Cryptosporidium spp. are a major cause of parasitic diarrhea in children under the age of five in developing countries (
17,
25). In these countries, cryptosporidiosis in early childhood can lead to persistent diarrhea, growth faltering, and impairment in physical and cognitive development (
9,
14). Very little is known about the molecular epidemiology or transmission dynamics of cryptosporidiosis in the developing world, where the burden of cryptosporidiosis is greatest. Two major
Cryptosporidium species infect humans;
C. hominis primarily infects humans and is the most frequently identified species in developing countries, and
C. parvum infects humans as well as animals (
29). Recently, subtyping at polymorphic loci has been used to characterize the genetic diversity of
Cryptosporidium spp. (
29). Currently, the most common locus for identifying subtype families and subtypes is that of the
gp40/15 or
gp60 gene, which we and others cloned (
8,
22,
26,
27). To date, at least 17 major subtype families, 6 from
C. hominis (Ia, Ib, Id, Ie, If, Ig) and 11 from
C. parvum (IIa, IIb, IIc, IId, IIe, IIf, IIh, IIi, IIj, IIk, IIl), have been identified in humans and animals worldwide (
18,
29). These include 78
C. parvum and 74
C. hominis subtypes, classified according to the number and type of serine-coding trinucleotide tandem repeats at the 5′ end of the gene (
29).
Although cryptosporidiosis is widely prevalent in Bangladesh (
2,
4,
16,
19,
20,
23,
24), there have been no studies on the genetic diversity of
Cryptosporidium spp. in this country. As part of a case control study on
Cryptosporidium (identified by screening 1,672 stool samples by microscopy) in children under the age of five presenting with diarrhea to the Dhaka Hospital of the International Centre for Diarrheal Disease Research (ICDDR) in Dhaka, Bangladesh (
19), we determined the species and identified the
gp40/15 subtype families and subtypes of
Cryptosporidium spp. infecting children in the study. The original study (
19) was approved by the Ethical Review Committee of the ICDDR, and the use of deidentified stool samples (which were stored at −80°C and shipped to Boston on dry ice) was approved by the Tufts Health Sciences Institutional Review Board.
Stool samples from 46 cases (diarrhea and stool microscopy positive for
Cryptosporidium) and 46 age-matched controls (diarrhea and microscopy negative for
Cryptosporidium) at presentation and from 30 cases and 23 controls at follow-up 3 weeks later were analyzed. DNA was extracted from stool samples as described previously (
21) or using the QIAamp stool minikit (Qiagen, Inc., Valencia, CA) and analyzed by nested PCR at the 18S rRNA locus (
30). Samples from all 46 cases and 7 of 46 controls at presentation and 12 of 30 cases and 2 of 23 controls collected at follow-up 3 weeks later were PCR positive for
Cryptosporidium. All 67 PCR-positive samples were analyzed by PCR restriction fragment length polymorphism (RFLP) at the 18S rRNA locus for species determination (
30). At presentation, 48 of the 53 (91%) samples were positive for
C. hominis, 4 were positive for
C. parvum (7%), and 1 was positive for
C. felis (2%) (). The same
Cryptosporidium sp. was identified at the initial and follow-up time points in all 14 of the samples that were PCR positive at follow-up.
| Table 1.Cryptosporidium species, subtype families, and subtypes identified from children in the studyd |
To identify subtype families, the
gp40/15 gene was amplified by nested PCR and subjected to PCR RFLP (
11). All 48
C. hominis samples displayed PCR RFLP profiles corresponding to that of five previously described subtype families (
11,
28), including Ia (8 samples), Ib (10 samples), Id (6 samples), Ie (13 samples), and If (11 samples) (). The
C. felis sample (number 76) displayed a subtype IIa profile. All 4
C. parvum samples (numbers 7, 9, 21, and 78) displayed an RFLP profile that has not been described previously (data not shown). At the follow-up visit, all 12 samples that remained PCR positive for
Cryptosporidium spp. had the same PCR RFLP profile as at the initial visit except for two that changed from Ib to Id (number 11) and Id to If (number 40) (), suggesting that these children were reinfected with a different subtype.
