E. coli CN13 is used to detect
somatic coliphages andE. coli Famp is used to detect male-specific coliphages (Figure 1). It was
later suggested that perhaps a single E. coli host could be used to simultaneously detect both
12
-------
somatic and male-specific coliphages present in a groundwater samples rather than having to use
two separate E. coli hosts to separately detect each coliphage group (Figure 1). If the presence of
either or both groups of coliphages indicates fecal contamination, simultaneous detection of both
on one host would reduce time, effort, materials and cost and provide appropriate data about
coliphage presence in a sample. As previously noted, E. coli C3000 is such a host. However use
of a single E. coli host bacterium capable of detecting both somatic and male-specific coliphages
had not been adequately tested for its performance characteristics in previous studies on the
development and evaluation of Methods 1601 and 1602 and their application to either seeded
samples or field samples of groundwater.
Survival of Coliphages in Groundwater. In the development and evaluation of methods for
coliphage detection in groundwater, the question has been raised as to how long samples can be
held before being subjected to analysis. It has been suggested that samples may have to be
collected and sent to a distant lab capable of coliphage analyses, but that the time between
sample collection and analysis may be more than 1 or 2 days. If the sample holding time is 2 or
more days will the coliphages still be present and be detectable? To address this question
additional experiments were done as an added task in Phase II of this study at the request of Dr.
Nena Nwachuku, the EPA project manager. Groundwater was seeded with known, low level
amounts of mixed populations of sewage-derived coliphages and aliquots of these samples were
subjected to coliphage analysis by Methods 1601 and 1602 on days 0, 2, 3 and 6. These assay
days were chosen to model those that might be used if samples were shipped to a lab for
coliphage analysis and even held overnight before analysis once received by the lab. The
13
-------
resulting data on coliphage concentrations were analyzed to determine if the coliphages were
stable and still detectable for periods ranging from 1 to 6 days.
Field Application of Methods 1601 and 1602 to Detection of Coliphages as Indicators of
Fecal Contamination in Vulnerable Groundwater. An important test of the newly developed
EPA methods to detect coliphages in groundwater, Methods 1601 and 1602, would be to validate
their performance for coliphage detection in vulnerable groundwater, in comparison with the
detection of fecal indicator bacteria and human enteric viruses in the same samples. Preferably
such studies would apply the methods to different, geographically representative groundwater in
order to make sure that the methods were not adversely affected by interfering constituents in the
groundwater, or so-called "matrix effects". Furthermore, the concurrent detection of coliphages
by Methods 1601 and 1602 in the same groundwater samples would provide an opportunity to
compare their relative detection sensitivities and lower limits of coliphage detection. In addition,
the concurrent detection of coliphages as well as fecal indicator bacteria and enteric viruses in the
same groundwater samples would make it possible to determine if coliphages were as good or
better than fecal indicator bacteria or enteric viruses in identify fecally contaminated ground
water. Such analysis would make it possible to determine if one of these microbe groups was a
superior indicator of fecal contamination because it was detected more frequently and/or at
higher concentrations. Such analyses were done in Phase II of this study.
14
-------
PHASE I STUDIES
PURPOSES, GOALS AND TASKS OF PHASE I STUDIES
The overall purposes and goals of Phase I studies were to determine the performance
characteristics of Methods 1601 and 1602 in detecting and quantifying somatic and male-specific
coliphages in ground water samples seeded with known quantities of natural, mixed populations
of coliphages obtained from municipal sewage. These studies were done using certain
modifications and additions to Methods 1601 and 1602 in order to address recommendations
suggested for the methods after their original development, evaluation and multi-laboratory
testing. Specifically, host E. coli C3000 was tested for simultaneous detection and quantification
of both somatic and male-specific coliphages in addition to testing the methods with the
individual hosts previously specified for separate detection of somatic (E. coli C3000) and male-
specific (E. coli Famp) coliphages. In addition, Methods 1601 and 1602 were tested using the
confirmation procedure, as proposed by EPA for all methods to detect microbes in ground water.
The key tasks and activities of the Phase I studies are listed below.
1. Recruit a total of 4 experienced laboratories, each from a different region of the country, to test
Methods 1601 and 1602 using the standard protocols with the modifications indicated: a)
include host E. coli 3000 for simultaneous detection of both somatic and male-specific
coliphages, and b) include confirmation of presumptive positive results obtained from samples.
15
-------
The 4 laboratories are:
University of North Carolina (UNC), under the direction of Mark D. Sobsey (southeast)
University of New Hampshire (UNH), under the direction of Aaron Margolin (northeast)
Texas A&M University (TAMU), under the direction of Suresh Pillai (southwest)
Wisconsin State Hygiene Lab (WSHL), under the direction of David Battigelli (upper
Midwest)
2. Develop bench sheets (bench laboratory aids or protocols in easy-to-follow format) to be used
by analysts performing the methods in these repeated, weekly experimental trials.
3. Perform weekly experimental tests (trials) of the methods using the developed bench sheets.
4. Test each method (1601 and 1602) simultaneously by the 4 laboratories on a weekly basis,
using locally collected ground waters seeded with the same stock of sewage-derived coliphages
prepared and distributed weekly by the lead or reference laboratory (UNC) and all three E. coli
hosts (CN13 for somatic, Famp for male-specifics and C3000 for both).
5. Perform repeated trials of each method and submit the results to the lead (UNC) laboratory for
compilation and data analysis in order to develop and evaluate a sufficient database to
characterize the performance of the methods.
16
-------
6. Identify any deficiencies or limitations encountered with the methods. If possible within a
short time period (no more than a few weeks), devise and implement modifications or corrective
measures to improve the performance characteristics of the methods.
7. Based on the compiled data from the 4 laboratories, determine if the performance
characteristics of the methods are of sufficient quality to recommend the use of the methods to
detect coliphages in ground water samples.
8. Save (archive) representative coliphages detected by each method on each E. coli host for
further characterization by the UNC laboratory to determine if the coliphage isolates have
properties consistent with a fecal origin. These properties include bacterial host range, growth
temperatures and taxonomic group (sub-set of representative isolates only).
17
-------
PHASE I METHODS AND MATERIALS
The methods and materials used in this project are those specified in the documents for US EPA
Methods 1601 and 1602. Stepwise procedural steps in the application of these methods for the
specific purpose of this study are also given in the laboratory bench sheets (laboratory bench
protocols) presented in the Appendix to this report. The only departures or modifications to
Methods 1601 and 1602 employed in this study are: (a) the addition of E. coli C3000 as a host
bacterium for the simultaneous detection of both somatic and male-specific coliphages, and (b)
the addition of the newly proposed confirmation procedure for plaques from plates of Method
1602 and from lysis zones of plates from Method 1601.
Method 1601
For Method 1601, the two-step enrichment method, the goal was for each of the 4 participating
laboratories to seed 30+ liters of ground water with a quantity of coliphage stock (filtered
sewage) to achieve between 1 and 2 infectious units of coliphages per liter of water. The seeded
water was then aliquotted into 30 1-liter volumes. Groups of 10 1-liter volumes were subjected
to the enrichment assay method using one of the three host bacteria, thereby testing each host
bacterium for coliphage detection using 10 replicate 1-liter volumes per host bacterium per
weekly experiment. As negative control samples, three additional 1-liter volumes of unseeded
ground water were also subjected to coliphage analysis by the two-step enrichment method using
each of the three different E. coli host bacteria. As negative controls, these samples were
intended to demonstrate no background level of coliphages were present in the ground water
18
-------
prior to seeding with sewage-derived coliphages. A total of 8 replicate experiments were
conducted, one experiment per week, between May and July, 2001.
Method 1602
For Method 1602, the single agar layer (SAL) method, the goal was for each of the 4
participating laboratories to seed replicate 300-mL volumes of water with a quantity of coliphage
stock (filtered sewage) to give about 100 infectious units of coliphages per 100 mL of ground
water. The seeded water was aliquotted as 3 100-mL volumes and each of these volumes was
assayed by the single agar layer method using one of the three different E. coli host bacteria. As
negative controls, 3 100-mL volumes of unseeded ground water were subjected to coliphage
analysis by the SAL method using each of the three different host bacteria. As negative controls,
these samples were intended to demonstrate no background level of coliphages were present in
the ground water prior to seeding with sewage-derived coliphages. A total of 10 replicate
experiments were performed, once experiment per week, during February and April, 2001.
19
-------
PHASE I RESULTS
Coliphage Recovery by Method 1602
Table 1 shows the recovery of seeded coliphages by method 1602 (single agar layer assay) for the
total of 10 successive trials performed weekly. In some initial weekly trials no data were
available from the WSLH laboratory. This was due to other obligations that precluded their
participation. In the interest of time, the initial three experiments were performed among the
other three laboratories in order to initiate the project and to begin addressing potential logistical
issues of coordination among laboratories. No serious logistical problems arose among the three
labs participating initially. This indicated a reliable system for concurrent method performance
among the labs using the same coliphage stocks prepared by UNC lab to seed test groundwater.
The WSLH also was unable to participate another week due to a state-mandated holiday.
20
-------
Table 1: Recovery of Seeded Coliphages in 100-mL Groundwater Samples by Method 1602
(Single Agar Layer Assay)
Date
21-Feb-Ol
27-Feb-Ol
6-Mar-Ol
13-Mar-Ol
20-Mar-Ol
27-Mar-Ol
3-Apr-Ol
10-Apr-Ol
17-Apr-Ol
24-Apr-Ol
Host
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
UNC
2%
80%
246%
9%
43%
22%
9%
37%
101%
9%
19%
72%
12%
20%
91%
4%
73%
48%
10%
30%
49%
5%
21%
26%
44%
139%
72%
33%
117%
36%
TAMU
14%
90%
160%
79%
43%
15%
6%
19%
18%
0%
0%
0%
8%
9%
68%
8%
30%
20%
16%
36%
47%
17%
63%
10%
no data
no data
no data
55%
88%
33%
WSLH
no data
no data
no data
no data
no data
no data
no data
no data
no data
20%
35%
28%
56%
94%
92%
39%
120%
40%
28%
70%
67%
34%
64%
32%
49%
77%
94%
32%
76%
26%
UNH
22%
65%
24%
131%
127%
53%
45%
50%
35%
26%
23%
33%
37%
39%
45%
81%
118%
70%
88%
77%
72%
93%
96%
77%
86%
84%
98%
77%
89%
84%
The percent coliphage recovery data were subjected to Analysis of Variance (ANOVA) to
discover if there were significant recovery differences among the hosts and/or the laboratories
(Table 2) . As shown in Table 2, there were significant differences in recovery among the 3 hosts
(p=0.00008), and significant differences in recovery among the four labs (p=0.000002). The
highest recovery (73%) was obtained using E coli CN13, the lowest (39%) was obtained using E
coli C3000 and an intermediate recovery of 46% was obtained with E coli Famp. The
differences among host bacteria were consistent (not significantly different) within each
21
-------
laboratory (p=0.42). This latter result suggests that recoveries among the three host bacteria are
generally similar within a lab and therefore, the hosts are equivalent on a within-lab basis. In
other words, the three different E. coli hosts will give similar recovery efficiencies when used by
an individual lab to analyze mixed populations of coliphages of sewage (fecal) origin in a
ground water matrix.
Table 2: Descriptive Statistics for Seeded Coliphage Recovery by Method 1602
Mean
Median
Mode
Std. Dev.
Var.
Minimum
Maximum
Count
95% CI
Mean
Median
Mode
Std. Dev.
Var.
Minimum
Maximum
Count
95% CI
OVERALL
UNC
49%
35%
NONE
52%
27%
2%
246%
30
19%
TAMU
35%
19%
0%
37%
14%
0%
160%
27
15%
WSLH
56%
49%
NONE
28%
8%
20%
120%
21
13%
UNH
68%
75%
NONE
31%
10%
21%
131%
30
12%
CN13
UNC
58%
40%
NONE
43%
18%
19%
139%
10
31%
TAMU
42%
36%
NONE
33%
11%
0%
91%
9
25%
WSLH
77%
76%
NONE
26%
7%
35%
120%
7
24%
UNH
77%
81%
NONE
33%
11%
23%
127%
10
24%
C3000
UNC
14%
9%
NONE
14%
2%
2%
44%
10
10%
TAMU
23%
14%
NONE
26%
7%
0%
79%
9
20%
WSLH
37%
34%
NONE
12%
2%
20%
56%
7
11%
UNH
69%
79%
NONE
35%
12%
22%
131%
10
25%
Famp
UNC
76%
61%
NONE
65%
43%
22%
246%
10
47%
TAMU
41%
20%
NONE
49%
24%
0%
160%
9
38%
WSLH
54%
40%
NONE
30%
9%
26%
94%
7
28%
UNH
59%
62%
NONE
25%
6%
24%
98%
10
18%
The differences in coliphage recovery efficiency among the laboratory groups led us to question
whether there were differences in the groundwater of each region (i.e., a "matrix" effect) which
might account for those observed differences in recovery. Further experiments were conducted
using regional groundwater and additionally reagent water (as a control measure) in an attempt to
answer this question. Four replicate experiments were conducted and these data are summarized
22
-------
in Table 3, with descriptive statistics in Tables 4a, 4b and 4c for hosts E. coli C3000, CN13 and
Famp, respectively..
Table 3: Coliphages Recovery Efficiency of Method 1602 Concurrently Applied to Seeded
Groundwater and Reagent Water
Date
3-Apr-Ol
10-Apr-Ol
17-Apr-Ol
24-Apr-Ol
Matrix
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
Host
C3000
C3000
CN13
CN13
Famp
Famp
C3000
C3000
CN13
CN13
Famp
Famp
C3000
C3000
CN13
CN13
Famp
Famp
C3000
C3000
CN13
CN13
Famp
Famp
UNC
10%
51%
30%
61%
49%
64%
5%
52%
21%
76%
26%
66%
44%
73%
139%
135%
72%
87%
33%
31%
117%
103%
36%
59%
TAMU
16%
22%
36%
49%
47%
88%
17%
13%
63%
66%
10%
18%
no data
no data
no data
no data
no data
no data
55%
13%
88%
48%
33%
49%
WSLH
28%
26%
70%
57%
67%
48%
34%
37%
64%
63%
32%
47%
49%
67%
76%
68%
94%
138%
32%
51%
76%
65%
26%
78%
UNH
88%
87%
77%
74%
72%
72%
93%
92%
96%
78%
77%
75%
86%
76%
84%
82%
98%
97%
77%
79%
89%
83%
84%
79%
23
-------
Table 4a: Descriptive Statistics: Coliphage Recovery by Method 1602 for Seeded Groundwater
and Reagent Water on Host E. coli C3000
Mean
Median
Std. Dev.
Var.
Minimum
Maximum
Count
95% CI
GROUNDWATER
UNC
23%
21%
18%
3%
5%
44%
4
29%
TAMU
29%
17%
22%
5%
16%
55%
3
55%
WSLH
36%
33%
9%
1%
28%
49%
4
14%
UNH
86%
87%
7%
0%
77%
93%
4
11%
REAGENT WATER
UNC
52%
52%
17%
3%
31%
73%
4
27%
TAMU
16%
13%
5%
0%
13%
22%
3
13%
WSLH
45%
44%
18%
3%
26%
67%
4
28%
UNH
84%
83%
7%
1%
76%
92%
4
12%
Table 4b: Descriptive Statistics: Coliphage Recovery by Method 1602 for Seeded Groundwater
and Reagent Water on Host E. coli CN-13
Mean
Median
Std. Dev.
Var.
Minimum
Maximum
Count
95% CI
GROUNDWATER
UNC
77%
73%
60%
36%
21%
139%
4
95%
TAMU
63%
63%
26%
7%
36%
88%
3
65%
WSLH
72%
73%
6%
0%
64%
77%
4
9%
UNH
86%
86%
8%
1%
77%
96%
4
12%
REAGENT WATER
UNC
94%
90%
32%
11%
61%
135%
4
52%
TAMU
54%
49%
10%
1%
48%
66%
3
25%
WSLH
63%
64%
5%
0%
57%
68%
4
8%
UNH
79%
80%
4%
0%
74%
83%
4
6%
Table 4c: Descriptive Statistics: Coliphage Recovery by Method 1602 for Seeded Groundwater
and Reagent Water on Host E. coli Famp
Mean
Median
Std. Dev.
Var.
Minimum
Maximum
Count
95% CI
GROUND
UNC
46%
43%
20%
4%
26%
72%
4
32%
TAMU
30%
33%
19%
4%
10%
47%
3
47%
WSLH
55%
49%
32%
10%
26%
94%
4
51%
UNH
83%
81%
11%
1%
72%
98%
4
18%
REAGENT
UNC
69%
65%
12%
2%
59%
87%
4
20%
TAMU
52%
49%
35%
12%
18%
88%
3
87%
WSLH
78%
63%
42%
18%
47%
138%
4
67%
UNH
81%
77%
11%
1%
72%
97%
4
18%
24
-------
These coliphage recovery data were subjected to ANOVA. No interaction effects were indicated
(p>0.09 in all cases), implying that any observed factor differences were consistent throughout
the experiment. As expected, there were significant differences in recovery among the host
bacteria (p= 0.00007), with recovery for E. coli C3000 (45%) being lower than recoveries for E.
coli CN13 (73%) and E. coli Famp (63%). There were also significant differences in recovery
among the laboratory groups (p=0.0000006), with UNH having the highest overall recovery
(83%), followed by UNC (60%), WSLH (58%), and TAMU showing the lowest overall recovery
(40%). But there was no significant difference due to a possible matrix effect (p=0.17), leaving
unexplained the previously observed differences among the laboratory groups.
Confirmation of Plaques in Method 1602
The laboratories also applied several modified versions of a confirmation procedure for plaques
isolated from the SAL plates of Method 1602 when applied to the detection of coliphages in
seeded water samples. Plaques were picked from the SAL plates using a variety of methods (i.e.,
with Pasteur pipettes, with Eppendorf pipettes, etc.), resuspended in Tryptic Soy Broth, and
spotted onto pre-poured gridded plates of Tryptic Soy Agar containing host bacteria (as used in
the Two-Step Enrichment procedure). The confirmation percentages are presented in Table 5a,
which summarizes all data by experiment date, lab and host and in Table 5b, and which
summarizes the descriptive statistics for the plaque confirmations. The results of these attempts
at coliphage plaque confirmation ranged from excellent (average 99-100% at UNH) to moderate
(38-68% at UNC). Overall, there are high likelihoods that the plaques detected on assay plates
for Method 1602 are indeed produced by coliphages, with a 78% average or a nearly 4 out of 5
25
-------
plaque confirmation rate. It is likely that confirmation rates can be further improved to give a
greater confirmation efficiency by minor modifications in the plaque recovery and re-spotting
procedure.
26
-------
Table 5a: Percent Confirmation of Picked Plaques Isolated using Method 1602
Date
21-Feb-Ol
27-Feb-Ol
6-Mar-Ol
13-Mar-Ol
20-Mar-Ol
27-Mar-Ol
3-Apr-Ol
10-Apr-Ol
17-Apr-Ol
24-Apr-Ol
Matrix
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
ground
reagent
Host
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
C3000
CN13
CN13
Famp
Famp
C3000
C3000
CN13
CN13
Famp
Famp
C3000
C3000
CN13
CN13
Famp
Famp
C3000
C3000
CN13
CN13
Famp
Famp
UNC
33%
85%
15%
13%
75%
15%
60%
45%
30%
21%
35%
15%
78%
50%
50%
40%
80%
35%
50%
65%
45%
75%
0%
0%
93%
100%
15%
90%
10%
60%
80%
95%
95%
70%
70%
70%
90%
90%
100%
95%
85%
80%
TAMU
not done
no data
not done
not done
no data
not done
80%
100%
88%
not done
no data
not done
no data
25%
0%
no data
0%
53%
70%
90%
80%
90%
20%
10%
80%
80%
90%
80%
50%
30%
not done
not done
no data
not done
not done
not done
100%
100%
100%
100%
80%
90%
WSLH
95%
100%
90%
100%
80%
45%
100%
60%
67%
100%
75%
80%
95%
90%
80%
100%
80%
95%
100%
No data
95%
No data
100%
No data
100%
100%
85%
95%
90%
100%
100%
100%
85%
100%
100%
95%
100%
100%
100%
100%
95%
95%
UNH
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
95%
95%
90%
100%
95%
95%
100%
95%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
100%
27
-------
Table 5b: Descriptive Statistics for Plaque Confirmation of Coliphages Isolated by Method 1602
(Percent Confirmation Overall and by E. coli Host for each Laboratory)
Mean
Median
Mode
Std. Dev.
Variance
Minimum
Maximum
Count
95% CI
Mean
Median
Mode
Std. Dev.
Variance
Minimum
Maximum
Count
95% CI
Overall
UNC
57%
63%
15%
31%
10%
0%
100%
42
10%
TAMU
67%
80%
80%
34%
11%
0%
100%
25
14%
WSLH
91%
95%
100%
13%
2%
45%
100%
39
4%
UNH
99%
100%
100%
2%
0%
90%
100%
42
1%
CN13
UNC
68%
75%
75%
26%
7%
15%
100%
14
15%
TAMU
74%
90%
100%
36%
13%
0%
100%
9
28%
WSLH
88%
90%
100%
12%
1%
60%
100%
13
7%
UNH
100%
100%
100%
1%
0%
95%
100%
14
1%
C3000
UNC
65%
71%
90%
29%
9%
13%
100%
14
17%
TAMU
86%
80%
80%
11%
1%
70%
100%
7
10%
WSLH
99%
100%
100%
2%
0%
95%
100%
13
1%
UNH
99%
100%
100%
3%
0%
90%
100%
14
2%
Famp
UNC
38%
33%
15%
30%
9%
0%
85%
14
18%
TAMU
47%
50%
NONE
34%
12%
0%
90%
9
26%
WSLH
87%
95%
95%
16%
3%
45%
100%
13
10%
UNH
99%
100%
100%
2%
0%
95%
100%
14
1%
Subsequent efforts to improve confirmation rates for plaques picked from SAL plates or lysis
zones picked from spot plates of the enrichment method compared the original EPA
conformation method described above to a modified method. In the modified method the picked
material from plaques or lysis zones was re-enriched by culturing in host bacteria again. The
picked plaque or lysis zone material was transferred to 5 mL of Tryptic Soy Broth to which had
been added host bacteria and the mixture was incubated overnight at 37°C. Volumes of 10
microliters were withdrawn from the resulting overnight enrichments and spotted onto prepoured
lawns of host cells in nutrient agar media. The spot plates were incubate at 37°C for a minimum
for 4 hours and then observed for lysis zones indicative of coliphage positivity. The results of a
side-by-side comparison of the original confirmation method with the modified method are
28
-------
summarized in Table 5c.
Summary of Method 1602 Results
Results from a series of 10 replicate experiments by all four participating laboratories on the
performance of the SAL method have been presented. The method was applied to replicate 100-
mL volumes of groundwater seeded with sewage coliphages and detected with each of three E.
coli host bacteria. The summarized results of these experiments (Table 6) show efficient
coliphage detection (average 53%) and confirmation (average 78%) in 100-mL volumes of
ground water. There were differences in recoveries based on the host used, and there were
unexplained differences in recovery among the laboratories. Confirmation of plaque isolates
gave success rates ranging from moderate to excellent among labs. Individual adaptations or
modifications of confirmation methods somewhat improved low confirmation rates. Method
1602 gave generally acceptable detection of coliphages in seeded ground water and the majority
of plaques detected by the method could be easily confirmed by a simple procedure. Overall, the
results of these studies indicate that there is high likelihood of detecting even low levels of
coliphages in 100-mL volumes of ground water using Method 1602.
