EPA/600/R-03/047
September 2002
CENTERS FDR DISEASE
CONTROL AND PREVENTION
Animal Source Identification Using A
Cryptosporidium DNA Characterization
Technique
by
Michael Royer
U.S. Environmental Protection Agency
Edison, New Jersey 08837
Lihua Xiao and Altaf Lai
Centers for Disease Control and Prevention
Atlanta, Georgia 30341
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
NATIONAL CENTER FOR INFECTIOUS DISEASES
DIVISION OF PARASITIC DISEASES
CENTERS FOR DISEASE CONTROL and PREVENTION
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
ATLANTA, GEORGIA 30341
-------
NOTICE
The U.S. Environmental Protection Agency through its Office of Research and Development partially funded and
collaborated in the research described here under EPA/NRMRL-HHS/CDC interagency agreement DW 75937984.
It has been subjected to the Agencies' peer and administrative review and has been approved for publication as an
EPA document.
Mention of trade names or commercial products does not constitute endorsement or recommendation for use.
11
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with protecting the Nation's land, air,
and water resources. Under a mandate of national environmental laws, the Agency strives to formulate and
implement actions leading to a compatible balance between human activities and the ability of natural systems to
support and nurture life. To meet this mandate, EPA's research program is providing data and technical support for
solving environmental problems today and building a science knowledge base necessary to manage our ecological
resources wisely, understand how pollutants affect our health, and prevent or reduce environmental risks in the
future.
The National Risk Management Research Laboratory is the Agency's center for investigation of
technological and management approaches for preventing and reducing risks from pollution that threatens human
health and the environment. The focus of the Laboratory's research program is on methods and their cost-
effectiveness for prevention and control of pollution to air, land, water, and subsurface resources; protection of water
quality in public water systems; remediation of contaminated sites, sediments and ground water; prevention and
control of indoor air pollution; and restoration of ecosystems. NRMRL collaborates with both public and private
sector partners to foster technologies that reduce the cost of compliance and to anticipate emerging problems.
NRMRL's research provides solutions to environmental problems by: developing and promoting technologies that
protect and improve the environment; advancing scientific and engineering information to support regulatory and
policy decisions; and providing the technical support and information transfer to ensure implementation of
environmental regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is
published and made available by EPA's Office of Research and Development to assist the user community and to
link researchers with their clients.
Hugh W. McKinnon, Director
National Risk Management Research Laboratory
111
-------
ABSTRACT
This document summarizes the application of a particular molecular method to improve detection and
differentiation of species and genotypes of Cryptosporidium oocysts found in environmental samples. Of particular
interest is the method's potential for determining the source animal types of oocysts in water samples. The molecular
method is a nested polymerase chain reaction (PCR)-restriction fragment length polymorphism (RFLP) procedure
that characterizes the small sub-unit (SSU) ribosomal RNA gene. The method was previously developed for
characterizing oocyst DNA from clinical samples. The current project explores the method's applicability to
environmental water samples, which have greater diversity of oocyst species and strains, lower concentrations of
oocysts, and different interferents than clinical samples. Results include demonstrating that the method is capable of
detection and differentiation of at least 10 species and 22 genotypes of Cryptosporidium; method sensitivity
demonstrated to a single oocyst with laboratory samples; and detection and differentiation of oocysts from oyster gill
washings and hemolymph, storm water, surface water, and raw waste water. The method's capability to determine
an oocyst's source animal type was demonstrated by identification in environmental water samples of host-adapted
Cryptosporidium species and genotypes that were consistent with the source animal types (i.e., humans, farm
animals, wildlife, and/or pets) inhabiting the sampled watersheds.
IV
-------
TABLE OF CONTENTS
NOTICE ii
FOREWORD iii
ABSTRACT iv
LIST OF FIGURES vi
ACKNOWLEDGMENTS vii
INTRODUCTION 1
Cryptosporidiosis and Cryptosporidium 1
Benefits of Identifying Host Range of Cryptosporidium Oocysts in Water 1
Determining the Host Range of Oocysts in Water Samples 2
Developing a Collection of Infection Data for Cryptosporidium-Host Pairs 2
Methods for Detailed Characterization of Oocysts 2
Discovery of Correlations Between Characteristics of Oocysts and Their Host Ranges 3
PROJECT RATIONALE, OBJECTIVES, AND TASKS 4
MATERIALS AND METHODS 4
Materials and Methods for Phylogenetic Analysis of Cryptosporidium Genus Based on SSU rRNA Genes
4
Development Process for SSU rRNA-based Nested PCR-RFLP Method for Cryptosporidium Detection and
Differentiation 5
Evaluation of SSU rRNA-based Nested PCR-RFLP Method for Cryptosporidium Detection and
Differentiation in Storm Water Samples 6
KEY RESULTS 7
Results of Phylogenetic Analysis of Cryptosporidium Genus Based on SSU rRNA Genes of Five Types of
Cryptosporidium 7
Results of Development of SSU rRNA Nested PCR-RFLP Diagnostic Tool 8
Evaluation of the SSU rRNA-based Nested PCR-RFLP Diagnostic Tool 10
Gill Washings and Hemolymph from Oysters 10
Storm Stream Flow Samples 10
Raw Surface Water Samples 11
Raw Wastewater Samples 11
Comparison of PCR Protocols for Species Detection, Differentiation, and Genotyping of Cryptosporidium
11
CONCLUSIONS AND RECOMMENDATIONS 12
Specific Conclusions 12
General Conclusions 14
Recommendations 15
Current problems in molecular detection of Cryptosporidium oocysts 15
Actions needed to enable routine use of molecular tools in water sample analysis 15
REFERENCES 17
APPENDIX 1-MolecilarTooli 19
-------
LIST OF FIGURES
Figure 1. Detection and Diagnosis of Cryptosporidium Parasites by Nested PCR-RFLP 5
Figure 2. Phylogenetic Relationships of Cryptosporidium Parasites to Other Apicomplexans(A) and Each Other(B)
(Xiao et al, 1999a) 7
Figure 3. Updated phylogenetic relationship of Cryptosporidium parasites 8
Figure 4. Detection of Cryptosporidium spp. by SSU rRNA-based Nested PCR 8
Figure 5. Differentiation of Cryptosporidium Species and Genotypes by SSU rRNA-based PCR-RFLP 9
Figure 6. Sensitivity of the SSU rRNA-based Cryptosporidium PCR-RFLP Genotyping Technique 10
Figure 7. Differentiation of the Cryptosporidium Parasites in Storm Water Samples by SSU rRNA-based PCR-
RFLP 11
VI
-------
ACKNOWLEDGMENTS
Much of the work described in this report on the SSU rRNA nested PCR-RFLP method was conducted
under EPA/NRMRL-HHS/CDC interagency agreement DW 75937984. The work conducted under this interagency
agreement built upon method development work previously conducted under EPA/Office of Water-HHS/CDC
interagency agreement 75937730, and a substantial amount of collaboration also occurred. Key collaborating
researchers include Ronald Payer, Agriculture Research Service of the U.S. Department of Agriculture; Kerri
Alderisio, New York City Department of Environmental Protection; Una Ryan and R.C. Andrew Thompson,
Murdoch University, Western Australia; Steve Gradus and Ajaib Singh, City of Milwaukee Public Health
Laboratories; Thaddeus K. Graczyk, Johns Hopkins School of Hygiene and Public Health, and Joseph Limor, and
Irshad Sulaiman, Centers for Disease Control and Prevention.
Vll
-------
INTRODUCTION
The purpose of this document is to summarize progress to date on the application to water samples of a
molecular method for improved detection and differentiation of species and genotypes of Cryptosporidium oocysts.
Of particular interest is the method's potential for determining the source animals of oocysts found in water samples.
The molecular method is a small sub-unit ribosomal RNA (SSU rRNA) gene-based nested polymerase chain reaction
(PCR)-restriction fragment length polymorphism (RFLP) procedure previously developed for characterizing oocyst
DNA from clinical samples. Exploring its applicability beyond clinical samples to water samples - where the
diversity of oocysts is greater, the concentration of oocysts is much lower, and the interferents are different -- was
undertaken as part of EPA-CDC interagency agreement 75937984. The development and testing of the method are
reported in greater detail in the referenced peer-reviewed journal articles. A number of the references are available
in full text on the Internet.
Cryptosporidiosis and Cryptosporidium
Cryptosporidiosis is a protozoan infection of humans, domestic animals and other vertebrates. In young
farm animals, especially pre-weaned diary calves, it causes severe enteritis resulting in significant morbidity,
mortality and economic loss. In humans C. parvum results in acute infection of the digestive system in
immunocompetent individuals, and chronic, life-threatening disease in immunocompromised patients. Several
transmission routes, including person-to-person, contamination of water or food, and zoonotic infection, are possible
(Payer et al., 1997).
