Medicago ruthenica, a wild legume forage widely distributed in the Eurasian steppe, demonstrates high genetic and phenotypic variation. M. ruthenica with a purely yellow flower (YFM), differing from the general phenotype of M. ruthenica with a purple flower (PFM), was recently discovered. The similar characteristics of YFM with Medicago falcata have led to conflicting opinions on its taxonomy using traditional morphological methods. The lack of chemotaxonomy information about M. ruthenica species and the unclear flower coloration mechanisms have hampered their study. Here, we investigated M. ruthenica using targeted metabolomics based on the chemotaxonomy method and elaborated the floral coloration mechanisms using transcriptomics. The identified flavonoids were the same types, but there were different contents in YFM and PFM, especially the contents of cyanidin-3-O-glucoside (C3G), an anthocyanin that causes the purple-reddish color of flowers. The over-accumulation of C3G in PFM was 1,770 times more than YFM. Nineteen anthocyanin-related genes were downregulated in YFM compared with their expression in PFM. Thus, YFM could be defined as a variety of M. ruthenica rather than a different species. The loss of purple flower coloration in YFM was attributed to the downregulation of these genes, resulting in reduced C3G accumulation. The taxonomic characteristics and molecular and physiological characteristics of this species will contribute to further research on other species with similar external morphologies.

Genome editing technology has revolutionized plant genetic breeding. However, Significant variability in editing activity has been observed across different genome editing systems and target sites, highlighting the importance of developing efficient evaluation systems for assessing genome editing efficiency in plants. In this study, we developed a simple, rapid, and efficient system based on hairy root transformation to evaluate somatic genome editing efficiency in plants. This system is easy to implement, does not require sterile conditions, and enables visual identification of transgenic hairy roots within two weeks. We first validated the system using the CRISPR/Cas9 genome editing platform, confirming its effectiveness. Subsequently, we applied this system to assess the somatic editing activity of the recently identified ISAam1 TnpB nuclease, which show considerable promise for plant genome editing applications. Furthermore, through protein engineering, we identified two variants, ISAam1(N3Y) and ISAam1(T296R), which exhibited a 5.1-fold and 4.4-fold enhancement in somatic editing efficiency, respectively. These findings demonstrate that the developed method provides an effective tool for optimizing genome editing system and screening potential target sites in plant genomes.