Table 6. Coliphage Detection in 100-mL Volumes of Seeded Ground Water by Method 1602
Coliphage Group
Somatic (E. co/z CN13)
Male-specific (E. coli Famp)
Both(£. co/zC3000)
Estimated
Phages/100 mL
100
100
100
Coliphage
Recovery (%)
64
58
38
Plaque Confirmation (%)
82
68
87
29
-------
Coliphage Recovery by Method 1601
The Two-Step Enrichment (SAL) validation study consisted of 8 replicate experiments
performed by the four participating laboratories. In each experiment a small volume of the
assayed sewage was added to a 30-liter volume of groundwater and mixed well to disperse the
inoculum evenly. This inoculated groundwater was then dispensed into 30 1-liter bottles to
which were added the enrichment media and the proper host bacteria (10 bottles per host). The
bottles were incubated overnight, and small portions were spotted onto gridded TSA plates as
described above. After incubation, these plates were examined for zones of lysis. Positive zones
of lysis were considered positive for coliphage, and these were counted and recorded for each
host. Based on the volume and the titer of the inoculated sewage, an expected coliphage titer per
bottle was calculated for each host. Based on the number of positive bottles, the MPN (Most
Probable Number) of coliphages per bottle was calculated using Thomas's MPN equation and
taken as the number of coliphages recovered (observed number of coliphages). Using this
observed MPN and the expected coliphage titer per bottle, percent recoveries were calculated.
The coliphage recovery rates for experiments in which replicate ten 1-liter volumes of
groundwater were seeded with about 1.5 to 3 infectious units of coliphages are presented in
Table 7a as percentage recoveries based on the observed (calculated) MPN coliphage
concentrations per liter and in Table 7b as a comparison of expected and observed number of
coliphage-positive 1-liter samples out of 10. Descriptive statistics presented in Tables 8a and
8b.
30
-------
Table 7a: Recovery of Seeded Coliphages in 1-Liter Groundwater Samples by Method 1601
(Two-step enrichment)
Date
9-May-Ol
14-May-Ol
21 -May -01
4-Jun-Ol
ll-Jun-01
18-Jun-Ol
25-Jun-Ol
9-Jul-Ol
Host
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
C3000
CN13
Famp
UNC
4%
134%
26%
30%
193%
>39%
10%
101%
26%
41%
372%
143%
20%
215%
44%
15%
>137%
13%
no data
18%
no data
261%
61%
<=518%
TAMU
no data
no data
no data
281%
92%
348%
no data
no data
no data
no data
no data
no data
79%
<=5%
<=14%
39%
163%
112%
125%
122%
45%
220%
191%
57%
WSLH
13%
>122%
19%
6%
>189%
7%
134%
148%
35%
23%
>359%
62%
24%
>204%
15%
28%
>124%
10%
6%
39%
56%
24%
187%
33%
UNH
>63%
131%
>70%
62%
9%
>21%
no data
no data
no data
122%
22%
99%
82%
304%
123%
130%
77%
21%
131%
220%
224%
108%
230%
2317%
31
-------
Table 7b: Descriptive Statistics for Recovery of Seeded Coliphages by Method 1601
Mean
Median
Std. Dev.
Variance
Minimum
Maximum
Count
95% CI
C3000
UNC
54%
20%
92%
84%
4%
261%
7
85%
TAMU
149%
125%
100%
100%
39%
281%
5
124%
WSLH
33%
24%
42%
18%
6%
134%
8
35%
UNH
109%
122%
27%
7%
62%
131%
7
25%
Mean
Median
Std. Dev.
Variance
Minimum
Maximum
Count
95% CI
CN13
UNC
171%
163%
117%
136%
18%
372%
8
97%
TAMU
114%
122%
74%
55%
0%
191%
5
92%
WSLH
296%
247%
208%
431%
39%
718%
8
174%
UNH
142%
131%
113%
127%
9%
304%
7
104%
Mean
Median
Std. Dev.
Variance
Minimum
Maximum
Count
95% CI
Famp
UNC
47%
26%
49%
24%
0%
143%
7
45%
TAMU
112%
57%
138%
190%
0%
348%
5
171%
WSLH
30%
26%
21%
4%
7%
62%
8
17%
UNH
424%
123%
838%
7018%
21%
2317%
7
775%
The results in Tables 7a and 7b indicate that when 1-liter volumes of ground water seeded with
1-2 PFU of coliphages are analyzed by the enrichment method, there is a very high likelihood
that coliphages will be detected with relatively high efficiency. Average coliphage recoveries
from 8 replicate trials per coliphage host per lab ranged were 86% for combined coliphages on
host E. coli C3000, 181% for somatic coliphages on host E. coli CN-13, and 153% for male-
specific coliphages on host E. coli Famp. The results in Tables 7a and 7b indicate variability in
coliphage recoveries from trial to trial. However, this extent of variability is to be expected
32
-------
because coliphage recoveries are based on MPN estimates of the number of positive 1-liter
enrichment culture bottles out often. Based on calculated 95% confidence intervals (CIs), the
observed degree of variability was within the range expected for a 10-culture, single dilution
Most Probable Number method. Examination of the 10-replicate single dilution MPN table in
Standard Methods for the Examination of Water and Wastewater indicates that MPN estimates
can have 95% CIs that vary by nearly 6-fold (600%) at low rates of positivity and almost always
3-fold (300%) or more at intermediate and high levels of positivity. Probably more important is
that of the total 82 trials in which 10 1-liter bottles of seeded ground water were used per trial,
coliphages were not detected in only 3 trials. Therefore, there is a very high probability of
detecting low levels of (1 to 3) coliphages in 1-liter volumes of groundwater when using this
method.
As in the statistical analyses for Method 1602 described earlier, the data for Method 1601 were
subjected to ANOVA. This analysis detected no significant differences in recovery among the
laboratories (p=0.38); nor did it detect any significant differences in recovery among the hosts
(p=0.41).
The results from this series of 8 replicate experiments per lab are also summarized in Table 8 and
Figure 1 based on the observed and expected number of positive 1-liter enrichment culture
bottles out of 10. These results indicate sensitive coliphage detection that is close to the
theoretical level of detection.
33
-------
Table 8. Comparison of Observed and Expected Coliphage Detection in 1-Liter Volumes of
Seeded Ground Water by Method 1601
Coliphage Group
Somatic (E. co/z CN13)
Male-specific (E. coli Famp)
Both(£. co/zC3000)
Estimated Coliphages/L
1.7
3.1
1.7
# Positive Bottles out of 10
Expected
8.2 (~8)
9.5 (9-10)
8.5 (8-9)
Observed (Average)
4.8(~5)
7.1 (~7)
5.7 (~6)
Figure 4. Coliphage Detection in 1-Liter Volumes of Seeded Ground Water by Method 1601
Somatic (E. coli CN13) Male-specific (E. coli Famp) Both (E. coli C3000)
nColiphage&'L • Expected Pos. (of 10} nObsened Pos. (at 10)
34
-------
Based on direct analysis (plaque assay) of the sewage-derived coliphage stocks added to the
ground water samples, the average concentrations of coliphages per 1-liter enrichment bottle
were: 1.7 for somatic coliphages (detected on host E. coli CN13), 3.1 for male-specific
coliphages (detected onE. coli Famp) and 1.9 for both groups of coliphages (detected on host E.
coli C3000). According to these estimated coliphage concentrations per liter of seeded ground
water, the expected numbers of positive enrichment bottles out of a total 10 enrichment bottles
per coliphage host are computed. These estimates of expected numbers of positive enrichment
bottles out of 10 were based on the principles of Poisson statistics (the Poisson Distribution and
Poisson probabilities). The estimates were computed directly from the Poisson probability
equation using the estimated coliphage concentrations per liter (based on direct assay of the
sewage coliphages seeded into groundwater as an estimate of the mean number of coliphages per
bottle. The expected numbers of positive bottles out of 10 were: about 8 for somatic coliphages,
9 to 10 for male-specific coliphages and 8 to 9 for both groups of coliphages. The
experimentally observed numbers of positive enrichment bottles out of 10 for each group of
coliphages are shown in Table 8 and Figure 1. Rounded to the nearest whole bottle, the numbers
of positive enrichment bottles out of 10 were: 5 for somatic coliphages, 7 for male-specific
coliphages and 6 for both groups of coliphages. These actual results indicate that the likelihood
of detecting coliphages when analyzing 1-liter volumes of ground water containing only about
1.5-3 coliphages per liter by the enrichment method are very high and very close to the
theoretically expected results. It is noteworthy that out of a total of 87 trials with 10 1-liter
enrichment bottles per trial there were only 4 occasions when all 10 bottles were negative. This
is well within the expected number of times when all 10 bottles would yield negative results,
35
-------
based on statistical considerations.
The expected probability of getting 10 negative bottles in a 10-bottle enrichment test when there
is an average of 1.5 to 3 infectious coliphages per bottle was actually higher than the observed
numbers of 10 negative enrichment bottles in a test. It should be noted that the detection of both
somatic and/or male-specific coliphages together on host E. coli C3000 was similar to the
detection of either somatic or male-specific coliphages on their respective E. coli hosts.
Although the numbers of positive enrichment bottles out of 10 were no higher on host E. coli
C3000 than on the other two hosts, the ability to simultaneously detect both groups of coliphages
using this host was not appreciably different than the detection of each coliphage group alone.
Based on the number of times that all 10 bottles in an enrichment test were negative, E. coli
C3000 was the same as or better than the other two hosts. The negativity rates per 30 trials of the
method were 1, 1 and 2 per 30 trials for E. coli C3000 (both), CN-13(somatic) and Famp (male
specific) respectively. These results indicate that E. coli C3000 can be successfully and reliably
used to simultaneously detect low levels of both somatic and male specific coliphages in 1-liter
volumes of ground water using the two-step enrichment method (Method 1601).
Summary of Method 1601 Results
In summary, recoveries of somatic, male-specific and total (somatic plus male-specific)
coliphages from 1-liter volumes of ground water were efficient using Method 1601. Coliphage
recoveries at input levels of about 1.5 to 3 infectious units per liter of ground water were
somewhat variable but close to those expected based on the infectivity titer of the sewage seed
36
-------
and to the expected number of positive 1-liter enrichment bottles out of a total of 10. The
observed variability of coliphage detection was no more variable than would be predicted for a
10-sample MPN test. Coliphage recoveries were not significantly different among E. coli hosts
and participating labs using ANOVA. Therefore, there is a high likelihood of detecting as few as
1-3 coliphages in 1-liter volumes of water using the two-step enrichment methods of Method
1601.
37
-------
PHASE I CONCLUSIONS
In comparing the two coliphage recovery and detection methods, the participating laboratory
groups tended to favor the Two-Step Enrichment over the Single Agar Layer method. This is
because the former test was considered easier to perform, sensitive in detecting low numbers of
coliphages and more consistent in its results. The Single Agar Layer (SAL) assay proved to be
cumbersome when assaying multiple samples, and the time constraints imposed by the method
were difficult to adhere to. The Two-Step Enrichment method is simpler to set up and much
easier to carry out. The statistical analyses showed it to be more consistent among different
laboratory groups. In addition, the results for the enrichment method showed that somatic and
male-specific coliphages can be detected simultaneously on a single host, E. coli C3000, at a
sensitivity comparable to detecting either somatic or male-specific coliphages individually. The
simultaneous detection of both somatic and male-specific coliphages simplifies the method as
well as lowers costs.
Further studies were done to characterize the performance of Methods 1601 and 1602 when
applied to the detection of coliphages in unseeded samples of fecally contaminated ground water
in Phase II of this study. The presence and concentrations of coliphages in field samples of
groundwater were compared and also were to also be compared to the presence and
concentrations of human enteric viruses in these fecally contaminated ground water samples.
This information was to be used to determine if somatic, male-specific and total coliphages are
38
-------
sensitive and reliable indicators of fecal contamination and the presence of human enteric viruses
in groundwater.
39
-------
PHASE II STUDIES
Statement of Work: Coliphage Method 1601 and 1602 Validation and Field Testing
Background. The US EPA's proposed Ground Water Rule may propose the examination of
ground waters for coliphages. Coliphages have been found to be reliable indicators of fecal
contamination of ground and surface water and of the fate of human enteric viruses in the
subsurface environment. Recently, two different EPA methods were developed to detect somatic
and male-specific coliphages in ground water. Method 1601 detects and quantifies coliphages by
liquid enrichment culture method and Method 1602 detects and enumerates coliphages by a
single agar layer (SAL) plaque assay. The original methods round robin studies analyzed for
somatic and male-specific coliphages separately, and did not evaluate E. coli C3000 as host
bacterium for detection and quantification of both somatic and male-specific coliphage. The
SAL method for coliphage detection did not require a confirmation step for the plaques that were
observed in the agar-host cell medium. There was some concern about the detection of "false
positives" based on simply scoring plaques. Therefore, the methods needed to be further
substantiated and validated and scientifically supportable to: (1) show that coliphages detected
by these methods are of likely fecal origin and have characteristics and properties consistent with
fecal origin, (2) reduce the cost and burden of using two different hosts to separately analyze
somatic and male-specific coliphage by measuring both simultaneously in a single E. coli host
bacterium, (3) include a simple confirmation step.
40
-------
Purpose and Objectives of the Study. The purpose of this study was to further validate EPA
method 1601 and 1602 and to test these methods in groundwater samples in four geographically
representative regional laboratories in the USA (North Carolina, Minnesota, New Hampshire and
Texas). The study was to determine if coliphages detected by Methods 1601 and 1602 can be
confirmed and show properties and characteristics consistent with a fecal origin. The labs that
conducted the studies were equipped and experienced to conduct the tasks required in the SOW.
The study was headed by a coliphage expert who has at least 20 years experience in coliphage
virology, who has experience in round robin testing, in developing research procedures and in
method 1601 and 1602. The expert participated in the original EPA round robin testing for
method 1601 and 1602, is knowledgeable about EPA programs, and the proposed Ground Water
Rule.
Specific Objectives. The specific objectives of the study were to: (1) conduct a field validation
of Methods 1601 and 1602 for coliphage detection, (2) determine the ability of coliphage
indicators in predicting the presence of human enteric viruses in the same ground water samples,
(3) confirm the presence of coliphages detected in ground water samples, (4) characterize the
properties of these coliphages to confirm that they are of likely fecal origin, (5) determine the
correlation and the reliability of such correlation in detecting fecal contamination and the
presence of human enteric viruses in ground water, and (6) address some key questions about
these coliphage methods and their use for coliphage detection in groundwater that arose in the
April 2004 EPA International Workshop on Coliphages as Indicators of Fecal Contamination in
Water and Other Environmental Media. An additional task was added to the study at the request
41
-------
of the EPA project manager. This task was: (7) to determine the stability or survival of
coliphages in groundwater samples that held for up to several days before analysis by Methods
1601 and 1602.
Task 1. Characterization and determination of properties of confirmed coliphage isolates.
A total of 800 coliphages (200 from each of the four participating laboratories) were to be
characterized of which 400 (100 from each laboratory) were to be from field samples and 400
(100 from each laboratory) were to be from sewage-seeded ground water samples used in
methods validation studies in Phase I. The contractor was to subject confirmed coliphage
isolates to the analyses described below.
(a) bacterial host range analyses. Determine the ability of test coliphage isolates to be grown in
both E. coli and non-E. coli coliform hosts and other bacteria by spotting onto pre-poured agar
medium-host cell lawns of the following 23 different host bacteria, if available: E. coli strains C,
CN13, C3000, K12F, K12F and Famp, S. typhimurium WG45 and WG49, Klebsietta
pneumoniae ATCC strains 23356 and 23357, Enterobacter cloacae ATCC strain 223355,
Citrobacter braakii (formerly Citrobacter freundil) ATTC strain6570 ATCC strain 12012,
Serratia marcescens ATCC strain 14764, Shigella sp. ATCC 23354, Shigella flexneri ATCC
12661, Yersinia pseudotuberculosis ATCC strain 23207, Proteus mirabilis ATCC strain 9921,
Yersinia enterocolitica ATCC strains 9610, 29913, Pseudomonas aeruginosa ATCC s train
12175, Aeromonas hydrophila ATCC strain 23211.
42
-------
(b) growth temperature range. Test the coliphage isolates for their ability to grow at
temperatures of 25, 36, 42 and 44.5°C on E. coli hosts.
(c )nucleic acid analyses. Examine coliphage isolates for taxonomy. Male specific coliphages
were be tested to determine the type of nucleic acid as either DNA or RNA.
The contractor shall analyze 32 geographically representative ground water samples (8 per
laboratory in four geographically representative laboratories) for coliphages and human enteric
viruses by combined cell culture and nucleic acid amplification methods.
Task 2. Cell culture RT-PCR or cell culture-PCR. The contractor shall analyze 32
geographically representative ground water samples (8 per laboratory in four geographically
representative laboratories) for coliphages and human enteric viruses by combined cell culture
and nucleic acid amplification methods.
Task 3. Coliphages and enteric viruses from groundwater. Each of the four participating
laboratories shall analyze an additional 8 samples of ground water for culturable human enteric
viruses and for coliphages. Four different ground waters shall be analyzed on two different
occasions. A total of 36 ground water samples shall be analyzed for coliphages and human
enteric viruses. The extent to which somatic and male-specific coliphages detected by Methods
1601 (in 1-liter sample volumes) and 1602 (in 0.1-liter sample volumes) are associated with the
occurrence of human enteric viruses in 100-1000-liter sample volumes of fecally contaminated
43
-------
ground water will be determined. The contractor shall statistically analyze data on the
occurrence of coliphages and human enteric viruses in field ground water samples to determine if
there is a co-occurrence and the extent to which they co-occur.
Task 4. Coliphage and bacterial analyses. Each lab will sample, process, and analyze ground
water by Methods 1601 and 1602 using hosts E. coli Famp, E. coli CN-13 and E. coli C3000
according to the established methods. In parallel to the coliphage analysis, each lab will sample
and process the same fecally contaminated ground waters by the EPA ICR methods for human
enteric viruses. Also, in parallel each lab will sample and process the same fecally contaminated
ground waters for E. coli and enterococci.
Task 5. Enteric virus analyses. Processed ground water samples will be analyzed for culturable
human enteric viruses by observation for cytopathogenic effects (CPE) in BGMK cells according
to EPA ICR method. In addition, the inoculated cell cultures also will be examined for non-
cytopathogenic enteroviruses, caliciviruses, adenoviruses, hepatitis A, rotaviruses, reoviruses by
combined cell culture and nucleic acid amplification methods, as previously described. The labs
also will analyze the samples for culturable human enteric viruses in CaCo2 cells. The data on
the occurrence and concentrations of coliphages and concentrations of human enteric viruses as
detected by CPE and by PCR will be statistically analyzed to determine co-occurrence with
coliphages.
Task 6. Report. The contractor shall prepare a consolidated draft report of all the data generated
44
-------
in all the tasks in all 4 laboratories and statistically analyzed. The report shall include an
interpretation of the results and recommendations to EPA. The report shall be submitted in 3
double spaced hard copies and a 31A diskette in WordPerfect, version 9/8.0 for Windows. A
summary fact sheet of the study and results shall accompany the draft report.
Task 7. Peer review. The contractor shall incorporate internal and external peer review
comments and a workshop input comments in a final report.
The EPA Work Assignment Manager will give technical direction in this study. The contractor
shall not cite, quote or distribute the results of this EPA study until EPA publishes it.
Publications from any aspect of this EPA research study will be subjected to EPA review and
will be published jointly with EPA. A monthly conference call shall be scheduled by the
contractor until completion of the study. An on site visit will be conducted by the EPAWAM on
a mutually acceptable date with the technical lead.
Schedule of Deliverables
The deliverables will include a consolidated 1st draft report with data from all the completed 7
tasks listed above and a final peer review report.
NOTE: These project objectives were not fully met due to extenuating circumstances.
Specifically, Task 1 could not be completed due to circumstances beyond the control of the
project investigators. Although more than 800 coliphage isolates were obtained during this
45
-------
study, the vast majority of the inventory of these coliphages were lost to due failures of ultracold
freezers in which the isolates were stored at the UNC laboratories for subsequent
characterization. These freezer failures were due to both freezer malfunctions and to power
outages at the laboratories that were due to natural disasters (ice storms and hurricanes) beyond
the control of the project investigators. Furthermore, these freezer failures caused the loss of the
majority of bacterial hosts on which the coliphages isolates were to be characterized as to host
range in order to fulfill Task la. Most of these bacterial strains had been previously purchased
from the American Type Culture Collection and the costs of replacing them were prohibitive and
had not been included in the project budget.
To compensate for the loss of these coliphage isolate samples and their further characterization,
the project labs undertook additional work in support of meeting other project objectives and
tasks. Specifically, the participating labs analyzed more samples of groundwater in the Phase II
studies than were originally specified. Contract specifications called for the analysis of a total of
64 samples (16 per laboratory) for coliphages, bacterial indicators and human enteric viruses.
The eventual number of samples analyzed was actually 106 samples (27 by three laboratories and
25 by the fourth laboratory). It was believed that the analysis of extra field samples would
provide more representative data for determining if coliphages were effective indicators of fecal
contamination of groundwater and of human enteric viruses. Additionally, the analysis of human
enteric viruses in groundwater samples was expanded to include astroviruses, which were not
originally included in the specifications for human enteric viruses to be analyzed. Hence, this
46
-------
additional task was taken on to further improve the opportunities to detect human enteric viruses
in groundwater as part of the effort to obtain more definitive data.
At the request of the UNC project manager, the UNC lab also took an additional task that was
not in the original scope of work or its budget. This additional task was to determine the survival
of coliphages in groundwater samples to be analyzed for viruses by Methods 1601 (two-step
enrichment spot plate) and Method 1602 (SAL plaque assay). Groundwater samples seeded with
mixed populations of sewage-derived viruses were analyzed for coliphages initially (on day zero)
and also after 2, 3 and 6 days of storage at 4°C. These survival experiments were done to
determine if collected samples held for several days prior to analysis due shipping and storage
still had most of their initial coliphages that still could be detected by Methods 1601 and 1602.
47
-------
PHASE II METHODS AND MATERIALS
Groundwater Samples and Wells
The original goal of this study was for each of the four, regionally representative laboratories
(southeast, northeast, upper midwest and southwest) to collect and analyze 27 ground water
samples from public water supply wells. Efforts were made to identify candidate public water
supplies that previously had coliform bacteria violations or other evidence of vulnerability to
fecal contamination. In some cases candidate wells were prescreened by bacteriological and
coliphage analyses for evidence of fecal contamination. Because not all participating labs could
identify and get access to 27 public water supply wells, some labs also included non-public and
private wells in their sampling. Three labs obtained 27 ground water samples and one lab
obtained a total of 25 samples for a total of 106 samples overall. The characteristics of the wells
that were sampled are presented below, by region.
Southeast. Of the 27 wells in the Southeast, 13 were in North Carolina and 4 were in Florida.
The Florida wells were all public water supply wells. Florida Well UNC #1 is in Orange County,
FL. There is no history of that well ever being disinfected. Florida Well UNC #2 is in Orange
County, Fl. The pump was taken out of service for repairs (rebuilt pump), and it was disinfected
in January - February of 2002. Prior to placing the well back into production it was disinfected
with chlorine. Approximately 30 gallons of 12% liquid bleach was placed into the well for 24
hours (lOOppm). Then water was discharged for a minimum of 4 hours and bacteriological
samples were taken to confirm their absence. Theses Florida wells were samples in June and
48
-------
September, 2002, which was 4 and 7 months after Well #2 had been chlorinated.
Two other Florida wells, designated UNC #3 and UNC #4 and located in Ocala County also
were sampled. Both wells had periodic coliform positivity during the and prior to the study
period (2002). They serve a population of about 57,000 in the Ocala area. UNC #3 and #4 were
not disinfected prior to or during sampling for this study. However, the utility currently (year
2004) adds calcium hypochlorite (granular chlorine) to control the total coliforms they are getting
(and have been getting a lot more since the hurricanes this year - 2004). The chlorine is now
added weekly to both wells to achieve the CT for 4 log virus removal through their treatment
process, including chlorination.
There were 13 wells in North Carolina, and the characteristics of these wells are summarized
below.
