Combined R2R3-MYB transcription factor mutants reveal the regulatory structure of the Arabidopsis thaliana flavonoid biosynthesis pathway.

Comparative metabolic profiling of new genetic multiple flavonoid r2r3-myb mutants show that different types of R2R3-MYBs activate the early flavonoid biosynthesis genes. A revised model for flavonoid biosynthesis in Arabidopsis thaliana is proposed that integrates the regulatory roles of these R2R3-MYBs across early and late biosynthetic steps. Flavonoids are a large group of specialized plant metabolites. Their biosynthesis is mainly transcriptionally regulated by a sophisticated network of different transcription factors from various families, with R2R3-MYB factors being the main determinant of specific flavonoid class formation. The early biosynthetic steps, leading to the formation of non-visible flavonoids, have been proposed to be regulated by three R2R3-MYBs, PRODUCTION OF FLAVONOL GYLCOSIDE1-3 (PFG1-3), while the later biosynthetic steps leading to the formation of visible anthocyanin and proanthocyanidin pigments are reported to be regulated by four R2R3-MYBs, PRODUCTION OF ANTHOCYANIN PIGMENT1-4 (PAP1-4) and TRANSPARENT TESTA2 (TT2), respectively. Several studies have indicated that this model for the transcriptional regulation of flavonoid biosynthesis may be incomplete. To address this issue, especially regarding the regulation of the early biosynthesis genes by PAP1-4 and TT2, we generated several multiple r2r3-myb mutant lines. We characterized the pfg1-3, pfg1-3 tt2 and pfg1-3 pap1-4 mutants and did comparative metabolite profiling. This revealed that only the pfg1-3 tt2 mutant was deficient in proanthocyanidins and only the pfg1-3 pap1-4 mutant was deficient in anthocyanins. We demonstrate that PAP and TT2 R2R3-MYBs are also capable of activating the early biosynthesis genes required for dihydroflavonol formation. Our results provide evidence that the traditional view of distinct branch-specific R2R3-MYB regulators in flavonoid biosynthesis is overly simplistic. We, therefore, propose a revised model for the transcriptional regulation of flavonoid biosynthesis.