Characteristics of a Novel Type of Bovine Cryptosporidium andersoni

Abstract

We isolated oocysts that resemble Cryptosporidium andersoni from cattle grazing on a farm in Japan. The partial sequences of genes from the isolate were coincident with published sequences of genes of C. andersoni. Since the isolate was able to infect SCID mice, the isolate appears to be a novel type of C. andersoni.

Before the proposal that Cryptosporidium andersoni might be involved, “large-type oocysts” from cattle were reported as Cryptosporidium muris or C. muris-like (1, 2). It is still unclear whether all large-type oocysts from cattle are C. andersoni and whether subpopulations of C. andersoni exist (6, 16).

In this study, we characterized Cryptosporidium oocysts isolated from cattle by morphological, biological, and genetic analysis.

Fecal samples were collected from grazing cattle on a farm in Miyagi Prefecture in the northern part of the main island of Japan. As reference strains, we used C. muris RN66, which was originally isolated from a house rat (7) and Cryptosporidium parvum HNJ-1, which was isolated from a Japanese woman (12). These strains had been passaged in SCID mice in our laboratory. Oocysts used in this study were purified by the sugar centrifugal flotation method. DNA from each sample was extracted with MagExtractor-Genome (TOYOBO, Osaka, Japan) after five rounds of freezing and thawing of oocysts. Three sets of primers were used to amplify fragments of genes, namely, 18S ribosomal DNA (rDNA) (5′-AACTTTACGGATCGCATCTCTGA-3′ and 5′-CCCATCACGATGCATACTCATAA-3′), the gene for heat shock protein 70 (HSP70; 5′-ACACATTGGCCATTCAAGGTT-3′ and 5′-GCTGGTGGAATACCATCTAAA-3′), and the gene for Cryptosporidium oocyst wall protein (COWP; 5′-TGATAAAATCTTGTCCTCCAGGT-3′ and 5′-GTTCGTTTGAACACATTGCAAC-3′). We designed these sets of primers using OLIGO 5.0 (National BioScience Inc., Plymouth, Minn.), and they were based on sequences of a bovine strain of C. muris and a hyrax strain of C. muris (GenBank accession numbers: AF093496, AF221542, and AF161579).

Subcloned product amplification by PCR of 18S rDNA and of genes for HSP70 and COWP were sequenced on an automated sequencer (ABI 310; Applied Biosystems Japan Ltd., Tokyo, Japan). The sequence accuracy of data was confirmed by two-directional sequencing. By using our sequence data and data from Cryptosporidium species in GenBank, we performed distance-based analysis using Kimura's distance formula (9) and then we constructed a phylogenetic tree using MEGA (version 2.1) (10).

To assess the infectivity of the isolate, we inoculated 10⁶ purified oocysts orally into eight 4-week-old SCID mice. The feces of each mouse were collected, and the discharge of oocysts was monitored for 28 days.

During the first survey of fecal samples from cattle after 23 days of grazing, we detected ovoid Cryptosporidium oocysts in feces from 6 of 113 cattle. Two of the six cattle died subsequently for unknown reasons. The four other cattle discharged oocysts for the entire grazing period (143 days). Long patent periods may be a characteristic of bovine Cryptosporidium, as reported previously (13, 14). We also detected oocysts in feces from 2 of 28 calves that were born during the grazing. The average dimensions of these oocysts ranged from 7.4 by 5.1 to 7.6 by 5.9 μm, whereas those of C. muris RN66 and C. parvum HNJ-1 are 8.1 by 5.1 μm and 4.8 by 4.2 μm, respectively. Postmortem examination of a 3-year-old cow revealed Cryptosporidium at various stages of development on the epithelial cells of the abomasum.