Type
Identification
Private industrial
Community water supply
Community water supply
Non-community water supply (private campgrounds)
Non-community water supply (private campgrounds)
Non-community water supply (private campground)
County
Cartaret
Cartaret
Carteret
Pamlico
Pamlico
Pamlico
Well
BF
BMHP
SBMHP
Camp DL
Camp SF #1
Camp SF #2
49
-------
Private Carteret TGUMC
Private Carteret GL
Private Pender SC
Private Pender VE
Private Duplin RH
Private Duplin KC
Private Duplin TH
Southwest. Only PWS wells were investigated in this study, and a total of eleven different PWS
wells were included. The sampling sites were located in the San Antonio region of Texas (wells
RS, KK, and HCR) and along the US-Mexico border in southern New Mexico (wells MHPa,
MHPb, MHPc, FVE, AVC, SME, and LME). The wells in the San Antonio region were part of a
karst aquifer and were previously implicated in a documented groundwater contamination event.
Also, during the initial pre-screening of the wells some of the samples were positive for somatic
and male-specific coliphages. The wells in southern New Mexico were identified as being
vulnerable to groundwater contamination based on parameters such as closeness to septic tanks,
proximity to the Rio Grande river and the aquifer in question. These wells were part of a
previous EPA-funded project on the microbiological quality of wells in the shallow aquifer along
the US-Mexico border during which some of the wells in the sampling area were positive for
enterococci, E. coli, male-specific coliphages and somatic coliphages. The wells were in the
100-150 feet depth range. The static water levels were around 10-20 feet and in terms of their
hydrogeologic setting, they were located in the Rio Grande alluvium/Hueco-Tularosa aquifers.
50
-------
Groundwater samples were collected between June 2002 and January 2003. Multiple samples
were collected from each of the wells to be representative of the aquifer and the sampling
location. During each sampling adequate volumes were collected for the coliphage analysis as
well as for the enteric virus analysis. Grab samples were collected for the coliphage and bacterial
analysis while the 1MDS filters were used for collecting the large volume enteric virus samples.
Upper Midwest. A total of 27 groundwater samples were collected from 25 wells. Two wells
were tested twice. Details of these wells are provided in the report of the Upper Midwest lab,
which is in the Appendix). All wells except 6 private ones in Minnesota were considered non-
community public water supplies by the State of Minnesota and none were disinfected.
Noncommunity transient public water supplies (i.e., groundwater) are monitored for nitrate and
total coliform bacteria as required by the SDWA. Private water systems including those places of
business not meeting the federal definition of PWS have no long term monitoring requirements.
Northeast. All sample sites were located in New England. Eight well sites were public water
sources and 17 were private wells. A total of 25 wells samples were collected instead of 27 due
to a very severe and harsh winter. NH had its first snowfall at the end of October and a second
snowfall at the beginning of November, 2002. Plans to sample two additional wells as soon as
the weather permitted could not be carried out because New England experienced one of the
snowiest winters ever. Therefore, only 25 well samples were collected and analyzed. Of the 25
wells, there were 12 sample sites in New Hampshire, two of which were from public wells that
were approximately 500 and 700 ft deep, respectively. None of these wells had any form of
51
-------
disinfection. The other wells from NH were all private wells. These wells also were not
disinfected. One well from NH was a private, very shallow well, less then 35 feet deep and lined
with stone. This was not considered a potable well but was used for farm irrigation. Four sites in
Maine were all privately owned wells and not disinfected. Three sites were in Vermont, and they
were all privately owned wells and not disinfected. All of the privately owned wells were drilled
wells, excepted for the one in NH as indicated above, and they were of varying depths that were
unknown to the homeowner at the time samples were collected. There were 6 samples from
public water supply wells in Massachusetts. The public water supplies in Massachusetts were
chosen due to positive results previously found for total and fecal coliforms, enterococci, and
male-specific coliphages. Additionally 3 of the 6 locations had positives previously reported for
rotavirus and enterovirus, by molecular methods.
Coliphage Analysis of Groundwater
Groundwater samples were analyzed by Method 1601, the two-step spot-plate enrichment
method and by Method 1602, the Single Agar Layer (SAL) plaque assay using sample volumes
of 1 liter and 100 mL, respectively, for each target group of coliphages (male-specific, somatic
and "total" coliphages). Host bacteria for the target groups of coliphages were E. coli CN-13 for
somatic coliphages, E. coli Famp for F+ coliphages and E. coli C3000 for "total" (somatic plus
F+) coliphages. Coliphage analyses were performed according to the EPA-approved methods,
except lysis zones from enrichment spot plates and plaques from SAL plates were confirmed
using the proposed EPA confirmation method. In this method, material from individual SAL
plaques or lysis zones on spot plates was removed (aspirated) with a Pasteur pipette, micropipette
52
-------
tip, or other device and the recovered material was resuspended in 0.5 mL of tryptic soy broth.
These suspensions were held briefly for coliphages to diffuse out of the agar and then the
samples were vortex mixed vigorously to disperse the coliphages. Then, 10|il aliquots were
removed from the suspension and spotted onto pre-poured spot-plates of the appropriate E. coli
host bacterium as in the enrichment procedure. The spot-plates were incubated overnight and
checked for zones of lysis. Any spots showing lysis were scored as confirmed coliphages.
Coliphage Isolate Characterization
A total of 800 coliphages (200 from each of the four participating laboratories) were to be
characterized for their properties to determine if they were of likely fecal origin. For each of the
four participating laboratories 100 coliphage isolates from the phase I studies with groundwater
samples seeded with sewage-derived coliphages and another 100 isolates from the unseeded field
groundwater samples of each laboratory were to be subjected to characterization by bacterial host
range analysis, growth temperature range analysis and determination of type of nucleic acid (for
F+ coliphages).
For bacterial host range analyses coliphage isolates were to be tested for their ability to grow in
both E. coli and non-E. coli coliform hosts and other bacteria by spotting onto pre-poured agar
medium-host cell lawns of the following 23 different host bacteria if available: E. coli strains C,
CN13, C3000, K12F, K12F and Famp, S. typhimurium WG45 and WG49, Klebsietta
pneumoniae ATCC strains 23356 and 23357, Enterobacter cloacae ATCC strain 223355,
Citrobacter braakii (formerly Citrobacter freundil) ATTC strain6570 ATCC strain 12012,
53
-------
Serratia marcescens ATCC strain 14764, Shigella sp. ATCC 23354, Shigella flexneri ATCC
12661, Yersinia pseudotuberculosis ATCC strain 23207, Proteus mirabilis ATCC strain 9921,
Yersinia enterocolitica ATCC strains 9610, 29913, Pseudomonas aeruginosa ATCC s train
12175, Aeromonas hydrophila ATCC strain 23211. Spotted plates are incubated at 37°C
overnight and observed for evidence of lysis of the host bacteria in each spot as evidence of
growth on each host bacterium.
For growth temperature range characterization, coliphage isolates were to be tested for their
ability to grow at temperatures of 25, 36, 42 and 44.5°C on E. coli hosts. Coliphage isolates were
to be serially diluted 10-fold and several dilution were to be spotted in 10 uL amounts onto
replicate pre-poured lawns of E. coli host bacteria in agar media Petri dishes. Each replicate
plate was to be incubated at the aforementioned temperatures overnight and then the spots n these
plates were to be observed and quantified for coliphage growth at each of the 4 test temperatures.
Coliphage growth at temperatures of not only 36°C but also growth at the temperatures of 42 and
or 44.5oC was considered evidence of thermotolerance and of a likely fecal origin.
For nucleic acid analyses of F+ coliphages, isolates were to be examined for taxonomy as F+
DNA or F+ RNA coliphages using previously described methods (Hsu et al, 1995). Briefly, 10
|il volumes of F+ coliphage suspensions were to be spotted onto duplicate pre-poured lawns of E.
coli host bacteria in agar medium. One plate contained Rnase at 100 |ig/mL and the other plate
did not. Plates were to be incubated overnight at 37°C and then they were to be observed for
lysis or the appearance of plaques in the spots of applied coliphage suspensions. Presence of a
54
-------
lysis zone or plaques in the spot on the plate without Rnase and the absence of such lysis or
plaques in the spot on the plate with Rnase were considered evidence of an RNA coliphage. The
presence of lysis or plaques in the spots of plates with and without Rnase was considered
evidence of an F+ DNA coliphage.
As indicated above, these coliphage characterization activities were not completed due to
extenuating circumstances beyond the control of the project investigators. Ultracold freezer
failures caused the loss of archived coliphage isolates to be characterized and also the loss of
most of the bacterial hosts that were to be used for host range characterization studies of these
coliphage isolates.
Bacteriological Analysis of Groundwater
Field groundwater samples were analyzed for E. coli and enterococci and in some cases for fecal
coliforms using EPA-approved methods. For E. coli, some labs used mFC agar for fecal
coliforms, with incubation at 44.5°C for 20-22 hours, followed by transfer or membranes to
nutrient agar-MUG medium, re-incubation for several hours, and observation for colonies
fiuorescing blue under long-wavelength UV light as evidence of E. coli colonies. Another lab
used mEC medium for E. coli with incubation at 44.5 °C (APHA, 1998). Another lab used
mColiBlue agar for simultaneous detection of total coliforms andE. coli, according to USEPA-
approved methods. For enterococcus analysis, labs used standard membrane filter methods and
with either modified ME agar or MEI agar and incubation conditions as specified in the EPA
method (APHA, 1995; Levin et al, 1975; USEPA, 2002). Samples for E. coli analysis were 100
55
-------
mL, although one lab also analyzed volumes of 1000 mL. Data for the results of this larger 1000
mL volume were not included in the compilation and analysis of data for all labs, as no other lab
analyzed this volume and it is not a standard volume used for bacteriological analysis of water.
The results for this larger volume are in the report from this participating laboratory, which is in
the Appendix to this report.
56
-------
Analysis of Groundwater for Human Enteric Viruses
Primary virus concentration from groundwater. Ground water from candidate wells was
sampled using the EPA ICR method, with minor modifications (USEPA, 1996). Groundwater
sample volumes of 1,500 liters (397 gallons) were to be filtered through a 1 MDS pleated
cartridge filter (CUNO) at pH 6-8. The filter was eluted with 1.5% beef extract (Becton
Dickinson #212303) buffered with 0.05 M glycine at pH 9.5. Viruses in the resulting beef
extract eluate were further concentrated by organic flocculation (acid precipitation) as specified
by the ICR methods. The only significant change to the ICR procedure was that the acid
precipitate was resuspended in 20 about mL of sodium phosphate rather than 30 mL in order to
reduce the concentrate volume and the number of cell cultures required to assay the concentrate.
The entire concentrate was filter-sterilized using a 0.2 micrometer pore size Gelman Serum
Acrodisc filter (#4525) which has been pretreated with a small volume of beef extract eluent to
minimize viral adsorption.
The filter-sterilized concentrate was subdivided into the following aliquots prior to being frozen
at -80°:
• Equivalent of 500 L of water sample for assay of viruses in Caco-2 cell cultures
• Equivalent of 500 L of water sample for assay of viruses in BGMK cell cultures,
further subdivided into a 1.5 mL subsample, and the remainder. The 1.5 mL
subsample was used in a pre-test for cytotoxicity in BGMK cultures.
• Equivalent of 100 L of water sample for assay of HAV in FRhK-4 cell cultures.
57
-------
• Equivalent of 100 L of water sample for assay by direct RT-PCR for caliciviruses
(noro viruses).
The remainder of the sample concentrate (20%, equivalent to 300 L) was archived at
- 80°C.
Virus isolation in cell cultures. Three cell lines were used to detect a range of infectious enteric
viruses. The BGMK cell line was used to propagate adenoviruses, enteroviruses, and reoviruses
according to the procedure of (Chapron et al. (2000). Caco-2 cells were used for the detection of
astroviruses and rotaviruses according to Chapron et al. (2000). The FRhK-4 cell line was used
to detect hepatitis A virus (HAV).
Cell culture infectivity assays were performed in a minimal number of 75 cm2 flasks (generally 4
or 5). A pretest to screen each sample concentrate for cytotoxicity was performed in 25 cm2
BGMK culture flasks, with one flask being inoculated with 1.0 mL of sample concentrate and a
second flask being inoculated with 0.5 mL of sample concentrate. The pretest cultures were
observed microscopically for evidence of cytotoxicity (or CPE) for one week, before the
remainder of the sample was inoculated into cell cultures.
Sample concentrates inoculated into BGMK and Caco-2 cultures were pre-activated by treatment
with the proteolytic enzyme trypsin prior to inoculation. This was done because previous studies
had indicated enhanced enteric viruses detection using this trypsin pre-treatment. Each
58
-------
concentrate was mixed with a solution of type IX trypsin (Sigma T-0303) yielding a final 10
|ig/mL concentration, then incubated 30 minutes at 37° C. Pre-activation was not necessary and
therefore not employed for HAV propagation in FRhK-4 cultures. Because divalent cations
enhance attachment of HAV and many other enteric viruses to cells, sample concentrates were
diluted with an equal volume of Dulbecco's PBS before being inoculated into FRhK-4 cultures.
Cell cultures in 75 cm2 (confluent monolayers for BGMK cells and 90-95% confluency for Caco-
2 and FRhK-4 cell cultures) were drained and rinsed three times with Dulbecco's phosphate
buffered saline (Sigma #D-8662, Gibco #14040 or equivalent) to remove residual serum. FRhK-
4 cell cultures were rinsed once. Replicate cultures were inoculated with sample concentrates,
and incubated at 37° C for 90 minutes, while being rocked every 15-20 minutes to re-distribute
the inoculum, to allow for virus adsorption to cells. One negative control flask was inoculated
with PBS before any flasks were inoculated with sample concentrates, and a second negative
control culture was similarly be inoculated at the end of the sample inoculation step. No enteric
virus positive control flasks were to be prepared at this time in order to avoid possible laboratory
virus contamination. Maintenance medium was then added to each flask, and the cultures were
incubated at 37° C. The maintenance medium for BGMK and Caco-2 cultures consisted of
serum-free Eagle's minimum essential medium with Earle's salts (MEM) supplemented with 5
|ig/mL trypsin. The maintenance medium used in FRhK-4 cultures consisted of Eagle's
minimum essential medium with Earle's salts supplemented with 2% serum and 30 mM MgCl2.
59
-------
Cultures were observed microscopically on days 1 and 2, then at least every other day following
inoculation. The occurrence of cytopathology or cytopathic effects (CPE) on the first two days
was tentatively assumed to be evidence of sample cytotoxicity or the release of cells from the
bottom surface of the tissue culture flask by the action of the trypsin. If cytotoxicity thought to
be associated with sample inocula was not too far advanced, the affected cultures were given a
change of maintenance medium to saved them from possible destruction by sample cytotoxicity.
Alternative approaches for cytotoxicity reduction included removing the inoculum following the
90 minute incubation period and then rinsing the cell layer with PBS, diluting the sample
concentrate in Dulbecco's PBS, or inoculating less concentrate into each cell culture. Every
reasonable effort was made to reduce sample cytotoxicity and maximize enteric virus detection.
BGMK and Caco-2 cultures were incubated for 7 at 37°C days following inoculation. All
cultures were freeze-thawed, and 10% of the lysate from each flask was inoculated into fresh
cultures for a second 7-day passage. If a flask exhibited possible viral cytopathology (CPE), the
lysate was passed through a 0.22|im pore size, sterilizing filter into a fresh culture to confirm the
presence of viruses and the absence of bacterial or fungal contamination. FRhK-4 cultures were
incubated for two 14-day passages, with the maintenance medium being changed after seven
days, to maximize the detection of typically slow-growing HAV.
At the end of the final cell culture passage, flasks were frozen and thawed twice. A 1-mL aliquot
from each of the first and second passage BGMK and Caco-2 flasks inoculated with a given
sample was pooled in a centrifuge tube. A half-volume of chloroform was added, and the tube
was vortex mixed at high speed for two minutes. The tube was centrifuged at 1,200-1,800 x g
60
-------
for 20 minutes, then the supernatant extract was removed and split into aliquots for viral analysis
or archiving.
Virus detection by nucleic acid amplification. Chloroform-extracted cell culture lysate pools
were examined for adenoviruses, astro viruses, enteroviruses, HAV, reoviruses and rotaviruses by
RT-PCR or PCR. Viral nucleic acids were extracted from lysates using the QIAamp Viral RNA
Mini Kit (Qiagen #52904). By modifying two steps of the QIAamp protocol, adenovirus DNA
could be efficiently recovered without compromising extraction of viral RNA:
In step #8 of the Qiagen protocol, The sample column was incubated for one minute after
adding buffer AW1, before centrifuging the column.
In step #9, the sample column was incubated for one minute after adding buffer AW2,
before centrifuging the column.
Prior to RNA extraction, enteric viruses in 2-milliliter aliquots of pooled, chloroform-extracted
cell culture lysates were concentrated using polyethylene glycol (PEG) precipitation.
Polyethylene glycol (Sigma #P2139, molecular weight = 8,000) was added to a final 8%
concentration. Sodium chloride was added to a 0.3 M concentration, and the sample was mixed
until the additives dissolved. The solution was incubated for two hours at room temperature and
then centrifuged at 6,700 x g for 20 minutes at 4° C. The pellet was resuspended in 300 |iL of
Dulbecco's PBS, then extracted with 300 |iL of chloroform.
61
-------
An aliquot of each water sample concentrate was assayed directly for human caliciviruses
(noroviruses) by nucleic acid amplification using RT-PCR because these viruses cannot be
propagated in cell cultures. Each sample was reconcentrated using polyethylene glycol
precipitation. PEG was added to a final 10% w/v concentration. Sodium chloride was added to a
final 0.3 M concentration. Since the sample has previously been adjusted to pH 7.0-7.5, no
further pH adjustment was necessary. The mixture was shaken until the additives had dissolved.
The solution was incubated at room temperature for two hours or at 4° C overnight. The solution
was centrifuged at 6,000 to 10,000 x g for 15 minutes, and the supernatant was removed by
aspiration and discarded. The pellet, which may not be visible, was resuspended in a maximum
of 140 |iL of Dulbecco's PBS containing magnesium and calcium ions.
Prior to RNA extraction, the resuspended pellet was extracted with chloroform. A 100-|iL
volume of chloroform was added, and the sample vortex mixed for one minute. The sample was
centrifuged at about 3,000 x g for 5-10 minutes. The supernatant was removed by aspiration and
recovered. Viral RNA was extracted from the recovered, chloroform-extracted supernatant using
the standard QIAamp Viral RNA Mini Kit protocol.
Nucleic acid amplification by (RT-)PCR
Introduction. Viruses in chloroform-extracted cell culture lysates that had been inoculated with
water sample concentrates and human caliciviruses (noroviruses) in aliquots of water sample
concentrates were analyzed by either PCR for adenoviruses or RT-PCR for astroviruses,
caliciviruses, enteroviruses, hepatitis A virus (HAV), reoviruses, and rotaviruses. The
62
-------
procedures for combined cell culture and (RT-)PCR were based on those previously used by
Chapron et al. (2000) with minor modifications. The nucleic acid extraction procedures and the
(RT-)PCR primers and amplification procedures applied to each virus or virus group are
described in more detail in the sections that follow and in the individual report of the other three
participating laboratories (see Appendix). RT-PCR for HAV and caliciviruses was done by the
University of North Carolina lab (Southeast), RT-PCR for enteroviruses was done by the
University of Minnesota lab (Upper Midwest), RT-PCR for reovirus and rotavirus was done by
the TAMU lab (Southwest), and (RT-)PCR for adenoviruses and astroviruses was done by the
University of New Hampshire lab (Northeast). The details of the virus (RT-)PCR procedures of
the participating laboratories are given below and also in more detail in the individual project
reports of the other three participating labs, which appear in the Appendix of this report.
Enterovirus RT-PCR
The primers for RT-PCR amplification of enteroviruses were:
3' pan-enterovirus primer: 5'-ACC GGA TGG CCA ATC CAA
5' pan-enterovirus primer: 5'-CCT CCG GCC CCT GAA TG
Random hexamers may also be used as the primer for enteroviruses for reverse transcription.
The reaction mixtures for a 3.5 |iL sample were as follows. (For larger sample volumes the
amounts were increased proportionally). (Note: These mixtures utilized reagents from the
GeneAmp RNA PCR Core Kit, Applied Biosystems #N808-0143.)
63
-------
RT master Mix
Stock cone.
reaction/reaction Final cone.
MgC12
1 Ox PCR Buffer II, pH 8.3
each dNTP
3' primer or random hexamers
MuLV reverse transcriptase
Rnase inhibitor
25 mM
lOx
10 mM
50 |iM
50 U/|iL
20 U/uL
4
2
2 each
0.5
0.9
0.9
5mM
Ix
ImM
1.26nM
45 units
1 8 units
The RT conditions were: 95° for 5 minutes, 42° for 60 minutes, and 95° for 5 minutes.
PCR master mix
MgC12
lOx PCR Buffer II, pH 8.3
Water (Sigma #W-4502)
Ampli-Taq DNA polymerase
5' primer
(and 3' primer if used random hexamers)
Stock cone.
25 mM
lOx
5U/^L
50 uM
reaction/reaction
4
8
66
0.5
0.5
Final cone.
2mM
Ix
2.5 units
0.25 uM
The PCR conditions were per cycle: 95° C for 1.5 minutes, 55° C for 1.5 minutes, and 72° C for
1.5 minutes, for a total of 40 cycles. The expected product (amplicon) size was 197 bp. The
internal oligonucleotide probe for hybridization was: 5'-TAC TTT GGG TGT CCG TGT TTC.
Hybridization was at 55° C.
64
-------
Hepatitis A virus RT-PCR
The primers for RT-PCR amplification of HAV were:
3' HAV primer: 5'-CTC CAG AAT CAT CTC CAA C
5' HAV primer: 5'-CAG CAC ATC AGA AAG GTG AG
(VP1-VP3 capsid protein interface region)
The RT-PVR reaction mixtures and reaction conditions were the same as for enteroviruses, and
the expected product (amplicon) size was 192 bp. The internal oligonucleotide probe for HAV
was: 5'- TGC TCC TCT TTA TCA TGC TAT G. and the hybridization temperature was 55° C
Rotavirus RT-PCR
The primers for RT-PCR amplification of rotaviruses were those for Group A, gene 9:
3' rotavirus primer: 5'-GGT CAC ATC ATA CAA TTC T
5' rotavirus primer: 5'-GAT ATA ACA GCT GAT CCA ACA AC
The reaction mixtures and reaction conditions were the same as for enteroviruses, and the
expected product (amplicon) size was 208 bp. The internal probe that could be used for product
confirmation by hybridization was as follows: 5'-AAT TGG AAA AAA TGG TGG CAA GT.
The hybridization temperature was 55° C.
65
-------
Adenovirus and Astrovirus (RT-)PCR
Nested PCR was performed on UNH, UNC, UMN and TAMU samples for both astrovirus and
adenovirus type 40 and 41. The equivalent volume of original water sample examined for each
virus was 500 liters. Positive controls were at the level of (RT-)PCR. Virus was added to cell
culture lysate to act as a positive control for (RT-PCR)PCR
Astrovirus. All molecular techniques were done as specified in the methods and materials
developed by the project team in communication with the EPA project manager. Astrovirus RT-
PCR was done according to Chapron et al. (2000). The primers used were specific for human
astrovirus:
RT primer 5'-GTAAGATTCCCAGATTGGT-3', and
PCR primer 5'-CCTGCCCCGAGAACAACCAAG-3'.
An 1 l-|iL sample of the combined (pooled) cell lysate was denatured with 0.5 |iL each of 0.05 M
EDTA and downstream primer at 99°C for 8 min. Eighteen |iL of the RT mixture was then
added and run for 42 min. at 42°C to reverse transcribe, followed by 5 min. at 99°C. The RT
mixture per sample consisted of 2.5 |iL 10X buffer n, 8.5 |iL of 25mM MgCl2 1.25 |iL of each
lOmM dNTP, 0.5 |iL of lOOmM DTT (Promega), 10 units of Rnasin, and 50 units of RT.
