Development of a framework for genotyping bovine-derived Cryptosporidium parvum, using a multilocus fragment typing tool.

BACKGROUND: There is a need for an integrated genotyping approach for C. parvum; no sufficiently discriminatory scheme to date has been fully validated or widely adopted by veterinary or public health researchers. Multilocus fragment typing (MLFT) can provide good differentiation and is relatively quick and cheap to perform. A MLFT tool was assessed in terms of its typeability, specificity, precision (repeatability and reproducibility), accuracy and ability to genotypically discriminate bovine-derived Cryptosporidium parvum. METHODS: With the aim of working towards a consensus, six markers were selected for inclusion based on their successful application in previous studies: MM5, MM18, MM19, TP14, MS1 and MS9. Alleles were assigned according to the fragment sizes of repeat regions amplified, as determined by capillary electrophoresis. In addition, a region of the GP60 gene was amplified and sequenced to determine gp60 subtype and this was added to the allelic profiles of the 6 markers to determine the multilocus genotype (MLG). The MLFT tool was applied to 140 C. parvum samples collected in two cross-sectional studies of UK calves, conducted in Cheshire in 2004 (principally dairy animals) and Aberdeenshire/Caithness in 2011 (beef animals). RESULTS: Typeability was 84 %. The primers did not amplify tested non-parvum species frequently detected in cattle. In terms of repeatability, within- and between-run fragment sizes showed little variability. Between laboratories, fragment sizes differed but allele calling was reproducible. The MLFT had good discriminatory ability (Simpson's Index of Diversity, SID, was 0.92), compared to gp60 sequencing alone (SID 0.44). Some markers were more informative than others, with MS1 and MS9 proving monoallelic in tested samples. CONCLUSIONS: Further inter-laboratory trials are now warranted with the inclusion of human-derived C. parvum samples, allowing progress towards an integrated, standardised typing scheme to enable source attribution and to determine the role of livestock in future outbreaks of human C. parvum.

Multilocus sequence typing of a global collection of Pasteurella multocida isolates from cattle and other host species demonstrates niche association.

BACKGROUND: Pasteurella multocida causes disease in many host species throughout the world. In bovids, it contributes to bovine respiratory disease (BRD) and causes haemorrhagic septicaemia (HS). Previous studies have suggested that BRD-associated P. multocida isolates are of limited diversity. A multilocus sequence typing (MLST) scheme for P. multocida was used to determine whether the low levels of diversity reported are due to the limited discriminatory power of the typing method used, restricted sample selection or true niche association. Bovine respiratory isolates of P. multocida (n = 133) from the UK, the USA and France, collected between 1984 and 2008 from both healthy and clinically affected animals, were typed using MLST. Isolates of P. multocida from cases of HS, isolates from other host species and data from the MLST database were used as comparison. RESULTS: Bovine respiratory isolates were found to be clonal (I(S)(A) 0.45) with 105/128 belonging to clonal complex 13 (CC13). HS isolates were not related to bovine respiratory isolates. Of the host species studied, the majority had their own unique sequence types (STs), with few STs being shared across host species, although there was some cross over between porcine and bovine respiratory isolates. Avian, ovine and porcine isolates showed greater levels of diversity compared to cattle respiratory isolates, despite more limited geographic origins. CONCLUSIONS: The homogeneity of STs of bovine respiratory P. multocida observed, and the differences between these and P. multocida subpopulations from bovine non-respiratory isolates and non-bovine hosts may indicate niche association.