Waterborne Cryptosporidiosis outbreaks can occur when large numbers of pathogenic Cryptosporidium
oocysts from the intestinal tracts of infected animals or humans are discharged into the environment, transported into
the water supply and through water treatment processes and the distribution system, then out the tap in
concentrations exceeding the infectious dose. The infectious dose may be as low as 10 oocysts for some strains for
healthy individuals, and presumably less for immunocompromised persons. Oocysts may remain viable for months
in the environment. They may be removed by filtration and are susceptible to ozone and ultraviolet light treatment,
but their oocyst wall helps them survive routine chlorine disinfection. Therefore, water systems that chlorinate, but
do not filter, must be particularly cautious about monitoring for oocysts and preventing their entry into source waters.
A waterborne Cryptosporidiosis outbreak in Milwaukee, WI in 1993 resulted in approximately 403,000 illnesses
(Working Group on Waterborne Cryptosporidiosis, 1997). Numerous other Cryptosporidiosis outbreaks in the U.S.
and other countries have occurred.
Benefits of Identifying Host Range of Cryptosporidium Oocysts in Water
When Cryptosporidium oocysts are found at levels of concern in source or treated drinking water, officials
responsible for watershed, utility, or public health management desire to quickly determine or confirm the source of
contamination. An important piece of evidence is the oocyst's host range, which is the range of animal types that can
be infected by the oocyst's Cryptosporidium species. While the host range does not directly identify the specific
source or its location, it does enable a more focused and efficient investigation of the most likely sources. Not only
does the host range indicate the potential upstream sources, but also the susceptible downstream hosts.
Cryptosporidium species that are adapted to only one or a small number of host animal types are valuable
indicators of the specific animal source of these oocysts in water. Recently it has been discovered that
Cryptosporidium parvum, which was thought to be extremely non-host-specific (i.e., it was thought to infect 79
different species of mammals (Payer et al., 1997)) may in fact be a multi-species complex that contains many
genotypes that are very host-specific. If so, detection of these C. parvum genotypes in water would help eliminate all
other source animal types from further investigation, except those known to be suitable hosts. The C. parvum human
genotype (now known as C. hominis) is one of the recently discovered host-specific genotypes. Host specificity data
indicate that this genotype, which is difficult to distinguish from other C. parvum genotypes based on morphology, is
almost exclusively infective for humans and non-human primates. Therefore, when it is found in U.S. waters, then
1
-------
with very few exceptions where non-human primates are potential contributors to the oocyst load, it can be
concluded that the oocysts came from human sewage.
For those Cryptosporidium species that have very broad host specificity (i.e., they can infect many members
of a taxonomic class, e.g., mammals or birds or reptiles), the benefits of host range determination are limited, but still
may be useful for decisions that only require knowledge of the host at the class level (e.g., mammals or birds) and for
the determination whether it belongs to a known human-pathogenic Cryptosporidium species.
Determining the Host Range of Oocysts in Water Samples
Determining the host range of the recovered oocysts is accomplished by (1) characterizing the oocyst, (2)
assigning the oocyst to its appropriate species or genotype, and (3) consulting available Cryptosporidium-host
infection data to determine the potential host animals for the Cryptosporidium species or genotype that was found.
Before the host range of oocysts recovered in a water sample can be determined, three requirements must be met: (1)
development of a collection of infection data for the pairs of hosts and Cryptosporidium species or genotypes of
interest in the watershed; (2) development of methods capable of detailed characterization of oocysts, if necessary to
the molecular level, to enable reliable identification and differentiation of oocyst types; and, (3) discovery of
correlations between measured characteristics of the oocysts' and their host ranges.
These three requirements are interdependent. The number ofhost-Cryptosporidium pairs in the infection
database is influenced by the number of Cryptosporidium species and genotypes as well as the number of hosts and
their immunocompetence categories (e.g. newborns, children, adults, elderly, AIDs, chemotherapy patients). The
Cryptosporidium species and genotype categories are affected by the capabilities and limitations of methods for
measuring oocyst characteristics. The oocyst characteristics are only useful for predicting host range if a reliable
correlation exists between the chosen characteristics and the host range of the oocyst. A brief discussion of the
individual requirements follows.
Developing a Collection of Infection Data for Cryptosporidium-Host Pairs
The host range of a particular Cryptosporidium species or genotype can be determined by experimental
infections of host animals where feasible. Identification of the Cryptosporidium species or genotype in naturally
infected animals also proves they are in the host range, but absence of Cryptosporidium in a host in a natural setting
does not confirm non-infectivity, since exposure is not strictly controlled. A considerable amount of host specificity
data has been collected. However, as Cryptosporidium species and genotype categories are split or joined together,
it will be necessary to re-assess previous conclusions about the host range of Cryptosporidium species and
genotypes. Undertaking human subjects testing is particularly rigorous, lengthy, and costly. The human infection
studies cannot be done if there is excessive risk to the participants, which would be the case for immunosuppressed
persons. The ideal, complete host specificity database would be a matrix with all of the relevant Cryptosporidium
species and genotypes in the row headings, all the host animals with their relevant immunocompetence levels in the
column headings, and results of infection studies in each corresponding cell. Generating all the data required to
populate the ideal database is probably not attainable, and the recent discoveries of new Cryptosporidium species
and the tentative identification by molecular tools of host-adapted genotypes indicates that the Cryptosporidium
species designations will remain in flux for some time. However, only a relatively small portion of the ideal database
is required for any particular watershed or situation, so the inability to fully populate the database is not a critical
problem. One type of infection data that is lacking is the characterization of Cryptosporidium species and genotypes
that infect wild animals.
Methods for Detailed Characterization of Oocysts
The second prerequisite for determining the host range of an oocyst in a water sample is the existence of
methods to determine Cryptosporidium species and genotype directly from the oocyst. This includes characterizing
the species based on the small number of oocysts likely to be present in the sample, eliminating or overcoming
interferences that may be present, and detecting unique features that define the Cryptosporidium species, including
its host specificity.
-------
Conventional Detection and Differentiation of Cryptosporidium Oocysts
The methods commonly used to detect and differentiate Cryptosporidium oocysts often do not enable
correlation of oocyst characteristics with its host range. Currently the identification of Cryptosporidium oocysts in
water samples is largely made by the use of immunofluorescent assay (IFA) after concentration processes (ICR
method, EPA method 1622/1623, flow cytometric method, solid-phase cytometric method, etc.) (Lindquist, et al.,
2001) (Xiao et al., 2002a). Because IFA detects oocysts from most Cryptosporidium parasites, the species
distribution of Cryptosporidium parasites in water samples cannot be assessed. In addition, diagnosis of
Cryptosporidium parasites to the species specific level is a challenge because many of the Cryptosporidium species
are morphologically similar. For example, it is very difficult for an experienced parasitologist to differentiate C.
parvum, C. wrairi, C. meleagridis, C.felis, C. canis , and C. saurophilum under a microscope. Morphometric
measurements are also needed for the differentiation of C. muris and C. andersoni from C. parvum and C. parvum-
related species, which can be problematic with water samples that normally have only a few oocysts. As a result,
conventional diagnostic practices rely on the presumed host specificity of Cryptosporidium parasites in addition to
morphology.
Molecular Tools for Detection and Differentiation of Cryptosporidium Oocysts
The use of molecular tools enables differentiation between oocysts that was not previously possible. The
term "molecular tools" is used here to refer to the procedures that are used, separately or in combination, to
characterize DNA sequences of selected portions of an organism's genome. The Cryptosporidium genome contains
approximately 10 to 20 million base pairs(bp) of DNA (Jenkins and Petersen, 1997). This project focused on
detection and differentiation of Cryptosporidium species and genotypes based on polymorphism in the SSU rRNA
gene, which is approximately 1733 to 1750 bp in length. Of particular relevance to this project are Polymerase chain
reaction (PCR) procedures, restriction fragment length polymorphism (RFLP) procedures, and DNA sequencing.
PCR rapidly generates thousands to billions of copies of targeted DNA sequences for use in other procedures (e.g.,
DNA sequencing and RFLP). In addition to copying target gene sequences for use in other methods, PCR can also
be used to detect the presence/absence of particular organisms by copying/not copying their DNA. DNA sequencing
determines the exact order of the nucleotides in the DNA molecule. The RFLP procedure is a faster, less costly, and
less detailed approach to characterizing DNA. The RFLP method cuts a selected segment of DNA into fragments
using restriction enzymes. The differing DNA fragment lengths are separated by electrophoresis and visualized by
staining procedures. If a unique number and length of fragments are formed for a particular species and genotype
then it can be identified by this method. Additional description of molecular tools is in Appendix 1.
