Genotypes of Cryptosporidium Species Infecting Fur-Bearing Mammals Differ from Those of Species Infecting Humans

Abstract

Of 471 specimens examined from foxes, raccoons, muskrats, otters, and beavers living in wetlands adjacent to the Chesapeake Bay, 36 were positive for five types of Cryptosporidium, including the C. canis dog and fox genotypes, Cryptosporidium muskrat genotypes I and II, and Cryptosporidium skunk genotype. Thus, fur-bearing mammals in watersheds excreted host-adapted Cryptosporidium oocysts that are not known to be of significant public health importance.

The enteric parasites in the genus Cryptosporidium can be transmitted through ingestion of contaminated water (5). However, the sources of contamination are not clearly identified. Cryptosporidium spp. have been reported to infect a wide range of wild mammals (15). Among them, wild rodents have received particular attention, and it has been suggested that they may serve as reservoirs of Cryptosporidium infection for domestic animals and humans (1-6, 13-15, 18). Since the 1993 cryptosporidiosis outbreak in Milwaukee, Wis., the water industry in the United States has been striving to provide Cryptosporidium-free drinking water through stringent treatment practices and source water protection. Although there has been speculation that wild mammals serve as potential sources of watershed contamination with Cryptosporidium oocysts infectious for humans (1-3, 7), the actual role of wildlife in the contamination of source water with human-pathogenic Cryptosporidium spp. remains unknown.

Results of recent studies indicate a strong host adaptation for Cryptosporidium (22). Eight species of Cryptosporidium have been identified as pathogens in humans: C. parvum, C. hominis, C. meleagridis, C. felis, C. canis, C. muris, and Cryptosporidium pig and cervine species (8-12, 20, 21). Of these, C. hominis, C. parvum, and C. meleagridis have been found most frequently, whereas the others have been identified mostly in clinical case reports involving a few persons. Thus, unless wild mammals can be shown to be a source of these three species, they do not represent a significant risk as a source of water contamination affecting humans under normal circumstances. The present study was conducted to determine if Cryptosporidium infections are present in wild mammals (beavers, muskrats, otters, raccoons, and foxes) living in Chesapeake Bay watersheds and, if so, to determine the prevalence and species of Cryptosporidium by molecular methods. Results of the study provide the first genetically based data on the role of wildlife in Cryptosporidium contamination in watersheds.

Wildlife fecal specimen collection and genomic DNA extraction.

A total of 471 fecal specimens were collected during January 2001 and January 2002 from 87 beavers, 237 muskrats, 20 otters, 51 raccoons, and 76 foxes trapped in the Caroline (Marshy Hope Creek, Federalsburg), Charles (Clifton Creek, Newburg), Dorchester (Hunting Creek, Hurlock), and Talbot (Choptank River, Easton) counties of Maryland. With few exceptions, most of the trapped animals were older than 12 months. Details of the animal sources were described previously in studies of Enterocytozoon spp. and Giardia spp. in these animals (16, 17). After removal of fecal debris from 15 g of feces from each animal by sieving followed by density gradient centrifugation over cesium chloride, 200 μl of purified parasite suspension was washed twice with distilled water by centrifugation at 12,000 × g for 15 min as previously described (16). The oocyst walls were lysed with 1 M KOH, and genomic DNA was extracted as previously described with a QIAamp DNA stool mini kit (QIAGEN Inc., Valencia, Calif.) (22).

SSU rRNA PCR-RFLP analysis.

The species and genotypic nature of Cryptosporidium in each fecal specimen was determined by a previously described PCR-restriction fragment length polymorphism (RFLP) method based on the small-subunit (SSU) rRNA gene (20, 21). DNA from each specimen was analyzed three times by a nested PCR using 0.5, 1.0, or 2.0 μl of DNA as templates. The secondary PCR products (10 μl) of about 830 bp were digested with SspI (New England BioLabs, Beverly, Mass.) and VspI (GIBCO BRL, Grand Island, N.Y.). Cryptosporidium species and genotypes were determined on the basis of banding patterns in agarose electrophoresis. Positive (DNA of C. serpentis) and negative (no templates) controls were used in each PCR run.

