An official website of the United States government.

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

What is the Potential for Human Isolates of Both Genotypes of C. parvum to Infect, Colonise and be Excreted by Farm Animals

Veterinary Laboratories Agency, UK
Start date
End date
This project will seek to answer the following key questions:
  • Whether both C.parvum genotypes and their various sub-types can infect a range of farm animals including calves, lambs, pigs?
  • Whether farm animals can excrete a sufficient quantity of oocysts to act as a reservoir/source for the infection of humans or other farm animals?
  • What is the risk of transmission of infection of both genotypes of C.parvum from farm animals to humans, within a defined localosed area?
More information
Objective 01: Validation of molecular methods
A multiplex allele specific Polymerase chain reaction (MAS-PCR) based on the Cryptosporidium parvum dihydrofolate reductase (dhfr) gene sequence was developed that differentiates genotype 1 (recently renamed C. hominis) from genotype 2 in a one step reaction. The MAS-PCR was validated on a panel of 34 microscopically positive and 37 microscopy negative C.parvum faecal samples of human and animal origin in comparison with 2 published PCR restriction fragment length polymorphism (RFLP) methods targeting dhfr and oocyst wall protein (cowp) genes. MASPCR was found to be as sensitive for species detection as the most sensitive of the other tests and detected more mixed genotype infections than the other two tests combined. In addition the MASPCR showed equivalent detection sensitivity in comparison with a published nested RFLP targeting the 18srRNA (18srRNA) gene on a panel of prepared mixed genotype samples. The one step reaction is simpler and less expensive to perform than the RFLP methods, while the C.parvum specific amplicon and those for genotype 1 and 2 (575bp, 357bp and 190bp respectively) can be easily distinguished on agarose gel.

A collaborative “blind” study of molecular genotyping techniques was performed at VLA Weybridge, HPA Swansea and HPA Colindale. Two hundred human faecal samples from single and mixed genotype infections were assessed for genotype using published PCR techniques on 3 different gene targets, including; dhfr, 18srRNA and cowp. Samples were genotyped at HPA Swansea using a nested cowp PCR and sent to VLA Weybridge to genotype using dhfr and 18srRNA PCR, the extracted sample DNA was then sent to HPA Colindale for use on the nested cowp PCR. After each batch of testing the results were compared by VLA Weybridge and recorded on a database. Results indicate 99% agreement of the different gene targets, with the exception of mixed genotype isolates where the dhfr and SSU-RNA PCR detected more of the mixed genotype samples.

C. parvum positive samples from the transmission experiments and from the human/ animal “linked” cases were subtyped using two different techniques. VLA Weybridge used a “novel” fingerprinting system (TR7/8) and the HPA Colindale by microsatellite analysis. Early results at the VLA showed that a reproducible, unique “fingerprint” was produced from both genotype 1 and 2, which when analysed by statistical software showed genetic linkage in related isolates compared to non-related isolates. Problems occurred when the samples became more complex i.e. mixed genotype 1 and 2 isolates; mixed infections with other protozoan e.g. Giardia, Cyclospora. Statistical analysis showed that cryptosporidium species produced fingerprints of similar size and complexity to Giardia isolates and could not distinguish between mixed cryptosporidium isolates. This technique although showed early promise has been shown to be unsuitable for demonstrating genetic linkage between isolates of unknown origin e.g. outbreak tracing. The TR7/8 fingerprinting technique may still have a role in monitoring of isolate stability within transmission models. Other published techniques were tried but involved a high degree of expensive sequencing and analysis to be performed.

DNA extracted from human or animal faeces where C. parvum genotype 1 (C. hominis) or genotype 2 had been detected was analysed at three polymorphic cryptosporidium-microsatellite loci named ML-1 (Cacciò et al., 2000), GP15 (Mallon et al., 2003) and MS5 (Mallon et al., 2003). ML1 and GP15 loci are tri-nucleotide repeat (GAG and TCA), while MS5 has a repeated unit of 24 nucleotides (CTCCCTCAGCTCCTCCGACTGCA). For C.parvum genotype 1 there has been 2, 5 and 4 different polymorphisms for ML-1, GP15 and MS5 respectively. For C. parvum genotype 2 there have been 4, 13 and 6 different polymorphisms for ML-1, GP15 and MS5 respectively. In these experiments PCR primers flanking the microsatellite markers were used, one of which was labelled with a fluorescent dye. PCR products were then accurately sized on a CEQ 8000 Genetic Analysis System (Beckman Coulter) using 33-75B CEQ Separation capillary array by comparison with the CEQ DNA size Standard Kit – 600 (labelled with Beckman Dye D1PA). The microsatellite techniques employed by HPA Colindale was applied to subtype both C.parvum genotype 1 and 2 and alleles remain stable on testing epidemiologically related isolated from cases of human infection.

