Massive invasion of organellar DNA drives nuclear genome evolution in Toxoplasma

Abstract

Toxoplasma gondii is a zoonotic protist pathogen that infects up to 1/3 of the human population. This apicomplexan parasite contains three genome sequences: nuclear (63 Mb); plastid organellar, ptDNA (35 kb); and mitochondrial organellar, mtDNA (5.9 kb of non-repetitive sequence). We find that the nuclear genome contains a significant amount of NUMTs (nuclear DNA of mitochondrial origin) and NUPTs (nuclear DNA of plastid origin) that are continuously acquired and represent a significant source of intraspecific genetic variation. NUOT (nuclear DNA of organellar origin) accretion has generated 1.6% of the extant T. gondii ME49 nuclear genome; the highest fraction ever reported in any organism. NUOTs are primarily found in organisms that retain the non-homologous end-joining repair pathway. Significant movement of organellar DNA was experimentally captured via amplicon sequencing of a CRISPR-induced double-strand break in non-homologous end-joining repair competent, but not ku80 mutant, Toxoplasma parasites. Comparisons with Neospora caninum, a species that diverged from Toxoplasma ~28 MY ago, revealed that the movement and fixation of 5 NUMTs predates the split of the two genera. This unexpected level of NUMT conservation suggests evolutionary constraint for cellular function. Most NUMT insertions reside within (60%) or nearby genes (23% within 1.5 kb) and reporter assays indicate that some NUMTs have the ability to function as cis-regulatory elements modulating gene expression. Together these findings portray a role for organellar sequence insertion in dynamically shaping the genomic architecture and likely contributing to adaptation and phenotypic changes in this important human pathogen.

Keywords: Coccidia; Non-Homologous End-Joining repair - NHEJ; Nuclear DNA of organellar origin - NUOT; Nuclear integrants of Mitochondrial DNA - NUMTs; Nuclear integrants of Plastid DNA - NUPTs.

Figure 1.. Characteristics of NUMTs and NUPTs in T. gondii ME49

(A) Cladogram of apicomplexan parasite relationships (B) Circos plot (Circos version 0.51) represents the distribution of NUMTs and NUPTs in the nuclear chromosomes of T. gondii ME49. The outer circle of yellow bands indicates the chromosomes as labeled and each tick on the band equals 100 kb. The ticks interior to the chromosomes are different features as indicated in the key. (C) The length and the percent identity of a NUMT and NUPT to mt and ptDNA respectively was calculated from RepeatMasker results and the distribution is plotted. (D) The age of each NUMT and NUPT was calculated based on the percent divergence from the mt and ptDNA of the corresponding species and the distribution of the age is plotted against the number of base pairs.

Figure 2.. Strain-specific presence and absence of NUMTs in T. gondii

(A, B) A likely strain-specific insertion in T. gondii (TG) GT1 (A) and deletion in ME49 (B) are shown. A multiple sequence alignment of these loci was performed for strains GT1, VEG and ME49 and the coordinates of the loci are indicated with reference with GT1 chromosome VI. The NUMT and flanking nuclear sequence are indicated in red and black respectively. (C) NUMTs in 16 T. gondii strains were compared to identify a strain-specific presence and absence. Strains are grouped and colored by clades as previously identified (7). The 14 chromosomes divided by black lines are represented on the x-axis. NUMTs that show differential presence/absence in the 16 T. gondii strains are numbered and ordered based on their location in ME49. The coordinates of each NUMT plus 200 bp flanking region on the ME49 chromosomes are shown at the bottom. The 200 bp regions were included to provide a location for NUMTs missing in ME49. The percent divergence of the NUMT from the mtDNA sequence is indicated above the coordinates. In the matrix, “+” and “−” indicate presence and absence of a NUMT in a particular strain. Each cell is colored based on local clade ancestry of that location (7) although, it does not imply that location originated from that clade. Yellow cells indicate clade E ancestry and were not included the NUMT analysis as an assembled genome sequence for a strain from this clade was not available

Figure 3.. Cis-regulatory activity of two selected NUMTs

(A, B) The structure of WT and NUMT deletion promoter constructs for U6 snRNA-associated Sm family protein gene TGME49_286560 (A) and myosin heavy chain gene TGME49_254850 (B). Nucleotide positions are referenced with respect to the start of translation (+1) of host gene and the red region indicates the position of the NUMT (Fig. S3). (C, D) Reporter assay results for the promoter of Sm-like protein (C) and mysoin heavy chain (D) genes. The graphs depict luciferase activity as ratios of Firefly:Renilla activity in relative luciferase units (RLU) from the different constructs containing either WT or mutagenized promoter. All luciferase readings are relative to an internal control (α-tubulin-Renilla). Error bars represent standard error calculated across the means of six independent electroporations. P-values were calculated using Student’s t-test.

Figure 4.. Capture of DNA insertions at CRISPR/Cas9-directed double-strand break at uprt locus in wildtype and Δku80 parasites.

Utilizing the CRISPR-Cas9 system, double-strand breaks were induced at the uprt locus in wildtype T. gondii and Δku80 parasites. The locus was then amplified and deep-sequenced to evaluate the extrachromosomal DNA insertions (mtDNA, ptDNA, CRISPR-Cas9 plasmid DNA) at the DSB site. (A) Apicomlexan cladogram with presence (+) or absence (−) of key NHEJ pathway genes and NUOT content in each species indicated. The coccidian branch is shown in bold. (B, C) The percent of amplicons with insertions from each DNA type (B) and the percentage of base pairs from each type present in the amplicons (C) are plotted for WT and Δku80 parasites. The amplicons analyzed comprise all read pairs that could be merged as well as unmerged reads. Error bars represent standard error of means of three replicates. Statistical significance was calculated using Student’s t-test (* p < 0.05, ** p < 0.01, *** p < 0.001). (D, E) Distribution of the insert length for each DNA type is plotted as a proportion (D) and as box plots (E). The data were generated from merged read pairs for one WT replicate. Color key is as shown in (B). Statistical significance was calculated using Student’s t-test (**** p < 0.0001). (F) Schematic representation of the modifications observed in the amplicon reads for WT parasites. The numbers under ‘WT amplicon’ indicate the length of the amplicon with the inverted triangle denoting the approximate site of the double strand break. The type of DNA inserted is indicated in the key in (A). Not all combinations of insertion patterns observed in the reads are represented. Examples of reads with the kind of modifications shown in this schematic are provided in Table S12. Amplicons with WT sequence (yellow) on both ends represent reads where the pairs could be merged. Numerous unmerged read with similar insertion patterns as shown in this figure exist but are not drawn here.