Discovery of Correlations Between Characteristics of Oocysts and Their Host Ranges
The third prerequisite for determining the host range of an oocyst in a water sample is the discovery of
correlations between measured characteristics of the oocysts and their host ranges. Previous attempts to establish a
fully reliable correlation have been unsuccessful. Numerous instances have occurred where presumably identical
Cryptosporidium species differed significantly in host specificity. These past failures occurred in large part due to
the inability to characterize the oocysts in sufficient detail to detect all differences relevant to host specificity. As
indicated above, molecular tools now enable the DNA sequences of oocysts to be characterized. The rationale for
using molecular tools to characterize oocyst DNA in order to determine its host animal range is that the oocyst is pre-
disposed, by the structural proteins and enzymes that are coded for in its DNA, to survive and reproduce in a limited
range of host environments. These host-adapted Cryptosporidium may develop over long periods of time through
co-evolution with the host animal. If Cryptosporidium species or genotypes differ in host specificity, then DNA
differences should occur between these species or genotypes at one or more gene loci. The challenge is to find one
or more genes where the presence of a particular set of DNA sequences is unique to all Cryptosporidium oocysts that
have the same host range. In this and other projects a number of different gene loci have been investigated for the
presence of these host-range-indicating sequences. The gene that codes for the small subunit ribosomal RNA has
been found to be a particularly promising site, since it contains DNA sequences that appear to be unique to the
Cryptosporidium genus, to particular Cryptosporidium species, and to host-adapted species and genotypes.
-------
PROJECT RATIONALE, OBJECTIVES, AND TASKS
The development and use of molecular tools for genetic analysis of Cryptosporidium oocysts is relatively
new, as is the documentation of the host-specific behavior of the various genotypes that have been (or will be)
discovered. Even less developed is the application of the new molecular tools to the investigation of the distribution
of Cryptosporidium species and genotypes in water samples, which is a much different matrix than clinical samples.
Specifically, water samples have low concentrations of oocysts, different interferents, and probably a wider range of
oocyst species and strains from not only humans, but also farm animals, companion animals, and wild animals.
Recognizing: (1) the investigative value of determining, directly from the oocyst, the host animals in which
waterborne oocysts were produced; (2) the promise of molecular tools for characterizing Cryptosporidium species
and genotypes of waterborne oocysts, and (3) the need for further development and evaluation of molecular tools
before they can be used for oocyst source animal determination, an interagency agreement was initiated between the
U.S. EPA/NRMRL and HHS/CDC. The ultimate objective of the project was to improve the techniques available to
investigate and prevent waterborne cryptosporidiosis. The particular focus of the interagency agreement was to
determine whether PCR-restriction enzyme digestion and sequencing assays developed by CDC (and partially
funded under EPA/Office of Water - HHS/CDC interagency agreement 75937730), which had been successfully
used on clinical samples, could also be applied to water samples and be incorporated into investigative approaches
for determining the sources of Cryptosporidium in water supplies.
Key tasks were: (1) confirmation that the genus Cryptosporidium is a multi-species complex by completion
of a phylogenetic analysis based on characterization and comparison of the SSU rRNA genes of four types of
Cryptosporidium', (2) development of an SSU rRNA-based nested PCR-RFLP method for detecting and
differentiating all known species of Cryptosporidium and multiple C. parvum genotypes; (3) testing the SSU rRNA-
based nested PCR-RFLP method on multiple samples from (a) hemolymph and gill washings from oysters, (b) storm
water, (c) surface water, and (d) wastewater to determine its ability to detect and differentiate oocysts and to
determine whether the oocyst types matched the probable source animal types; and (4) evaluate and briefly
summarize the promise and the challenges for molecular tools for detection of Cryptosporidium oocysts in water.
-------
MATERIALS AND METHODS
The materials and methods used for the phylogenetic analysis of the Cryptosporidium genus and the
development and evaluation of the SSU rRNA-based nested PCR-RFLP method for identification and differentiation
of Cryptosporidium species and genotypes are described in detail in the referenced articles (Xiao et al., 1998, 1999 a
and b, 2000, 2001, and 2002a; and, Sulaiman et al., 1999). In a similar or identical manner the method was also
applied to gill washings and hemolymph from oysters, surface water from several states, and raw wastewater. Other
gene loci (e.g., beta-tubulin, actin, 70kDa heat shock protein, and thrombospondin anonymous protein genes) were
also investigated (Sulaiman et al., 1998 and 1999a, 2000, and 2002) under the interagency agreement and other
projects for species detection and differentiation, but these efforts are not discussed further here. The review of the
SSU rRNA method and other molecular methods for oocyst DNA characterization was completed based on review
of the literature and summarization of relevant research by the review authors.
Materials and Methods for Phylogenetic Analysis of Cryptosporidium Genus Based on SSU
rRNA Genes
The SSU rRNA gene was selected for the phylogenetic analysis for several reasons (Xiao et al., 1999a and
b). Ample sequence data are available for this gene since it has been extensively studied because it is present in all
eukaryotic organisms and it plays an important role in protein synthesis. The gene has conserved regions, which
may contain genus/species specific detection sites. The gene also has polymorphic regions, which may contain
species/genotype differentiation sites. There are five copies of the rRNA gene per sporozoite, which increases
sensitivity of detection by PCR.
Isolates of Cryptosporidium parasites for SSU rRNA gene sequencing were obtained from humans, cattle,
calf, snakes, lizards, a guinea pig, a camel, a hyrax, a chicken, a rhesus monkey, a ferret, a pig, a dog, a kangaroo, a
turkey, and a cat. Partial sequences covering the most polymorphic regions were obtained from C. parvum human
genotype isolates, bovine isolates, and one C. baileyi isolate. Oocyst type determination was by morphology and/or
host specificity. DNA extraction was extracted by freeze-thaw and adsorption procedures.
The full-length SSU rRNA gene was amplified from each sample by conventional PCR by using forward
and reverse primers with 25 nucleotides each. The SSU rRNA gene is 1733 to 1750 bp long, depending on species
and genotype. Each PCR consisted of 35 cycles of denaturation at 94«C for 45 s, annealing at 60«C for 45 s, and
extension at 72«C for 60 s; an initial denaturation step consisting of incubation at 94«C for 5 min and a final
extension step consisting of incubation at 72«C for 10 min were also included. The copied DNA segments were
sequenced with an ABI377 autosequencer (Perkin Elmer, Foster City, Calif.)
The SSU rRNA sequences of Cryptosporidium from this study were aligned and compared to other
apicomplexan parasites by a neighbor joining (NJ) tree analysis to assess the genus status of Cryptosporidium. The
SSU rRNA sequences of Cryptosporidium species from this study were also aligned and compared to each other by
a neighbor joining tree analysis. The analysis checked for evolutionary distances indicating separate species,
clustering within species, and clustering between species with similar host specificity, and similarities in locations of
mutations.
Development Process for SSU rRNA-based Nested PCR-RFLP Method for Cryptosporidium
Detection and Differentiation
The conceptual design of the SSU rRNA-based nested PCR-RFLP method is illustrated in Figure 1. The
SSU rRNA gene was selected as the target gene for the reasons previously described. DNA sequences from four
Cryptosporidium species (i.e., C. parvum, C. serpentis, C. muris, and C. baileyi), which were obtained during the
previously described phylogenetic analysis, were used in the initial method development process (Xiao et al.,
1999a).
-------
18S rRNA
1733 1750
Primary PCK
Secondary PCR
1325 bp
820 bp
Restriction Digestion
^" ^^^" Various fragments
Figure 1. Detection and Diagnosis of Cryptosporidium Parasites by Nested PCR-RFLP
The DNA sequences were aligned and examined for restriction sites (i.e., sites at which available restriction
enzymes will cut the DNA). Sspl restriction enzyme was selected for species diagnosis. Vspl restriction enzyme
was selected for genotyping of C. parvum. A DNA sequence (~ 1325 bp) in the SSU rRNA gene was identified for
amplification by primary PCR and within that sequence an -820 bp target segment was selected for secondary PCR.