Amplification by PCR of 18S rDNA of C. muris RN66 and the isolate yielded products that were 1,253 and 1,255 bp long, respectively, and C. muris RN66 and the isolate yielded products of amplification of the genes for HSP70 (1,145 bp) and COWP (448 bp) of the same respective lengths. There were no differences in the respective nucleotide sequences of these three genes among isolates obtained from eight cattle. The nucleotide sequence of 18S rDNA from the isolate was identical to that from a bovine strain of C. muris (GenBank accession no. AF093496 [19]). Although only a sequence of 265 bp that was within the region we sequenced was registered for C. andersoni in GenBank (accession no. AJ275963 [16]), this sequence was homologous to that obtained from the new isolate. Partial sequences of genes for HSP70 and COWP from the isolate were coincident with published data for bovine isolates of C. muris (AF221542 [17]) and C. andersoni (AF266262 [20]).

For phylogenetic analysis, we constructed trees with the unweighted pair group method with arithmetic mean from aligned sequences of 18S rDNA and sequences of genes for HSP70 from various isolates of Cryptosporidium. In the case of 18S rDNA, two distinct clusters were formed in the genus Cryptosporidium: the first cluster consisted of two genotypes of C. muris strains plus Cryptosporidium serpentis; and the second cluster consisted of isolates of Cryptosporidium baileyi, Cryptosporidium felis, Cryptosporidium meleagridis, Cryptosporidium wrairi, and eight different genotypes of C. parvum. In the case of the gene for HSP70, two distinct clusters were also formed: the first consisted of two genotypes of C. muris strains and C. serpentis and C. baileyi; and the second cluster consisted of C. felis, C. meleagridis, C. wrairi, and six different genotypes of C. parvum.

All of eight SCID mice discharged oocysts after oral inoculation of the isolate. The prepatent period for the isolate was 14 days in all mice, while that of C. muris RN66 was 6 days. We confirmed that discharged oocysts were identical to inoculated oocysts by DNA sequence analysis using the primers described by Awad-el-Kariem et al. (3) (Table 1).

TABLE 1.

Transmission study of inoculation with 10⁶ oocysts of C. muris strain RN66 and the C. andersoni strain of Kawatabi^a

Mouse no.	Inoculated oocyst	Prepatent period (days)	Similarity of 18S rDNA sequences between isolated and discharged oocysts (%)
1	C. muris RN66	6	100
2	C. muris RN66	6	100
3	C. muris RN66	6	100
4	C. muris RN66	6	100
1	Kawatabi strain	14	100
2	Kawatabi strain	14	100
3	Kawatabi strain	14	100
4	Kawatabi strain	14	100
5	Kawatabi strain	14	100
6	Kawatabi strain	14	100
7	Kawatabi strain	14	100
8	Kawatabi strain	14	100

^aValues of sequence similarity were calculated by alignment of sequences of inoculated and discharged oocysts of each strain.

In a previous paper (8), we proposed that the Cryptosporidium oocysts that we had isolated from cattle were C. muris because they were infective to laboratory mice and rats. However, the earlier oocysts might belong to the same strain as the present isolate or to a similar strain, because the sizes of oocysts and the nucleotide sequence of the 18S rDNA of these earlier oocysts were the same as those of the isolate (unpublished data). The present isolate was infective to SCID mice, but it has been reported that C. andersoni did not infect SCID mice. Thus, the present isolate differed from published strains of C. andersoni and might be a novel type of C. andersoni. We shall refer to this strain as the C. andersoni strain of Kawatabi. The similarity in terms of the nucleotide sequences of C. andersoni Kawatabi and of other strains of C. muris (GenBank accession nos. AF093497 and AF093498) was 98.81%. The sequence similarity between the C. parvum isolate from deer (GenBank accession no. AF093494) and the C. parvum isolate from marsupial (GenBank accession no. AF108860) was 98.89%. Nevertheless, the value for C. parvum (marsupial strain) and C. wrairi (GenBank accession no. AF115378), which are considered to be distinct species, was 98.93%. Thus, we could not conclude from the sequence similarity alone that the Kawatabi strain was distinct from C. muris. Further study may be needed for strict discrimination between C. andersoni and the Kawatabi strain, with analysis of other genes that have been examined in previous studies of genotypes of C. parvum (4, 5, 15, 18).