Reevaluation of the Toxoplasma gondii and Neospora caninum genomes reveals misassembly, karyotype differences, and chromosomal rearrangements

Abstract

Neospora caninum primarily infects cattle, causing abortions, with an estimated impact of a billion dollars on the worldwide economy annually. However, the study of its biology has been unheeded by the established paradigm that it is virtually identical to its close relative, the widely studied human pathogen Toxoplasma gondii By revisiting the genome sequence, assembly, and annotation using third-generation sequencing technologies, here we show that the N. caninum genome was originally incorrectly assembled under the presumption of synteny with T. gondii We show that major chromosomal rearrangements have occurred between these species. Importantly, we show that chromosomes originally named Chr VIIb and VIII are indeed fused, reducing the karyotype of both N. caninum and T. gondii to 13 chromosomes. We reannotate the N. caninum genome, revealing more than 500 new genes. We sequence and annotate the nonphotosynthetic plastid and mitochondrial genomes and show that although apicoplast genomes are virtually identical, high levels of gene fragmentation and reshuffling exist between species and strains. Our results correct assembly artifacts that are currently widely distributed in the genome database of N. caninum and T. gondii and, more importantly, highlight the mitochondria as a previously oversighted source of variability and pave the way for a change in the paradigm of synteny, encouraging rethinking the genome as basis of the comparative unique biology of these pathogens.

Figure 1.

Comparative analysis of genome assemblies of Neospora caninum and Toxoplasma gondii using third-generation sequencing data reveals misassembly and karyotype differences. (A) Comparative analysis of the T. gondii type II (TgME49) genome assembly and the N. caninum Liverpool (NcLiv) strain genome assembly, obtained based on Sanger technology sequencing data. (B) Comparative alignment of the NcLiv genome assemblies using Sanger and third-generation (long-read) technology. (C) Comparative alignment of the T. gondii type II (TgME49) genome assemblies based on Sanger technology sequencing data or third-generation (long-read) technology of T. gondii type I (TgRH). (D) Comparative alignment of the T. gondii type I (TgRH) and the NcLiv genome assemblies based on third-generation (long-read) sequencing technology. (E) Chromosomal layout of N. caninum. Karyotype, chromosome length, telomeres, putative centromeres, and large repeats are shown.

Figure 2.

Regions of synteny breaks between N. caninum and T. gondii are populated by three conserved domains. (A) Sequence identity of domains identified at regions where chromosomal rearrangements have occurred. (B) Graphical representation of Chromosome VIII of NcLiv. Comparative alignment to the T. gondii chromosomes. Percentages of sequence identity are shown. Regions examined for the presence of motifs are indicated (light green). The position of the putative centromere is indicated in orange. Note that large repetitive regions were not identified in this chromosome. 5′ (light purple) and 3′ (dark purple) telomeres are indicated. The identity and number of domains found per region, in Chromosome VII, are indicated.

Figure 3.

Gene annotation reveals previously unknown genes in the genome of N. caninum. (A) Graphical representation of the position of novel genes in the newly assembled chromosomes. Red lines mark the position of novel genes along chromosomes. Alignment to the previously assembled NcLiv genome is partially shown for reference. Three insets are shown to highlight the annotation of three new genes in three newly assembled genomic regions. (B) Putative function of six out of the 38 newly identified genes. The remaining 32 genes are annotated as hypothetical. (C) Representative example of the improvement in annotation in regions that had been previously collapsed owing to the presence of tandem repeats. Several new genes were annotated, all of whose annotation is supported by RNA-seq data.

Figure 4.

N. caninum apicoplast genome structure and annotation. (A) Schematic representation of the circularized apicoplast genome and annotation. The presence of an inverted repeat sequence is highlighted by a shaded area. %GC is shown in gray. GC skew is represented in variable colors. Open reading frames present within the 35-kb circular apicoplast genome are shown (for details, see Supplemental Table 5). (B) Read coverage count along the length of the 35-kb apicoplast genome is shown. (C) The presence of an inverted repeat sequence is graphically represented in a YASS plot. (D) Alignment of long-read sequences to the apicoplast genome. A few reads spanning the end regions, supporting its circular topology, are highlighted.

Figure 5.

Comparative analysis of mitochondrial genome structures and annotations of Neospora and Toxoplasma reveals gene fragmentation and reshuffling between species and strains. (A) The repetitive nature of the gene structure in a 32-kb mitochondrial DNA contig of NcLiv is graphically represented in a YASS plot. (B) The repetitive nature of the gene structure in a 16-kb mitochondrial DNA contig of NcLiv is graphically represented in a YASS plot. (C) Comparative alignment between two NcLiv mitochondrial contigs of 16 and 32 kb, respectively. (D) Comparative alignment between a NcLiv mitochondrial contig of 32 kb and a NcUru1 mitochondrial contig of 38 kb. (E) Comparative alignment between two NcUru1 mitochondrial contigs of 16 and 38 kb, respectively. (F) The repetitive nature of the gene structure in a 16-kb mitochondrial DNA contig of NcUru1 is graphically represented in a YASS plot. (G) Comparative alignment between a NcLiv mitochondrial contig of 32 kb and a T. gondii mitochondrial contigs of 39 kb.