The forward and reverse primers (~ 20 bp in length) selected for primary and secondary PCR are: (1) common to the
Cryptosporidium genus, but not present in the DNA of other microorganisms, and (2) bracket the target region (~
820 bp) of the SSU rRNA gene that contains unique sequences that enable species and genotype differentiation by
RFLP and/or DNA sequencing. Nested PCR was chosen to maximize sensitivity of detection. Nested PCR first
amplifies the larger (-1325 bp) segment, and then, starting with some of the primary PCR product, amplifies the
smaller (~ 820 bp) segment, which improves the overall efficiency of target DNA amplification compared to single-
round PCR. Primer selectivity was confirmed by detection (i.e., PCR amplification) when Cryptosporidium DNA
(i.e., C. parvum bovine and human genotypes; C. muris, C. serpentis, and C. baileyi) were present and non-
detection (i.e., no PCR amplification) when non-Cryptosporidium parasite DNA (i.e., Eimeria and Giardia) was
present. DNA from Cryptosporidium species and genotypes were also digested with Sspl and Vspl restriction
enzymes to determine whether they produced the predicted and unique restriction fragment patterns upon
electrophoresis and visualization by ethidium bromide. Sensitivity of the S SU PCR-RFLP method was confirmed by
testing on serial dilutions of DNA to an equivalent of one oocyst.
For Cryptosporidium species and strains not detected or differentiated by this PCR-RFLP approach there
are several options to address the problem. Cryptosporidium oocyst types not detected indicates sequence
differences in primer regions, and this problem may be addressed by different primers to detect those
Cryptosporidium species or genotypes. No examples of non-detectable Cryptosporidium species have been found to
date. If Cryptosporidium oocyst types are detected, but not differentiated, then these oocyst types may be
differentiated by (1) use of different restriction enzymes or (2) direct sequencing. Cryptosporidium oocyst types that
are known to be different (e.g. morphology or host specificity), but do not have SSU rRNA sequence differences will
not be differentiable by RFLP at the SSU rRNA gene locus, but may be detectable and differentiable by PCR-RFLP
at other gene loci. No examples of this situation have been found to date.
Evaluation of SSU rRNA-based Nested PCR-RFLP Method for Cryptosporidium Detection
and Differentiation in Storm Water Samples
-------
The key animal sources of oocysts were identified by the environmental setting and general knowledge of
the sampling sites (Xiao et al., 2000). Grab or composite samples were taken during storm flows. The Information
Collection Rule (ICR) method was used for storm water sample collection/processing. (In subsequent evaluations
EPA Method 1623 was used for surface water sample collection and processing, and centrifugation of grab samples
was employed for wastewater processing). Oocysts in water samples that were concentrated by filtration, Percoll-
sucrose floatation or centrifugation were further purified by immunomagnetic separation (IMS). Direct DNA
extraction without IMS interfered with PCR because of the presence of PCR inhibitors. IMS-purified oocysts were
subjected to 5 freeze-thaw cycles, incubated with 1 mg/ml of proteinase K at 56 °C for at least 1 h, and diluted with
equal volume of pure ethanol. Oocyst DNA was extracted by passing the oocyst-ethanol suspension through
QIAamp DNA Mini isolation columns.
For the PCR-RFLP analysis a primary PCR product of about 1,325 bp was amplified. Thirty-five (35)
replication cycles were completed at about 2.5 minutes per cycle. Secondary PCR product of 826-864 bp
(depending on isolates) was then amplified from 2 t of the primary PCR reaction, using different primers. For
restriction fragment analysis, 20 1 of the secondary PCR product was digested in a total of 50 t reaction mix,
consisting of 20 units of Ssp I or Vsp I and 5 1 of respective restriction buffer at 37 C for 1 hr. The digested
products were fractionated on 2.0% agarose gel and visualized by ethidium bromide staining and/or characterized by
DNA sequencing. A modified procedure was employed for multiple species in a single sample. Each sample was
analyzed at least 3 times by PCR-RFLP, using different volumes of DNA preparation (0.25, 0.5, and 1 |al) for PCR.
Where multiple species occur in a single sample, then multiple additional bands occur after electrophoresis, which
causes difficulty in interpreting the results. Since one oocyst type usually predominates in a sample, dilution of
DNA prior to PCR and multiple PCR assays can increase detection of the non-dominant type. If suitable separation
cannot be obtained, then the fragments can be cloned and multiple clones sequenced. Both approaches increase time
and expense.
For confirmation, the secondary PCR products were sequenced using an ABI377 autosequencer.
Nucleotide sequences generated were aligned with each other and with known Cryptosporidium species and C.
parvum genotypes previously obtained, using computer software Wisconsin Package Version 9.0 (Genetics
Computer Group, Wisconsin) and manual adjustment. Phylogenetic analysis (i.e. construction of evolutionary tree
based on gene sequence similarity) was performed on the aligned sequences to assess relationship among isolates.
-------
KEY RESULTS
Results of Phylogenetic Analysis of Cryptosporidium Genus Based on SSU rRNA Genes of
Five Types of Cryptosporidium
Although biological data support the hypothesis that there are multiple species in the genus
Cryptosporidium, a previous analysis of the available genetic data suggested that there is insufficient evidence for
species differentiation, and hence that it would be infeasible to determine animal source types by genetic
characterization. However, this study (Xiao et al., 1999a) revealed that the genus Cryptosporidium is distinct from
other apicomplexan parasites. Also, Cryptosporidium contains phylogenetically distinct species such as C. parvum,
C. muris, C. baileyi, and C. serpentis, which is consistent with the biological characteristics and host specificity
data. The Cryptosporidium species formed two clades (i.e., groups), with C. parvum and C. baileyi belonging to one
clade and C. muris and C. serpentis belonging to the other clade (Figure 2). Another study (Xiao et al., 1999b)
extended the phylogenetic analysis to include C. felis, C. meleagridis, and some additional host-adapted C. parvum
genotypes (dog, pig, kangaroo, ferret, mouse, and monkey). Subsequent SSU rRNA gene sequencing and
phylogenetic analyses have confirmed the groupings described above and produced a more detailed characterization
of the phylogenetic relationship of Cryptosporidium parasites as shown in Figure 3 (Xiao et al., 2002b).
A
0.05
100
100
B
C. parvum
C. wmiri
C. baileyi
101
\C. serpentis
100
100
- Theileri&patva.
Cytauxzoonfetts
- Ba&esui flivergens
S5j Eimeria tenella
'oa p Eimeriapmecox
Cyctospotv, sp.
Safcacystis tenella
Frenfalia micrvti
101
j F,
T-l /:
(fl[rffi«
100 'Kens
_ . i f Toxaplasma gondti
100 \Neospom conation
Perfansts marinas
98 r Gymnodinsan &nptex
"lf£i Gytodinium impmliaim
C. parvum HFL5
C. pan/urn HFL2
C.jBramHCNV4 '
CparvwmDRt
C. parvum BOH6
C,wrairi
C. serpentis CSP02
CserpenfeCSP04
99
Cserpen»sCSP01
C.serpenftsCSP05
»OOC.munslDRH 13
|amunsCMUD3
CfnunslDVS-811
C. haileyi'CBA01
Figure 2. Phylogenetic Relationships of Cryptosporidium Parasites to Other
Apicomplexans(A) and Each Other(B) (Xiao et al., 1999a)
-------
0.02 substitutions/site
I 1
C mean (517)
Oyptosporidium ferret genotype (351)
C pa.yvv.7n mouse genotype (411)
C parvum, bovine genotype (6)
C meleagridis (295)
C patvuni rabbit genotype (2244)
C patvum human genotype (120)
C pawum monkey genotype (44)
dyptesporidium skunk genotype (1170)
I as. ff? r C^pfo^pjit&wm opossum genotype I (1041)
' Qypfospotidiiiiti marsupial genotype (428)
Ctyptospotidium pig genotype (427)
Oyptospondiiitn fox genotype (2041)
Gypfc>jsjj0«(£zim muskrat genotype (5490)
Oyptosporidium opossum genotype II (1040)
Qypt&spari-dium deer mouse genotype (1457)
Ctyptospo fidium bear genotype (961)
C careJs coyote genotype (2011)
C catiis fox-genotype (5035)
C canis dog genotype (244)
yphilnjn (340)
Cfelis&W)
Ctypt&spt} fidiufn, bovine genotype B (2622)
Qypiospofidiiim. deer genotype (2040)
" ' ' goose genotype (886)
G^pto^'C'n-t&um. tortoise genotype (750)
T *t /- M+^ «i"i
Ctyptospofidium, snake genotype (938)
C baiieyi (39)
OyptodjJoni&JiTra Hiard genotype (1665)
* t C mwfis (34)
" C andersoni (20)
Figure 3. Updated phylogenetic relationship of Cryptosporidium parasites inferred by the neighbor-
joining analysis of the SSU rRNA gene based on genetic distances calculated by the Kimura two-
parameter model. The tree was rooted with an SSU rRNA sequence from Eimeria tenella, and the
root was removed to show the details of the relationship among Cryptosporidium parasites.
