BackgroundHuman cyclosporiasis is caused by Cyclospora cayetanensis. It is endemic in many developing countries, responsible for significant morbidity in children and AIDS patients [1, 2]. It is also a major cause of foodborne diarrheal illnesses in industrialized nations, especially in North America. Since the mid-1990s, numerous outbreaks of cyclosporiasis have occurred in the United States and Canada, mostly associated with fresh produce imported from Mexico and Central America [2, 3]. The lack of molecular diagnostic tools for case-linkage and trace-back investigations of outbreaks has hampered responses from public health and regulatory agencies such as the U.S. Food and Drug Administration and Centers for Disease Control and Prevention. This was exemplified in two recent large multistate outbreaks of cyclosporiasis in 2013 and 2014 in the United States [4] (http://www.cdc.gov/parasites/cyclosporiasis/outbreaks/index.html).
One major reason for the lack of molecular diagnostic tools and poor understanding of the basic biology of C. cayetanensis is the limited availability of nucleotide sequences from the parasite. At the end of 2014, among the ~500 entries of Cyclospora sequences in GenBank, almost all of them were from the rRNA (SSU rRNA, 5.8S rRNA, ITS-1, and ITS-2) and 70 kDa heat shock protein (HSP70) genes. The conserved sequence nature of rRNA and HSP70 genes and intra-isolate variations among different copies of ITS-1 and ITS-2 make the development of genotyping tools for the parasite difficult [5–9]. There is a need for genomic sequencing to identify polymorphic markers and promote biologic studies of C. cayetanensis [10].
Apicoplasts and mitochondria are two important intracellular organelles of apicomplexan parasites [11]. The apicoplast is the result of secondary endosymbiosis by ancient apicomplexan parasites, and harbors metabolic pathways that are necessary for the survival of parasites but are very distinct from those of host species [12]. Likewise, mitochondria play key roles in energy metabolism in apicomplexan parasites [11]. Most apicomplexan parasites, with the notable exception of Cryptosporidium spp., have both apicoplast and mitochondrial genomes. Because of the prokaryotic origins and biological importance of both apicoplasts and mitochondria, they have been widely used as drug targets against apicomplexan parasites [12–15]. The non-recombining and co-inherited nature of apicoplast and mitochondrial genomes has recently been used in the development of barcoding tools for tracking worldwide migration of Plasmodium spp. [16–18]. Even at the species level, a recent barcoding study of Eimeria spp. has shown that partial mitochondrial cytochrome c oxidase subunit I (cox1) sequences are more reliable species-specific markers than complete nucleic SSU rRNA sequences, as they provide more synapomorphic characters [19]. The near complete genome of C. cayetanensis has been published recently [20].
As there are no data on the apicoplast genome and only partial data on the mitochondrial genomes of C. cayetanensis, we have characterized in this study the complete apicoplast and mitochondrial genomes of an isolate from China. Data generated have shown a close relatedness in apicoplast and mitochondrial genomic sequences between C. cayetanensis and cecum-infecting avian Eimeria spp. and nucleotide sequence polymorphism in mitochondrial genomes between the isolate in this study and the one in the recent publication. These sequence data should be useful in improving our understanding of the biology of C. cayetanensis and developing advanced molecular detection tools.
MethodsDNA preparation, sequencing and de novo assemblyDNA was extracted from a C. cayetanensis specimen collected in July 2011 from a patient in Kaifeng, Henan, China. The identification of C. cayetanensis was done by acid-fast microscopy and confirmed by PCR and sequence analyses of the SSU rRNA gene as previously described [5]. DNA was extracted from sucrose and cesium chloride gradient-purified oocysts using the QIAamp®DNA Mini kit (Qiagen, Valencia, CA). The genome of C. cayetanensis was sequenced using the Roche 454 GS-FLX Titanium (Roche, Branford, CT), with average read lengths of 400–450 bp. Supplemental 100 × 100 bp paired-end sequencing was completed using the Illumina Genome Analyzer IIx (Illumina, San Diego, CA). After sequencing, 454 sequence reads were assembled using Roche Newbler (gsAssembler 2.3). Illumina reads were trimmed using clc_quality_trim within the CLC Assembly Cell v 4.1.0 (http://www.clcbio.com/) with minimum quality score of 20 and minimum read length of 70. Trimmed reads were assembled de novo with the clc_assembler of the CLC Assembly Cell. Contigs for C. cayetanensis apicoplast and mitochondrion genes were identified by blasting the assemblies against the GenBank database.
