Fecal samples were collected from diarrheal patients from a hospital in Nairobi, Kenya, as part of a larger study. The patient described was an HIV-infected man who had clinical tuberculosis and diarrhea. Isospora belli was also detected in a fecal sample from the patient.

The fecal samples were preserved in both sodium acetate formalin and 2.5% potassium dichromate and kept at 4°C. They were stained with Kinyoun’s carbol fuchsin modified acid-fast stain and examined by oil immersion microscopy . An aliquot of 400 μL of the sample suspension in 2.5% potassium dichromate was processed for genotypic analysis. Potassium dichromate was washed 5 times with cold, distilled water until the yellow color cleared. Oocysts were ruptured by freeze-thaw, and DNA was extracted by using a QIAamp DNA Mini Kit (Qiagen, West Sussex, UK)) for stool DNA purification as per protocol.

A section of the SSU (18S) rRNA gene was amplified by nested polymerase chain reaction (PCR) as described (14), using the forward primers 5´-TTCTAGAGCTAATACATGCG-3´ and the reverse primer 5´-CCCTAATCCTTCGAAACAGGA-3´ for primary PCR. Secondary PCR used primers 5´-GGAAGGGTTGTATTTATTAGATAAAG-3´ and reverse primer 5´-AAGGAGTAAGGAACAACCTCCA-3´, employing the Techne (FTGENE2D Techne, Cambridge Ltd., UK) thermal cycler. Restriction fragment analysis of the secondary PCR product was done by digesting 15 μL of product in a 40-_L total reaction volume consisting of 15 U of Ssp1 and 3 μL of restriction buffer (Boehringer Mannheim Biochemicals, Indianapolis, IN) for species identification and Asn1 (Boehringer Mannheim) for genotyping in the same concentration at 37°C for 1 hour. Digestion products were separated on 2% agarose gel and visualized by ethidium bromide staining. The internal (secondary) fragment was purified by using the Prep-A-Gene DNA purification kit and cloned into PGEM-T Easy plasmid vector (Promega Corporation, Madison, WI) as described by the manufacturer. The cloned product was sequenced and aligned with previously published sequences of the 18S rRNA gene of Cryptosporidium species by using the CLUSTALX (EMBL, Heidelberg, Germany) program and manual adjustments. Multiple alignment was done with the Phylogeny Inference Package (PHYLIP version 3.5c, J. Felsenstein and the University of Washington, Seattle, WA). Sequences were analyzed by using DNADIST followed by neighbor joining (NEIGHBOR, PHYLIP package). One hundred replica samplings were analyzed for percentage bootstrap values. Accession numbers for Cryptosporidium 18S rRNA genes used were AF093498, AF093497, AF093496, AF108866 and AF093489, AF093499, AF112569 AF115377.

Microscopic examination of the acid-fast stained fecal smear revealed ovoid oocysts that were an average size of 7.5-9.8 x 5.5-7.0 μm (Figure 1). Cysts of I. belli were also identified in the stained smear.

Restriction endonuclease digestion with Ssp1 of the secondary PCR 18S rRNA product yielded two fragments of 385 bp and 448 bp in size, while Asn1 digestion yielded two visible bands that were 102 bp and 731 bp. The results match restriction fragment patterns observed following similar digestions of C. muris amplicons from a rock hyrax and Bactrian camel isolates (14).

The resulting sequence 18S r RNA gene fragment of the C. muris human isolate was deposited in the EMBL Nucleotide Sequence Database (Accession no. AJ307669). Sequence analysis with ClustalX showed this human C. muris isolate had a 100% nucleotide identity to that of a C. muris isolate from a rock hyrax and a Bactrian camel (EMBL Accession no. AF093498, AF093497), 98.8% identity to a C. muris “calf” isolate (AF093496), 96.5% with C. serpentis (AF108866), and only 87.8% identity to C. parvum human type (AF093489). C. muris calf isolate (AF093496) has since been shown to be a different species from C. muris (“mouse” type, Accession no. AF093498 ) and has been given a new name, C. andersoni. The phylogenetic tree showed topology similar to that already published for Cryptosporidium, with C. parvum clustering in one clade, and our patient’s sample and published sequences of C. muris (rock hyrax isolate), C. andersoni (calf isolate), and C. serpentis clustered in another group (Figure 2).

Our study used genotypic analysis to confirm microscopic detection of Cryptosporidium oocysts in fecal samples and indicated that C. muris can indeed infect humans. Although immunosuppression has been observed to produce an increased susceptibility to cryptosporidiosis, the range of Cryptosporidia that can cause human cryptosporidiosis is still being elucidated (8,13,15). Lately, novel genotypes and non-C. parvum species such as C. meleagridis , C. felis, and C. parvum “dog” type have been identified not only in HIV-infected persons but also in HIV-uninfected patients (7,9,10). Genotypic analysis of Cryptosporidium organisms in fecal samples in the United Kingdom showed the occurrence of C. meleagridis, C. felis, and C. parvum “dog” type in immunocompetent and immunosuppressed persons (10,16). In another study in Peru, C. felis, C. parvum “dog” type, and C. meleagridis were identified in children not infected with HIV. In that study, C. meleagridis was as common as C. parvum “bovine” type; it appeared to be a stable part of the enteric pathogen mix causing cryptosporidiosis, perhaps only being identified with current definitive molecular methods (9).

C. muris infects the gastric glands of immunocompetent or immunocompromised (nude and SCID) mice (17); however, since our patient was co-infected with I. belli, the role of C. muris in our patient’s gastroenteritis and its possible site of infection in this patient are unclear.

A report of possible asymptomatic C. muris infection in healthy persons (11) and our finding of it in an immunosuppressed patient suggest that this may be yet another Cryptosporidium species with a zoonotic potential. The range of animal reservoir hosts in which C. muris has been identified or experimentally transmitted adds to the importance of Cryptosporidium species as a public health concern (3,4,15). The current genotypic analyses are making it possible to make more conclusive diagnoses and to speculate on possible sources of infection (14–16). These techniques will need to be applied more widely to identify and characterize isolates of Cryptosporidium for more definitive epidemiologic mapping.