Comparative genomic analysis of the IId subtype family of Cryptosporidium parvum

Abstract

Host adaptation is known to occur in Cryptosporidium parvum, with IIa and IId subtype families preferentially infecting calves and lambs, respectively. To improve our understanding of the genetic basis of host adaptation in Cryptosporidium parvum, we sequenced the genomes of two IId specimens and one IIa specimen from China and Egypt using the Illumina technique and compared them with the published IIa IOWA genome. Sequence data were obtained for >99.3% of the expected genome. Comparative genomic analysis identified differences in numbers of three subtelomeric gene families between sequenced genomes and the reference genome, including those encoding SKSR secretory proteins, the MEDLE family of secretory proteins, and insulinase-like proteases. These gene gains and losses compared with the reference genome were confirmed by PCR analysis. Altogether, 5,191-5,766 single nucleotide variants were seen between genomes sequenced in this study and the reference genome, with most SNVs occurring in subtelomeric regions of chromosomes 1, 4, and 6. The most highly polymorphic genes between IIa and IId encode mainly invasion-associated and immunodominant mucin proteins, and other families of secretory proteins. Further studies are needed to verify the biological significance of these genomic differences.

Keywords: Cryptosporidium parvum; Genomics; Host adaptation; Transmission; Whole genome sequencing.

Fig. 1

Insertion of one Cryptosporidium-specific (conserved sequence motif) SKSR gene at the 3′ end of chromosome 3 in genomes of Cryptosporidium parvum sequenced in this study. Compared with the published IIa reference IOWA isolate, IId specimens from China (31727) and Egypt (34902) had a 4,135 bp and 4,158 bp insertion at the 30 end of chromosome 3, which encodes a 289 and 292 amino acid (aa) peptide with up to 63% and 64% sequence identity to the Cryptosporidium-specific SKSR gene cgd3_10, respectively. IIa specimen 35090 appears to have the same insertion in this region. The vertical red (black) line in the IOWA reference isolate represents the border of the 3′ end of chromosome 3 and the 5′ end of chromosome 4, while the vertical black box (to the left) indicates the same sequence location in genomes under comparison. The coding regions of two genes at the 5′ end of chromosome 4, cgd4_10 and cgd4_20, are shown as horizontal bars. The sequence insertions in chromosome 3 are shown in red (grey), chromosome 3 sequences upstream from the insertion are shown in green (black) on the left, and chromosome 4 sequences are shown in blue (black) on the right. In the reference IOWA genome, the chromosomes 3 and 4 have numerous copies of the telomeric repeat sequences (TTTAGG at the 3′ end of chromosome 3 and CCTAAA at the 5′ end of chromosome 4), which are shown in white. Sizes of the full insertion in the 31727 and 34902 genomes are specified. Only the first 283 bp of the insertion is shown for the 35090 genome, but another 2,368 bp fragment of the insertion is present at the end of the genome sequence alignment. White peaks within each block are sequence divergence between the reference (IOWA) genome and genomes obtained in this study, while black peaks within the insertions are sequence differences among the latter. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Deletion of one gene encoding the (conserved sequence motif) MEDLE family of secretory protein and insertion of one gene encoding the insulinase-like peptidase at the 3′ end of chromosome 6 in genomes of Cryptosporidium parvum sequenced in this study. Compared with the IIa reference IOWA gene, the three C. parvum specimens sequenced, including the IIa specimen from Egypt (35090) and IId specimens from and Egypt (34902) and China (31727), have a 5,273 bp insertion (shown in red (grey)) after gene cgd6-5490. The insertion has up to 66–71% nucleotide sequence identity to cgd3_4260 or cgd3_4270 (genes encoding insulinase-like peptidases similar to cgd6_5510 and cgd6_5520 downstream from the insertion). Upstream from the insertion, the reference IOWA genome has two genes encoding the MEDLE family of secreted proteins (cgd6_5480 and cgd6_5490), whereas genomes sequenced in this study have only one such gene (cgd6_5490) in this region (sequences shown in white blocks are not present in genomes sequenced in this study, including cgd6_5480 in the first white block in the IOWA reference isolate). Within the panels, the single vertical red (black) line represents the border of two contigs in the genome assembly, while the vertical black box indicates the relative location of the cgd6_5500 gene. White peaks within each block are sequence divergence between the reference (IOWA) genome and genomes obtained in this study. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