PCR confirmationPCR amplification and sequencing of the amplicons were used to confirm the inverted repeats in the plastid genome. Two pairs of primers were used to amplify the regions that join the inverted repeat to the main body of the apicoplast. The primer sets were F1: 5′-AGT CGC TAA GTA GCC AAG TTT-3′ and R1: 5′-TTG TCT TGC CTG TGC TAT AGT AAT-3′, and F2: 5′- ACT ACA TCA ACG GCT AAC T-3′ and R2: 5′-GTA CGA GAG GAC CAA AGA AA-3′, which targeted a 634 (between rps4 and LSU rRNA genes) and 743 bp fragment (between ycf24 and LSU rRNA genes), respectively. The reactions contained 1 μl of DNA, 1× GeneAmp PCR buffer (Applied Biosystems, Foster City, CA), 1.5U of GoTaq polymerase (Promega, Madison, WI), 250 μM dNTPs (Promega), 3 mM MgCl2, 250 nM of each primer, and 400 ng/μl of non-acetylated bovine serum albumin (Sigma-Aldrich, St. Louis, MO). The amplification was performed on a GeneAmp PCR 9700 thermocycler (Applied Biosystems), consisting of an initial denaturation at 94 °C for 5 min; 35 cycles at 94 °C for 45 s, 55 °C for 45 s, and 72 °C for 1 min; and a final extension at 72 °C for 7 min. The secondary PCR products were detected by 1.5 % agarose gel electrophoresis, and sequenced in both directions using the BigDye Terminator V3.1 Cycle Sequencing Kit (Applied Biosystems) on an ABI 3130xl Genetic Analyzer (Applied Biosystems).
Genome annotation and comparisonAs the apicoplast and mitochondrial genomes of C. cayetanensis were in complete synteny with those of Eimeria spp., we used the Rapid Annotation Transfer Tool (RATT) [21] to transfer the annotations from the E. tenella apicoplast genome (AY217738) and E. dispersa mitochondrion genome (KJ608416). Potential genes and open reading frames (ORFs) were also identified using GeneMark [22]. The annotation was then manually improved by comparing the predicted ORFs and RATT annotations.
Detection of sequence variationsThe 454 and Illumina sequence reads were mapped to reference genomes and assembled sequence contigs using the CLC Genomics Workbench 7.0.3 (http://www.clcbio.com/). The probabilistic variant detection tool in the software package was used to detect single nucleotide variations (SNVs) among mapped reads, using the parameters of min coverage 10, probability 90, required variant count 2, and all other filters unchecked. In-house perl scripts were developed to generate SNV data for display. The assembled apicoplast and mitochondrial genomes were further compared with the corresponding reference genomes using Nucmer within the MUMmer package [23].
Phylogenetic analysisFull genome sequences for apicoplasts and mitochondria of other apicomplexans were extracted from GenBank and aligned with C. cayetanensis sequences by using the sequence alignment function within the CLC Genomics Workbench 7 (http://www.clcbio.com/). Poorly aligned positions in the alignments were eliminated by using Gblocks [24]. The maximum likelihood phylogenies of these sequences were inferred from the data using the CLC Genomics Workbench under the general time reversible model of nucleotide substitution. The confidence of the cluster formation was assessed by using bootstrap analysis with 1,000 replicates.
Nucleotide sequence accession numbersNucleotide sequence data from the whole genome sequencing, including the SRA data and assembled contigs, were submitted to NCBI BioProject PRJNA256967. Sequences of the annotated apicoplast and mitochondrial genomes were deposited in the GenBank database under accession numbers KP866208 and KP796149.
DiscussionIn this study, the complete apicoplast genome of C. cayetanensis has been sequenced for the first time. We have also obtained the last 45-bp sequence to complete the mitochondrial genome published recently [20]. The apicoplast and mitochondrial genomes of C. cayetanensis are both highly similar to those of Eimeria spp. in genome sizes and gene contents, with a complete synteny in genome organization. As seen in Eimeria spp. and other apicomplexans, the relatively large apicoplast genome of C. cayetanensis codes for the entire array of machinery needed for the expression of the apicoplast genome, including two copies of SSU rRNA and LSU rRNA, various ribosomal proteins, three RNA polymerases, and tRNAs for all 20 amino acids. Thus, C. cayetanensis should have a functional apicoplast organelle, which as in other apicomplexans likely plays key roles in the biosynthesis of fatty acids, isoprenoids, iron-suffer clusters and heme [12, 25]. Like in other coccidian parasites [26], the TGA codon for tryptophan is common in the apicoplast genome of C. cayetanensis. In contrast, hemosporidians such as Plasmodium spp., Theileria spp., Babesia spp., and Leucocytozoon spp. do not have any in-frame TGA codons in their apicoplast genomes [25, 27]. One in-frame TGA codon was also seen in the mitochondrial cox3 gene in C. cayetanensis.
