Genome-wide SNP marker discovery and phylogenetic analysis of mulberry varieties using double-digest restriction site-associated DNA sequencing.

There has been long taxonomic debate on mulberry species (genus Morus) because the classification of mulberry species has relied on morphological characteristics. Although attempts for classifying mulberry species using molecular markers have been performed, phylogenetic relationships among diploid mulberry species remain unclear. In this study, we aim to investigate the genetic relationship between 54 diploid mulberry varieties belonging to seven different Morus species (M. alba, M. indica, M. bombycis, M. acidosa, M. latifolia, M. kagayamae, and M. rotundiloba) and one unspecified Morus species ('Enbu') using genome-wide SNP discovery and phylogenetic analysis via double-digest restriction site-associated DNA sequencing (ddRAD-seq). Genome-wide 2229 homozygous SNPs of 54 mulberry varieties in the eight species were identified by ddRAD-seq. Results of the phylogenetic analysis identified only three clear monophyletic clades in two Japanese native species, M. acidosa and M. kagayamae, which are found on different geographically isolated islands and a Thai species, M. rotundiloba, whereas the other species were non-monophyletic. Varieties of M. bombycis, another Japanese native species, were roughly classified into three groups. Of these, two M. bombycis groups were monophyletic with M. acidosa and M. kagayamae, respectively, while another M. bombycis group was not monophyletic. Varieties of M. indica, an Indian native species, were classified into two different monophyletic clades. Of these, one clade was clearly monophyletic with an indigenous variety in Kenya, 'Enbu', while another clade was monophyletic with M. rotundiloba and one M. latifolia variety. There were no clear monophyletic clades within M. alba and M. latifolia varieties, which could be a result of several hybridization events after their introductions from China to Japan. Our results suggested that it was difficult to clearly classify the hybridized mulberry varieties even with genome-wide DNA markers. In addition to phylogenetic analysis, we also evaluated morphological characteristics of mulberry leaves for each variety. The results of morphological evaluation indicated that leaf tip ratio may correlate to genetic difference among the two M. bombycis groups in monophyletic clades and another M. bombycis group in non-monophyletic clades. These results suggested that leaf tip ratio might be used for evaluating hybridization of M. bombycis varieties. Over all, our results may provide new insights into taxonomic debate of mulberry species.