After the RT step, 28.5 |iL of a PCR master mix was added. The PCR mixture per sample
consisted of: 3 |iL of 10X buffer n, 1 |iL of the PCR primer, 0.5 |iL of the RT primer, 24 |iL of
66
-------
molecular grade water, and 2.5 units of Ampli-Taq DNA polymerase. The PCR amplification
parameters were 95°C for 5 minute hot start, followed by 35 cycles of: 95 °C for 30 seconds,
56°C for 30 seconds, 72°C for 30 seconds, with a final extension at 72°C for 5 minutes. These
primers yielded a 193 and/or 243 bp amplicon.
For nested PCR, 1 |uL from each RT-PCR reaction was added to a new tube containing 90 |uL of
a nested PCR reaction mixture, which contained 8 mM MgCl2 10 |iL lOx buffer, ImM of each
dNTP, 2.5 units of Ampli-Taq DNA polymerase and 1 |iM of each primer. The primers used
were: 5'-CCTTGCCCCGAGCCAGAA-3' and 5'-TTGTTGCCATAAGTTTGTGAATA-3'.
These primers yield a 143 and/or 183-bp amplicon. Twelve |ul of each RT-PCR product as well
as 12 |iL of the nested PCR product was resolved and sized by electrophoresis on an 1.8%
agarose gel, stained with ethidium bromide. Molecular weights were determined by comparison
with a 1 Kb DNA ladder (Life Technologies). Astro virus serotype 2 was used as a positive
control.
Adenovirus. All molecular techniques were done as specified in the methods and materials
developed by the project team and communicated to the EPA project manager. Adenovirus
Hexon PCR was done generally according to the procedures of Xu et al. (2000).
The PCR primers used were:
Adi 5'-CCCTGGTA(G/T)CC(A/G)AT(A/G)TTGTA-3' and
Ad25'-TTCCCCATGGC(Inosine)CA(C/T)AACAC-3'.
67
-------
A 5|iL sample of the combined cell lysates was added to 47.5 |iL final volume PCR master mix.
Final concentrations in the PCR master mix per sample were 1.5mM MgCl2, Ix (lOx Buffer II),
0.2mM dNTP mix, 0.6|iM of each primer, and 2.5 units of Ampli-Taq DNA polymerase. The
PCR parameters were 95°C for 5 minutes, followed by 40 cycles of: 94°C for 1 minute, 55°C for
1 minute, and 72°C for 2 minutes, with a final extension at 74°C for 5 minutes. These primers
yielded a 482 bp amplicon.
For nested PCR, 1 |ul from each PCR reaction was added to a new tube containing 90 fo.1 of a
nested PCR reaction mixture, which contained 8 mM MgCl2 10 |iL lOx buffer, ImM of each
dNTP, and 1 |iM of each primer. The primers used were:
5'-GCCACCGAGACGTACTTCAGCCTG-3'and
5'-TTGTACGAGTACGCGGTATCCTCGCGGTC-3.
These nested primers were specific for Adenovirus type 40 and 41. Samples were run for 35
cycles of: 95°C for 30 seconds, 55°C for 30 seconds, 72°C for 30 seconds, yielding a 142 bp
amplicon. Twelve |iL of each nested PCR product was resolved and sized by electrophoresis on
1.8% agarose gels and stained with ethidium bromide. Molecular weights were determined by
comparison with a 1 Kb DNA ladder (Life Technologies). Adenovirus 40 and 41 were used as
positive controls.
68
-------
Reovirus RT-PCR
The primers for RT-PCR amplification of reoviruses and kindly provided by Shay Fout of the US
EPA, Cincinnati, were:
3' pan-reovirus primer: 5'-GTG CTG AGA TTG TTT TGT CCC AT
5' pan-reovirus primer: 5'-ACG TTG TCG CAA TGG AGG TGT
Reaction mixtures for 5 |iL samples were:
RT master mix
MgCl2
1 OX PCR Buffer II
dNTP mix
3' primer
Water
Stock cone.
25 mM
10X
10 mM each
10 nM
reaction/reaction
1.8
3
2
5
11.45
Final cone.
1.5mM
IX (Applied Biosystems)
0.7 mM
1.7jiM
The initial RT-PCR reaction conditions were: 99° C for 5 minutes; then tubes were placed in ice.
Enzyme mix
RNasin
MuLV RT
Stock cone. reaction/reaction Final cone.
30 units/|iL 0.75 22 units (Promega N2511)
50 units/|iL 1 50 units (Applied Biosystems)
The RT reaction conditions were: 43° C for 60 minutes, 95° C for 5 minutes, and then tubes were
placed in ice.
69
-------
PCR master mix
MgCl2
1 OX PCR Buffer II
5' primer
AmpliTaq Gold
Stock cone.
25mM
10X
10 nM
reaction/reaction
4.2
7
5
1
Final cone.
1.5mM
IX
0.5 [iM
(pH 8.3 buffer only)
The PCR reaction conditions were per cycle: 95° C for 1 minute, 55° C for 1.5 minutes, and 72°
C for 1.5 minutes, for a total of 40 cycles. The expected product (amplicon) size was!25 bp.
Internal oligonucleotide probes for the individual reovirus types 1, 2 and 3 were kindly provided
by Shay Fout, and the hybridization temperature was 51° C.
Calicivirus direct RT-PCR
Calicivirus (Norovirus) RT-PCR analysis of concentrated virus samples from groundwater was
done with the modified generic primers designated JV12/JV13 (Vinje et al., 2001; Hamidjaja et
al. 2004).
3' RegA primer: 5'-CTC (A/G)TC ATC (Inosine)CC ATA (A/G)AA (Inosine)GA
5' MJV12 primer: 5'-TA(C/T) CA(C/T) TAT GAT GC(A/C/T) GA(C/T) TA
The RT-PCR reaction mixture for 5|il samples had the following composition:
Antisense mix
RegA primer
Water
Stock cone.
50
reaction/reaction
1.2
2.8
Final cone.
60 pM
70
-------
The initial RT reaction conditions were as follows: 94° C for 2 minute and then chilling in ice.
RT master mix
MgC12
1 OX PCR Buffer II
dNTP mix
Water
AMV-RT
Rnase inhibitor
Stock cone.
25mM
10X
10 mM each
10 units/|iL
40 units/|iL
reaction/reaction
1.8
1.5
1.5
0.2
0.5
0.5
Final cone.
3 mM
IX (pH 8.3)
1 mM
5 units
20 units
The subsequent RT reaction conditions were: 42° C for 60 minutes and then 94° C for 5 minutes,
followed by chilling in ice.
PCR master mix
MgC12
1 OX PCR Buffer (pH 9.0)
dNTPs
MJV12 primer
RegA primer
Taq polymerase
Water
Stock cone.
25 mM
10X
10 mM each
50
50
5 units/|iL
reaction/reaction
2.4
4.5
0.5
1
0.6
0.5
35.5
Final cone.
1.5mM
0.2 mM
50 pM
50 pM
2.5 units
71
-------
The PCR reaction conditions were an initial 94° C for 3 minutes, followed by 40 cycles with
each cycle consisting of: 94° C for 1 minute, 50° C for 1.5 minutes, and 74° C for 1 minute. This
was followed by 74° C for 7 minutes. The expected product (amplicon) size was 327 bp.
The internal oligonucleotide probe for human caliciviruses was a mixture of one Group 1 probe
(GGI) and three Group II probes as follows:
GGI probe: 5'-ATG GA(CT) GTT GG(CT) GA(C/T) TAT GT (20 pM)
GGIId probe: 5'-TGG AAC TCC ATC GCC CAC TGG (40 pM)
GGIIe probe: 5'-TGG AAC TCC ATC ACA CAT TGG (80 pM)
GGLeeds probe: 5'-TCA CCA GAT GTT GTC CAA GC
The hybridization temperature was 42° C.
MS-2 RT-PCR
Coliphage MS2 was used as the positive control for RT-PCR analyses by some laboratories to
avoid introduction of potential human enteric virus contamination. About 100 pfu of MS2 per
reaction tube was to be used and RT-PCR was done according to Meschke and Sobsey (1998).
The RT-PCR primers for MS2 were as follows:
3' (downstream) MS2 primer: 5'-CCC TAC AAC GAG CCT AAA TTC
5' (upstream) MS2 primer: 5'-GCA ACC TCC TCT CTG GCT AC
72
-------
Random hexamers could be used as the primer MS2 reverse transcription. Reaction mixtures and
reaction conditions for MS2 were the same as for enteroviruses. The expected PCR product size
was 220 bp.
RT-PCR and PCR controls
To reduce the possibility of cross-contamination of field samples during nucleic acid
amplification, RNA coliphage MS-2 was selected to act as a positive control for RT-PCR
procedures. This positive control with its own pair of primers was included in each set of
reactions done by some participating labs. Other labs already had their own RT-PCR and PCR
controls and they used those existing QA/QC control procedures and reagents that were already
in place. The details of those measures can be found in the reports of the other 3 participating
laboratories, which are in the Appendix to this report.
The minimal negative control samples that were to be run as part of each set of RT-PCR and
PCR samples included the following: (1) combined master mixes, enzymes and water done
twice, one tube placed at the beginning of the set, and one tube placed at the end of the set, and
(2) a cell culture negative control. Because the same pooled cell cultures were tested for multiple
groups of human enteric viruses, cell culture negative control RNA extracts needed to be assayed
using all of the appropriate virus primer pairs used by a given participating laboratory. RT-PCR
for HAV and Caliciviruses was done by the UNC lab (Southeast), RT-PCR for enteroviruses was
done by the University of Minnesota lab (Upper Midwest), RT-PCR for reovirus and rotavirus
was done by the TAMU lab (Southwest), and (RT-)PCR for adenoviruses and astroviruses was
done by the University of New Hampshire lab (Northeast). Further details of the virus (RT-)PCR
73
-------
methods of the participating laboratories are given in their individual project reports, which
appear in the Appendix of this report.
Confirmation of presumptive (RT-)PCR positive samples
Amplified PCR products from non-nested protocols were to be examined by agarose gel
electrophoresis. If cDNA bands of the appropriate size were detected, the presence of enteric
virus sequences was to be confirmed using a labeled oligonucleotide probe internal to the
original amplicon, as specified above. If enteric virus cDNA was detected, it was to be preserved
for possible nucleotide sequencing.
74
-------
PHASE II RESULTS AND DISCUSSION
Introduction
The Phase II studies of the project consisted of both additional lab studies as well as field studies.
Lab studies were performed to further characterize and improve the method to confirm coliphage
isolates from plaques on SAL plates and lysis zones on spot plates from the enrichment method.
Lab studies also were conducted to determine the survival of coliphages in groundwater held at
4oC for up to 6 days prior to coliphage assay by Methods 1601 and 1602. Field studies consisted
of the analysis of groundwater samples from wells for F+, somatic and "total" coliphages (by
Methods 1601 and 1602), fecal indicator bacteria (E. coli and enterococci), and human enteric
viruses by each of the four regional labs. Each lab collected and analyzed groundwater samples
in its region. The samples concentrated for recovery of human enteric viruses by each of the 4
labs were divided into aliquots so that individual aliquots could be sent to other participating labs
for centralized analysis of one or two the different target groups of human enteric viruses. The
data for coliphages and fecal indicator bacteria in groundwater were analyzed to determine if the
analysis of both coliphages and fecal indicator together in the same sample of groundwater gave
greater detection of fecally contaminated groundwater than the analysis of only one indicator,
either a bacterium or a coliphage.
Results of Field Sample Analysis of Coliphage and Bacterial Indicators in Groundwater
Table 9 contains all of the data for the presence and concentrations coliphages and fecal indicator
bacteria in samples of groundwater from all four laboratories.
75
-------
Table 9. Coliphages Detected by Methods 1601 and 1602 and Indicator Bacteria in Groundwater
Lab and Samples
TAMU-RS (1)
TAMU-HCR(l)
TAMU-RS (2)
TAMU-BM(l)
TAMU-KK(l)
TAMU-RS (3)
TAMU-KK (2)
TAMU-HCR (2)
TAMU-RS (4)
TAMU-RS(S)
TAMU-MHPla
TAMU-MHPlb
TAMU-AVC1
TAMU-FVE1
TAMU-AVC2
TAMU-FVE2
TAMU-FVE3
TAMU-AVC3
TAMU-MHPlc
TAMU-MHP2a
TAMU-MHP2c
TAMU-MHP3a
TAMU-MHP2b
TAMU-SME1
TAMU-SME2
TAMU-LME1
TAMU-MHP3b
UNH-1
UNH-2
UNH-3
UNH-4
UNH-5
UNH-6
UNH-7
UNH-8
UNH-9
UNH-10
UNH-11
UNH-1 2
UNH-1 3
UNH-1 4
UNH-1 5
UNH-1 6
UNH-1 7
UNH-1 8
UNH-1 9
SAL (#/100mL)
Famp
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
CN-13
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
C3000
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Enrichment (1 L)
Famp
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
CN-13
1
0
0
0
0
0
1
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
C3000
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Bacteria/100 mL)
Fee. Colif.
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
1
0
0
0
0
0
0
0
0
0
0
0
0
0
35
0
0
0
2
E. coli
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Ent.
0
0
0
0
0
0
0
1
0
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
13
76
-------
UNH-20
UNH-21
UNH-22
UNH-23
UNH-24
UNH-25
MN-01 Amu
MN-02 Ger
MN-03 Rou
MN-04 Tur
MN-05 Bro
MN-06 OG
MN-07 KM
MN-08 KM
MN-09 Ham
MN-lONor
MN-11 Pre
MN-12 Imm
MN-13 Cen
MN-14His
MN-15Nor
MN-16Lak 1
MN-17LakM
MN-18LakM
MN-19 Al
MN-20 Day
MN-21 LakM
MN-22 GF
MN-23 TA
MN-24 Mil
MN-25 Jay
MN-12 ChR
MN-16 CemR
MN-24 Mil R
UNC-1-BMH
UNC-2-GL
UNC-3-VE
UNC-4-KC
UNC-5-OC-FL#l
UNC-6-OC-FL#2
UNC-7-KC
UNC-8-VE
UNC-9-BF
UNC-10 SB MHP
UNC-11-BMH1
UNC-12 GL
UNC-13_OC-FL#1
UNC-14 OC-FL#2
UNC-15_GL
UNC-16-CDL
0
0
0
0
0
0
0
0
4
2
0
0
1
40
0
9
0
0
0
0
0
234
0
0
0
0
0
0
3
11
0
3
6
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
TNTC
4
2
2
0
0
58
12
0
28
0
0
0
0
0
574
0
0
0
0
9
2
2
6
6
3
5
5
0
0
0
0
0
0
0.4
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
4
3
0
0
4
7
0
0
0
0
0
0
0
0
0
0
0
0
1
2
7
1
3
4
6
0
0
0
0.4
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
No data
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
30
0
1
0
1
0
0
0
0
0
0
1
17
3
0
0
248
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
15
1
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
2
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
12
1
0
0
20
0
0
0
0.5
0
1.5
0
0
0
0
0
0
0
0
0
0
77
-------
UNC-17-BF
UNC-18-SB MHP
UNC-19-CDL
UNC-20-CSF #1
UNC-21-CSF#2
UNC-22-OcaFL#l
UNC-23-OcaFL#2
UNC-24-CSF #1
UNC-25-CSF #2
UNC-26-OcaFL#l
UNC-27-OcaFL#2
0
0
0
0
0
0
0.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
*TAMU = Texas Agricultural and Mechanical University, UNH = University of New Hampshire,
MN = University of Minnesota and UNC = University of North Carolina
Table 10 summarizes these data on the presence of fecal indicator microbes, including somatic
coliphages, male-specific coliphages, "total" coliphages (detected on host E. coli C3000), fecal
coliform bacteria, E. coli and enterococci in groundwater samples in this study on the basis of
positive samples, regardless of microbe concentration. A total of 107 samples were analyzed and
these samples correspond to the samples that were also analyzed for human enteric viruses.
Additional groundwater samples were analyzed by some laboratories in the initial screening of
groundwater wells for possible inclusion in the study. However, these samples are not included
in the table because not all microbial indicators were measured by all methods during this pre-
screening analysis effort and there was no concurrent analysis of human enteric viruses for
possible comparison.
78
-------
Table 10. Frequency of Occurrence of Fecal Indicator Microbes in Field Ground Water Samples
Indicator
Somatic Coliphage - SAL
F+ Coliphage by SAL
"Total" Coliphage" - SAL
Somatic Coliphage Enrichment
F+ Coliphage Enrichment
"Total" Coliphage Enrichment
Fecal Coliform
E. co li
Fecal Coliform and/or E. coli
Enterococci
#. Pos./# Tested at:
TAMU
1/27
0/27
0/27
5/27
1/27
2/27
Not done
2/7
2/27
2/27
UMN
16/28
11/28
12/28
2/28
2/28
3/28
7/28
3/28
7/28
6/28
UNC
2/27
1/27
2/27
1/27
0/27
0/27
0/27
0/27
0/27
2/27
UNH
0/25
1/25
0/25
0
/25
1/25
1/25
4/25
0/25
4/25
4/25
Total # Pos./ Total #
Tested, All Labs
19/116
13/116
14/116
8/116
4/116
6/116
11/80
5/116
13/116
14/116
%
Positive
16.4%
1 1 .2%
12%
6.9%
3.4%
5.2%
13.8%
4.3%
1 1 .2%
12.1%
As shown in Table 10, The frequency of detection of any single fecal indicator microbe was
highest for somatic coliphages as measured by the SAL method at 16.4% and second highest for
fecal coliform at 13.8%. However, the frequency of detection of somatic coliphage by the SAL
method and of fecal coliform was not significantly different (P = 0.768 by Mann-Whitney li-
test). Interpretation of these statistical results for comparative detection of somatic coliphages by
SAL and fecal coliforms is limited. This is because not all samples were analyzed for both of
these fecal indicators and therefore a paired statistical analysis of the results was not possible.
Enterococci and "total coliphage" by the SAL method were tied for third in detection frequency
at 12.1%. Overall, these results indicate the rate of detection of any single fecal indicator was
higher for coliphages, specifically somatic coliphage detected by SAL, than any other single
indicator tested. It is also noteworthy that the simultaneous detection of both somatic and male-
specific coliphages as "total coliphages" by the SAL method on a single host bacterium, E. coli
79
-------
C3000, gave a high frequency of detecting fecal contamination at 12.1%, making it one of the
best indicators tested.
Examination of the results of coliphage analyses in Table 10 indicate that each coliphage group
was detected more frequently by the single agar layer (SAL) method than by the two-step
enrichment spot plate method. This finding is striking given that the sample volume for the SAL
method was only 100 mL and for the enrichment method it was 1 liter. The comparative
detection of coliphages by SAL and enrichment methods was 16.4% versus 6.9% for somatic
coliphage, 11.2% versus 3.4% for F+ coliphages and 12.1% versus 5.2% for "total" coliphages.
The results for the frequency of detection of the different coliphage groups by the SAL or
enrichment method were statistically compared by a non-parametric, paired t-test (Wilcoxon
matched-pairs signed-ranks test). The detections frequencies between SAL and enrichment
methods were significantly different for somatic coliphages (P = 0.137) and for F+ coliphages (P
= 0.0351) and they were nearly significant for "total" coliphages (P = 0.580). Overall, these
results indicate that the SAL method gave significantly better detection coliphages than did the
enrichment method. The reasons for this are not known and probably deserve further
investigation. It should be remembered that both methods were highly efficient in detecting
coliphages when tested in phase I studies on seeded samples of groundwater.
Comparative Detection of Two Indicators in Groundwater Samples
It was of interest to consider the simultaneous detection of two indicators in groundwater
samples. This is because the proposed groundwater rule has considered the possibility of
measuring only one indicator in a sample (either bacterium or a coliphage indicator) versus
80
-------
measuring both a bacterial and a coliphage indicator. Therefore, a fundamental consideration is
whether or not the dual measurement of two indicators improves the detection of fecal
contamination in groundwater samples by increasing the frequency of fecal indicator (a positive
sample). The results for selected pairs of fecal indicators that gave the highest detection
frequencies are summarize in Table 11.
Table 11. Frequency of Occurrence of Dual Indicators in Field Ground Water Samples
Indie ator P air
Somatic and/or F+ Coliphage - SAL
Somatic and/or F+ Coliphage - Enrichment
Enterococci and Fecal Coliform and/or E. coli
Somatic Coliphage - SAL and/or Enterococci
Somatic Coliphage - SAL and/or Fecal Coliform
or E. coli
Fecal Coliform and/or E. coli
Enterococci
Enterococci and Fecal Coliform and/or E. coli
Somatic Coliphage - SAL and/or Enterococci
Somatic Coliphage - SAL and/or Fecal Coliform
or E. coli
#. Pos./# Tested at:
1/27
5/27
3/27
2/27
3/27
2/27
2/27
3/27
2/27
3/27
16/28
4/28
9/28
16/28
17/28
7/28
6/28
9/28
16/28
17/28
3/27
1/27
2/27
2/27
2/27
0/27
2/27
2/27
2/27
2/27
1/25
1/25
6/25
4/25
4/25
4/25
4/25
6/25
4/25
4/25
Total # Pos./ Total
# Tested, All Labs
21/116
11/116
20/116
24/116
26/116
13/116
14/116
20/116
24/116
26/116
%
Positive
18.1%
9.5%
17.2%
20.7%
22.4%
11.2%
12.1%
17.2%
20.7%
22.4%
Dual versus individual detection of somatic and F+ coliphages by SAL. As a first case of
comparing the detection of positive samples with pairs of indicators versus single indicators is
the SAL detection of somatic and/or male-specific coliphages in groundwater samples. This
coliphage indicator pair was considered because the measurement of both of these two groups of
coliphages in an option in the proposed groundwater rule. Currently, there is no clear basis for
choosing one coliphage group over the other and therefore, the measurement of both coliphage
groups in a sample on their respective E. coli hosts is an option. The frequency of detecting a
positive sample (positive for one or the other or both) was 18.1% (Table 11). For each group
81
-------
alone, the SAL detection frequency was 16.4% for somatic coliphages and 11.2% for F+
coliphages (Table 10). The SAL detection of either or both of these coliphage indicators in a
sample by dual analysis (18.1%) was compared to the frequency of detection of each of them
alone (F+ SAL = 11.2% and somatic SAL = 16.4%) using the Friedman test, a non-parametric
Analysis of Variance (ANOVA). The P value was very significant (0.0055), indicating the
detection frequencies were significantly different and were highest for the detection of either or
both coliphage groups when both are measured in a sample.
When the individual SAL detection frequencies of F+ coliphages (11.2%) and somatic
coliphages (16.4%) were compared by anon-parametric t-test (Wilcoxon matched-pairs signed-
ranks test), there was no significant difference (P= 0.105), indicating equivalent detection of
either of these two coliphage groups alone. Furthermore, when the SAL detection frequencies of
F+ coliphages alone (11%) were compared to the dual detection of either F+ coliphages and/or
somatic coliphages (18.1%) by the Wilcoxon matched-pairs signed-ranks test, the difference was
very significant (P = 0.0078). This indicates that SAL detection of both coliphage groups is
better than detecting F+ coliphages alone. A similar comparison for SAL detection of somatic
coliphages alone (16%) versus the SAL detection of either or both coliphages when both are
measured in a sample (18.1%) indicated no significant difference because the sample size was
too small. Overall, SAL detection of both groups of coliphages (F+ and somatic) is better than
SAL detection of either group alone.
82
-------
Dual versus individual detection of bacterial indicator pairs. When the detection of two fecal
indicator bacteria such as fecal coliforms and/or E. coli versus enterococci is considered because
both groups are measured simultaneously, the frequency of detecting a positive sample (positive
for one or the other or both) was also high at 17.2% (Table 11). For each group alone, the
detection frequency was 12.1% for enterococci and 11.2% for fecal coliforms and/or E. coli
(Table 10). The dual detection of either or both of these bacterial indicator groups in a sample
(18.1%) was compared to the frequency of detection of each of them alone (enterococci = 12.1%
and fecal coliforms and/or E. coli =11.2%) using the Friedman test. The P value was significant
(0.0366), indicating the detection frequencies were significantly different and were highest for
the detection of either or both indicator bacteria groups when both are measured in a sample.