Phylogenetic analysis of both apicoplast and mitochondrial genomes from a diverse group of apicomplexans supports the close relationship between C. cayetanensis and avian Eimeria spp., thus confirming the results of direct comparison of organelle genome organization and those from previous phylogenetic analyses of the partial nucleic SSU rRNA and HSP70 genes [28–31]. However, some minor differences in results among phylogenetic studies have been observed. In the present study, phylogenetic analyses of the complete apicoplast and mitochondrial genomes both suggest that C. cayetanensis is more closely related to the cecum-infecting avian Eimeria spp. such as E. tenella and E. necatrix than to the small intestine-infecting avian Eimeria spp. such as E. brunetti. This is in agreement with the results of an analysis of the partial SSU RNA gene [30]. Other analyses of the SSU rRNA gene, however, have shown that C. cayetanensis formed a sister cluster with avian Eimeria spp. [29, 32]. Like in this study, most phylogenetic analyses of the SSU rRNA gene have shown a clear separation of the cecum-infecting avian Eimeria spp. from the small intestine-infecting Eimeria spp. [19, 29, 30, 32]. A recent analysis of the complete mitochondrial genomes of Eimeria spp. has also shown the formation of separate clusters by the cecum-infecting and small intestine-infecting avian Eimeria spp. [33]. Comparisons of the nucleic genomes of these parasites are needed to better understand the evolutionary position of Cyclospora spp. and the biologic significance of the relatedness between C. cayetanensis and cecum-infecting avian Eimeria spp.
The structure of the mitochondrial genome of C. cayetanensis in this study is different from the one proposed recently. Based largely on amplification failures using primers that have worked reliably for various Eimeria spp. and new primers designed based on the assumption of a circular or linear concatenated genome, Ogedengbe and colleagues have suggested recently that C. cayetanensis likely has a linear monomeric genome [20]. This is very differently from the linear concatenated genome of E. tenella identified based on restriction fragment length polymorphism analysis [34]. Results of our genome assembly and sequence mapping indicate that C. cayetanensis probably has the same mitochondrial genome structure as E. tenella. In the published study, the high sequence divergence between Cyclospora and Eimeria near the joint of two mitochondrial genome units was probably responsible for the amplification failures in the analysis of the sequence at the joint for C. cayetanensis.
Some sequence differences have been observed between mitochondrial genomes of this and the published studies. Altogether, there are eight SNVs and one MNV between genomes of the Chinese isolate examined in this study and the 2013 Texas outbreak isolate in the previous report [20]. Four of the SNVs are in the cox1 and cox3 genes. Previously, analysis of the full cox1 and cox3 genes showed no sequence differences among five C. cayetanensis isolates from limited areas (Southeast Asia and United States) [20]. Comparative analysis of the apicoplast and mitochondrial genomes has been used in tracking the spread of P. falciparum and P. vivax [16–18]. With further verifications, this approach can be potentially used in geographic tracking of C. cayetanensis during investigations of cyclosporiasis outbreaks in North America.
With the whole apicoplast and mitochondrial genome sequences available, better intervention measures and diagnostic tools can potentially be developed for C. cayetanensis. The close relatedness between avian Eimeria spp. and Cyclospora spp. indicates that avian Eimeria spp., especially those infecting the cecum, may serve as an alternative model for studying Cyclospora spp. Thus far, there are no feasible animal models for C. cayetanensis [35], and feeding sporulated C. cayetanensis oocysts to human volunteers did not produce any infection [36]. The genetic similarity between avian Cyclospora spp. and Eimeria spp. suggests that many of the drugs used in the treatment of poultry coccidiosis may be effective against C. cayetanensis infection. Other therapeutics specifically targeting the mitochondrial and apicoplast metabolism can be developed [13, 14, 37].