Production of yellow-flowered gentian plants by genetic engineering of betaxanthin pigments.

Genetic engineering of flower color provides biotechnological products such as blue carnations or roses by accumulating delphinidin-based anthocyanins not naturally existing in these plant species. Betalains are another class of pigments that in plants are only synthesized in the order Caryophyllales. Although they have been engineered in several plant species, especially red-violet betacyanins, the yellow betaxanthins have yet to be engineered in ornamental plants. We attempted to produce yellow-flowered gentians by genetic engineering of betaxanthin pigments. First, white-flowered gentian lines were produced by knocking out the dihydroflavonol 4-reductase (DFR) gene using CRISPR/Cas9-mediated genome editing. Beta vulgaris BvCYP76AD6 and Mirabilis jalapa MjDOD, driven by gentian petal-specific promoters, flavonoid 3',5'-hydroxylase (F3'5'H) and anthocyanin 5,3'-aromatic acyltransferase (AT), respectively, were transformed into the above DFR-knockout white-flowered line; the resultant gentian plants had vivid yellow flowers. Expression analysis and pigment analysis revealed petal-specific expression and accumulation of seven known betaxanthins in their petals to c. 0.06-0.08 µmol g FW-1 . Genetic engineering of vivid yellow-flowered plants can be achieved by combining genome editing and a suitable expression of betaxanthin-biosynthetic genes in ornamental plants.

Efficient double-flowered gentian plant production using the CRISPR/Cas9 system.

Japanese cultivated gentians are highly valued ornamental flowers in Japan, but the flower shape is mostly limited to the single-flower type, unlike other flowers such as roses and carnations. To overcome this limitation, we used the CRISPR/Cas9 genome editing system to increase double-flowered genetic resources in gentians. Our approach targeted an AGAMOUS (AG) floral homeotic gene (AG1), which is responsible for the natural mutation that causes double flowers in gentians. We designed two targets in exon 1 of AG1 for genome editing and found that 9 of 12 herbicide-resistant shoots had biallelic mutations in the target regions of AG1. These nine lines all produced double flowers, with stamens converted into petaloid organs, similar to the natural mutant. We also analyzed the off-target effects of AG2, which is homologous to AG1, and found that such effects occurred in gentian genome editing but with low frequency. Furthermore, we successfully produced transgene-free genome-edited plants (null segregants) by crossing with wild-type pollen. F1 seedlings were subjected to PCR analysis to determine whether foreign DNA sequences, two partial regions of the CaMV35S promoter and Cas9 gene, were present in the genome. As a result, foreign genes were segregated at a 1 : 1 ratio, indicating successful null segregant production. Using PCR analysis, we confirmed that four representative null segregants did not contain transfer DNA. In summary, our study demonstrates that the CRISPR/Cas9 system can efficiently produce double-flowered gentians, and null segregants can also be obtained. These genome-edited plants are valuable genetic resources for future gentian breeding programs.