Tracking host sources of Cryptosporidium spp. in raw water for improved health risk assessment

Abstract

Recent molecular evidence suggests that different species and/or genotypes of Cryptosporidium display strong host specificity, altering our perceptions regarding the zoonotic potential of this parasite. Molecular forensic profiling of the small-subunit rRNA gene from oocysts enumerated on microscope slides by U.S. Environmental Protection Agency method 1623 was used to identify the range and prevalence of Cryptosporidium species and genotypes in the South Nation watershed in Ontario, Canada. Fourteen sites within the watershed were monitored weekly for 10 weeks to assess the occurrence, molecular composition, and host sources of Cryptosporidium parasites impacting water within the region. Cryptosporidium andersoni, Cryptosporidium muskrat genotype II, Cryptosporidium cervine genotype, C. baileyi, C. parvum, Cryptosporidium muskrat genotype I, the Cryptosporidium fox genotype, genotype W1, and genotype W12 were detected in the watershed. The molecular composition of the Cryptosporidium parasites, supported by general land use analysis, indicated that mature cattle were likely the main source of contamination of the watershed. Deer, muskrats, voles, birds, and other wildlife species, in addition to sewage (human or agricultural) may also potentially impact water quality within the study area. Source water protection studies that use land use analysis with molecular genotyping of Cryptosporidium parasites may provide a more robust source-tracking tool to characterize fecal impacts in a watershed. Moreover, the information is vital for assessing environmental and human health risks posed by water contaminated with zoonotic and/or anthroponotic forms of Cryptosporidium.

FIG. 1.

Geographic location of the South Nation watershed and distribution of sampling locations within the watershed. Water flow occurs in a northerly direction, toward MST-2 (the most downstream site).

FIG. 2.

Molecular forensic profiling of Cryptosporidium parasites by repetitive nested PCR-RFLP analysis with microscope slides containing three water samples processed by method 1623. The top panel shows repetitive nested PCR results for three water samples in which five oocysts (lanes 1 to 5, MST-15), seven oocysts (lanes 6 to 10, MST-7), and eight oocysts (lanes 11 to 15, MST-11) were enumerated from the microscope slide. The lower panels show results of RFLP analysis with restriction enzymes SspI, VspI, and DdeI for positive PCRs. RFLP patterns in lanes 1 and 4 for site MST-15 represent two variants of Cryptosporidium muskrat genotype I verified by sequence analysis. Lane 2 represents a mixture of genotypes with the predominant genotype in the reaction product suspected to be the W1 genotype on the basis of RFLP comparison. RFLP patterns for the five positive reaction mixtures of repetitive PCR for sample MST-7 (lanes 6 to 10) indicate that all of the reaction products are composed of C. andersoni. DNA sequence analysis verified two sequences (differing by a single T addition; see Fig. 5) of C. andersoni in this water sample. RFLP results for the sample from MST-11 indicate two genotypes in this sample; lanes 11 and 12 show the single C. andersoni species, while lane 15 was a variant of the W12 genotype (determined by DNA sequence analysis). Lane 13 shows a mixture of these two genotypes.

FIG. 3.

Differentiation of Cryptosporidium species and genotypes by RFLP analysis. Nested PCR products are displayed in the top panel. Secondary PCR products were digested with restriction enzymes SspI, VspI, and DdeI, and the results of the digestions are shown (bottom three panels). Lanes 1 and 2 represent two isolates with 100% identity match to the previously described cervine genotype (accession no. AY262328) and the W4 genotype (accession no. AY737592). Lanes 3 and 4 represent two genetically different isolates which were not an exact match to sequences in GenBank but clustered as variants of Cryptosporidium muskrat genotype I (see Fig. 4, isolates 186 and 190). Lanes 5 and 6 represent the two sequences of C. andersoni. The sequence from lane 8 was a 100% identity match to W12 (accession no. AY007254), while lane 7 showed variation with respect to the sequence under accession no. AY007254 but clustered with W12 by phylogenetic analysis (see Fig. 4 and 5, isolate 168). Lanes 9 (isolate 148) and 10 (isolate 147) display the RFLP profiles of two sequences with no GenBank match. Phylogenetic analysis placed these and two GenBank sequences into a cluster with the Cryptosporidium fox genotype (accession no. AY120907) (Fig. 4). The sample displayed in lane 11 was a 100% identity match to the sequence with accession no. AY737567 (muskrat genotype II), while the samples in lanes 12 and 13 matched the sequence with accession no. AY545548 (muskrat genotype II) and lane 14 matched C. baileyi (accession no. AF093495). The DNA sequence from the sample in lane 15 was a 100% identity match to the W1 genotype (accession no. AF262330) and displayed minor RFLP variation with respect to that in lane 16, which was a 100% identity match to the C. parvum KSU-1 isolate (accession no. AF308600). Lane 17 is the laboratory C. parvum control.

FIG. 4.

Phylogenetic relationships among Cryptosporidium isolates from the South Nation watershed and known sequences in GenBank. Distances were calculated with the Kimura two-parameter model, and the phylogenetic tree was inferred by neighbor-joining analysis. The outgroup (E. tenella) was used to root the tree. Numbers on the branches are percent bootstrap values (>75%) obtained by using 1,000 resamples. GenBank accession numbers are in brackets, and boxes with dotted borders indicate the branches where isolates from the South Nation watershed clustered.

FIG. 5.

Sequence variation for isolates from the South Nation watershed that do not have 100% identity matches to GenBank sequences. Dots indicate nucleotides identical to the first sequence. Dashes were inserted to accommodate insertions or deletions in the aligned sequences. Nucleotide positions are in relation to the specified GenBank sequence.