When the individual detection frequencies of enterococci (12.1%) and fecal coliforms and/or E.
coli (11.2%) were compared by the non-parametric Wilcoxon matched-pairs signed-ranks test,
there was no significant difference (P= 0.839), indicating equivalent detection of either of these
bacterial indicator groups alone. Furthermore, when the detection frequencies of enterococcus
alone (12.1%) was compared to the dual detection of either or both enterococcus and/or fecal
coliforms and/or E. coli (17.2%) by the Wilcoxon matched-pairs signed-ranks test, the difference
was significant (P = 0.031). This indicates that dual detection of both bacterial indicator groups
is better than detecting enterococci alone. A similar comparison for detection of fecal coliforms
and/or E. coli alone (11.2%) versus the dual detection of either or both enterococcus and/or fecal
coliforms and/or E. coli (17.2%) by the Wilcoxon matched-pairs signed-ranks test also was
significant (P = 0.016). This indicates that detection of both bacterial indicator groups is better
than detecting fecal coliforms and/or E. coli alone. Overall, detection of both groups of bacteria
is better than detection of either group of bacteria alone.
83
-------
Dual versus individual detection of bacterial and coliphage indicator pairs. The extent to
which dual detection of a coliphage indicator and bacterial indicator versus individual detection
of either one alone was considered. This is because the detection of both a coliphage indicator
and a bacterial indicator in a groundwater sample has been an option for the proposed
groundwater rule. The first pair of coliphage and bacteria indicators to compare was somatic
coliphages detected by the SAL method (16.4%) and enterococci (12.1%) (Table 10). This pair
was chosen because these were the individual coliphage and bacterial indicators measured in all
samples and detected most frequently. For SAL somatic coliphages and/or enterococcus being
measured together in samples, the frequency of detecting a positive sample (positive for one or
the other or both) was 20.7% (Table 11), which was higher than either indicator alone or any
coliphage pair or bacterial pair. Furthermore, a statistical comparison of measuring either
indicator alone or both indicators together showed significant improvement in detecting fecal
contamination of groundwater. The detection of either or both of these indicator groups in a
sample (20.7%) was statistically compared to the frequency of detection of each of them alone
(enterococci = 12.1% and SAL somatic coliphages = 16.4%) using the Friedman test. The P
value was significant (0.028), indicating detection frequencies were significantly different and
were highest for the detection of either or both indicator groups (enterococci and/or SAL somatic
coliphages) when both are measured in a sample. When the individual detection frequencies of
enterococci (12.1%) and SAL somatic coliphages (16.4%) were compared by the non-parametric
Wilcoxon matched-pairs signed-ranks test, there was no significant difference (P= 0.355),
indicating equivalent detection of either of these indicator groups alone. Furthermore, when the
detection frequencies of enterococcus alone (12.1%) were compared to the detection of either or
both enterococcus and/or SAL somatic coliphages (20.7%) by the Wilcoxon matched-pairs
84
-------
signed-ranks test, the difference was very significant (P = 0.002). This indicates that detection of
both indicator groups is better than detecting enterococci alone. A similar comparison for
detection of SAL somatic coliphages alone (16.4%) versus the detection of either or both
enterococcus and/or SAL somatic coliphages (20.7%) by the Wilcoxon matched-pairs signed-
ranks test was not significantly different (P = 0.206). Overall, the detection of both groups of
indicators (SAL somatic coliphages and enterococci) was generally better than detection of either
indicator group alone.
When the coliphage with the highest detection frequency, which was SAL somatic coliphages
(16.4%), and the bacterial indicator with the highest detection frequency (and measured in any
project samples), which was fecal coliforms and/or E. coli (11.2%), were being measured
together in samples, the frequency of detecting a positive sample (positive for one or the other or
both) was 22.4%% (Table 11). This indicator detection frequency was even higher than either
indicator alone or any coliphage pair or any bacterial pair. Furthermore, a statistical comparison
of measuring either indicator alone or both indicators together showed significant improvement
in detecting fecal contamination of groundwater. The detection of either or both of these
indicator groups in a sample (SAL somatic coliphages and/or fecal coliforms and/or E. coli =
22.4%) was statistically compared to the frequency of detection of each of them alone (SAL
somatic coliphages = 16.4% and fecal coliforms and/or E. coli =11.2%) using the Friedman
test). The P value was very significant (0.003), indicating the detection frequencies were
significantly different. When the individual detection frequencies of fecal coliforms and/or E.
coli (11.2%) and SAL somatic coliphages (16.4%) were compared by the non-parametric
Wilcoxon matched-pairs signed-ranks test, there was no significant difference (P= 0.276),
85
-------
indicating equivalent detection of either of these indicator groups alone. Furthermore, when the
detection frequency of SAL somatic coliphages alone (16.4%) was compared to the combined
detection of either or both fecal coliforms and/or E. coli and/or SAL somatic coliphages (22.4%)
by the Wilcoxon matched-pairs signed-ranks test, the difference was significant (P = 0.039).
This indicates that detection of both indicator groups is better than detecting SAL somatic
coliphages alone. A similar comparison for detection of fecal coliforms and/or E. coli alone
(11.2%) versus the combined detection of either or both fecal coliforms and/or E. coli and/or
SAL somatic coliphages (22.4%) by the Wilcoxon matched-pairs signed-ranks test was
extremely significantly different (P = 0.0002). Therefore, the dual detection of both a coliphage
indicator and a bacteria indicator in groundwater samples is better than detecting either indicator
group alone.
Because there is interest in using a single coliphage host to simultaneously detect both somatic
and male-specific (or "total") coliphages, it also was of interest to examine the dual detection of
total coliphages and a bacterial indicator. When "total" coliphages detected by the SAL method
(detection frequency =12.1%) and the bacterial indicator of enterococci (detection frequency =
12.1 %) are considered together, the frequency of detecting a positive sample (for either or both
indicators) was 19.8% (Table 11). A statistical comparison of measuring either indicator alone
or both indicators together showed significant improvement in detecting fecal contamination of
groundwater. The detection of either or both of these indicator groups in a sample (19.8%) was
statistically compared to the frequency of detection of each of them alone (SAL total coliphages
= 12.1% and enterococci = 12.1%) using the Friedman test. The P value was significant (0.011),
indicating the detection frequencies were significantly different. When the individual detection
86
-------
frequencies of enterococci (12.1%) and SAL total coliphages (12.1%) were compared by a non-
parametric t-test (Wilcoxon matched-pairs signed-ranks test), there was no significant difference
(P>0.999), indicating equivalent detection of either of these indicator groups alone. Furthermore,
when the detection frequencies of SAL total coliphages alone (12.1%) was compared to the dual
detection of either or both SAL total coliphages and enterococci (19.8%) by the Wilcoxon
matched-pairs signed-ranks test, the difference was very significant (P = 0.0039). This indicates
that detection of both indicator groups is better than detecting SAL total coliphages alone. A
similar comparison for detection of enterococci alone (12.1%) versus the dual detection of either
or both enterococcus and/or SAL total coliphages (19.8%) by the Wilcoxon matched-pairs
signed-ranks test the difference also was very significant (P = 0.0039). Therefore, dual detection
of both a coliphage indicator ("total" coliphages by SAL) and a bacteria indicator (enterococci) in
groundwater samples is better than detecting either indicator group alone. This is the case even
for the single coliphage indicator capable of detecting both male-specific and somatic coliphages
("total" coliphages).
Detection of "total" coliphages by SAL (detection frequency = 12.1%) and the bacterial indicator
of fecal coliforms and/or E. coli (detection frequency = 11.2%) also were considered together,
and the frequency of detecting a positive sample (for either or both indicators) was 20.7% (Table
11). A statistical comparison of measuring either indicator alone or both indicators together
showed significant improvement in detecting fecal contamination of groundwater. The detection
of either or both of these indicator groups in a sample (20.7%) was statistically compared to the
frequency of detection of each of them alone (SAL total coliphages = 12.1% and fecal
coliforms/^. coli =11.2%) using the Friedman test. The P value was very significant (0.005),
87
-------
indicating the detection frequencies were significantly different and were highest for the
detection of either or both indicator groups (fecal coliforms/£. coli and SAL total coliphages)
when are both measured in a sample. When the individual detection frequencies of the bacterial
indicator (fecal coliforms/£. coli =12.1%) and SAL total coliphages (11.2%) were compared by
the non-parametric Wilcoxon matched-pairs signed-ranks test, there was no significant difference
(P = 0.865), indicating equivalent detection of either of these indicator groups alone.
Furthermore, when the detection frequencies of SAL total coliphages alone (12.1%) was
compared to the detection of either or both SAL total coliphages and/or fecal coliforms/£. coli
(20.7%) by the Wilcoxon matched-pairs signed-ranks test, the difference was very significant (P
= 0.002). This indicates that detection of both indicator groups is better than detecting SAL total
coliphages alone. A similar comparison for detection of fecal coliforms/£. coli alone (11.2%)
versus the dual detection of either or both fecal coliforms/£. coli and/or SAL total coliphages
(20.7%) by the Wilcoxon matched-pairs signed-ranks test also was extremely significant (P =
0.001). Therefore, these results again show that detection of both a coliphage indicator ("total"
coliphages) and a bacterial indicator (fecal coliforms and/or E. coli) in groundwater samples is
better than detecting either indicator group alone.
It is noteworthy that the frequency of detecting fecal contamination in a groundwater sample was
nearly as high with "total" coliphages detected by SAL and a bacteria indicator (with either
enterococci or fecal coliforms/^. coli at 19.8 and 20.7%, respectively) as with somatic coliphages
detected by SAL with a bacteria indicator (with either enterococci or fecal coliforms/^. coli;
20.7% and 22.4%, respectively). Overall, these results indicate that the measurement of both a
coliphage indicator and a bacterial indicator together in a groundwater sample gives a higher
88
-------
frequency or likelihood of detecting fecal contamination than measuring any single indicator
alone or even measuring pairs of bacterial indicators or pairs of coliphage indicators.
Statistical Comparisons of Fecal Indicators in Groundwater Samples
Additional statistical analyses of data on indicator occurrence in groundwater samples were
performed for the data set of all groundwater samples for which human enteric viral analysis was
done. Initially, statistical analysis was done on samples for which there were results for all of the
fecal indicators tested by all methods. Because fecal coliforms were not analyzed by one of the 4
participating labs, they were excluded from some analyses that required complete sample and
indicator pairing or matching (no missing data for any indicator in any sample). An analysis of
all indicator data was done by a repeated measures ANOVA (Friedman Test), which assumes
that the data in each row (which represents a water sample) is matched (a reasonable assumption
because it is for a specific sample). The analysis gave a P-value of <0.0001, which is highly
significant, and therefore, indicates that variation among column medians (for microbial
indicators and coliphage methods) is significantly greater than expected by chance. Hence, the
different indicators and coliphage methods gave significantly different results in detecting fecal
contamination. Interestingly all post-tests for significant differences between each possible
combination of indicator pair were not significant, with all P-values >0.05.
The data for all indicators, including fecal coliforms, also were analyzed by the Kruskal-Wallis
Test (Nonparametric one-way ANOVA), which assumes no matching and does not require a
complete matrix, thereby allowing for missing data for some samples. This analysis can be
justified on the basis of simply considering an unknown distribution of microbial indicators of
89
-------
fecal contamination in groundwater with unknown variability over time and space. The P value
was significant at 0.0086, indicating that variation among column medians was significantly
greater than expected by chance. This analysis again shows that different indicators and
coliphage methods gave significantly different results in detecting fecal contamination.
These indicator data for groundwater samples were re-analyzed based on sample positivity
instead of microbe concentrations by dichotomizing the data as positive and negative samples
and assigning a value of 1 to a positive sample and keeping 0 for a negative sample. This
analysis was again done by a repeated measures ANOVA (Friedman Test), which assumes that
the data in each row (which represents a water sample) is matched (a reasonable assumption
because it is results for the same specific sample). The analysis gave a P-value of <0.0001,
which is highly significant, and therefore, indicates variation among column medians (for
microbial indicators and coliphage methods) is significantly greater than expected by chance.
Hence, the different indicators and methods gave significantly different results in detecting fecal
contamination. Interestingly all post-tests for significant differences between each possible
combination of indicator pair were not significant, with all P-values >0.05. When these data
were re-analyzed by the Kruskal-Wallis Test (Nonparametric one-way ANOVA), which assumes
no matching and does not require a complete matrix, thereby allowing for missing data in some
samples, the P value also was significant at 0.0086. This result indicates that variation among
column medians (for the different fecal indicators and coliphage methods) was significantly
greater than expected by chance.
90
-------
These indicator data for sample positivity in groundwater samples also were re-analyzed by a
non-parametric repeated measures Analysis of Variance (Friedman Test) by excluding the fecal
coliform data and including only those samples that were analyzed for human enteric viruses.
The P value for this test was 0.0001, considered extremely significant, and indicating that
variation among column means (for the different fecal indicators and coliphage methods) was
significantly greater than expected by chance. As a post-test to the Friedman Test, the Tukey-
Kramer multiple comparisons test was used to determine if there were significant differences in
sample positivity between pairs of indicators. Significant differences in sample positivity were
observed for the indicator pairs of: SAL somatic versus Enrichment F+ (P<0.01), SAL somatic
versus Enrichment "total" coliphage (E. coli C3000) (P<0.05), and SAL somatic versus E. coli
(P<0.01). None of the other indicator pairs were significantly different (P>0.05).
Analysis for Enteric Viruses in Groundwater
A total 106 samples of groundwater were analyzed for enteric viruses and no enteric viruses were
detected in any of the samples analyzed. All but three (3) of the groundwater samples were 1500
liters in volume. There were 3 well samples from the Northeast for which the sample volume
was less than 1500 liters due to a lack of water in the well or to clogging of the filter used to
concentrate viruses from water. These Northeast wells were: Well #2 = 592 L, Well #4 = 229 L,
and Well #14 = 400 L.
None of the 106 groundwater samples collected and analyzed were positive for human enteric
viruses by cell culture and (RT-)PCR for adenoviruses, astroviruses, enteroviruses, reoviruses,
rotaviruses or hepatitis A virus or, in the case of caliciviruses, by direct RT-PCR analysis.
91
-------
Presumptive positive results observed in one participating lab for detachment of cells from flasks
of all samples were due to non-viral effects such as the action of the trypsin in the medium or to
cytotoxicity from the sample concentrate inocula. Trypsin can cause the cells to dislodge from
the surface of the flask and appear abnormal. Trypsin effects are the most likely explanation
because cell detachment was also observed in the negative control cultures. Regardless of the
cause of this effect, it was not due to the presence of any enteric viruses, based on the negative
results from sample analysis by nucleic acid amplification methods. Therefore, despite well
developed protocols and the analysis of large sample volumes for a range of human enteric
viruses, none were found in any of the samples analyzed. It is also noteworthy that no laboratory
experienced any episode of false positive viral contamination in negative control samples or
other types of virus-free control samples. Hence, no viral contamination occurred that could
have compromised the interpretation of positive results had there been any virus-positive field
samples.
Survival of Coliphage in Seeded Groundwater
Coliphage survival in groundwater after storage at 4°C for times periods of 0-6 days was
determined for samples seeded with sufficient coliphages (as filtered raw sewage) to give about
30-70 PFU per 100 mL when assayed by the SAL method. Duplicate experiments were done and
the average results of these experiments are summarized in Figure 5 below.
92
-------
Figure 5. Survival of F+, Somatic and "Total" Coliphages in Seeded 100-mL Samples of
Groundwater Held at 4°C and Assayed by the SAL Method on Days 0, 2, 3 and 6.
F+ Q Somatic s Total
2 3
Time (Days)
As shown in Figure 5, F+, somatic and "total" coliphages survived relatively well for 2 or 3 days,
with average survivals of >70% compared to day zero when detected by the SAL method
(Method 1602). By day 6, SAL coliphage titers were somewhat lower, with average survivals of
about 40-60% compared to day zero. Overall, these results indicate that samples of groundwater
for coliphage analysis by SAL can be held for periods of 2 or 3 days with only relatively minor
losses in coliphage titer (<30%) and with high probabilities of detecting coliphages that were
initially present when the samples were collected. These coliphage survival data were subjected
to statistical analyses by both parametric and non-parametric analysis of variance (ANOVA). The
coliphage titers were not significantly different at the 5% level (p > 0.05) among the days of
analysis (days 0, 2, 3 and 6). These results indicate that coliphage titers in groundwater as
measured by SAL did not significantly decrease over the 6-day holding period.
93
-------
Coliphage survival in groundwater after storage at 4°C for times periods of 0-6 days was also
determined for groundwater samples seeded with sufficient coliphages (as filtered raw sewage) to
give about 5 infectious units per sample bottle mL of 1000 mL when assayed by the enrichment
method (Method 1601). As shown in Table 12, F+, somatic and "total" coliphages survived
relatively well for as long as 6 days, with the number of positive enrichment bottles out of 10
remaining at high levels of 7 to 10.
Table 12. Survival of Coliphages in Seeded Groundwater Held at 4°C and Assayed by the Two-
Step Enrichment Spot-plate Method
Time
(Days)
0
2
3
6
No. of Positive Enrichment Bottles of 10 for Indicated Coliphage Group
F+ Coliphages
8
10
9
9
Somatic Coliphages
10
8
8
8
"Total" Coliphages
10
10
10
7
Overall, these results indicate that samples of groundwater for coliphage analysis by the
enrichment method can be held for periods of as long as 6 days with no appreciable loss losses in
coliphage titer and with high probabilities of detecting coliphages that were initially present
when the samples were collected. These coliphage survival data were subjected to statistical
analyses by both parametric and non-parametric analysis of variance (ANOVA). The coliphage
titers were not significantly different at the 5% level (p > 0.05) among the days of analysis (days
0, 2, 3 and 6). These results indicate that low coliphage titers in groundwater (about 5 infectious
94
-------
units per liter) as measured by enrichment did not significantly decrease over the 6-day holding
period.
Comparison of Coliphage, Bacterial Indicator and Enteric Virus Detection in This Study
and in Previous Studies in the USA
Only a few previous studies have examined coliphages and bacteria in groundwater of the USA.
In one study by Abbaszadegan et al. (1999), coliphages were analyzed in the equivalent of about
15-liter samples of water using eluates from adsorbent filters used to concentrate enteric viruses
from groundwater samples. Coliphages were assayed on the following host bacteria: E. coli C
for somatic coliphages (this host is similar t E. coli CN-13), Salmonella WG49 for F+ coliphages
and E. coli C3000 for both somatic and male-specific ("total" coliphages). Of the 444 samples
analyzed the percentages of positive samples were: 4.1% onE. coli C, 10.8% onE. coli C3000
and 9.5% on Salmonella WG49. The rates of coliphage positivity in this previous study are
lower than the rates of positivity in this current study. In the current study 16.4% of samples
were positive for somatic coliphages detected in 100-mL sample volumes by the SAL method on
E. coli CN-13 (Method 1602) compared to 4.1% positive for somatic coliphages detected onE.
coli C. In the current study 11.2% of samples were coliphage positive for F+ coliphages on host
E. coli Famp by SAL compared to 9.5% positive for F+ coliphages on Salmonella WG49. In the
current study 12.1% of samples were positive for "total" coliphages on E. coli C3000 compared
to 10.8% positive on this host in previous studies. The percent of samples positive for any of the
three coliphage hosts was 20.7% and for all three hosts together it was 0.2%. In the current study
the percent of samples positive for the coliphage host pair of E. coli CN-13 (somatic coliphages)
and E. coli Famp (F+ coliphages) was 20.7%. Thus the rate of positivity of two hosts in the
95
-------
current study was the same as the rate of positivity for 3 hosts in this previous study.
Furthermore, these results for the current study employed 100-mL samples assayed by SAL
compared to 15-liter samples assayed as filter eluate concentrates by the double agar layer plaque
assay in sample concentrate volumes of 2.5 or 5 mL per plate. The recovery efficiency and lower
detection limit of the coliphage assay method used in the previous study was not reported.
In the same study by Abbaszadegan et al. (1999), culturable human enteric viruses were analyzed
in the equivalent of 160-gallon (605-liter) samples of water by CPE in BGM cell cultures. In 442
samples, 4.8% of sample sites and 4.1% of total samples were positive for culturable viruses by
CPE. In comparison, no culturable human enteric viruses were detected in any of the 106
groundwater samples, each of 1500-liter (400-gallon) volume, analyzed in the current study.
In a later study Karim et al. (2004) sampled 20 groundwater wells monthly for 12 months from
11 states for coliphages, bacterial indicators and human enteric viruses. Sixteen of the wells were
known to be fecally contaminated. Wells were monitored for the presence of culturable viruses,
enteric virus nucleic acid (enterovirus, hepatitis A, norwalk virus, rotavirus, and adenovirus) by
(RT-)PCR, coliphages using USEPA Methods 1601 and 1602, double agar layer method (DAL),
and RT-PCR, and indicator bacteria (total coliforms, E. coli, enterococci, and Clostridium
perfringens spores). A total of 231 to 235 samples were analyzed per well. The percentages of
(RT-)PCR-positive samples for enteric viruses were: 2.1% for enteroviruses, 0% for HAV, 5.6%
for rotavirus, 4.3% for Norwalk Virus, and 0.4% for adenovirus. For culturable viruses by CPE,
positivity was 3.9%. As previously indicated, no human enteric viruses were detected in this
current study.
96
-------
For coliphage indicators detected by the enrichment method (1601) the percentage of positive
samples was 0% for somatic coliphages in 100 mL and 1000 mL volumes, and 0.4 and 2.2% for
F+ coliphages in 100 mL and 1000 mL, respectively. These are lower rates of sample positivity
than were obtained in this current study, which were 6.9% for somatic and 3.4% for F+
coliphages in 1000 mL volumes. For coliphage indicators detected by the SAL method (1602) in
100-mL sample volumes, the percentage of positive samples was 0.9% for somatic coliphages
and 5.6% for F+ coliphages. These also are lower rates of positivity than in this current study,
which were 16.4% and 11.2% for somatic and F+ coliphages, respectively. For enterococcus, the
percentage of positive samples was 0.4 and 5.5% for 100 mL and 1000 mL volumes respectively.
In this current study enterococcus positivity in 100- mL samples was much higher at 12.1%. For
E. coli, the percentage of positive samples was 4.3% and 11.1% for 100 mL and 1000 mL
samples, respectively. In this current study, the frequency of E. co/z'-positive 100-mL samples
was 4.3%, which is the same E. co/z-positivity the as the in previous study.
The results of the previous study suggested that dual monitoring for both a bacterial indicator and
coliphage would be useful for detecting fecal contamination of groundwater. As in this current
study, monitoring coliphages and bacteria together detected fecally contaminated wells more
frequently than either a coliphage or a fecal bacterial indicator alone. As a single indicator, total
coliforms in 1-L sample volumes were found to occur most frequently (80% of the wells and
38.3% of the samples). However, total coliforms are not fecal indicator bacteria and in our
opinion would not seem to be appropriate or useful as a single fecal indicator organism. The
authors of the previous study concluded that the dual measurement of both a coliphage and a
bacterial indicator would increase the detection of fecally contaminated groundwater. No single
97
-------
fecal indicator alone was as effective in detecting fecal contamination as the dual use of a
coliphage and a bacterial indicator. These previous findings are consistent with those of this
current study, which also found that the dual measurement of two indicators and especially a
coliphage and a bacterial indicator increased the likelihood of detecting fecally contaminated
groundwater.
