Christoph Sperisen, Willy Tinner, Sarah Schmid, Rémi Genevest, Tomasz Suchan, Nadir Alvarez, Erika Gobet
Over the last decade, protocols aimed at reproducibly sequencing reduced-genome subsets in non-model organisms have been widely developed. Their use is, however, limited to DNA of relatively high molecular weight. During the last year, several methods exploiting hybridization capture using probes based on RAD-sequencing loci have circumvented this limitation and opened avenues to the study of samples characterized by degraded DNA, such as historical specimens.
Here, we present a major update to those methods, namely hybridization capture from RAD-derived probes obtained from a reduced eXome template (hyRAD-X), a technique applying RAD sequencing to messenger RNA from one or few fresh specimens to elaborate bench-top produced probes, that is, a reduced representation of the exome, further used to capture homologous DNA from a samples set. In contrast to previous hybridization capture methods, the reference catalogue on which reads are aligned does not rely on de novo assembly of anonymous RAD-sequencing loci, but on an assembled transcriptome obtained from RNAseq data, thus increasing the accuracy of loci definition and single-nucleotide polymorphisms (SNP) call, and targeting, specifically, expressed genes. Finally, the capture step of hyRAD-X relies on RNA probes, increasing stringency of hybridization, making it well suited for low-content DNA samples. As a proof of concept, we applied hyRAD-X to subfossil needles from the coniferous tree Abies alba, collected in lake sediments (Origlio, Switzerland) and dating back from 7200 to 5800 years before present (bp). More specifically, we investigated genetic variation before, during and after an anthropogenic perturbation that caused an abrupt decrease in A. alba population size, 6500–6200 years bp.
HyRAD-X produced a matrix encompassing 524 exome-derived SNPs. Despite a lower observed heterozygosity was found during the 6500–6200 years bp time slice, genetic composition was nearly identical before and after the perturbation, indicating that re-expansion of the population after the decline was most likely driven by local specimens.
To the best of our knowledge, this is the first time a population genomic study incorporating ancient DNA samples of tree subfossils is conducted at a moderate cost using reproducible exome-reduced complexity.