DNA sequence and phylogenetic analyses.

To confirm the RFLP results, all secondary PCR products were sequenced with an ABI 3100 genetic analyzer (Applied Biosystems, Foster City, Calif.). The nucleotide sequences obtained were aligned with each other and those from the GenBank database with the program ClustalX (ftp://ftp-igbmc.u-strasbg.fr/pub/ClustalX/). A neighbor-joining phylogenetic tree was generated with TREECON version 1.3b (http://www.psb.rug.ac.be/bioinformatics/psb/Userman/treeconw.html), based on evolutionary distances calculated with the Kimura two-parameter model. The confidence of grouping was accessed by bootstrapping, using 1,000 replicates.

Prevalence of Cryptosporidium spp. in wildlife.

Fecal specimens from 471 wild mammals were analyzed for Cryptosporidium spp. by PCR. Cryptosporidium spp. were detected in 36 animals (8%), including 28 of 237 muskrats (12%), 6 of 76 foxes (8%), 2 of 51 raccoons (4%), none of 87 beavers, and none of 20 otters (Table 1). Infected animals, of which 17 were females and 19 were males, ranged from 12 to 36 months.

TABLE 1.

Cryptosporidium species and genotypes in wild mammals in watersheds in the Chesapeake Bay area

Animal	No. of SSU rRNA-positive specimens/total no. of specimens (%)	RFLP pattern (no. of specimens)	Species and genotype(s) (no. of positive specimens)
Beaver	0/87
Fox	6/76 (7.9)	C. canis (5) muskrat genotype (1)	C. canis fox genotype (4), C. canis dog genotype (1), Cryptosporidium muskrat genotype I (1)
Muskrat	28/237 (11.8)	Cryptosporidium muskrat genotype (22), new Cryptosporidium genotype (4), muskrat genotype + new genotype (2)	Cryptosporidium muskrat genotype I (22), Cryptosporidium muskrat genotype II (4), m uskrat genotype I + muskrat genotype II (2)
Otter	0/20 (0)
Raccoon	2/51 (3.9)	Cryptosporidium skunk genotype (2)	Cryptosporidium skunk genotype (2)
Total	36/471 (7.6)	4 RFLP patterns	5 Cryptosporidium spp.

Cryptosporidium genotypes in wildlife.

RFLP analyses of the PCR products with restriction enzymes SspI and VspI showed four banding patterns among positive specimens (Fig. 1). Most of the muskrat specimens had RFLP banding patterns identical to those of the previously described Cryptosporidium muskrat genotype, with two bands visible after both SspI and VspI digestion (lanes 1 and 2 in Fig. 1). Some muskrats had a PCR product very similar to that of the muskrat genotype, but the sizes of the two SspI bands were smaller (lanes 3 and 4 in Fig. 1). Specimens from two muskrats (specimens 3579 and 5538), however, had both RFLP banding patterns. In contrast, the two positive specimens from raccoons had an RFLP pattern identical to that of the previously described Cryptosporidium skunk genotype (lanes 5 and 6 in Fig. 1), whereas most of the positive fox specimens had the RFLP pattern for C. canis. One fox specimen had the RFLP pattern of the Cryptosporidium muskrat genotype.

FIG. 1.

Presence of four Cryptosporidium genotypes in fur-bearing wild mammals in the Chesapeake Bay area as indicated by RFLP analyses of PCR products of the SSU rRNA gene with restriction enzymes SspI (the upper lanes) and VspI (the lower lanes). Lanes 1 and 2, muskrat genotype I (specimens 5526 and 5559); lanes 3 and 4, muskrat genotype II (specimens 3568 and 3665); lanes 5 and 6, skunk genotype (specimens 6006 and 6001); lanes 7 and 8, C. canis (specimens 6008 and 5977); lane 9, blank; and lane 10, positive control (C. serpentis). Molecular size markers are 100-bp ladders.