Confirmation of major insertions at the 3′ end of chromosomes 3 and 6 in Cryptosporidium parvum IId genomes by PCR. (A) Confirmation of the 4,135–4,189 bp insertion in chromosome 3 by PCR targeting on conserved sequences upstream of the telomeric repeats in the reference IOWA genome and 5′ end of the insertion in three genomes obtained from this study. Among the IIa and IId specimens analysed, only three IId specimens from Egypt, Spain and Greece produced the expected 781 bp PCR product. (B) Confirmation of the 5,273 bp insertion in chromosome 6 by PCR targeting conserved nucleotide sequences flanking the insertion (3′ end of cgd6_5490 and 5′ end of cgd6_5500). Among the IIa and IId specimens analysed, only three IIa specimens from the United States produced the expected 781 bp PCR product, as the large insertion in IId genomes between cgd6_5490 and cgd6_5500 had prevented the amplification of the target sequence. M, size marker in 100 bp; lane 1, IIaA15G2R2 from USA; lane 2, IIaA15G2R1 from USA; lane 3, IIaA18G3R1 from USA; lane 4, IIdA20G1 from Egypt; lane 5, IIdA21G1 from Spain; lane 6, IIdA16G1 from Greece; lane 7, negative control and primary PCR; and lane 8, negative control for secondary PCR.

Fig. 4

Phylogenetic relationship of Cryptosporidium parvum genomes characterised in this study as indicated by neighbour-joining analysis of whole genome single nucleotide variant data (A) and concatenated sequences (B) of 11 highly polymorphic genes (cgd1_120, cgd1_470, cgd4_10, cgd4_20, cgd4_30, cgd6_40, cgd6_1080, cgd6_5460, cgd6_5470, cgd6_5490, and cgd6_5500). In the former, genetic distances were computed using R package ‘APE’ and are in the number of single nucleotide variants, whereas in the latter, they were computed using the Kimura 2-parameter model and are in the units of the number of nucleotide substitutions per site. Numbers of single nucleotide variants and percentages of bootstrapping (1000 replicates) are shown on branches in A and B, respectively. Both trees are outgrouped with sequences from the Cryptosporidium hominis TU502 isolate.

Fig. 5

Distribution of single nucleotide variants in Cryptosporidium parvum genomes sequenced in this study in comparison with the published reference IOWA genome. To detect single nucleotide variants, sequence reads from each specimen were mapped to the 9.1 Mb reference IOWA genome and single nucleotide variants present were detected using the Basic Variant Detection tool in CLC Genomic Workbench 8.5. The Y-axis values are numbers of single nucleotide variants per 1,000 nucleotides and X-axis values are nucleotide positions of the single nucleotide variant peaks across the eight linked chromosomes (chrom). The X-axis labels for (A–C) are not drawn at the same scale, resulting in a slight shift of single nucleotide variant peaks. Highly polymorphic genes (single nucleotide variants values >20/1,000 nucleotides) are labelled in (A) 31727: IId specimen from China; (B) 34902: IId specimen from Egypt; (C) 35090: IIa specimen from Egypt. Genes only highly polymorphic between IId and the reference IOWA genomes are labelled with an asterisk.

Fig. 6

Distribution of single nucleotide variants among Cryptosporidium parvum genomes sequenced in this study. To detect single nucleotide variants, sequence reads from one specimen were mapped to the assembled genome of another specimen and single nucleotide variants present were detected using the Basic Variant Detection tool in CLC Genomic Workbench 8.5. The Y-axis values are numbers of single nucleotide variants per 1,000 nucleotides and X-axis values are nucleotide positions of the single nucleotide variant peaks across the eight linked chromosomes. Highly polymorphic genes (single nucleotide variants values > 20/1,000 nucleotides) are labelled. (A) 34902, IId specimen from Egypt versus 35090, IIa specimen from Egypt; (B) 31727, IId specimen from China versus 35090; (C) 34902 versus 31727.