Bootstrap values (in percentage) above 50 from 1,000 pseudo-replicates are shown for both the
neighbor-joining (the first value) and maximum parsimony analyses (the second value). Figures in
parentheses are isolates' designation numbers. (Xiao et al., 2002b)
Results of Development of SSU rRNA Nested PCR-RFLP Diagnostic Tool
Based on the genetic information obtained and analyzed during the phylogenetic analysis, a species- and
strain-specific PCR-RFLP diagnostic tool was developed. The diagnostic tool reveals the presence of
Cryptosporidium oocysts by amplification of the primary and secondary PCR target DNA sequences. If the target
sequences are not produced, then Cryptosporidium DNA are not present in the sample. Figure 4 illustrates the
positive response for 5 types of Cryptosporidium oocysts and absence of target DNA replication when two non-
Cryptosporidium parasites' DNA are tested (Xiao et al., 1999a). In the RFLP procedure, digestion of the
secondary PCR products with Sspl and Vspl restriction enzymes produces unique patterns that often enable
differentiation of Cryptosporidium species and C. parvum genotypes. Differentiation of some species and genotypes
may require either a third digestion with a different restriction enzyme or direct sequencing (Xiao et al., 1999b; Xiao
et al., 2001).
-------
3456
1. Molecular markers
2. C, parvum bovine genotype
3. C. pan'ui/t human genotype
4, C tnurii
-'. I j. A tLI ItC.Il 1 i.}
6. C. bailtyi
7. E. papillala
8, E. niescbuly
9. G. duadenalis
10. Negetive control
Upper panel: primary PCR products
Lower panel: secondary PCR products
Figure 4. Detection of Cryptosporidium spp. by SSU rRNA-based Nested PCR
Figure 5 shows the results of the SSU rRNA-based PCR-RFLP procedure on 14 types of Cryptosporidium oocyst
DNA. There were seven non-C. parvum species and seven C. parvum genotypes. Eight of the types, comprised of
four species (i.e., C. muris, C. serpentis, C. baileyi, and C. felis) and four C. parvum genotypes (i.e., pig, marsupial,
bovine, and human) produced unique restriction fragment patterns that enabled them to be differentiated. The
remaining six types fell into two groups that could be differentiated from the first eight, but not from each other, by
electrophoresis and ethidium bromide staining. These remaining types could be differentiated by DNA sequencing
of the secondary PCR fragments. One group contained C. meleagridis, C. parvum ferret, and C. parvum mouse and
the other group was C. parvum dog, C. lizard sp., and C. wrairi.
12 34 5 6 7 B 9 10 1112 13 14
1.500
1,500
«00
300
2.
3.
4.
5-
6.
7.
8.
9,
ID.
11
12.
13.
14.
C. muris
C. serpentis
C iailaji
C trnxri
G parvum pig geitolype
d parvum dog genotype
Cryptosporidium sp .: fizard
. C. parvum ferret genotype
C. parvum marsupial genotype
. d parvatn mo us*? genotype
C, parvum bovine genotype
C. parvum human genotype
Upperpanel: Sspl. digestion
Lowerpanel: Vspl digestion
Figure 5. Differentiation of Cryptosporidium Species and Genotypes by
SSU rRNA-based PCR-RFLP
10
-------
The SSU-rRNA nested PCR-RFLP method was demonstrated to detect a single oocyst in laboratory
dilution tests (Figure 6). In one field test it indicated presence of oocysts with slightly higher frequency than
microscopic examination (Sulaiman et al., 1999b).
8 9 10 11
1. 10,000 oocysts
2. 4,000 oocysts
3. 400 oocysts
4. 100 oocysts
5. 20 oocysts
6. 10 oocysts
7. 5 oocysts
8. 2 oocysts
9. 1 oocyst
10. 0.5 oocyst
11.0 oocyst
Figure 6. Sensitivity of the SSU rRNA-based
Cryptosporidium PCR-RFLP Genotyping Technique
Initially oocysts concentrated by the ICR procedure could not be amplified by PCR, but it was then found
that the immunomagnetic separation (IMS) technique utilized in the newer protozoa concentration/separation
methods (EPA 1622 and 1623) successfully removed interferents and produced oocysts that can be amplified by
PCR. An exhaustive examination of the effects of a wide range of interferents on detection and differentiation of
oocysts was not within the scope of the project, although it was shown that IMS-PCR detects and differentiates
oocysts in surface water, storm water, and wastewater (Xiao et al., 2000).
Discrimination of different genotypes within a single sample was done by dilution and multiple PCR-RFLP
analyses. If discrimination is not possible by PCR-RFLP, then it may be possible by DNA sequencing (Xiao et al.,
2000 and 2001).
Evaluation of the SSU rRNA-based Nested PCR-RFLP Diagnostic Tool
The SSU rRNA-PCR-RFLP method was successfully used to differentiate Cryptosporidium species and C.
parvum genotypes in gill washings and hemolymph from oysters, storm water, raw surface water, and wastewater.
Gill Washings and Hemolymph from Oysters
The diagnostic tool was used to analyze oocysts recovered from the hemolymph and gill washings of
oysters collected from the Chesapeake Bay. Oysters are filter feeders that concentrate and accumulate oocysts from
surface water, thus enabling the investigator to avoid these tasks. Sixty-five pooled oyster samples were analyzed.
Cryptosporidium oocysts were present in 26 samples. Twenty-four of the samples contained C. parvum and each of
the other samples contained C. baileyi (typically found in birds) and C. serpentis (typically found in snakes). Of the
C. parvum positive samples, 22 of 24 were the bovine genotype (also known as genotype 2), which is typically from
cattle, humans, and other ruminants. Two samples were positive for C. parvum human genotype (also known as
genotype 1), which only circulates among humans. (Xiao et al., 1998)
Storm Stream Flow Samples
When the molecular tool was applied to water samples from storm stream flows in the New York City
Watershed, 12 genotypes were found in 27 of 29 samples. Four of the 12 genotypes matched sequences from known
Cryptosporidium parasites: C. baileyi (from birds), Cryptosporidium from snakes, and two Cryptosporidium
genotypes from opossums. No genotypes found in the storm samples matched those from humans, farm animals, or
companion animals (i.e., C. fells, C. meleagridis, C. andersoni, and the human, bovine, and pig genotypes of C.
parvum), indicating that genotypes in storm water were probably from wildlife (Figure 7). This conclusion is
consistent with the environmental setting of the sampling sites and the presence of the four genotypes with known
11
-------
(wild) animal sources (Xiao et al., 2000). The presence of unknown genotypes indicates the existence of, and need
to fill, gaps in the genotyping of Cryptosporidium oocysts in wildlife.
1 234 56 78 9 10 11
bp
Figure 7. Differentiation of the Cryptosporidium
Parasites in Storm Water Samples by SSU rRNA-based
PCR-RFLP.
Lanes 1,2,4,8,10, and 11, unknown Cryptosporidium
spp.;
lane 3, Cryptosporidium from snakes;
lane 5, C. baileyi;
lane 6, Cryptosporidium opossum genotype 2;
lanes 7 and 9, C. parvum bovine-like genotype.
Raw Surface Water Samples
When the molecular tool was applied to raw surface water samples that were collected from several states,
25 of the 55 surface water samples were positive for Cryptosporidium. The species and/or genotypes found in raw
surface water were: C. parvum human and bovine genotypes, C. baileyi, and C. andersoni. C. parvum (both human
and bovine genotypes) was the predominant (21 of 25 samples) Cryptosporidium species found in surface water.
C. andersoni, which occurs in juvenile and adult cattle, was found at a moderate frequency (5 samples). Many
surface water samples, particularly those from the Chesapeake Bay sampling sites, contained more than one
genotype. The distribution of host-adapted Cryptosporidium species detected was consistent with the observed
potential sources (i.e., cattle farm runoff or wastewater discharges) of water contamination (Xiao et al., 2001).