Responses to Questions and Comments of the April 2004 Coliphage Workshop
Several questions about coliphage methods were identified by participants at an "International
Workshop on Coliphages as Indicators of Fecal Contamination in Water and Other
Environmental Media," that was sponsored by US EPA and held in Arlington, VA, April 20-21,
2004. These questions and our responses to them are given below in this section of the report.
1. Costs of the coliphage tests?
Response. The four participating laboratories have estimated the costs of coliphage testing and
these costs are listed in the Table below.
98
-------
Table 13. Costs of Coliphage Analysis by the Four Study Laboratories - November 2004
Coliphage Assay Costs
UNC
SAL 100 mL water sample assayed for each host
C3000 Famp CN13 Famp+CN13
Labor and materials $54 $55 $55 $84
46% indirect costs $25 $25 $25 $38
Total $79 $80 $80 $122
Enrichment C3000 Famp CN13 Famp + CN13
Labor and materials $67 $68 $68 $92
46% indirect costs $31 $31 $31 $42
Total $98 $99 $100 $134
TAMU
Actual Costs of doing coliphage analysis.
Single Agar Layer (per host bacterium) per sample
Total Labor time: (3.5 hours @ $ 20.00/hour): $70.00
Material costs: $10.00
Total cost: $ 80.00
2-step Enrichment
Total Labor Time: 3 hours @ $20.00/hour: 20.00
Material costs: $ 10.00
Total Cost: $ 70.00
These are the costs per sample, per host bacterium and does not include "overhead" or other costs.
The labor includes media preparation, analysis time and data recording.
UNH -All Methods
C3000 Famp CN13 Famp + CN13
$68 $68 $68 $103
$32 $32 $32 $47
$100 $100 $100 $150
U of Minn
SAL 100 mL water sample assayed for each host
C3000 Famp CN13
Labor and materials $80 $80 $80
49% indirect costs $39 $39 $39
Total $119 $119 $119
Enrichment
C3000 Famp CN13
Labor and materials $100 $100 $100
49% indirect costs $49 $49 $49
Total $149 $149 $149
2. The need for and effectiveness of the method of confirming coliphage-positives in the tests?
In this study the plaque conformation procedure was carefully studies. I was found that the
plaque confirmation rate based on the development of lysis or plaques on spots of host lawns in
99
-------
agar medium averaged nearly 80%. It was concluded this was a simple and sufficiently reliable
confirmation method and would be adequate for routine use by labs doing coliphage analysis
3. The issue of using a singe indicator, such as a bacterial indicator or a coliphage indicator,
versus using both a bacterial indicator and a coliphage indicator in detecting a positive ground
water sample? Specifically, the extent to which there is increased detection of positives when
using only one indicator such as a bacterial indicator versus two indicators - a coliphage indicator
and a bacterial indicator?
The results of the phase II field studies of this project show quite clearly that the dual measure of
two indicators, especially a coliphage indicator and a bacterial indicator significantly increases
the frequency of getting a positive sample. Measuring either a coliphage or a bacterial indicator
alone gave significantly lower detection of positive samples. Therefore, the results of this study
support the use of both coliphage and bacterial indicators together in the analysis of groundwater
samples for evidence of fecal contamination.
4. The choice of coliphages to detect: somatic, male-specific or "total" coliphages?
The results of the current study provide data showing that the frequency of detecting a coliphage
in ground water is highest for somatic coliphages and nearly as high for "total" coliphages using
either the SAL (Method 1602) or enrichment (Method 1601) methods. Therefore, it is concluded
from these results that either of these coliphage groups are likely to give greater detection of
coliphages than the measurement of F+ coliphages. However, F+ coliphages also are important
indicators of fecal contamination. Because they have the ability to distinguish human from
animal fecal contamination F+ coliphages and especially F+ RNA coliphages also have merit as
100
-------
coliphage indicators of fecal contamination of groundwater.
5. Whether or not study wells were subjected to treatment (disinfection)?
The wells of the current study were not disinfected. Most were non-community public water
supplies, some were private wells and some were public water supplies. These wells were not
required to disinfect or otherwise treat in the States where the wells were located. Because two
of the wells in one state had periodic coliform violations, they are now routinely chlorinated.
However, at the time of the study they were not being chlorinated.
101
-------
SUMMARY AND CONCLUSIONS
In initial studies Methods 1601 and 1602 were evaluated for their ability to detect somatic, male-
specific (F+) and total (somatic plus F+) coliphages in groundwater samples seeded with mixed,
natural populations of coliphages from sewage. The SAL method (Method 1601) was applied in
10 experiments to replicate 100-mL volumes of groundwater seeded with sewage coliphages for
coliphage detection with each of three E. coli host bacteria: E. coli CN-13 for somatic
coliphages, E. coli Famp for male-specific (F+) coliphages and E. coli C3000 for somatic plus F+
("total") coliphages. There was efficient coliphage detection (average 53%) and plaque
confirmation (average 78%) in 100-mL volumes of ground water. Overall, the results of these
studies indicate that there is high likelihood of detecting even low levels of coliphages in 100-mL
volumes of ground water using Method 1602.
For evaluation of the enrichment method (Method 1601), recoveries of somatic, F+ and total
coliphages from 10 replicate 1-liter volumes of seeded ground water in eight replicate
experiments were efficient at coliphage input levels of about 1.5 to 3 infectious units/L.
Recoveries were somewhat variable but close to those expected based on the expected number of
positive 1-liter enrichment bottles out of a total of 10. There is a high likelihood of detecting as
few as 1-3 coliphages in 1-liter volumes of water using the two-step enrichment methods of
Method 1601. For both Method 1601 and 1602, the results of studies with seeded samples of
groundwater showed that holding samples at 4°C for up to 3 days did not significantly reduce the
ability to detect low levels of coliphages. Hence samples can be collected, shipped and stored
prior to assay.
102
-------
These improved and validated versions of Methods 1601 and 1602 were further validated and
evaluated studies of coliphage detection and occurrence in more than 100 field samples of
groundwater from wells located in 4 different geographic regions of the USA (Northeast,
Southeast, Southwest and upper Midwest). Overall, the results for fecal indicator occurrence in
the field groundwater samples analyzed in this study indicated that coliphages are reliable
indicators for detecting fecal contamination and can detect fecal contamination as frequently or
more frequently than do bacterial indicators. In more than 100 groundwater samples collected
from wells coliphages were detected with greater or similar frequency than were fecal indicator
bacteria. The percentages sample positivity for coliphages were 11 to 16% by the SAL method
(Method 1602) and 6.9 to 3.4% by the two-step enrichment spot plate method (Method 1601).
By comparison, the percentages of sample positivity for bacteria (fecal coliforms, E. coli or
enterococci) ranged from 13.8 to 4.3%.
Coliphage detection in groundwater was higher using the SAL assay (Method 1602) on 100-mL
sample than using the two-step enrichment spot plate method (Method 1601) on 1-liter samples.
Additionally, coliphage detection by either method was highest for somatic coliphages, next
highest for "total" coliphages and lowest for F+ coliphages. The relatively high detection of
"total" coliphages by the SAL method indicates that a single host, E. coli C3000, can be used to
detect either somatic or male-specific coliphage or both with a high degree of sensitivity.
The results from the analyses of these groundwater samples indicate that there is a significantly
greater likelihood of detecting fecal contamination if two indicators are analyzed in the same
sample than if only one indicator is analyzed. Detection of two indicators was higher with a
103
-------
coliphage and a bacteria indicator pair (as high as 20.7 and 22.4% positivity) than with either
pairs of coliphage indicators (as high 18% positivity) or pairs of bacterial indicators (17.2%
positivity). Therefore, rates or frequencies of detecting fecal contamination in groundwater are
higher when using two fecal indicators than using a single fecal indicators, and highest when
using a coliphage and a bacterial indicator together. These findings clearly support the position
of determining groundwater vulnerability to fecal contamination by measuring both a coliphage
indicator and a bacterial indicator, rather than measuring either one alone.
Human enteric viruses, including adenoviruses, astroviruses, enteroviruses, hepatitis A virus,
reoviruses and rotaviruses were not detected by combined cell culture and (RT-)PCR in any of
the 106 samples analyzed. Human Caliciviruses (Noroviruses) were not detected by direct RT-
PCR of virus concentrates from the same 106 samples groundwater. Therefore, it was not
possible to compare or look for associations in occurrence of coliphages and/or bacteria relative
to human enteric viruses in groundwater samples.
It is recommended that EPA adopt these improved methods for coliphage detection for the
forthcoming Groundwater Rule. It is also recommended that the Groundwater Rule require the
analysis of both coliphages and fecal indicator in the same sample of groundwater in order to
significantly increase the likelihood of detecting fecal contamination. Examination of
groundwater samples for a single indicator, either a virus or a bacterium, will significantly reduce
the chances of detecting fecal contamination.
104
-------
REFERENCES
Abbaszadegan, M., P.W. Stewart, M.W. LeChevallier, J.S. Rosen and C.P. Gerba (1999)
Occurrence of Viruses in Ground Water in the United States. American Water Works Service
Company. Belleville, IL and Voorhees, NJ.
American Public Health Association. (1995) Section 9230. Standard Methods for the
Examination of Water and Wastewater. 19th ed. American Public Health Association,
Washington DC.
Chapron, C.D., Ballester, N.A., Fontaine, J.H., Frades, C.N. & Margolin, A.B. 2000 The
Detection of Astro virus, Entero virus and Adeno virus Type 40 and 41 in Surface Waters
Collected and Evaluated by the Information Collection Rule and Integrated Cell Culture/Nested
PCR Procedure. Appl. and Environ. Microbiol, 60 (6), 2520-2525.
De Leon, R., Y. S. C. Shieh, R. S. Baric, and M. D. Sobsey. 1990. Detection of enteroviruses and
hepatitis A virus in environmental samples by gene probes and polymerase chain reaction. Proc.
Water Quality Technology Conference, 1989, pp. 833-853. Amer. Water Works Assoc., Denver,
Co.
Environmental Protection Agency. 1995 Virus Monitoring Protocol for the Information
Collection Requirements Rule. U.S. Environmental Protection Agency, publication EPA/814-B-
95-002. Government Printing Office, Cincinnati, Ohio.
105
-------
Fout, Shay (2001) personal communication. Not yet published.
Gouvea, V., R. I. Glass, P. Woods, K. Tanuguchi, H. F. Clark, B. Forester, and Z. Y.
Fang. 1990.Polymerase Chain Reaction Amplification and Typing of Rotavirus Nucleic Acid
From Stool Specimens. J. Clin. Microbiol. 59:276-282.
Hsu FC, Shieh YS, van Duin J, Beekwilder MJ, Sobsey MD. (1995) Genotyping male-specific
RNA coliphages by hybridization with oligonucleotide probes. Appl Environ Microbiol. 1995
Nov;61(ll):3960-6.
Karim, M.R., M.W. LeChevallier, M. Abbaszadegan, A. Alum, J. Sobrinho and J. Rosen (2004)
Field Testing of USEPA Methods 1601 and 1602 for Coliphage in Groundwater Awwa Research
Foundation, American Water Works Association, Denver, CO.
Levin, M.A., J.R. Fischer and V.J. Cabelli. 1975. Membrane Filtration Technique for
Enumeration of Enterococci in Marine Waters. Applied Microbiology 30: 66-71.
Meschke, J. S., and M. D. Sobsey. 1998. Comparative adsorption of Norwalk virus, poliovirus 1
and F+ RNA coliphage MS2 to soils suspended in treated wastewater. Water Sci. Tech. 38:187-
189.
Schwab, K. J.,R. De Leon, and M. D. Sobsey. 1995. Concentration and purification of beef
extract mock eluates from water samples for the detection of enteroviruses, hepatitis A virus and
106
-------
Norwalk virus by reverse transcription-PCR. Appl. Environ. Microbiol. 61:531-537.
Schwab, K. J.,R. De Leon, and M. D. Sobsey. 1995. Concentration and purification of beef
extract mock eluates from water samples for the detection of enteroviruses, hepatitis A virus and
Norwalk virus by reverse transcription-PCR. Appl. Environ. Microbiol. 61:531-537.
Sobsey, M. D., K. S. Schwab, R. De Leon, and Y. S. C. Shieh.l996.Enteric Virus Detection by
Nucleic Acid Methods. American Water Works Association Research Foundation, Denver.
USEPA (2002) Method 1600: Enterococci in Water by Membrane Filtration Using membrane-
Enterococcus Indoxyl-p-D-Glucoside Agar (mEI), 15pp. EPA-821-R-02-022. U.S.
Environmental Protection Agency, Office of Water, Washington, DC 20460.
Vinje, J. et al. 2001. Evaluation of different RT-PCR primer pairs for the detection of low
numbers of "Norwalk-like viruses". Poster presented at American Society for Microbiology
annual meeting, Orlando, Florida, May, 2001.
Vinje J, Hamidjaja RA, and Sobsey MD. 2004 Detection of novel genotypes using VP1 typing
for genetic classification of genogroup I and II noroviruses. J. Virological Methods. 116(2): 109-
117.
Xu, W., M. C. McDonough, and D. D. Erdman. 2000. Species-specific identification of human
adenoviruses by a multiplex PCR assay. J. Clin. Microbol. 38:4114-4120.
107
-------
APPENDICES
APPENDIX I
EPA Coliphage Groundwater Study
Summary Report of the Northeast Region
Mark D. Sobsey, Study Principal Investigator, University of North Carolina
and
Aaron Margolin, Co-Principal Investigator, and Nicola Ballester
University of New Hampshire, Durham, NH
November, 2004
Introduction
The purpose of this study was to validate and apply US EPA Methods 1601 and 1602 for
detection coliphages in water by applying them to field samples of groundwater. The goal was to
examine 27 samples of groundwater, preferably from public water supply wells, for somatic, F+
and total coliphages, fecal indicator bacteria and human enteric viruses in the Northeast United
States.
108
-------
Methods and Materials
Sampling sites
All sample sites were located in New England. Eight well sites were public water sources and 17
were private wells. A total of 25 wells were sampled instead of 27 due to a very severe and harsh
winter. NH had its first snowfall at the end of October and a second snowfall at the beginning of
November, 2002. Plans to sample two additional wells as soon as the weather permitted could
not be carried out because New England experienced one of the snowiest winters ever. Therefore,
only 25 well samples were collected and analyzed. Of the 25 wells, there were 12 sample sites in
New Hampshire, two of which were from public wells that were approximately 500 and 700 ft
deep, respectively. None of these wells had any form of disinfection. The other wells from NH
were all private wells. These wells also were not disinfected. One well from NH was a private,
very shallow well, less then 35 feet deep and lined with stone. This was not considered a potable
well but was used for farm irrigation. Four sites in Maine were all privately owned wells and not
disinfected. Three sites were in Vermont, and they were all privately owned wells and not
disinfected. All of the privately owned wells were drilled wells, excepted for the one in NH as
indicated above, and they were of varying depths that were unknown to the homeowner at the
time samples were collected. There were 6 samples from public water supply wells in
Massachusetts. The public water supplies in Massachusetts were chosen due to positive results
previously found for total and fecal coliforms, enterococci, and male-specific coliphages.
Additionally 3 of the 6 locations had positives previously reported for rotavirus and enterovirus,
by molecular methods.
109
-------
Sampling
All groundwater samples were collected using the EPA ICR method between June and
November 2002 (US EPA, 1996). At each well site, 1500 L of well water was collected through
a sterile 1MDS filter setup. For a few samples, the filter clogged or the well ran dry, before 1500
liters could be processed. In these cases the successfully filtered volume is reported.
Additionally a 10 L grab sample was collected from each well in a sterile container for
bacteriological and coliphage analyses. Enteric virus sampling equipment was sterilized between
well sites with 0.1% Bleach solution followed by successive 2% sodium thiosulfate and distilled
water rinses. All samples were kept at 4°C and analyzed within 48 hours.
Bacteriological analysis
All bacterial analysis of fecal coliforms and enterococcus was done as specified in EPA-
approved methods. The samples were analyzed by membrane filtration using mFC and MEI
agars.
Coliphage analysis
All analysis of male-specific (F+), somatic coliphages and total coliphages was done as specified
in EPA-approved methods. The US EPA Methods 1601 (enrichment) and 1602 (single agar
layer) were used with the host bacterial .coli Famp for F+ coliphages, E. coli CN-13 for somatic
coliphages, andE. coli C3000 for "total" coliphages.
Enteric virus recovery and analysis
The 1MDS filters used to adsorb viruses from samples of ground were eluted and concentrated as
110
-------
specified in the EPA ICR method. The only change was that the final resuspended acid
(flocculated beef extract) precipitate was suspended in 20 mL of phosphate buffered saline
(Sigma D8662) rather than 30 mL of sodium phosphate. Samples were filtered through a 37 mm
diameter, 0.2 micrometer pore size, beef extract pre-treated Gelman Serum Acrodisc filter
(4525). The filter-sterilized concentrate was divided into 5 aliquots. Aliquots were: 2/ 6.7 mL
portions for UNH Caco-2 and BGMK cell cultures, 211.3 mL portions for UNC FRhK-4 cell
culture and Calicivirus (norovirus) analysis, and 4 mL was archived. These sample concentrate
volumes are equivalent to 500 liters of initial water for inoculation into Caco-2 and BGMK cells,
respectively, the equivalent of 100 liters of initial water for FRhK-4 (HAV) and Calicivirus
(norovirus) analysis, and 300 liters of initial water for archiving. All aliquots were frozen at -
80°C prior to shipment and analysis.
Tissue culture protocol for virus isolation in BGMK and CaCo-2 cells
UNH screened concentrates for BGMK cytotoxicity on 25 cm2 flasks prior to inoculation of
samples onto 75 cm2 flasks. Sample concentrates were pre-activated for 30 minutes at 37°C with
10 |ig/mL of type IX trypsin (Sigma T-0303) for both Caco-2 and BGMK inoculates. BGMK and
Caco-2 cell cultures were rinsed three times with PBS before inoculation. Inoculated cultures
were incubated at 37°C for 90 minutes with rocking every 15-20 minutes. Only negative controls
were run. Serum free maintenance media containing 5 |ig/mL trypsin was added to cultures after
incubation. Cultures were incubated at 37°C for 7 days. The cultures were checked
microscopically daily for the first two days and then every other day thereafter. After 7 days all
cultures were freeze thawed and 10% of the lysate was filtered through a 0.22 um filter and
inoculated onto new cells for a second passage. At the end of the second passage cultures were
111
-------
freeze thawed twice. Lysates were pooled and divided into aliquots for further analysis and
shipping. Samples were not chloroform extracted. Sample aliquots were sent to UMN for
enterovirus analysis and TAMU for rotavirus and reovirus analysis.
RT-PCR Analysis for Astrovirus and Adenovirus
Nested PCR was performed on UNH, UNC, UMN and TAMU samples for both Astrovirus and
Adenovirus type 40 and 41. The equivalent volume of original water sample examined for each
virus was 500 liters. Positive controls we did were at the level of (RT-)PCR. Virus was added to
cell culture lysate to act as a positive control for (RT-PCR)PCRNested PCR was performed on
UNH, UNC, UMN and TAMU samples for both Astrovirus and Adenovirus type 40 and 41. The
equivalent volume of original water sample examined for each virus was 500 liters. Positive
controls we did were at the level of (RT-)PCR. Virus was added to cell culture lysate to act as a
positive control for (RT-PCR)PCRNested PCR was performed on UNH, UNC, UMN and
TAMU samples for both Astrovirus and Adenovirus type 40 and 41. The equivalent volume of
original water sample examined for each virus was 500 liters. Positive controls we did were at
the level of (RT-)PCR. Virus was added to cell culture lysate to act as a positive control for (RT-
PCR)PCRNested PCR was performed on UNH, UNC, UMN and TAMU samples for both
Astrovirus and Adenovirus type 40 and 41. The equivalent volume of original water sample
examined for each virus was 500 liters. Positive controls we did were at the level of (RT-)PCR.
Virus was added to cell culture lysate to act as a positive control for (RT-PCR)PCR
112
-------
Astrovirus. All molecular techniques were done as specified in the methods and materials
developed by the project team in communication with the EPA. Astrovirus RT-PCR was done
according to Chapron et al. 2000. The primers used were specific for human astrovirus, RT
primer 5'-GTAAGATTCCCAGATTGGT-3' and PCR primer 5'-
CCTGCCCCGAGAACAACCAAG-3'. An ll-|iL sample of the combined cell lysates was
denatured with 0.5 |iL each of 0.05 M EDTA and downstream primer at 99°C for 8 min.
Eighteen |iL of the RT mixture was then added and run for 42 min at 42 °C to reverse transcribe
and then 5 min at 99°C. The RT mixture per sample consisted of 2.5 |iL 10X buffer n, 8.5 |iL of
25mM MgCl2 1.25 ^L of each lOmM dNTP, 0.5 ^L of lOOmM DTT (Promega), 10 units of
Rnasin, and 50 units of RT. After the RT step 28.5 |iL of a PCR master mix was added. The
PCR mixture per sample consisted of 3 |iL of 10X buffer n, 1 |iL of the PCR primer, 0.5 |iL of
the RT primer, 24 |iL of molecular grade water, and 2.5 units of Ampli-Taq DNA polymerase.
The parameters were 95°C, 5 minute hot start, followed by 35 cycles of 95°C for 30 seconds,
56°C for 30 seconds, 72°C for 30 seconds, with a final extension at 72°C for 5 minutes. These
primers yielded a 193 and/or 243 bp amplicon.
For nested PCR, 1 |iL from each RT-PCR reaction was added to a new tube containing 90 |iL of
a nested PCR reaction mixture, which contained 8 mM MgCl2 10 |iL lOx buffer, ImM of each
dNTP, 2.5 units of Ampli-Taq DNA polymerase and 1 |iM of each primer. The primers used
were 5'-CCTTGCCCCGAGCCAGAA-3' and 5'-TTGTTGCCATAAGTTTGTGAATA-3'. These
primers yield a 143 and/or 183-bp amplicon. Twelve |iL of each RT-PCR product as well as 12
|iL of the nested PCR product was run and sized by electrophoresis on 1.8% agarose gel, stained
with ethidium bromide. Molecular weights were determined by comparison with a 1 Kb DNA
113
-------
ladder (Life Technologies). Astrovirus serotype 2 was used as a positive control.
Adenovirus. All molecular techniques were done as specified in the methods and materials
provided by EPA. Adenovirus Hexon PCR was done according to Xu et al. 2000. The primers
used were Adi 5'-CCCTGGTA(G/T)CC(A/G)AT(A/G)TTGTA-3' and Ad2 5'-
TTCCCCATGGC(Inosine)CA(C/T)AACAC-3'. A 5|iL sample of the combined cell lysates was
added to 47.5|iL final volume PCR master mix. Final concentrations in the PCR master mix per
sample were 1.5mM MgCl2, Ix (lOx Buffer II), 0.2mM dNTP mix, 0.6|iM of each primer, and
2.5 units of Ampli-Taq DNA polymerase. The PCR parameters were 95°C for 5 minutes,
followed by 40 cycles of 94°C for 1 minute, 55°C for 1 minute, 72°C for 2 minutes, with a final
extension at 74°C for 5 minutes. These primers yielded a 482 bp amplicon.
For nested PCR, 1 |iL from each PCR reaction was added to a new tube containing 90 |iL of a
nested PCR reaction mixture, which contained 8 mM MgCl2 10 |iL lOx buffer, ImM of each
dNTP, and 1 |iM of each primer. The primers used were 5'-
GCCACCGAGACGTACTTCAGCCTG-3' and 5'-
TTGTACGAGTACGCGGTATCCTCGCGGTC-3'. These nested primers were specific for
Adenovirus type 40 &41. Samples were run for 35 cycles of 95°C for 30 seconds, 55°C for 30
seconds, 72°C for 30 seconds yielding a 142 bp amplicon. Twelve |iL of each nested PCR
product was run and sized by electrophoresis on 1.8% agarose gels and stained with ethidium
bromide. Molecular weights were determined by comparison with a 1 Kb DNA ladder (Life
Technologies). Adenovirus 40 & 41 were used as positive controls.
114
-------
Data management
Results of all analyses were entered into the attached excel spreadsheet as well as a laboratory
notebook.