DNA sequencing analyses of the PCR products showed the presence of five Cryptosporidium spp., four of which have been reported before (22). The 25 specimens with the RFLP pattern of the Cryptosporidium muskrat genotype generated SSU rRNA sequences typical of the genotype; 24 were from muskrats and one was from a fox. The SSU rRNA sequences from four fox specimens were identical to the C. canis fox genotype, but the sequence of one fox specimen was identical to that of the C. canis dog genotype. The SSU rRNA sequences from the two raccoons were of the Cryptosporidium skunk genotype. As expected, SSU rRNA sequences from six muskrat specimens belonged to a new Cryptosporidium genotype. To differentiate the two genotypes in muskrats, the previously named Cryptosporidium muskrat genotype was renamed muskrat genotype I and the new genotype was named genotype II. DNA sequence analysis confirmed that two muskrat specimens (3579 and 5538) had both Cryptosporidium muskrat genotypes I and II.

Genetic uniqueness of Cryptosporidium muskrat genotype II.

New Cryptosporidium muskrat genotype II showed significant differences from the SSU rRNA sequences from other known Cryptosporidium spp. or genotypes. The difference in the nucleotide sequences of muskrat genotype II and muskrat genotype I was about 5%. A similar level of difference was found between muskrat genotype II and the two C. canis genotypes. The genetic distance between muskrat genotype II and the skunk genotype was 1.4%. This distance was still significantly greater than the differences between some other established Cryptosporidium species, such as between C. muris and C. andersoni and between C. parvum and C. wrairi (22). The genetic uniqueness of the Cryptosporidium muskrat genotype II was also reflected in the phylogenetic analysis; all isolates of this genotype formed a distinct cluster. Again, muskrat genotype II was more closely related to the skunk genotype than to other genotypes found in this study (Fig. 2).

FIG. 2.

Relationship of five wildlife Cryptosporidium spp. inferred by a neighbor-joining analysis of SSU rRNA sequences. The Kimura two-parameter model was used in distance calculation. Numbers on branches are percent bootstrap values from 1,000 resamplings.

Public health significance of Cryptosporidium spp. in wild mammals.

Results of this study support previous findings on the host-specific nature of Cryptosporidium spp. (22). In general, each of the four Cryptosporidium-positive wildlife species in this study had its own host-specific Cryptosporidium genotype. Muskrats were infected only with Cryptosporidium muskrat genotypes I and II, and foxes were infected mostly with the C. canis fox genotype. Both positive raccoons had the Cryptosporidium skunk genotype, which was previously described to occur only in a few skunks (22). The only exception to host specificity of the parasites in this study is the finding of muskrat genotype I in one fox. It is not known, however, whether this rare finding represents established infection or merely passage of ingested Cryptosporidium oocysts, as coprophagia is common in foxes.

From feces of 471 animals examined in this study, the C. canis dog genotype was the only known human pathogen found, and this was found in less than 3% of all the Cryptosporidium-positive specimens. Findings from this study have clearly demonstrated for the first time that nearly all Cryptosporidium oocysts from a large number of diverse species of fur-bearing wild mammals are host-adapted Cryptosporidium species and genotypes that have never been found in humans or farm animals. These findings are in agreement with a previous report in which only wildlife Cryptosporidium genotypes were identified in runoff (storm water) from rural nonagricultural areas (19). The most common Cryptosporidium parasite found in fur-bearing mammals in this study, muskrat genotype I, is also one of the most common genotypes (W7) found in storm water in the previous study (19). Therefore, adult, wild, fur-bearing animals in wetlands can be a source of Cryptosporidium oocysts in watersheds, but this contamination poses little risk to public health or to livestock.

Nucleotide sequence accession numbers.

The nucleotide sequences of the SSU rRNA gene of the Cryptosporidium muskrat genotype II have been deposited in GenBank under accession numbers AY545546 to AY545548.