Raw Wastewater Samples
When the molecular tool was applied to raw wastewater samples from a wastewater treatment plant in
Milwaukee, WI, 12 of the 49 samples were positive for Cryptosporidium. C. parvum human, bovine, and dog
genotypes (now known as C. canis (Payer, et al., 2001)); C. felis, C. andersoni, C. muris, and an unknown genotype
(now known to be Cryptosporidium deer genotype) were found in the wastewater samples. C. andersoni was found
at the highest frequency in the wastewater samples, which is consistent with (1) known occurrence of C. andersoni
in juvenile and adult cattle and (2) the presence of a large slaughterhouse (1800 beef cattle per day) that discharged
pre-treated effluent into the sewer system several miles upstream of the treatment plant. The appearance of C. muris,
probably from rodents, and C. parvum dog genotype is also consistent with potential fecal microorganism
contributors to an urban sewer system. The low observed frequency of C. parvum human genotype was surprising,
but may be explained by the sampling period occurring in April to July, when the incidence of human
cryptosporidiosis tends to be low (Xiao et al., 2001).
Comparison of PCR Protocols for Species Detection, Differentiation, and Genotyping of
Cryptosporidium
In 1999 the specificity and sensitivity of 11 PCR protocols were evaluated for species detection,
differentiation and genotyping of Cryptosporidium parasites in clinical samples. Although many of the protocols
performed well in their particular niches, the SSU rRNA nested PCR offered a wider range of detection,
differentiation, and genotyping capability and better sensitivity than the other tools. Ten protocols amplified C.
parvum genotypes l(human) and 2(bovine), and the expected fragment sizes were obtained. Two species-
12
-------
differentiating protocols were not Cryptosporidium specific, as the primers used in these protocols also amplified the
DNA ofEimeha species. Six C. parvum genotyping protocols that were based on six different gene loci were
limited to C. parvum, since the primers used in these protocols amplified only the DNA of genotype 1 and/or
genotype 2 isolates of C. parvum, but not the DNA of non-C.parvum oocysts. Sensitivity studies revealed that two
nested PCR-RFLP protocols - the one based on the SSU rRNA gene described in this summary and the other on the
dihydrofolate reductase genes - are more sensitive than single-round PCR or PCR-RFLP protocols (Sulaiman et al.,
1999b). A literature review of molecular detection of Cryptosporidium oocysts in water was also completed in 2002
(Xiao et al., 2002a). Twelve protocols were identified. These methods include PCR, nested PCR, IMS-PCR, IMS-
nested PCR, IMS-nested PCR-RFLP, Reverse Transcription (RT)-PCR, and cell culture (CC)-PCR. Other method
differences include the target genes (e.g., SSU rRNA, Oocyst wall protein, HSP 70, undefined sequences, and
TRAP-C2) and target gene regionis, the DNA extraction method, and the amount of testing to date on environmental
samples. Again, the SSU rRNA nested PCR methods, of which there now are several, offer a wider range of
detection, differentiation, and genotyping capability, as well as improved sensitivity compared to the other tools.
13
-------
CONCLUSIONS AND RECOMMENDATIONS
Specific Conclusions
A molecular method (SSU rRNA-based nested PCR-RFLP) has been developed and demonstrated to
be capable of detecting and differentiating, in clinical or environmental samples, at least 10 species and 22
genotypes of Cryptosporidium. The species detected and differentiated to date by this method are: C. andersoni,
C. baileyi, C. canis, C.felis, C. meleagridis, C. muris, C. parvum, C. saurophilum, C. serpentis, and C. wrairi. The
Cryptosporidium genotypes detected and differentiated are: bovine (2 genotypes), ferret, human (C. hominis), pig,
marsupial, rabbit, mouse, deer (2 genotypes), deer mouse, bear, skunk, opossum (2 genotypes), fox (two genotypes),
muskrat, goose, snake, lizard, and tortoise.
The SSU rRNA-based nested PCR-RFLP method has been used to show the presence of five (5)
species or genotypes of Cryptosporidium that have been found in human patients. The confirmed human-
infective species and genotypes are: C. parvum human genotype, C. parvum bovine genotype, C. felis, C. canis, and
C. meleagridis. Other Cryptosporidium parasites such as C. andersoni, C. muris, Cryptosporidium cervine and pig
genotypes have also been found in humans, but much less frequently than the 5 common Cryptosporidium parasites.
The SSU rRNA-based nested PCR-RFLP method can help prevent over-estimation of the human-
pathogenic potential of oocysts found in water samples by enabling oocysts to be grouped, depending on
available genetic and host-specificity data as: (1) known human pathogenic, (2) suspected human pathogenic,
and (3) known non-human pathogenic. This capability is of particular value where the oocysts are difficult to
distinguish by typical microscopic techniques, and the conservative assumption that all oocysts are human
pathogenic is much costlier than a more accurate assessment.
For Cryptosporidium species and genotypes that are known with some confidence to strictly or
primarily infect particular animal species, then it is possible to determine the source animal species based on
DNA characterization. Based on available host-specificity information, plus the ability of the SSU rRNA-based
nested PCR-RFLP method to detect and differentiate these species and genotypes, the following Cryptosporidium -
host animal species pairs can be determined from DNA characterization of the oocyst: C. parvum human genotype -
humans and non-human primates; C. wrairi - guinea pigs; C. felis - primarily cats; C. andersoni - primarily juvenile
and adult cattle; and, C. canis - primarily dogs.
For Cryptosporidium species and genotypes that are known to strictly or primarily infect a particular
animal class, then it is possible to determine the source animal class based on DNA characterization. Based on
available host-specificity information, plus the ability of the SSU rRNA-based nested PCR-RFLP method to detect
and differentiate these Cryptosporidium species and genotypes, the following crypto-host animal class pairs can be
determined from DNA characterization of the oocyst: C. baileyi - birds; C. meleagridis - birds; C. parvum bovine
genotype - mammals (including humans); C. saurophilum - reptiles (lizards); and C. serpentis - reptiles (snakes).
Exceptions are possible for C. meleagridis, which has been found in humans in a limited number of cases.
Although their host specificity is not yet well-documented, the following Cryptosporidium genotypes
have been detected, some for only a limited number of times, in the same hosts. If one accepts that these
genotypes are tentatively host-specific, then the following additional source animal species can be tentatively
identified by the SSU rRNA-based nested PCR-RFLP method or DNA sequencing of the oocysts: C. parvum
mouse genotype - mice; Cryptosporidium ferret genotype -ferrets; Cryptosporidium fox genotype - foxes; C. parvum
monkey genotype - monkeys; Cryptosporidium skunk genotype - skunks; Cryptosporidium opossum genotypes -
opossums; Cryptosporidium deer mouse genotype - deer mice; Cryptosporidium deer genotype - deer;
Cryptosporidium goose genotype - geese; Cryptosporidium bear genotype - bear; C. canis fox genotype - foxes;
C. canis coyote genotype - coyotes; C. parvum pig genotype - pigs; Cryptosporidium muskrat genotype - muskrats.
14
-------
The existence of Cryptosporidium species and genotypes that have limited host-specificity makes it
impossible, when they are present in a sample, to identify a unique source-animal species based on DNA
characterization alone. Cross-transmission studies have demonstrated that there are non-host specific
Cryptosporidium species and genotypes. Specifically, it has been demonstrated at the genotype level by the
C. parvum bovine genotype, which has been found in both infected humans and infected cattle.
From a public health standpoint, it would be very useful to be able to identify from the oocysts the
following four sources of mammalian oocysts: humans, cattle, wildlife, and pets. Based on data collected to
date, the current capability of making the distinctions between several key pairs of potential sources (human
vs. cattle, human vs. wildlife, human vs. pets, and cattle vs. wildlife) is as follows:
Human vs. Cattle
-D If C. parvum human genotype is found, then humans are the likely sources of these oocysts.
-D If C. parvum bovine genotype is found, then the source of these oocysts could be either humans,
cattle, sheep or perhaps some other mammals as well. Subgenotype analysis can be useful because
some bovine genotype isolates have only been found in humans.
-D If C. andersoni is found, this could indicate that the source is juvenile or adult cattle. C. andersoni
is very rarely found in humans.
D Humans vs. Wildlife
-D If the C. parvum human genotype is found, then only humans or non-human primates are known
sources of these oocysts.
-D If C. parvum bovine genotype is found, then the source could be either humans or cattle, or
perhaps other mammals, including wildlife.
- If the tentative wildlife host-specificity of several Cryptosporidium genotypes is accepted, then
oocysts can be tentatively linked to specific wildlife species (e.g., deer, ferret, fox, mouse,
opossum, raccoon, bear, muskrats, birds, reptiles).
- Numerous wildlife Cryptosporidium types have not yet been characterized.
Human vs. Pets (Cats or Dogs)
-D If the C. parvum human genotype is found, then only humans or non-human primates are the
known source of these oocysts.
-D If the C. parvum bovine genotype is found, the source of these oocysts could be humans or pets.
