Quantifying the Number of Independent Organelle DNA Insertions in Genome Evolution and Human Health

Abstract

Fragments of organelle genomes are often found as insertions in nuclear DNA. These fragments of mitochondrial DNA (numts) and plastid DNA (nupts) are ubiquitous components of eukaryotic genomes. They are, however, often edited out during the genome assembly process, leading to systematic underestimation of their frequency. Numts and nupts, once inserted, can become further fragmented through subsequent insertion of mobile elements or other recombinational events that disrupt the continuity of the inserted sequence relative to the genuine organelle DNA copy. Because numts and nupts are typically identified through sequence comparison tools such as BLAST, disruption of insertions into smaller fragments can lead to systematic overestimation of numt and nupt frequencies. Accurate identification of numts and nupts is important, however, both for better understanding of their role during evolution, and for monitoring their increasingly evident role in human disease. Human populations are polymorphic for 141 numt loci, five numts are causal to genetic disease, and cancer genomic studies are revealing an abundance of numts associated with tumor progression. Here, we report investigation of salient parameters involved in obtaining accurate estimates of numt and nupt numbers in genome sequence data. Numts and nupts from 44 sequenced eukaryotic genomes reveal lineage-specific differences in the number, relative age and frequency of insertional events as well as lineage-specific dynamics of their postinsertional fragmentation. Our findings outline the main technical parameters influencing accurate identification and frequency estimation of numts in genomic studies pertinent to both evolution and human health.

Keywords: cancer genomics; mitochondria; numts; nupts; organelle insertions.

Fig. 1.

—Number of inferred (A) numts and (B) nupts obtained by different clustering stringencies and concatenation distances. Clustering stringency is shown on the x-axis, different concatenation distances are depicted in colors (blue 50 bp, red 500 bp, yellow 3 kb, and purple 10 kb). Species with at least 400 BLAST hits are shown.

Fig. 2.

—(A) A schematic phylogenetic tree and distribution of BLAST identity scores for permissive counting showing at least 20 inferred insertions with a concatenation distance of 500 bp for (B) numts and for (C) nupts. Distribution of BLAST identity scores for permissive counting showing at least 20 inferred insertions with a concatenation distance of 500 bp. Concatenated numt and nupt scores were calculated as a weighted mean. Scales are shown between 60% and 100% identity and the distribution is shown up to 30%. The sum of the cumulative distribution is one. Some species include columns above 30% as Chlorella (40%), Coccomyxa (40%) and Emiliania (40%) for nupts, and Cyanophora (40%) and Nematostella (70%) for numts.

Fig. 3.

—Difference in percent identity between BLAST hits belongs to the same insertion. (A) For each inferred insertion, the maximum identity score-difference between separated BLAST hits was calculated. (B) Histograms of the difference of % identity for numts and nupts. Data from permissive concatenation with a distance up to 500 bp.