Probing the genomic limits of de-extinction in the Christmas Island rat

Abstract

Three principal methods are under discussion as possible pathways to "true" de-extinction; i.e., back-breeding, cloning, and genetic engineering.^1,2 Of these, while the latter approach is most likely to apply to the largest number of extinct species, its potential is constrained by the degree to which the extinct species genome can be reconstructed. We explore this question using the extinct Christmas Island rat (Rattus macleari) as a model, an endemic rat species that was driven extinct between 1898 and 1908.^3-5 We first re-sequenced its genome to an average of >60× coverage, then mapped it to the reference genomes of different Rattus species. We then explored how evolutionary divergence from the extant reference genome affected the fraction of the Christmas Island rat genome that could be recovered. Our analyses show that even when the extremely high-quality Norway brown rat (R. norvegicus) is used as a reference, nearly 5% of the genome sequence is unrecoverable, with 1,661 genes recovered at lower than 90% completeness, and 26 completely absent. Furthermore, we find the distribution of regions affected is not random, but for example, if 90% completeness is used as the cutoff, genes related to immune response and olfaction are excessively affected. Ultimately, our approach demonstrates the importance of applying similar analyses to candidates for de-extinction through genome editing in order to provide critical baseline information about how representative the edited form would be of the extinct species.

Keywords: Christmas Island rat; Rattus macleari; ancient DNA; de-extinction; evolutionary divergence; genomic sequencing; molecular dating.

Figure 1

The phylogenetic placement and evolutionary timescale of R. norvegicus, Rattus nitidus, R. macleari, Rattus rattus, and Rattus tanezumi Numbers following the species names indicate the coverage of genomic sequencing data for the corresponding species when mapped to the Norway brown rat reference genome. Related to Table S6.

Figure 2

Numbers of genes found at different coverage levels after mapping Christmas Island rat genomic sequencing data to the Norway brown rat reference genome Coverage levels (on a scale of 0–1) are shown next to the figure. Related to Tables S2 and S3 and Figure S2.

Figure 3

Annotation of genes with unrecoverable regions in of Christmas Island rat genome (A) GO enrichment (q < 0.05) of Christmas Island rat genes obtained at coverage lower than 0.9. Numbers following bars: the number of genes; x axes: −log10(q value); y axes: the GO terms enriched in genes with coverage lower than 0.9. (B) KEGG enrichment (q < 0.05) in genes with coverage lower than 0.9; x axes: rich factor, number of genes with coverage lower than 0.9/total genes in KEGG terms; y axes: the KEGG pathways enriched in genes with coverage lower than 0.9. (C and D) The coverage of genomic sequencing data on major histocompatibility complex (MHC) and vomeronasal receptor (VoR) genes in the Norway brown rat reference genome. Related to Tables S4 and S5.