-D If C.felis and C. canis are confirmed host-specific to cats and dogs respectively, then it will be
possible to link the oocysts to these pets, although not directly determine if they are domesticated
or feral, which may be apparent from the watershed.
-D Since C.felis and C. canis have been found in clinical samples from small numbers of humans,
this potential source would need to be investigated as well.
D Cattle vs. Wildlife
-D If C. parvum bovine genotype is found, then the source could be either humans or cattle, or
perhaps other mammals as well.
-D If C. andersoni is found, this could indicate that the source is juvenile or adult cattle, but this may
change if C. andersoni is found to be human pathogenic.
- If several Cryptosporidium genotypes that have been tentatively identified as wildlife are host-
specific, then oocysts can be tentatively be linked to specific wildlife species (e..g, deer, raccoon,
opossum, ferret, fox).
- Numerous wildlife Cryptosporidium oocyst DNA types are not yet characterized. Further
characterization of wildlife samples are needed to pinpoint this source.
The feasibility of identifying source animal contributors of oocysts collected in water samples using
the SSU rRNA-based nested PCR-RFLP method was successfully demonstrated with multiple samples from
four different water matrices and combinations of source animals.
15
-------
This method was not expected to and does not provide the number of oocysts of each species and
genotype that is present in a water sample. Quantitative accuracy was not a capability expected or explored in the
project, although method sensitivity (i.e., the ability to detect the presence of as few as 1 oocyst/sample) is important
and was examined. Also, the total and species-specific recovery efficiency of the concentration and separation
methods has a bearing on conclusions that can be drawn about oocyst distributions that are found. As recovery,
separation, and concentration methods improve in speed, economy, and effectiveness, the feasibility and value of
molecular tools will also improve.
This method was not expected to and does not provide information about the viability of the oocysts,
nor does it provide routine on-line monitoring of Cryptosporidium oocyst contamination.
General Conclusions
The effective application of molecular methods to determination of source animal types from oocysts
in water samples is dependent on the existence and understanding of the host-specificity of the current and
growing number of Cryptosporidium genotypes in humans, cattle, pets, and wildlife.
D If more host-specific Cryptosporidium genotypes are identified, then it may be possible to identify
additional host animals and human pathogenicity by oocyst DNA characterization, as was the case when,
for examples: C. parvum was found to consist of a human-specific genotype (now a named species, C.
hominis) and a non-specific bovine genotype; and C. wrairi was found to be host-specific for guinea pigs.
D Conversely, it may be found that some of the current species and genotypes considered to be host-specific
are not, and therefore molecular techniques may have to be used in combination with conventional
analytical techniques to determine host animals.
D Even if a method was available to determine all Cryptosporidium species and genotypes, this would not
help determine the source (e.g., human vs cattle vs companion animals vs. wildlife) for Cryptosporidium
oocysts from non-host-specific species and genotypes.
Assuming that the number of host-specific genotypes increases significantly, there will probably need
to be a concurrent increase in the speed, specificity, and economy of methods capable of detecting and
differentiating the oocyst DNA to determine the source animal type. There are currently efforts underway to
address these needs (e.g., Limor et al., 2002; McDonald et al., 2002).
The discovery of host-specific genotypes and the availability of methods for species and genotype
detection and differentiation indicates a possible need to re-evaluate Cryptosporidium characterization
practices for feeding studies. For example, comparative evaluations of the infectivity and pathogenicity of the 5
human-pathogenic Cryptosporidium species and genotypes are needed.
At the present time the Cryptosporidium species and genotype detection and differentiation methods
and the supporting host specificity data described in this document and elsewhere are best suited to
generating, supporting, or refuting preliminary hypotheses or conclusions about the type of source animals
(i.e., humans, farm animals, pets, wild animals) in which the oocysts were produced and their human-
pathogenic potential. Preliminary hypotheses about source type or pathogenicity can be very important for
orienting investigation or response resources to the most likely contamination sources. Also, preliminary
conclusions based on other data about the potential source animals or pathogenicity can be checked against the
molecular evidence. Sufficient resources are required for representative sampling and analysis; collection and
sequencing of fecal samples may be required if Cryptosporidium genotype data are not available for animals known
to be in the watershed; and adequate time and lab capacity must be available from the limited number of laboratories
that perform the procedures and data analysis.
At the present time the Cryptosporidium species and genotype detection and differentiation methods
and the supporting host specificity data described in this document are not well-suited for the following
applications.
16
-------
D Conclusively proving the source animal type of reputed host-specific oocysts found in water. The difficulty
here stems from a lack of comprehensive Cryptosporidium host-specificity data and Cryptosporidium
genotype characterization data.
D Routine water monitoring - the time, cost, equipment, and experience required to successfully perform
these methods makes them inappropriate for routine water monitoring. A rough estimate of the present
cost/sample is $ 400 for materials (including filters and IMS and sequencing) plus labor and equipment.
D Quantitative characterization of the distribution of Cryptosporidium oocyst genotypes in water. Even
assuming that an assembly of samples are analyzed that are representative of the water body, the PCR-
RFLP process involves steps that are not at present readily quantifiable. For example, the PCR process
makes thousands to billions of copies of the target gene and the replication efficiency is not suitably defined
to enable accurate calculation of the initial number of DNA molecules. Efforts are underway to develop
real-time, quantitative PCR (Limor et al., 2002; McDonald et al., 2002)
It is considered unlikely at any time that Cryptosporidium species and genotype detection and
differentiation methods and supporting host specificity data as described in this document or elsewhere will
be suitable for:
D Identifying the particular animal species that produced non-host-specific Cryptosporidium genotypes
D Matching the oocysts in water to a specific individual animal.
Recommendations
While molecular detection and characterization of Cryptosporidium oocysts has made substantial progress
and shows considerable promise, there are some current and future issues that should be addressed to enable and
accelerate the use of molecular tools to their full potential for generating data that are useful in the risk assessment of
various waters in different environmental settings, and for watershed management and source water protection
(Xiao et al., 2002a).
Current problems in molecular detection of Cryptosporidium oocysts
D Only a limited number of tools for species differentiation, most of which are based on the small subunit
rRNA gene
D Nonspecificity of some species differentiation tools
D Misinterpretation of data because of outdated knowledge of the evolving research field
D Existence of erroneous data in the database and publications
D Lack of laboratory and field data on host specificity of Cryptosporidium species and genotypes.
Actions needed to enable routine use of molecular tools in water sample analysis
D Rigorous standardization and testing have yet to be carried out in order to develop quality assurance and
quality control procedures
D Development of protocols that allow the extraction of PCR-quality DNA without using the expensive and
pathogen-specific IMS
D Turnaround times have to be reduced to allow close to real-time detection for routine monitoring
D Quantitative and high resolution typing procedures (i.e., subgenotyping) need to be incorporated for
analysis of samples in special situations (such as outbreaks or bioterrorism)
D Utilization of new techniques such as real-time PCR, biosensors, and microarrays.
D Continuous molecular characterization of Cryptosporidium parasites from various wildlife to expand
current data on host-specifity
D More extensive use of SSU rRNA PCR-RFLP tools in the analysis of different environmental samples to
allow more confidence in the association of common genotypes with environmental settings.
17
-------
REFERENCES
Payer, Ronald, C.A. Speer, and J.P. Dubey. 1997. The General Biology of Cryptosporidium. In: Cryptosporidium
and Cryptosporidiosis. Ed., Ronald Payer. CRC Press. New York. pp. 1-41.
Payer, R., J.M. Trout, L. Xiao, U.M. Morgan, A.A. Lai, and J.P. Dubey. 2001. Cryptosporidium canis n.sp. from
domestic dogs. J. Parasitol. Dec., 87(6):1415-1422.
Payer, R. , T.K. Graczyk, E.J. Lewis, J.M. Trout, C.A. Farley. 1998. Survival of infectious Cryptosporidium parvum
oocysts in seawater and eastern oysters (Crassostrea virginica) in the Chesapeake Bay. Appl. Env. Microbiol. 64:
1070-1074.; http://aem.asm.Org/cgi/content/full/64/3/1070
Jenkins, M.C. and C. Petersen. Molecular Biology of Cryptosporidium . 1991. In: Cryptosporidium and
Cryptosporidiosis. Ed., Ronald Payer. CRC Press. New York. pp. 225-232
Limor, J.R., A.A. Lai, and L. Xiao. 2002. Detection and differentiation of Cryptosporidium parasites that are
pathogenic for humans by real-time PCR. J. Clin. Microbiol. 40(7):2335-8
Lindquist, H.D. J.W. Bennett, M. Ware, R.E. Stetler, M. Gauci, and F.W. Schaefer. 2001. Testing methods for
detection of Cryptosporidium spp. in water samples. Southeast Asian J Trop Med Public Health 32 (Suppl 2): 190-4.
