Historical Plant Embryos as Alternative Sources of Ancient DNA for Whole Genome Sequencing

Natural history and agricultural collections contain hundreds of millions of specimens classified in terms of time, space and taxonomy. Since the first success of ancient DNA (aDNA) isolation in the 1980s, these repositories, including herbaria for plants, have been intensively used to support studies in taxonomy, macroevolution and genetic responses to anthropogenic activities. For herbarium specimens, aDNA is usually extracted from leaf samples and is highly fragmented (typically length of 50-100 bp). To optimise the amount of data retrieved and minimise destructive sampling, we isolated DNA from an unconventional plant tissue type-seed embryos. We carried out whole-genome sequencing and compared sequenced DNA quality between embryo and leaf tissue from the same specimen. We evaluated endogenous DNA proportion, median fragment length, damage fraction per site (λ), decay rates (k), nucleotide misincorporations and library complexity for three species: cultivated rice Oryza sativa, wild rice O. rufipogon and wild barley Hordeum spontaneum. In O. sativa, embryos exhibited significantly higher endogenous content and median fragment length than leaves. The superior DNA preservation was likely due to the protective role of the seed husk, which might play an important role in DNA preservation in plants collected in the tropics. By contrast, in temperate H. spontaneum, tissue type had minimal impact on DNA quality. Despite the minuscule size of the embryos, all derived genomic libraries were highly complex, sufficient for deep whole genome sequencing. These results highlight seed embryos as a promising alternative aDNA source for herbarium specimens, economic botany and anthropological museum collections.