Results and Discussion Summary
The objectives of this project were: 1) evaluate EPA methods 1601 and 1602 for the recovery and
detection of male specific coliphage, somatic coliphage and total )somatic and male-specific)
coliphage from well water; 2) Compare the efficiency of using a single host, C3000 for the
detection of both phages; 3) Compare EPA methods 1601 and 1602 for phage detection using all
three hosts; 4) Determine if there is a correlation between the detection of coliphage using either
EPA Method 1601 or 1602 and indicator bacteria (fecal coliforms and Enterococcus) and 5)
Determine if there is any correlation between the detection of indicator bacteria, coliphage (using
either method) and certain enteric viruses detected by the Polymerase Chain Reaction Assay
(PCR). For this work, 25 wells in the Northeast, some from New Hampshire, Vermont, Maine
and Massachusetts, were sampled and evaluated for each organism.
An entire summary of the results can be found in the accompanying data spreadsheet. All
samples were negative for Adenovirus and Astrovirus using an integrated cell culture Polymerase
Chain Reaction Assay followed by a nested PCR assay.
Overall, 9 of 25 samples or 36% were positive for one or more fecal indicator microbe, either a
coliphage or a bacterial indicator. Only one well of 25 (4%) was positive for coliphage (Table 1).
115
-------
This coliphage-positive was by the enrichment assay on hosts E. coli Famp and C3000. The
same well was positive for 4 plaques/100 mL using Method 1602 on F+ host E. coli Famp, but it
was negative for coliphage plaques on E. coli C3000. The one well that was positive for
coliphage was negative for both indicator bacteria and enteric viruses. No well was positive for
coliphage by either method with any other microorganism tested.
Eight wells of 25 (32%) were positive for indicator bacteria. Two wells were positive for both
fecal coliforms and Enterococcus, 3 wells were positive for fecal coliforms only and negative for
Enterococcus and 3 wells were positive for Enterococcus only but negative for fecal coliforms.
Since so many samples from community wells were negative for all microorganisms, the
decision was made to include private drilled wells in the study. It was hoped that because these
wells, on average, were probably less deep then the community wells, that there would be an
increased chance of detecting indicator organisms as well as enteric viruses. While some of the
wells were positive for indicator bacteria and one was positive for coliphage (though both
coliphage and bacteria were not the same well), no wells were positive for enteric viruses. Two
of these wells in VT were less then 100 ft deep. These wells were included on the study in
further efforts to increase the probability of detecting enteric viruses. Both of these wells were
negative for viruses and coliphages while one of the wells was positive for 1 fecal coliform
colony in the 1 L volume assayed. To further increase the probability of detecting enteric
viruses, a stoned lined, non-potable well, which was only approximately 35 feet deep was also
sampled. This well was negative for all indicator bacteria, coliphage and virus.
116
-------
There were 6 samples from public wells in Massachusetts. The public water supplies in
Massachusetts were chosen due to positive results previously found for total and fecal coliforms,
enterococci, and male-specific coliphages. Additionally 3 of the 6 locations had positives
previously reported for rotavirus and enterovirus by molecular methods. All of these 6 samples
were negative for coliphage and enteric virus. One of these samples was positive for both fecal
coliforms and enterococcus, 2 of the remaining 5 wells were positive for enterococcus but for no
other microorganism.
One of the key study objectives was to evaluate coliphage occurrence, by either method, as an
indicator for the presence of enteric virus. The results of this study indicate that, overall, the
wells were not contaminated by enteric viruses at the time they were sampled. However, the
results did yield positive results for coliphage presence in groundwater in the absence of
detectable bacterial indicators in that sample. However, bacterial indicators were found more
frequently than coliphages (8 samples versus 1 sample) and they were found in the absence of
detectable coliphages in these samples. Therefore, coliphages and bacteria were not detected
together in any of the samples analyzed. This finding of a lack of co-occurrence of coliphages
and bacteria in the same sample supports the measurement of both coliphages and bacteria in
groundwater samples as a way to increase the likelihood of detecting fecal contamination.
117
-------
Table 1. UNH Coliphage Results for Northeast Groundwater
Well Sample Number and Type
Well # 3, Private
Wells 1,2 and 4-25
Samples positive for phage by SAL
(100 mL)
Famp
0
0
Cn-13
0
0
C3000
0
0
Samples positive for phage by
enrichment (1 L)
Famp
1
0
Cn-13
0
0
C3000
1
0
Table 2. UNH Groundwater Wells Positive for Bacterial Indicators
Positive UNH Bacterial Results for
Groundwater Samples
Well Type and Number
Community, Well # 1
Private; shallow, Well #4
Private Well #9
Private, Well #15
Private, Well # 18
Private, Well # 19
Community, #23
Community, #23
Fecal Coliforms
100 ML
1
0
0
35
0
2
0
5
1 L
200
1
0
TNTC
0
10
0
69
Enterococcus
100 ML
0
0
0
0
1
13
2
2
1 L
0
0
3
0
93
89
87
32
118
-------
Table 3. Results of UNH Groundwater Samples Analyzed for Adenovirus and Astrovirus
Well#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Results
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
for
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
all
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
viral
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
analyses
119
-------
Table 4. Summary of UNH Samples Positive for Coliphages, Bacterial Indicators, Adenoviruses
and/or Astroviruses
Well Number and Type
Well # 3, Private
Well # 1, Community
Well # 9, Private
Well # 15, Community
Well# 18, Community
Well# 19, Community
Well #21, Private
Well #22, Private
Well #23,
All other wells
Samples positive for:
Coliphage
1
0
0
0
0
0
0
0
0
0
Bacterial Indicators
FC
0
1
0
1
0
1
0
1
1
0
Ent
0
1
0
1
1
1
1
1
0
Any
1
1
1
1
1
1
1
1
0
Phage and
Bacterial
Indicators
0
0
0
0
0
0
0
0
0
0
Coliphage
and Virus
0
0
0
0
0
0
0
0
0
0
Bacterial
Indicators
and Virus
0
0
0
0
0
0
0
0
0
0
References
Chapron, C.D., Ballester, N.A., Fontaine, J.H., Frades, C.N. & Margolin, A.B. 2000 The
Detection of Astro virus, Entero virus and Adeno virus Type 40 and 41 in Surface Waters Collected
and Evaluated by the Information Collection Rule and Integrated Cell Culture/Nested PCR
Procedure. Appl. and Environ. Microbiol, 60 (6), 2520-2525.
Environmental Protection Agency. 1995 Virus Monitoring Protocol for the
Information Collection Requirements Rule. U.S. Environmental Protection
Agency, publication EPA/814-B-95-002. Government Printing Office, Cincinnati,
120
-------
Ohio.
Environmental Protection Agency. 2000 Method 1601: Male-specific (F+) and Somatic
Coliphage in Water by Two-step Enrichment Procedure. Draft April 2000. Office of Water,
Washington, B.C.
Environmental Protection Agency. 2001 Method 1602: Male-specific (F+) and Somatic
Coliphage in Water by Single Agar Layer (SAL) Procedure. Draft January 2001. Office of
Water, Washington, D.C.
Xu, W., McDonough, M.C., & Erdman D.D. 2000. Species-specific identification of human
adenoviruses by a multiplex PCR assay. J. Clin. Microbiol. 39, 4114-4120.
121
-------
APPENDIX II
EPA OST Groundwater Coliphage Project
Coliphage, Bacteria and Human Enteric Virus Isolation from Ground Water-
Southwest Region Laboratory
Texas A&M University - Prof. Suresh Pillai, Co-PI
Introduction and Background
Ground water samples for microbial analyses.
The original goal of this study was for each of the four, regionally representative laboratories
(southeast, northeast, upper Midwest and southwest) to collect and analyze 27 ground water
samples from public water supply wells. Efforts were made to identify candidate public water
supplies that previously had coliform bacteria violations or other evidence of vulnerability to fecal
contamination. In some cases candidate wells were prescreened by bacteriological and coliphage
analyses for evidence of fecal contamination. Because not all participating labs could identify and
get access to 27 public water supply wells, some labs also included non-public and private wells
in their sampling. Three labs obtained 27 ground water samples and one lab obtained a total of 25
samples for a total of 106 samples overall. The characteristics of the wells that were sampled are
presented in data tables in the Results section of this report. This report describes the methods,
materials, coliphage and bacterial indicator and enteric virus results for samples from the
southwest region. The results for bacteria and coliphage analyses of these samples are also
122
-------
presented in an Excel spreadsheet that accompanies this document and in the PowerPoint
presentation that was delivered at the April, 2004 EPA Coliphage Workshop.
Methods and Materials
Sampling sites and wells in Texas and New Mexico: Only PWS wells were included in this study,
and a total of eleven different PWS wells were identified for this study. The sampling sites were
located in the San Antonio region of Texas (wells RS, KK, and HCR) and along the US-Mexico
border in southern New Mexico (wells MHPa, MHPb, MHPc, FVE, AVC, SME, and LME). The
wells in the San Antonio region were part of a karst aquifer and were previously implicated in a
documented groundwater contamination event. Also, during the initial pre-screening of the wells
some of the samples were positive for somatic and male-specific coliphages. The wells in
southern New Mexico were identified as being vulnerable to groundwater contamination based on
parameters such as closeness to septic tanks, proximity to the Rio Grande river and the aquifer in
question. These wells were part of a previous EPA-funded project on the microbiological quality
of wells in the shallow aquifer along the US-Mexico border during which some of the wells in the
sampling area were positive for enterococci, E. coli, male-specific coliphages and somatic
coliphages. The wells were in the 100-150 feet depth range. The static water levels were around
10-20 feet and in terms of their hydrogeologic setting, they were located in the Rio Grande
alluvium/Hueco-Tularosa aquifers.
Sampling: Groundwater samples were collected between June 2002 and January 2003. Multiple
samples were collected from each of the wells to be representative of the aquifer and the sampling
123
-------
location. During each sampling adequate volumes were collected for the coliphage analysis as
well as for the enteric virus analysis. Grab samples were collected for the coliphage and bacterial
analysis while the 1MDS filters were used for collecting the large volume enteric virus samples.
Microbiological Analysis: The USEPA methods 1601 and 1602 were used for the coliphage
analysis. The host bacteria used in these analyses included E. coli Famp for F+ coliphage, E .coli
CN-13 for somatic coliphages andE. coli C-3000 for "total" coliphages. The samples were also
analyzed for E. coli and Enterococcus spp using the membrane filtration protocol and M-coli
Blue and MEI agars, respectively.
Virus concentration from ground water samples. Viruses were concentrated from 1500-liter
ground water samples by filtering the water at its ambient pH using standard 1-MDS
electropositive cartridge filters (ZetaPor Virosorb, Cuno Product No.45144-01-1MDS) and using
procedures described by the US Environmental Protection Agency (2001). In cases where the
sample size is not 1500 liters, the filtered volume is specified in the results section of this report.
Filter elution and concentration. Human enteric viruses were eluted from the Cuno 1-MDS
cartridge filters with a solution of 1.5% beef extract (Becton Dickinson) plus 0.05 M glycine at
pH 9.5. The eluent was allowed to contact the filter cartridge for a minimum of six minutes.
Viruses in the beef extract-glycine eluate were subsequently concentrated into a smaller volume
by acid precipitation (organic flocculation). Briefly, the eluates were adjusted to pH 3.5, stirred
slowly for 30 minutes, then centrifuged at 6,200 x g for 20 minutes. The resulting pellets were
resuspended with 15 mL of 0.15 M NajHPO,,, adjusted to pH 9.0-9.5, and centrifuged at 6,200 x g
124
-------
for 15 minutes to remove residual particulates. The resulting supernatants were adjusted to pH
7.0-7.5, and filtered through 0.2|im pore size, serum Acrodisc syringe filters (Pall) to remove
bacterial and fungal contaminants. Each concentrated sample of 18-20 mL was subdivided into
aliquots for subsequent detection of specific enteric virus groups, then stored at -80°. Aliquots of
the concentrate were shipped to participating labs for their respective viral analyses by integrated
cell culture and (RT-)PCR (for hepatitis A virus [HAV], enteroviruses, adenoviruses, rotaviruses,
reoviruses and astroviruses) and direct RT-PCR for noroviruses (human, Norwalk-like
caliciviruses).
Infectivity assays in BGMK and Caco-2 cell cultures. One-third of each sample concentrate,
equivalent to 500 L of source water, was inoculated into cultures of the Buffalo Green Monkey
Kidney (BGMK) continuous cell line, and another third was inoculated into cultures of the Caco-2
continuous cell line. Another portion, corresponding to 100 liters of ground water , was used for
cell culture plus RT-PCR analysis of hepatitis A virus (HAV), and another portion, also
corresponding to 100 liters of ground water, was used for direct RT-PCR of human caliciviruses
(noroviruses). The remaining one-fifth of the sample was archived as a contingency for possible
future analysis. Sample concentrates for cell culture inoculation were pre-activated by adding
type IX trypsin (Sigma T-0303) to a 10 |ig/mL concentration, and incubating at 37° for 30 minutes
prior to inoculation. Newly confluent layers of each cell type in 75 cm2 tissue culture flasks were
rinsed three times with Dulbecco's phosphate buffered saline (PBS) supplemented with
magnesium and calcium (Gibco) to remove residual calf serum associated with the cell growth
medium. The cultures were inoculated with trypsin pre-activated concentrate, and incubated at
37° for 80 minutes. Serum-free maintenance MEM medium with Earle's salts supplemented with
125
-------
5|ig/mL type IX trypsin was added. Inoculated cultures were incubated for 7 days at 37° with
periodic microscopic examination for evidence of viral cytopathology.
After seven days, the inoculated cultures were frozen and thawed twice. Newly confluent PBS-
rinsed layers of the same cell line were inoculated with 10% of the cell culture lysate from each
first passage (initial) culture, calf serum-free medium supplemented with trypsin was added, and
the cultures were incubated at 37° for a second passage of the sample material. The second
passage cultures were periodically observed microscopically, then frozen seven days after
inoculation.
All first and second passage cell cultures were frozen and thawed twice. A single lysate pool of
about 35 mL was prepared for each ground water sample by combining 10% of the lysate from
both first and second passage BGMK and Caco-2 cultures that had been inoculated with a specific
water sample concentrate. A 10-mL portion of each lysate pool was extracted with 5 mL of
chloroform, and centrifuged at 1,800 x g for 15 minutes. Each sample extract was subdivided into
aliquots for isolation of viral nucleic acid and viral nucleic detection using the nucleic acid
amplification methods of either polymerase chain reaction (PCR) for DNA viruses (adenoviruses)
or reverse transcription PCR (RT-PCR) by other participating laboratories, and stored at -80°.
Tissue culture protocol for virus isolation in BGMK and CaCo-2 cells at Texas A&M
University. The groundwater concentrates (equivalent to 500L) were initially pre-tested for
cytotoxicity after an initial pre-activation. (No cytotoxicity tests were done prior to the CaCo-2
cell cultures since none of the samples were positive for cytotoxicity on BGMK cells).
126
-------
Preactivation was done using 0.5 mL and 1.0 mL of the groundwater concentrate. The 0.5ml
sample was added to 5|iLof trypsin and the 1.0 mL sample was added to lOul of trypsin. The
samples were incubated for 30 minutes and then refrigerated prior to the cytotoxicity tests. The
T25 flasks (having 80% confluency) were washed twice with 5 mL of Hanks Balanced Salt
Solution (HBSS). The cells were inoculated with 0.5ml of the preactivated sample, and incubated
for 90 minutes with cells being rocked every 15 minutes. After the 90-minute incubation, 5ml of
MEM complete (serum free with 0.25|iL/mL of trypsin) was added and the cells were observed
for 2 days. Cytotoxicity was evaluated using a sterile HBSS -inoculated "negative control."
Each of the T75 flasks were washed with 15 mL of HBSS two times. The HBSS was siphoned off
and the flasks were inoculated with the remainder of the sample across 3 flasks. Two negative
controls (1 before inoculation of sample and 1 after inoculation of sample) were also included.
The flasks were incubated for 90 minutes at 37°C with 5% CO2 and rocking every 15 min. After
the 90-minute incubation, 15 mL of MEM (serum free with 0.25|iL/mL of trypsin) was added.
The flasks were incubated at 37°C for 7 days and observed every day for cytopathic effects (CPE).
The same procedure was followed for CaCo-2 cells as well
The samples were passaged a second time by freeze thawing once and removing approximately
10% of the lysate from the original fiaks and placed in new 100% confluent flasks that were
washed as mentioned previously. The samples were incubated for 90 minutes, rocking every 15
minutes and 15 mL of MEM (serum free containing 25|iL/mL of trypsin) was subsequently added.
The samples for incubated for another 7 days and observed by microscopy daily.
127
-------
The samples were passaged a third time by removing the lysate from the second passage, filter
sterilized through a 100 mm diameter, 0.22 |im pore size cellulose ester filter into T75 flasks that
were prepared as before. The flasks were placed in the incubator at 37°C for 5-7 days and
observed for cytopathic effects.
Table 1. Tissue Culture Results for Virus Isolation from Ground Water Samples Based on
Microscopic Observation Only
Sample ID
RS(1)
HCR(l)
RS (2)
BM (1)
KK(1)
RS(3)
KK(2)
HCR (2)
RS(4)
RS(5)
MHPla
MHPlb
AVC1
FVE1
AVC2
FVE2
FVE3
AVC3
MHPlc
MHP2a
MHP2c
MHPSa
MHP2b
SME1
SME2
LME1
MHPSb
Sample volume
BGMK
cells
4.75
CaCo-2
cells
5.5
CPE Results (BGMK and CaCo-2)
Passage # 1
+
Passage # 2
+
Passage # 3
+
Groundwater concentrate sample lost due to centrifuge tube breakage
Groundwater concentrate sample lost due to centrifuge tube breakage
5.0
4.75
5.0
5.0
6.0
5.0
5.0
5.0
5.0
5.0
5.0
4.0
4.5
6.5
4.5
5.5
4.75
4.8
5.3
6.1
4.8
6.0
4.6
4.3
6.0
6.0
5.5
5.5
6.5
5.5
4.5
5.5
6.5
6.0
6.0
4.5
6.0
7.75
5.0
6.0
5.5
5.0
5.5
6.5
5.0
6.0
5.25
4.5
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
3
= indicates possible cytopathic effect.
128
-------
Viral RNA Extraction for Rotavirus and Reovirus detection by RT-PCR
The cell culture extracts from the BGMK cells (from passage # 1 and passage # 3) (1 mL each)
were combined with 2 mL from CaCo-2 cell lysates and to this was added to 2 mL of chloroform.
The mixture was vortexed for 2 min at high speed, then centrifuged at 18K rpm for 20 minutes.
The top layer was pipetted out and aliquotted into 4 cryo-tubes (1 mL each). (Samples 1-9 that
was sent from Texas A&M University contained only extracts from CaCo-2 cells due to a
laboratory error). One cryo-tube of each sample was shipped to UNC, Univ. of Minnesota and
UNH.
The QiAmp viral RNA extraction kit was used for RNA extraction from the cell culture lysates
per the manufacturer's recommended protocols (Qiagen, Valencia, CA). The final extract was
resuspended in 80 |uL of buffer, which was stored at -80C until the RT-PCR analyses.
RT-PCR Analysis for Rotaviruses and Reoviruses
Rotavirus analysis. For Rotavirus, 3.5 |iL of the RNA extract was used. Separate RT and PCR
amplifications were performed with 10 and 50 |iL total reaction volumes, respectively. The final
concentrations in the RT step were 5mM (IX PCR Buffer n), 5mM MgCl2, ImM of each dent,
1.26 |iM of 3' rotavirus primer, 45 units of Reverse Transcriptase, RNAse inhibitor (18 units). The
sample was "hot-started" (95°C for 5 min) and when the temperature reached 60°C, reverse
transcriptase, RNAse inhibitor and dNTP were added. The RT step was conducted at 42°C for 60
minutes. A wax layer was used to prevent accidental aerosolization of samples when the tubes
were subsequently opened. The samples were heated at 95°C for 5 minutes. The sample was
maintained at 80°C. The PCR master mix was then added to this sample. The final concentration
129
-------
in the sample after the addition of the PCR master mix was 2mM MgCl2, IX PCR Buffer II, 2.5
units of Taq DNA polymerase, and 0.25|iM of 5' Rotavirus primer. The cycling conditions were
95°C for 1.5 min, 55C for 1.5 min and 72C for 1.5 minutes. Forty PCR cycles were performed.
The PCR products were run on a pre-made (6 %) Novex TBE gels (Invitrogen, Valencia, CA) for
detection of the 208 bp product.
The controls included a Rotavirus RNA-spiked positive control and a water negative control.
Additionally, MS2 RNA was spiked into a select number of samples to detect any possible sample
inhibition. Primers directed to the capsid gene of the MS2 RNA were used for this purpose
(Valenzuela and Pillai, 1998).
Reovirus Analysis. A volume of 5ul of the RNA extract was used for RT-PCR analysis. Separate
RT and PCR amplifications were performed, with 10 |il and 50 |il reaction volumes respectively.
The final concentration of the RT components were 1.5 mM Mgcl2, IX of PCR Buffer n, 0.7 mM
of each dNTP and 1.7|iM of the 3' Reovirus primer. A wax layer was used to prevent accidental
aerosolization during subsequent handling. The samples were heated at 99°C for 5 minutes and
then placed on ice. Once the samples were cooled, RNAse inhibitor (22 units), and 50 units of
Reverse Transcriptase were added. The RT conditions were 43°C for 60 minutes. The samples
were subsequently heated for 5 minutes at 95°C and placed on ice. The PCR master mix was then
added to this sample, giving a final volume of 50 micro liters. The final concentrations of the PCR
mix ingredients were 1.5mM of MgCl2, IX PCR Buffer II, 05 |iM of the 5' Reovirus primer and
5.0 units of the Taq DNA polymerase. The PCR amplification conditions were 95°C for 1 minute,
55°C for 1.5 minutes, 72°C for 1.5 minutes. Forty PCR cycles were performed. The PCR products
130
-------
were resolved on a 6% TBE premade Novex gels.
RESULTS
Enteric viruses
None of the 27 groundwater samples from either Texas or New Mexico were positive by cell
culture and (RT-)PCR for adeno viruses, astro viruses, enteroviruses, reoviruses, rotaviruses or
hepatitis A virus or, in the case of caliciviruses, by direct RT-PCR analysis. The presumptive
positive results for cytopathic effects shown in Table 1 must have been due to non-viral effects
such as the action of the trypsin in the medium. Trypsin can cause the cells to dislodge from the
surface of the flask and appear abnormal, or to cytotoxicity from the sample concentrate inocula.
Regardless of the cause of this effect, it was not due to the presence of any of the viruses for
which samples were analyzed by nucleic acid methods.
Bacterial and Coliphage Indicators in Groundwater.
The results for bacterial and coliphage indicators in positive samples are summarized in Table 2.
In all, 7 of 27 samples (26%) were positive for at least one indicator microbe.
131
-------
Table 2: Summarized data showing groundwater wells that were positive for bacterial and/or viral
(coliphage) indicators.
Sample
RS(1)
KK(2)
HCR(2)
MHPa(l)
RS(5)
AVC(l)
AVC(3)
Entero-
cocci
(Number/
100 mL)
0
0
1
5
0
0
0
E. coli
Number/
100 mL
0
0
0
1
1
0
0
Method 1602
Number/100 mL
Famp
0
0
0
0
0
0
0
CN-13
0
0
1
0
0
0
0
C3000
0
0
0
0
0
0
0
Method 1601
Positive (+) or Negative (-) per
Indicated Volume
Famp
100
mL
-
-
-
-
-
-
+
1000
mL
-
-
-
-
-
-
-
CN-13
100
mL
+
-
+
-
-
-
+
1000
mL
-
+
+
-
-
+
-
C3000
100
mL
-
-
+
-
-
-
-
1000
mL
-
-
-
-
-
-
1000
Bacterial Indicators. Out of 27 samples that were analyzed, only 2 sample (7.4%) were positive
for E. coli and 2 samples (7.4%) were positive for Enterococci. There was only 1 sample that was
positive for both E. coli and Enterococci. The maximum density of E. coli in a sample was 1
CFU/100 mL compared to Enterococci, which showed a maximum density of 5 CFU/100 mL.