To confirm the PCR RFLP findings and identify subtypes, the
gp40/15 PCR amplicons from samples obtained at presentation were sequenced (
11). In addition, amplicons from the four samples displaying the new PCR RFLP profile were cloned into the pCR 2.1-TOPO vector (Invitrogen Corp., Carlsbad, CA), and the inserts from two or three clones were sequenced. Sequences were obtained from all but three (sample numbers 1, 6, and 70) of the 53 samples (
29). Phylogenetic analysis of sequences compared to those deposited in GenBank () supported the classification of the samples into six previously described
gp40/15 subtype families (Ia, Ib, Id, Ie, If, and IIa). Sequences of the four samples with the new PCR RFLP pattern were identical to each other and were most similar to the subtype family IIe sequence (). However, in addition to five additional tandem TCA repeats in these sequences compared to the IIe sequence, there were several polymorphisms outside the serine-coding region which resulted in amino acid changes (not shown). We therefore propose that these 4 sequences be classified into a new subtype family named “IIm” according to currently accepted nomenclature (
29).
Subtype families Ie and If were the most common and were present in 25% and 23% of all samples, respectively, although the overall dominance rates of these subtype families in humans worldwide are only 5.1 and 1.1%, respectively (
18). Within subtype family Ie, as reported previously from other areas (
18), subtype IeA11G3T3 was dominant. However, all 11 subtypes within the If family were IfA13G1, which has not been reported previously. Subtype families Ib and Ia were each present in 15% of samples in our study. However, the most common subtype worldwide that is present in the Ib subtype family, i.e., IbA10G2 (
18), was not identified. As previously reported from other areas (
18), subtype family Ia was the most diverse and included five different subtypes. However, again, the most common Ia subtypes, IaA12G1R1 and IaA21G1R1 (
18), were not present. Within subtype family Id, subtype IdA15G1 was the most common, as reported previously (
18). However, one sample from this subtype family displayed an IdA24 subtype which, again, has not been previously reported. Subtype family Ig was not identified in any of the samples in our study.
All four
C. parvum samples in our study were of the IIm subtype family. This subtype family has not been identified in any animal thus far and may therefore be another of the so-called “human-adapted” or “anthroponotic”
C. parvum subtype families, similar to the IIc subtype family (
29), which, interestingly, was not identified in our study. Together with the finding that all but one (which was
C. felis) of the other species identified were
C. hominis, this suggests that transmission in children in this area is predominantly anthroponotic, i.e., transmitted from human to human, and is similar to that in other developing countries where
C. hominis and anthroponotic
C. parvum subtype families predominate (
1,
6,
21). Interestingly, although the
gp40/15 locus is not thought to be amplifiable from
C. felis DNA (
29), DNA from the
C. felis sample from this study and two others from India (
1) did amplify with the
gp40/15 primers used in these studies, and all three were of the subtype IIa family. This finding could also be the consequence of mixed infections, as previously described (
5).
In conclusion, this is the first study to characterize genetic diversity at the subtype level in Cryptosporidium spp. from Bangladesh. Further molecular, clinical, and epidemiological studies of Cryptosporidium infections in vulnerable human populations as well as in domestic animals and environmental samples are required to investigate the transmission dynamics of cryptosporidiosis and design effective strategies to block transmission and prevent spread of the disease in developing countries, where the burden of this disease is greatest.
Nucleotide sequence accession numbers. The gp40/15 sequences obtained in this study have been deposited in GenBank under accession numbers AY700385 to AY700401 and JF927169 to JF927200. Accession numbers of the IIm sequences include AY700401 (number 9), AY700385 (number 21), AY700395 (number 78), and AY700396 (number 7).
![[down-pointing small open triangle]](/corehtml/pmc/pmcents/x25BF.gif)
This article has been 
