Comparative genomic analysis reveals occurrence of genetic recombination in virulent Cryptosporidium hominis subtypes and telomeric gene duplications in Cryptosporidium parvum

Abstract

Background: Cryptosporidium hominis is a dominant species for human cryptosporidiosis. Within the species, IbA10G2 is the most virulent subtype responsible for all C. hominis-associated outbreaks in Europe and Australia, and is a dominant outbreak subtype in the United States. In recent yearsIaA28R4 is becoming a major new subtype in the United States. In this study, we sequenced the genomes of two field specimens from each of the two subtypes and conducted a comparative genomic analysis of the obtained sequences with those from the only fully sequenced Cryptosporidium parvum genome.

Results: Altogether, 8.59-9.05 Mb of Cryptosporidium sequences in 45-767 assembled contigs were obtained from the four specimens, representing 94.36-99.47% coverage of the expected genome. These genomes had complete synteny in gene organization and 96.86-97.0% and 99.72-99.83% nucleotide sequence similarities to the published genomes of C. parvum and C. hominis, respectively. Several major insertions and deletions were seen between C. hominis and C. parvum genomes, involving mostly members of multicopy gene families near telomeres. The four C. hominis genomes were highly similar to each other and divergent from the reference IaA25R3 genome in some highly polymorphic regions. Major sequence differences among the four specimens sequenced in this study were in the 5' and 3' ends of chromosome 6 and the gp60 region, largely the result of genetic recombination.

Conclusions: The sequence similarity among specimens of the two dominant outbreak subtypes and genetic recombination in chromosome 6, especially around the putative virulence determinant gp60 region, suggest that genetic recombination plays a potential role in the emergence of hyper-transmissible C. hominis subtypes. The high sequence conservation between C. parvum and C. hominis genomes and significant differences in copy numbers of MEDLE family secreted proteins and insulinase-like proteases indicate that telomeric gene duplications could potentially contribute to host expansion in C. parvum.

Figure 1

Structural organization of two Illumina-sequenced genomes of Cryptosporidium hominis comparing to eight chromosomes (numbered and separated by vertical red lines) of published Cryptosporidium parvum genome. The color blocks (known as Locally Collinear Blocks) are conserved segments of sequences internally free from genome rearrangements, whereas the inverted white peaks within each block are sequence divergence between the reference C. parvum (IOWA) genome and C. hominis genome under analysis. A. Coverage of two C. hominis genomes showing possible sequence rearrangements in chromosomes 2, 4, 5 and 6. Assembled contigs are bordered by vertical red lines. For specimens 30976, only Cryptosporidium contigs were used in mapping. B. Possible sequence rearrangements at the 5′ end of chromosome 2. C. Possible sequence rearrangements in chromosomes 4 and 5.

Figure 2

Deletion of genes in Cryptosporidium hominis genomes in comparison with Cryptosporidium parvum. A. Deletion of four genes (cgd6_5480, cgd6_5490, cgd6_5510, and cgd6_5520) at the 3′ end of chromosome 6 (probably should be the 5′ end of chromosome 5) in C. hominis. B. A major 19,048-bp deletion in C. hominis genome in chromosome 8, including the cgd8_680 and cgd8_690 genes. Note the ~10 kb sequence gap in C. parvum.

Figure 3

Cryptosporidium hominis-specific nature of Chro.50011. A. Insertion of ~4,860 bp containing the Chro.50011 gene at the 3′ end of chromosome 3 in C. hominis. B. Confirmation of the absence of the ortholog of Chro.50011 in four specimens of C. parvum by PCR analysis of three regions of the Chro.50011 gene. The faint band in PCR analysis of the 3′ end of the gene in C. parvum specimen 38416 produced a nucleotide sequence identical to Chro.50011 in C. hominis.

Figure 4

Distribution of SNPs in Cryptosporidium hominis genome by chromosome. The number of SNPs in a sliding window of 2,000 bp with 200 bp steps across each of the eight chromosomes is shown. A. Sequence divergence between specimen 30976 of the IaA28R4 subtype and the published isolate TU502 of the IaA25R3 subtype. B. Sequence divergence between specimen 37999 of the IbA10G2 subtype and specimen 30976 of the IaA10G2 subtype.