NewtCap: An Efficient Target Capture Approach to Boost Genomic Studies in Salamandridae (True Salamanders and Newts)

Abstract

Salamanders have large and complex genomes, hampering whole genome sequencing. However, reduced representation sequencing provides a feasible alternative to obtain genome-wide data. We present NewtCap: a sequence capture bait set that targets c. 7 k coding regions across the genomes of all true salamanders and newts (the family Salamandridae, also known as "salamandrids"). We test the efficacy of NewtCap, originally designed for the Eurasian Triturus newts, in 30 species, belonging to 17 different genera that cover all main Salamandridae lineages. We also test NewtCap in two other salamander families. We discover that NewtCap performs well across all Salamandridae lineages (but not in the salamander families Ambystomatidae and Hynobiidae). As expected, the amount of genetic divergence from the genus Triturus correlates negatively to capture efficacy and mapping success. However, this does not impede our downstream analyses. We showcase the potential of NewtCap in the contexts of; (1) phylogenomics, by reconstructing the phylogeny of Salamandridae, (2) phylogeography, by sequencing the four closely related species comprising the genus Taricha, (3) hybrid zone analysis, by genotyping two Lissotriton species and different classes of interspecific hybrids, and (4) conservation genetics, by comparing Triturus ivanbureschi samples from several wild populations and one captive-bred population. Overall, NewtCap has the potential to boost straightforward, reproducible, and affordable genomic studies, tackling both fundamental and applied research questions across salamandrids.

Keywords: Caudata; Urodela; exon capture; high throughput sequencing; hyb‐seq; target enrichment.

FIGURE 1

NewtCap‐based phylogeny of the Salamandridae family. The phylogeny is based on Maximum Likelihood inference of concatenated data of 204,600 informative SNPs using RAxML. Overall layout and clade labels conform to a previous transcriptome‐based phylogeny (Rancilhac et al. 2021). The tree is rooted on the branch separating the newts and the clade containing the true salamanders and Salamandrina (see also Figure S1 for the same tree, but with original labels, and Figure S2 for the additional, extended tree, including Ambystoma, Paradactylodon and Mertensiella, that confirms the root position adopted here). All nodes have a bootstrap support of 100%.

FIGURE 2

A Taricha phylogeny obtained with NewtCap‐derived data. The phylogeny is based on Maximum Likelihood inference of concatenated data of 9,730 informative SNPs using RAxML. Notophthalmus is used to root the tree. All nodes have a bootstrap support of 100% (not shown).

FIGURE 3

Triangle plot of different Lissotriton hybrid classes based on NewtCap‐derived data. The plot, based on 666 informative SNPs, shows the relationship between the hybrid index (the fraction of the alleles per individual that derived from each of the two parental species, also known as the ancestry) and the interclass heterozygosity (the fraction of the alleles per individual that is heterozygous for alleles from both parental species). The L. vulgaris individuals are in the bottom left corner, the L. montandoni individuals in the bottom right corner, and the F1 hybrid offspring in the top corner. The F2 and Bx (‘backcross’) hybrids are placed inside the triangle, with two Bx samples almost fully overlapping (marked with *).

FIGURE 4

Genetic differentiation between wild and captive Triturus ivanbureschi populations based on NewtCap‐derived data. (A) The wild population localities (details in Table S1; four postglacial populations are represented by dark blue, light blue, yellow, and pink colors in Bulgaria, Greece and Turkey, and three populations from the glacial refugial area are represented by red, orange and green colors in Turkey). (B) A plot of the first versus the second Principal Component (PC) places the captive individuals closest to a population from just west of Istanbul. (C) The dendrogram produced by the HCA analysis, showing the Individual Dissimilarity as well as the Coancestry Coefficient, again shows that captive samples cluster with a population just west of Istanbul.