Viral (Coliphage) Indicators. Out of 27 samples, 5 samples were positive for coliphages. There
was only 1 sample that was positive for male-specific coliphages (based on detection of a plaque
on E. coli host Famp). This is in contrast to 5 samples that were positive for somatic coliphages
(based on plaques on E. coli host CN-13 or growth in enrichment cultures) while 2 samples were
positive for "all" coliphages based on E. coli host C-3000. Four samples were positive for
coliphages when 1000 mL was analyzed compared to 3 samples that were positive when only 100
mL samples were analyzed. Two of the samples were positive when 100 mL and 1000 mL
132
-------
aliquots of the sample were screened for coliphages.
Comparison of Bacterial and viral Indicator Results: Table 2 shows the results from the bacterial
and viral (coliphage) indicator analyses so that the two types of indicators can be compared. Out
of 27 samples that were analyzed for bacterial and viral indicators, 7 (25.9%) were positive for
either bacterial or viral indicators. Only 3 of the samples (11.1%) were positive for either of the
bacterial indicators (E. coli or enterococci) while 5 samples (18.5%) were positive for coliphages
(either by Method 1601 or Method 1602). Four samples (14.8%) were positive for coliphages but
negative for bacterial indicators. This is in comparison to only 2 samples (7.4%) that were
positive for bacterial indicators but negative for viral indicators.
These results suggest that coliphages can be used as a tool for screening ground water samples for
the presence of fecal contamination. The results strongly suggest that coliphage analysis should be
conducted along with or in addition to conventional bacterial indicator analysis. This is because
the inclusion of coliphages increases the likelihood of detecting a contaminated samples, based on
the presence of either bacteria or coliphage indicators. The total absence of human enteric viruses
in the presence of the selected indicator organisms suggest that it is highly unlikely that pathogens
would be detected routinely. It is possible that only under heavily contaminated conditions would
there be a direct correlation or co-occurrence between the presence of viral pathogens and fecal
indicator organisms.
133
-------
References
Chapron, C.D., Ballester, N.A., Fontaine, J.H., Frades, C.N., and A.B. Margolin (2000) Detection
of astroviruses, enteroviruses, and adenovirus types 40 and 41 in surface waters collected and
evaluated by the information collection rule and an integrated cell culture-nested PCR procedure.
Appl. Environ. Microbiol, 66(6):2520-5.
Schwab, K. J.,R. De Leon, and M. D. Sobsey. 1995. Concentration and purification of beef extract
mock eluates from water samples for the detection of enteroviruses, hepatitis A virus and Norwalk
virus by reverse transcription-PCR. Appl. Environ. Microbiol. 61:531-537.
US EPA (2001) USEPA Manual of Methods for Virology, Chapter 14. EPA 600/4-84/013 (N14),
April 2001, Office of Research and Development, Washington DC 45260
Xu, W., McDonough, M.C., and D.D. Erdman (2000) Species-specific identification of human
adenoviruses by a multiplex PCR assay. J. Clin. Microbiol., 38(11):4114-20. Erratum in: J. Clin.
Microbiol., 2001, Apr.; 39(4):1686
Valenzuela, R.B., and S.D. Pillai. 1998. Persistence of naked viral RNA molecules in groundwater.8th
Intl symposium on Microbial Ecology. Nova Scotia. August
Vinje J, Koopmans MP. (1996) Molecular detection and epidemiology of small round-structured
viruses in outbreaks of gastroenteritis in the Netherlands. J Infect Dis. 1996 Sep;174(3):610-5.
134
-------
APPENDIX III
EPA Coliphage Method Validation Project Report:
Detection of Coliphages, Indicator Bacteria and Enteric Viruses in Groundwater
AUTHORS
Sagar M. Goyal, DVM, PhD
Yashpal Malik, DVM, PhD
Baldev R. Gulati, DVM, PhD
Sunil Maherchandani, DVM, PhD
Sigrun Haugerud, BS
University of Minnesota
And
Mark D. Sobsey
University of North Carolina
STUDY COMPLETED ON
June 30, 2003
PERFORMING LABORATORY
Department of Veterinary Diagnostic Medicine
College of Veterinary Medicine, University of Minnesota
1333 Gortner Avenue, St. Paul, MN 55108, USA
Contact Information
Sagar M. Goyal
Department of Veterinary Diagnostic Medicine
College of Veterinary Medicine, University of Minnesota,
1333 Gortner Avenue, St. Paul, MN 55108, USA.
Phone: 612-625-2714; Fax: 612-624-8707
Email: goyalOO 1 @umn.edu
Mark D. Sobsey
University of North Carolina
CB# 7431, McGavran-Greenberg Hall, Room 4114a, Chapel Hill, NC 27599-7431
Telephone : 919-966-7303 Email: Mark_Sobsey@unc.edu
135
-------
Purpose: To determine if FRNA phages are useful indicators of fecal contamination and human
enteric viruses by testing well water samples for the presence of fecal coliforms, Escherichia coli,
Enterococcus, somatic coliphages, FRNA coliphages, "total" coliphages and human enteric
viruses.
Materials and Methods
Source of samples. Ground water samples were collected from 27 candidate wells (address with
contact numbers are provided in Table 1). All wells except 6 private ones in Minnesota are
considered public water supplies by the State of Minnesota and none are disinfected.
Recovery of Enteric Viruses. From each well 1,500 liters of water was pumped through a 1-
MDS filter cartridge followed by virus elution in 1.5% beef extract-0.05 M glycine solution.
Another 5 liter sample of water was collected from each well in a sterile container for
bacteriological and coliphages analysis. These samples were maintained at 4°C until analyzed,
usually within 24 hrs of collection. The results are given in Table 2.
Bacteriological evaluation. Grab samples of water were analyzed for fecal coliforms, E. coli and
Enterococcus using Membrane Filter (MF) technique as recommended in chapter 9 of Standard
Methods for the Examination of Water and Wastewater (American Public Health Association,
1998). Briefly, a 100 mL volume of a water sample was filtered through a 0.45 |im pore size, 47
mm diameter membrane filter. These filters were then placed on plates of selective mFC agar for
fecal coliforms, mEC for E. coli and mE media for Enterococcus. For fecal coliforms, the plates
136
-------
were incubated at 44.5°C for fecal coliforms andE. coli and at 41.5°C for Enterococcus.. The
number of characteristic colonies was counted following incubation for 24 hrs and concentrations
are expressed as colony-forming units per 100 mL. All media were obtained from Becton
Dickinson, Cockeysville, MD.
Coliphages analyses. All 27grab samples were analyzed for the presence of FRNA (male-
specific) coliphages, somatic coliphages and "total" coliphages using single agar layer procedure
and enrichment method (Methods 1601 and 1602; Environmental Protection Agency, 200la;
200 Ib). The host bacteria were E. coli Famp (ampicillin and streptomycin resistant mutant of E.
coli; ATCC 700891) for FRNA coliphages, CN13 (nalidixic acid resistant mutant of E coli;
ATCC 700609) for somatic coliphages, andE. coli C3000 (ATCC 15597) for "total" coliphages.
A log phase culture of the host bacterium was prepared by inoculating a stock of the bacteria into
30 mL of trypticase soy broth followed by incubation for 4 hrs at 37°C on a shaker platform. To
100 mL aliquots of water samples were added 0.5 mL of 4 M MgCl2, 10 mL of log phase culture
of host bacteria, and 100 mL of molten and cooled double strength tryptic soy agar. The sample
was thoroughly mixed and poured into four 150-mm Petri plates followed by incubation at 37°C
for 24 hrs. Positive results were indicated by circular zones of lysis in contrast to opaque lawn of
host bacterial growth. Plaques from all four plates were counted for each sample. Plaques were
confirmed by picking them, resuspending the picked material in 100 ul of TSB, spotting onto
prepoured lawns of the respective host bacterium, incubating for 4 hours at 37°C, and observing
the spots for evidence of coliphage presence as lysis zones or plaques.
137
-------
In the other method of coliphages testing, the enrichment-spot plate method, 12.5 mL of MgCl2,
50 mL of 10X TSB and 10 mL of ampicillin/streptomycin or nalidixic acid and 5 mL of host
culture E. coli Famp, CN13 or C3000 were added to 1-liter aliquots of water. After incubation at
37°C for 24 hrs, 10 |iL of the culture was spotted on freshly prepared Spot plates of the respective
host culture (E. coli Famp, N13 or C3000). Positive results were indicated by circular zones of
lysis in contrast to opaque lawn of host bacterial growth. . ((Plaques were confirmed by picking
them, resuspending the picked material in 100 |iL of TSB, spotting onto prepoured lawns of the
respective host bacterium, incubating for 4 hours at 37°C, and observing the spots for evidence of
coliphage presence as lysis zones or plaques. This method is qualitative in nature because it scores
sample volumes as either positive or negative for coliphages.
Enteric virus isolation. Viruses were isolated from groundwater samples using the US EPA ICR
Method with minor modifications (US EPA, 1996). After filtering 1,500 liters of water through
the CUNO 1-MDS filter, adsorbed viruses were eluted from the filter with 1.5% beef extract-0.05
M glycine solution (pH 9.5). The eluate was further concentrated using the acid precipitation
method. All concentrates were suspended in the same volume (22 mL) of sodium phosphate
buffer. The final sample was filter sterilized using a 25 mm diameter 0.22 micrometer pore size
Gelman Acrodisc filter. Two aliquots of 2 mL each were sent to UNC for detection of human
caliciviruses (noroviruses) and Hepatitis A viruses. Two aliquots of 7.5 mL each, corresponding
to 500 liters of groundwater, were used for culturable virus isolation by inoculation of BGM and
Caco-2 cell lines. All samples were passaged twice in BGM and Caco-2 cells, with incubation
periods of one week per passage. The culture fluids were pooled separately (one sample passaged
twice in BGMK and Caco-2 was pooled). Pooled lysates were chloroform extracted and aliquots
138
-------
of 2 mL each, corresponding to 100 liters of groundwater, were sent to UNH and TAMU for
detection of adenoviruses, reovirus, rotavirus and astrovirus. Cell culture lysates from UNC,
UNH, and TAMU were also received for detection of enteroviruses by the cell culture and RT-
PCR methods described here. A volume of concentrated sample corresponding to 100 liter of
groundwater was also examined for human caliciviruses (noroviruses) by direct RT-PCR at UNC
RT-PCR. Approximately 5 mL volumes of all cell culture lysates were concentrated to 300 |uL
using PEG 8000. Of this, 140 |iL was used for RNA extraction using Qiagen RNA extraction kit.
The remaining 160 |iL was archived. The primers used for amplification of enterovirus nucleic
acid are shown below (Schwab et al., 1996).
3' Primer: 5' ACC GGA TGG CCA ATC CAA 3'
5' Primer: 5' CCT CCG GCC CCT GAA TG 3'
RT-PCR conditions were according to those previously used and were: RT - 42°C for 60 min,
followed by inactivation of RT at 95°C for 15 min. Denaturation - 95°C for 90 sec; Annealing -
55°C for 1.5 min; extension - 72°C for 1.5 min; final extension - 72°C for 10 min. No. of cycles-
40 (3). The RT-PCR products were analyzed by agarose gel electrophoresis and confirmed by
ethidium bromide staining for observation of DNA amplicons of the correct size. For positive
amplification, an amplicon of 197 bp was expected.
Results Summary. Of the 27 wells tested, fecal coliforms were detected in 7 (26%), E. coli in 3
(11%) and Enterococcus in 6 (8 positive samples) (22%). Three of 27 wells contained fecal
139
-------
coliforms, E. coli and Enterococci while one well was positive for both fecal coliforms and E.
coli. Somatic coliphages were detected in 16 wells (59%) male-specific FRNA phages in 11 wells
(41%) (12 positive samples), and "total" coliphages in 12 samples. None of the samples showed
cytopathological effects (CPE) characteristic of enteric viruses during their passages in BGM and
Caco-2 cell lines. None of the samples was positive for enteric viruses by RT-PCR or PCR.
References
US EPA (1996) ICR Microbial Laboratory Manual. Office of Research and Development, EPA
Number: 600R95178. Pages: 233, Washington, DC
Environment Protection Agency (200la). Method 1601: Male-specific (F+) and somatic
coliphages in water by two-step enrichment procedure. United States, Environment Protection
Agency, Office of Water, Washington, D.C. 2001. http://epa.gov/nerlcwww/1601ap01.pdf
Environment Protection Agency (200Ib). Method 1602: Male-specific (F+) and somatic
coliphages in water by single agar layer procedure. United States, Environment Protection
Agency, Office of Water, Washington, D.C. 2001. http://epa.gov/nerlcwww/1602ap01.pdf
American Public Health Association. Standard Methods for Examination of Water and
Wastewater, 20th Edition, Washington, DC. 1998.
Schwab, K.J, De Leon R., and M.D. Sobsey (1996) Immunoaffinity concentration and
140
-------
purification of waterborne enteric viruses for detection by reverse transcriptase PCR. Appl
Environ Microbiol, 62(6):2086-94.
Table 1. Details of Groundwater Wells Selected/Screened During the Study
Date on Wells for EPA groundwater study on coliphage methods
Sample
1
2
3
4
5
6
7
8
9
10
Well
01 Amundson
02 Gervais
03 Round
04 Turtle
05 Brookdale
06 Oak Groove
07 Keller Golf
08 Keller Golf
09 Hamilton
10 Norwood
Well Name
Amundson
Farms
Lake Gervais
Round Lake
Park
Turtle Lake
Brookdale Park
Oak Groove
Park
Keller Go If
Course
Keller Main
Park
Hamilton Park
Norwood Park
Date of
sampling
4/11/2002
4/23/2002
4/23/2002
4/23/2002
4/30/2002
4/30/2002
5/3/2002
5/3/2002
5/7/2002
5/7/2002
Well Address
Amundson Farms, RR1 Box
25, Chattfield, MN 55923
Lake Gervais, 2500 Ederton
St., Maplewood, MN
Round Lake Park, 910 Frost,
St. Paul, MN
Turtle Lake, 4079 Hodgson
Rd., Shoreview, MN
Brookdale Park, 7650 June
Ave. North, Brooklyn Park,
MN55443
Oak Grove Park, 6941 102nd
Avenue N., Brooklyn Park,
MN55443
2166 Maplewood Drive,
Maplewood, MN 55109
Keller Main Park, Hwy. 61,
Maplewood, MN 55109
6101 Candlewood drive,
Brooklyn Park, MN
8100 Newton Ave. N.,
Brooklyn Park, MN
Contact
person
Brad
Richard
(Dick)
Haus
Richard
(Dick)
Haus
Dick
Layne
Dick
Contact Phone
507-867-3396
651-748-2500
651-748-2500
651-748-2500
763-493-8350
763-493-8350
651-766-4173
651-748-2500
763-493-8350
763-493-8350
141
-------
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
1 1 Presbyterian
12 Immanuel
13 Central
14 Historical
15 Northwood
16 Lakewood
17 Lakewood
Mausoleum
18 Lakewood
Maintenance
19 Al
20 Dayne
21 Lake Maria
22 Gibbs Farm
23 Tom Arendt
24 Milena
25 Jay
12 Immanuel
Presbyterian
Church Maple
Plain
Immanuel
United
Methodist
Church
Central Park
Brooklyn Park
Historical Farm
Northwood
Park
Lakewood
Lakewood
Cemetery
Lakewood
Cemetery
Alan
Ducommun's
Father
Dayne
Ducommun
Lake Maria
Gibbs Farm
Museum
Tom Arendt
Home
Milena's house
Jay Keil home
Immanuel
United
Methodist
5/9/2002
5/9/2002
5/13/2002
5/13/2002
5/13/2002
5/13/2002
5/16/2002
5/16/2002
5/20/2002
5/20/2002
5/29/2002
5/29/2002
6/1/2002
6/3/2002
6/3/2002
6/14/2002
558 County Rd. 110, Maple
Plain, MN 55359
10095 County Rd. 101,
Cocoran, MN
8440 Regent Ave. Brooklyn
Park, MN
4345 101st Avenue N.,
Brooklyn Park
1 07th Quebeck Ave. N.,
Brooklyn Park, MN
3600 Hennepin Ave.
3600 Hennepin Ave.
3600 Hennepin Ave.
5435 152nd Ave., Anoka, MN
55303
4841 Salish Circle, Ramsey,
MN55303
11411 Clementa, Monticello,
NN 55362
Larpentar - Cleveland Av.
9871 John Trail, Chisago
City, MN55193
58585 222nd Street,
Litchfield, MN55355
1 8076 68th Ave, Darwin, MN
55355
3600 Hennepin Ave.
Ron
Gj erde
Ron
Gj erde
Ron
Gj erde
Alan
Ducommu
n
Dayne
Ducommu
n
Tom/Mark
Tom
Arendt
Milena
Milena
Ron
Gj erde
763-479-2158
763-420-2585
763-493-8350
763-493-8350
763-493-8350
612-822-2171
612-822-2171
612-822-2171
763/753/5090
763-878-2325
651-257-2295
320-693-6754
320-693-6754
612-822-2171
142
-------
27
24 Milena
Church
Milena's house
6/17/2002
58585 222nd Street,
Litchfield, MN55355
Milena
320-693-6754
143
-------
Table 2. Bacteriological, Coliphage and Virological Analysis of 27 Well Water Samples from
Minnesota
EPA PROJECT - COMPLETE RESULTS OF WATER SAMPLE TESTING
Bacteriological analysis
F.
Well#
01
Amundson
02 Gervais
03 Round
04 Turtle
05
Brookdale
06 Oak
Groove
07 Keller
Main
08 Keller
Main
09
Hamilton
10
Norwood
11
Presbyteria
n
12
Immanuel
13 Central
14
Historical
15
Fee.
Col.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Pos.,
30/10
0
Neg.
Pos.
1/100
Neg.
coli
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Enter-
ococci
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Pos.,
1/100
Neg.
Neg.
Neg.
Pos.,
1/100
Neg.
Pos.
1/100
Neg.
COLIPHAGES (Method 1601, 1602)
SOM-
ATIC
Pos.,
(tntc)
Pos.,
4/100
Pos.,
2/100
Pos.,
4/100
Neg.
Neg.
Pos.,
1/100
Pos.,
12/100
Neg.
Pos.,
28/100
Neg.
Neg.
Neg.
Neg.
Neg.
F+
Neg.
Neg.
Pos.,
4/100
Pos.,
2/100
Neg.
Neg.
Pos.,
58/100
Pos.,
40/100
Neg.
Pos.,
9/100
Neg.
Neg.
Neg.
Neg.
Neg.
TOTA
L
NT
Pos.,
5/100
Pos.,
4/100
Pos.,
3/100
Neg.
Neg.
Pos.,
4/100
Pos.,
7/100
Neg.
Neg,
0/100
Neg.
Neg.
Neg.
Neg.
Neg.
144
Method
1601 &
1602
1601 &
1602
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
Virus Isolation in cells
BGMK
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
3 passages,
no cpe
2 passages,
no cpe
3 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
3 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
Caco-2
2 passage, no
cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
3 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
RT-PCR
on Pooled
cell lysate
for Entero-
virus
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
-------
Northwood
16
Lakewood
1
17
Lakewood
Mausoleum
18
Lakewood
Mausoleum
19 Al
20 Dayne
21 Lake
Maria
22 Gibbs
Farms
23 Tom
Arendt
24 Milena
25 Jay
12 Church
Repeat
16
Cemetery
Repeat
24 Milena
Repeat
Pos.
1/100
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Pos.
1/100
Pos.
17/10
0
Pos.
3/100
Neg.
Neg.
Pos.
248/1
00
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Pos.,
12/1
00
Pos.,
1/10
0
Neg.
Neg.
Pos.,
3/10
0
Pos.,
1/100
Neg.
Neg.
Neg.
Neg.
Neg.
Pos.,
1/100
Neg.
Pos.,
2/100
Pos.,
15/100
Neg.
Neg.
Pos.,
20/100
Pos.,
574/100
Neg.
Neg.
Neg.
Neg.
Pos.,
9/100
Pos.,
2/100
Pos.,
574/100
Pos.,
574/100
Pos.,
574/100
Pos.,
574/100
Pos.,
574/100
Pos.,
574/100
Pos. ,
234/10
0
Neg.
Neg.
Neg.
Neg.
Neg.
Neg.
Pos.,
3/100
Pos.,
11/100
Neg.
Pos.,
3/100
Pos.,
6/100
Pos.,
2/100
Neg.
Neg.
Neg.
Neg.
Neg.
Pos.,
1/100
Pos.,
2/100
Pos.,
7/100
Pos.,
1/100
Pos.,
3/100
Pos.,
4/100
Pos.,
6/100
Neg.
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
Method
1601 &
1602
no cpe
2 passages,
no cpe
3 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
no cpe
2 passages,
no cpe
3 passages,
no cpe
2 passages,
no cpe
3 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
2 passages,
no cpe
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
145
-------
Positive UNH Bacterial Results for Groundwater Samples
Well Type and Number
Community, Well # 1
Private; shallow, Well #4
Private Well #9
Private, Well #15
Private, Well # 18
Private, Well # 19
Community, #23
Community, #23
Fecal Coliforms
100 ML
1
0
0
35
0
2
0
5
1L
200
1
0
TNTC
0
10
0
69
100 ML
0
0
0
0
1
13
2
2
Enterococcus
1L
0
0
3
0
93
89
87
32
Table 3. Results of UNH Groundwater Samples Analyzed for Adenovirus and Astrovirus
Well#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Results
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
for all viral analyses
146
-------
Table 4. Summary of UNH Samples Positive for Coliphages, Bacterial Indicators, Adenoviruses
and/or Astroviruses
Well Number and Type
Well # 3, Private
Well # 1, Community
Well # 9, Private
Well # 15, Community
Well# 18, Community
Well# 19, Community
Well #21, Private
Well #22, Private
Well #23,
All other wells
Samples positive for:
Coliphage
1
0
0
0
0
0
0
0
0
0
Bacterial Indicators
FC
0
1
0
1
0
1
0
1
1
0
Ent.
0
1
0
1
1
1
1
1
0
Any
1
1
1
1
1
1
1
1
0
Phage and
Bacterial
Indicators
0
0
0
0
0
0
0
0
0
0
Coliphage
and Virus
0
0
0
0
0
0
0
0
0
0
Bacterial
Indicators
and Virus
0
0
0
0
0
0
0
0
0
0
References
Chapron, C.D., Ballester, N.A., Fontaine, J.H., Frades, C.N. & Margolin, A.B. 2000 The
Detection of Astro virus, Entero virus and Adeno virus Type 40 and 41 in Surface Waters Collected
and Evaluated by the Information Collection Rule and Integrated Cell Culture/Nested PCR
Procedure. Appl. and Environ. Microbiol, 60 (6), 2520-2525.
Environmental Protection Agency. 1995 Virus Monitoring Protocol for the Information
Collection Requirements Rule. U.S. Environmental Protection Agency, publication EPA/814-B-
95-002. Government Printing Office, Cincinnati, Ohio.
147
-------
Environmental Protection Agency. 2000 Method 1601: Male-specific (F+) and Somatic
Coliphage in Water by Two-step Enrichment Procedure. Draft April 2000. Office of Water,
Washington, B.C.
Environmental Protection Agency. 2001 Method 1602: Male-specific (F+) and Somatic
Coliphage in Water by Single Agar Layer (SAL) Procedure. Draft January 2001. Office of
Water, Washington, D.C.
Xu, W., McDonough, M.C., & Erdman D.D. 2000. Species-specific identification of human
adenoviruses by a multiplex PCR assay. J. Clin. Microbiol. 39, 4114-4120.
148
------- |