Objective 02 - Transmission experiments
Transmission experiments were based upon administering human origin C.parvum oocysts, of known quantity and genotype, to four calves, four pigs and four lambs. The animals were then monitored by microscopy and PCR for signs of infection, colonisation and excretion of oocysts in the faeces. The transmission studies have shown than C.parvum genotype 2 originating from human samples are capable of infecting, colonising and being excreted by lambs, calves and piglets. Previous studies at the VLA have shown a 100% infectivity rate, with all human genotype 2 isolates used infecting all lambs, calves and piglets. Each isolate showed a strong pathogenicity, with all animals exhibiting two or more signs of depression, anorexia, increased temperature and diarrhoea. Oocyst output in all animals was >5 million oocysts per gram of faeces for over 6 days within a 21 day period.

The transmission studies using human C.parvum genotype 1 have shown that a reservoir of C. parvum genotype 1 is capable of infecting, colonising and being excreted in lambs, calves and piglets. Interestingly, each genotype 1 isolate was not capable of infecting all the three different species (lambs, calves, piglets), with typically only one species infected. Different subtypes may have different animal host specificity. Subgenotyping techniques with higher levels of discrimination are currently being assessed in an International trial, the results of which will be available shortly. All isolates have been kept and will be used as historical samples to test and validate these higher discriminatory subtyping methods at the VLA Weybridge as they become available [this procedure has been costed in to the next Cryptosporidium ROAME project]. In addition, not all the animals within the species were infected typically one or two animals remained uninfected despite being dosed with the same isolate at the same time. This may suggest a level of resistance to C.parvum genotype 1 displayed in some animals. VLA Weybridge published the first report worldwide of a successful transmission into a lamb. This was then expanded upon at the Cryptosporidium Conference held in Perth Australia (October 01) and created high levels of interest. The microsatellite method detected stable combinations of alleles during passage through experimentally infected animals, and that some of the same allele combinations were detected in material from unrelated human cryptosporidiosis infections.

Objective 3 - Epidemiological study in SW England
Within a cryptosporidiosis study in South West England, patients with the disease were asked about contact with farms and farm animals within the two weeks before illness. When farm contact was identified, recently voided faeces were randomly sampled from a cross-section of the animals on that farm and examined for the presence of Cryptosporidium by microscopy. All positives and a proportion of negative samples were then genotyped using MASPCR and 18srRNA PCR and subtyped using TR7/8 and microsatellite analysis. Faeces from the human isolates were genotyped at HPA Swansea and Colindale by nested cowp RFLP and at VLA Weybridge by MASPCR and 18srRNA PCR. Animal faeces were sent directly to VLA Weybridge and screened by microscopy and PCR. It must be remembered that the “linked” animal samples are not necessarily genetically linked but the odds are increased.

The genotype and fingerprint profiles of human cases with the linked animal samples, and the animal transmission samples has shown that all isolates produced individual fingerprint profiles but that related isolates have a higher degree of homology. However, the problems stated in objective 1 that occur when the samples become more complex also confused some of the results obtained.

Of the 12 cases identified with a possible linkage, 5 cases (A, B, C, D and E) were identified where the genotype was obtained and there was sufficient material to perform “clean up” procedures to be able to provide a clear fingerprint from both the human case and the “linked” animal samples. Of those five, two human cases were genotype 1 and three were genotype 2. An additional mixed genotype human sample also showed a clear fingerprint. When analysed by the Gel ComparII software packages, using Ward and Dice statistical analysis, the human fingerprints and their related “linked” animal fingerprints showed a theoretical linkage, with two or more of the animal samples, in 3 of the 5 cases.

Objective 4 - Risk analysis
The risk assessment developed has addressed the question of whether the daily risk of human infection from both genotypes of C.parvum following contact with farm animals is significant in comparison with the daily risk from other known routes (person-to-person, and waterborne). The assessment indicates that for both adults and children, the daily risk of human infection with C. parvum (genotype 2) from farm animals is slightly smaller than that from person-to-person contact, but larger than that associated with drinking tap water. However, the model developed is preliminary, and the results should be treated as initial estimates of the daily risk associated with the transmission of C. parvum from each of the known routes.

Funding Source
Dept. for Environment, Food and Rural Affairs
Project number
Bacterial Pathogens
Sanitation and Quality Standards