Proanthocyanidins (PAs), also called condensed tannins, can protect plants against herbivores and are important quality components of many fruits. Two enzymes, leucoanthocyanidin reductase (LAR) and anthocyanidin reductase (ANR), can produce the flavan-3-ol monomers required for formation of PA polymers. We isolated and functionally characterized genes encoding both enzymes from grapevine (Vitis vinifera L. cv Shiraz). ANR was encoded by a single gene, but we found two highly related genes encoding LAR. We measured PA content and expression of genes encoding ANR, LAR, and leucoanthocyanidin dioxygenase in grape berries during development and in grapevine leaves, which accumulated PA throughout leaf expansion. Grape flowers had high levels of PA, and accumulation continued in skin and seeds from fruit set until the onset of ripening. VvANR was expressed throughout early flower and berry development, with expression increasing after fertilization. It was expressed in berry skin and seeds until the onset of ripening, and in expanding leaves. The genes encoding LAR were expressed in developing fruit, particularly in seeds, but had low expression in leaves. The two LAR genes had different patterns of expression in skin and seeds. During grape ripening, PA levels decreased in both skin and seeds, and expression of genes encoding ANR and LAR were no longer detected. The results indicate that PA accumulation occurs early in grape development and is completed when ripening starts. Both ANR and LAR contribute to PA synthesis in fruit, and the tissue and temporal-specific regulation of the genes encoding ANR and LAR determines PA accumulation and composition during grape berry development.

(12)(13)(14)(15). The core PRC2 complex is Ϸ600 kDa, and it can be associated with additional proteins including Polycomb-like (PCL) and the histone deacetylase RPD3 (16) in a complex of 1,000 kDa. The E(Z) protein has histone 3 lysine 27 (H3K27) methyltransferase activity; addition of this mark of inactive chromatin is thought to be the basis of the repression of gene expression by PRC2 (17). Arabidopsis has homologues of PRC2 proteins that are required for the regulation of various devel- VIN3 is a member of a plant-specific protein family with plant homeodomain (PHD) and fibronectin 3 (FNIII) domains (7). VIN3 protein binds to regions of the promoter and first intron of FLC. Unlike the constitutively expressed VRN2 mRNA, the VIN3 mRNA is present at very low abundance during growth at warm temperatures, with expression increasing progressively during a vernalization treatment and returning to prevernalized levels when the plant is returned to normal temperatures (7). This cold-driven accumulation of VIN3 mRNA may be part of a mechanism to time the duration of vernalization and ensure that short cold periods do not promote flowering.The repression of FLC expression after vernalization is accompanied by modifications to histones associated with the FLC locus. In nonvernalized plants, FLC chromatin has high histone H3 acetylation (H3Ac) and H3K4 trimethylation (me3), marks of active chromatin but low levels of the inactive marks H3K9me2 and H3K27me2. After vernalization, H3Ac and H3K4me3 are reduced and H3K9me2 and H3K27me2 are increased (7,21,22). These changes suggest that the formation of a repressed chromatin state at FLC after vernalization is the basis of the epigenetic regulation of FLC. Loss of VIN3 function prevents loss of H3Ac and methylation of H3K9 and H3K27 in vernalized plants. In vrn2 mutants, vernalization gives a transient loss of H3Ac but there is no methylation of H3K9 or H3K27 after vernalization (7). These data suggest that VIN3 may recruit a histone deacetylase to FLC and that VRN2 acts as part of a PRC2-like complex to methylate histone H3 to epigenetically repress FLC expression in vernalized plants.In this paper, we use epitope-tagged proteins to show that VRN2 is associated with the polycomb group protein homologues FIE, SWINGER (SWN; also known as EZA1; ref. 23), and CURLY LEAF (CLF) in a PRC2-like complex and that these proteins are required for the repression of FLC by vernalization. We also show that VRN2 and VIN3 can be part of the same protein complex, suggesting a physical link between these two components of the vernalization response mechanism.

A protocol using HCl‐acidified 4‐dimethylaminocinnamaldehyde (DMACA) was developed for screening proanthocyanidins (PA) in plants. Individual PA‐containing cells could be observed following staining of decolourised leaves with DMACA. The quantitative assay using DMACA‐HCl to detect soluble PA and butanol–HCl to detect insoluble PA could detect PA content lower than 0·4 mg g−1 dry matter (DM). The DMACA–HCl protocol is recommended for PA detection in low‐PA plants because of its higher sensitivity than vanillin–acid protocols. Twenty‐two forage legume species, most of which had been reported free of leaf PA but bloat‐safe, were re‐examined using the DMACA–HCl protocol. PA‐positive cells were observed in leaves of 15 species. Measurable amounts of PA were detected in leaves of seven species, with the PA content ranging from 0·6 to 11 mg g−1 DM. White clover (Trifolium repensL), red clover (T pratenseL) and lucerne (Medicago sativaL) were also found to have PA‐positive leaf trichomes. By analysis of these data and the literature, the threshold PA content for bloat safety in forage legumes was estimated to be 1–5 mg PA g−1 DM.