Authors

Discussion

Cryptosporidium parvum is a protozoan parasite that infects a wide range of animals and humans typically causing a diarrhoeal disease, which can lead to death especially in immunocompromised individuals.  The robust nature of Cryptosporidium oocysts enables the parasite to be transmitted through multiple routes from diverse hosts. Thus, in order to assess risks of infection, investigate outbreaks, and apply appropriate interventions, tracking sources of contamination and routes of transmission is of paramount importance. In the absence of a multilocus genotyping scheme, subtyping C. parvum isolates has been mainly restricted to the DNA sequence analysis of a gene encoding a 60 KDa glycoprotein (gp60). Although this marker provides some discrimination and inference of linkage to point source contamination, the recombinant nature of this parasite observed both experimentally and in nature demands a multilocus-based method for more accurate discrimination. Whole genome sequencing would provide the ultimate determination of variation, but for Cryptosporidium it is a time consuming and expensive approach to be implemented routinely in clinical laboratories and for inter-laboratory surveillance. It has been shown that multilocus genotyping based on genetic loci containing a variable number of tandem repeats, can enable rapid characterization of outbreak isolates and infer linkage. 

During an expert workshop on Cryptosporidium genotyping hosted by the Robert Koch Institute, Berlin, 2016, the criteria to develop a harmonised approach to intra-species differentiation in Cryptosporidium were established. Based on these criteria we interrogated the C. parvum Iowa II reference genome with Tandem Repeat Finder software in order to identify variable number tandem repeats loci (VNTR) suitable for fragment sizing by most fragment sizing platforms: VNTR loci containing repeats ≥ 6 bp in tandem and providing an amplicon size < 300 bp including 50bp flanking the repeat region at both sides for the location of primers. Twenty eight markers were identified initially (Tandem Repeats Finder, New York) and subsequently showed multiple alleles in seven of our own C. parvum genomes. From these, seven markers were selected for in vitro evaluation based on their higher variability. PCR primers were designed and fragment sizes estimated for 20 C. parvum samples using a QIAxcel Advanced (Qiagen) machine, run under optimised conditions, and compared with sequencing. Estimated fragment sizes were confirmed in most samples. When repeatability and reproducibility was investigated using triplicate PCR reactions of four samples on two occasions, discrepancies were observed in a small number of reactions but were resolved readily. However,  when applied to a larger validation panel of 268 C. parvum samples representing spatio-temporal variation, outbreaks and sporadic infections of humans and animals, 5% PCR reactions were not detectable, and a further 5% provided inconsistent and unresolvable ambiguous sizes especially in the 6 bp repeat markers; consequently, MLGs could not be assigned to 48 % samples. To improve accuracy and reliability, the validation panel is under re-evaluation using labelled primers for fragment sizing on a different platform (3500 Genetic Analyzer, Applied Biosystems^TM), using multiplexed PCRs for an economic approach. This should enable full-scale validation according to internationally accepted methods for evaluation of microbial typing schemes: typability, discriminatory power, epidemiological concordance and efficiency, and the results will be presented at the conference.