McDonald, L.M., K. Sargent, A. Armson, R.C. Tompson, J.A. Reynoldson. 2002. The development of a real-time
quantitative PCR method for characterization of Cryptosporidium parvum in vitro culturing system and assessment
of drug efficiency. Molecular and Biochemical Parasite logy. (2):279-282.
*Sulaiman, I.M., L. Xiao, C. Yang, L. Escalante, A. Moore, C.B. Beard, M.J. Arrowood, and A.A. Lai. 1998.
Differentiating Human from Animal Isolates of Cryptosporidium parvum. Emerging Infectious Diseases, Vol 4, No.
4, Oct-Dec, 1998. pp. 681-685; http://www.cdc.gov/ncidod/eid/vol4no4/sulaiman.htm
*Sulaiman, I.M., A.A. Lai, M.J. Arrowood, and L. Xiao. 1999a. Biallelic Polymorphism in the Intron Region of P-
tubulin Gene of Cryptosporidium Parasites. Journal of Parasitology, Volume 85, Issue 1, 1999, pp. 154-157
*Sulaiman, I.M., L. Xiao, A.A. Lai. 1999b. Evaluation of Cryptosporidium parvum Genotyping Techniques. Journal
of Applied and Environmental Microbiology, Vol 65, Issue 10, pp. 4431-4435^
http://aem.asm.org/cgi/content/full/65/10/4431
Sulaiman, I.M., U.M. Morgan, R.C. Thompson, A.A. Lai, L. Xiao. 2000. Phylogenetic Relationships of
Cryptosporidium Parasites Based on the 70-Kilodalton Heat Shock Protein (HSP70) gene. Journal of Applied and
Environmental Microbiology, Vol 66, Issue 6, pp. 2385-91; http://aem.asm.Org/cgi/content/full/66/6/2385
Sulaiman, I.M., A.A. Lai, L. Xiao. 2002. Molecular Phylogeny and Evolutionary Relationships of Cryptosporidium
Parasites at the Actin Locus. J. Parasitology. Vol 88, pp. 388094
Working Group on Waterborne Cryptosporidiosis. 1 997. Cryptosporidium and Water: A Public Health Handbook.
Atlanta, Georgia; http://www.cdc.gov/ncidod/diseases/crvpto/crypto.pdf
*Xiao, L., I. Sulaiman, R. Payer, A.A. Lai. 1998. Species and Strain-Specific Typing of Cryptosporidium Parasites
in Clinical and Environmental Samples. In: Mem Inst Oswaldo Cruz, Rio De Janeiro, Vol 93(5), Sep/Oct 1998;
http://www.dbbm.fiocruz.br/www-mem/935/23m.pdf
17
-------
*Xiao, L., L. Escalante, C. Yang, I. Sulaiman, A.A. Escalante, R.J. Montali, R. Payer, and A.A. Lai. 1999a.
Phylogenetic Analysis of Cryptosporidium Parasites Based on the Small Sub unit Ribosomal RNA Gene Locus.
Journal of Applied and Environmental Microbiology, Volume 65, Issue 4, 1999, pp. 1578-1583;
http://aem.asm.Org/cgi/content/full/65/4/1578ffT2
Xiao, L., U.M. Morgan, J. Limor, A. Escalante, M. Arrowood, W. Shulaw, R.C.A. Thompson, R. Payer, and A.A.
Lai. 1999b. Genetic Diversity within Cryptosporidium parvum and Related Cryptosporidium species. Applied and
Environmental Microbiology. Vol 65, No. 8, pp. 3386-3391). http://aem.asm.Org/cgi/content/full/65/8/3386
*Xiao, L., K. Alderisio, J. Limor, M. Royer, and A.A. Lai. 2000. Identification of species and sources of
Cryptosporidium oocysts in storm waters with a small- subunit rRNA-based diagnostic and genotyping tool. Appl.
Environ. Microbiol. 66:5492-5498; http://aem.asm.org/cgi/content/full/66/12/5492
*Xiao, L. A. Singh, J. Limor, T. Graczyk, S. Gradus, and A. Lai. 2001. Molecular characterization of
Cryptosporidium Oocysts in Samples of Raw Surface Water and Wastewater. Appl. Env. Microbiol. 67:1097-1101;
http://aem.asm.Org/cgi/content/full/67/3/1097
*Xiao, L. and A.A. Lai. 2002a. Molecular detection of Cryptosporidium oocysts in water: the challenge and promise.
(Draft) WQTC Conference, 11/02.
Xiao, L., I.M. Sulaiman, U.M. Ryan, L. Zhou, E.R. Atwill, M.L. Tischler, X. Zhang, R. Payer, A.A. Lai. 2002b.
Host Adaptation and Host-Parasite Co-evolution in Cryptosporidium: Implications for Taxonomy and Public Health.
2002. International Journal for Parasite logy. In Press.
Work funded partially or completely under EPA/NRMRL-HHS/CDC interagency agreement
75937984
18
-------
APPENDIX 1 - Molecular Tools
DNA sequencing determines the exact order of the nucleotides in the DNA molecule. DNA sequencing is
important for a variety of reasons, including the identification or confirmation of particular genotypes. DNA
sequencing is commonly performed by variations on the Sanger method, which involves producing a sequence of
DNA fragments that are progressively longer by one, known nucleotide. By ordering the fragment lengths by
electrophoresis, which is sufficiently sensitive to separate DNA fragments differing in length by one nucleotide, and
then matching the corresponding terminal nucleotide, the DNA sequence is determined. Sequencing gives the most
detailed information, but it is a relatively slow and expensive method.
Restriction Fragment Length Polymorphism (RFLP) is a faster, less costly, and less detailed approach to
characterizing DNA. The RFLP method cuts a selected segment of DNA into fragments using restriction enzymes,
which only cut the DNA at locations straddled by specific 4- or 6-nucleotide sequences. Hence, the number and
length of the fragments generated in the RFLP process depend on (1) the selection of the restriction enzyme and its
particular target sequence, and (2) the existence, number, and location on the original DNA segment of the target
sequence for the restriction enzyme. The differing DNA fragment lengths are separated by electrophoresis and
visualized by staining procedures. An RFLP process can be designed to produce a known number and length of
fragments by first sequencing the target gene, then identifying the locations of the target sequences for the available
restriction enzymes, and finally, selecting the restriction enzyme that gives the best fragment combination for
detecting or differentiating particular genotypes.
DNA sequencing and RFLP procedures both require very large numbers of copies of the target gene
sequence. These copies are prepared by another critical molecular tool - the polymerase chain reaction (PCR).
The PCR procedure uses heating and cooling cycles, specific DNA primer sequences, and a supply of nucleotides to
mimic the DNA replication process and rapidly generate multiple copies of the desired DNA sequence. The number
of copies produced is 2 , where n is the number of cycles. Starting with one double-strand of DNA, 20 cycles of
100% efficient PCR will produce about 1 million copies, and 35 cycles will produce about 35 billion copies. In
addition to copying target gene sequences for use in other methods, PCR can also be used to detect the presence of
particular organisms. When used for detection purposes, DNA primers are selected that will bind to unique DNA
sequences of the target organism. If the target DNA is present, then replication of the DNA primer sequences and
the intervening DNA sequence will occur. If the target DNA is not present, then the primers will not attach and the
replication process will not begin. Nested PCR is a technique that uses two rounds of PCR on the same target
sequence to increase the sensitivity of the method by increasing the number of copies of the target DNA sequence
that can be produced, hi the first round of PCR the DNA primers are chosen so that the DNA segment copied
includes the target sequence as well as a substantial amount of DNA on both sides of the target sequence. In the
second round of PCR the primers are selected to copy only the target sequences on a portion of the copies of the
first-round PCR products.
Cloning is another approach for making copies of a portion of a genome. Cloning involves the cutting and
insertion of the target DNA sequence into a vector, which is then inserted into a bacterium where the target DNA
sequence is replicated along with the vector DNA and the bacterium. Following replication the target DNA is
separated from the bacteria and vector and the desired manipulations or analyses are performed. Cloning has the
advantage of not requiring prior knowledge of the cloned sequence, whereas PCR requires that an approximately 20-
nucleotide sequence be known for the primer sequence. The molecular tools for determining species and genotype
of Cryptosporidium oocysts are not quantitative methods, so they characterize the DNA sequences, but do not
provide accurate data on the initial concentration of each genotype present.
19
------- |