Biosynthesis and regulation of flavonoids in buckwheat.

Buckwheat contains an abundance of antioxidants such as polyphenols and is considered a functional food. Among polyphenols, flavonoids have multiple functions in various aspects of plant growth and in flower and leaf colors. Flavonoids have antioxidant properties, and are thought to prevent cancer and cardiovascular disease. Here, we summarize the flavonoids present in various organs and their synthesis in buckwheat. We discuss the use of this information to breed highly functional and high value cultivars.

1. Introduction

Common buckwheat (Fagopyrum esculentum L.) is an annual crop grown widely around the world. Flour, made from the seeds, is used in foods such as noodles and pancakes, and sprouted seeds are used as a vegetable. Buckwheat accumulates several kinds of flavonoids, such as flavonols, flavones, anthocyanins and proanthocyanidins (PAs). Flavonoids are synthesized and accumulate in a variety of tissues in many plant species and can help to protect against various stresses (Winkel-Shirley 2001, 2002). As part of a balanced diet, flavonoids reduce inflammation, enhance immunity, and lower the risks of cardiovascular disease and certain cancers (Santos-Buelga and Scalbert 2000, Yildizogluari ).

Many researchers have identified the types and chemical structures of these flavonoids in buckwheat and have tried to alter the type and concentration by changing the growing conditions (e.g. Kalinova and Vrchotova 2009, Kim , 2013, Kiprovski , Nam , Park , Watanabe , 1998). Improvements in technology, such as next generation sequencing, gene expression analysis, transient assay experiments, and analysis of transgenic plants, have advanced the study of their synthesis and regulation (e.g. Gupta , Li , Matsui ). In particular, the creation of the Buckwheat Genome Data Base (BGDB; Yasui ) has accelerated molecular analyses.

In this review, we describe the different classes of flavonoids present in buckwheat. We also describe enzymes, both known and putative, involved in the synthesis, transport, and regulation of each flavonoid class. Finally, we discuss the use of this information in breeding high value cultivars with improved flavonoid content.

2. Flavonoids in buckwheat: types and distribution

Distribution pattern of flavonoids in buckwheat

Many researchers have investigated the types and contents of flavonoids in buckwheat plants, seeds, flowers, and food products (e.g. Kalinova and Vrchotova 2009, Kim , 2013, Kiprovski , Nam , Park , Watanabe , 1998; Fig. 1). The flavonoids present depend on growth stage, organ, cultivar, growing area and season, and are synthesized in response to biotic and abiotic signals such as environmental conditions (e.g. Kalinova and Vrchotova 2009, Watanabe and Ito 2002).

Fig. 1.

Distribution of flavonoids in buckwheat based on previous reports. This figure is based on Supplementary Figure S1 of Matsui with added information. References: [1] Matsui , [2] Li , [3] Matsui , [4] Kalinova and Vrchotova (2009), [5] Kiprovski , [6] Nam , [7] Watanabe , [8] Watanabe (1998), [9] Watanabe (2007), [10] Kim , [11] Park , [12] Nagatomo , [13] Kim . Plant materials, growing conditions, and methods of measurement were different in each paper.

Chemical characteristics of flavonoids

Basic chemical structure of flavonoid groups

Flavonoids are divided into several groups, including flavones, flavonols, anthocyanins and PAs, by their chemical structure and the degree of saturation and oxidation of rings (Fig. 2A). They are classified as flavans when all positions in the C ring are saturated; as flavones when position 4 is a ketone and positions 2 and 3 are not saturated; and as anthocyanidins when the charge is positive and all rings are aromatic (Fig. 2A). When the C ring has an OH substitution, the suffix becomes -ol; for example, an OH at position 3 of the C ring of flavone creates flavan-3-ol, which is a flavanol (Fig. 2A). Flavonols are further classified by the substitution pattern on the B ring: position 4 and optionally positions 3 and 5 can have an OH: one at position 4 is called kaempferol, two at positions 3 and 4 is named quercetin, and three at positions 3, 4, and 5 is named myricetin (Fig. 2C).

Fig. 2.

Chemical structures of flavonoids. (A) Basic chemical structures. (B) Buckwheat flavones. Apigenin is an aglycone of vitexin and isovitexin; luteolin is an aglycone of orientin and isoorientin. However, both apigenin and luteolins themselves are not contained in buckwheat (the box with the dashed line suggests). (C) Representative flavonols and a flavonol glycoside, rutin. (D) Representative anthocyanidins. Buckwheat has only cyanidin 3-glycosides. (E) Catechins and epicatechins. Buckwheat, like most other plant species, has only (+)-catechin and (–)-epicatechin.

Flavones

Four flavone glycosides have been identified in buckwheat: vitexin (apigenin-8-C-glucoside), isovitexin (apigenin-6-C-glucoside), orientin (luteolin 8-C-glucoside), and isoorientin (luteolin 6-C-glucoside) (Fig. 2B). These flavone C-glycosides accumulate in cotyledons and seeds of common buckwheat (Fig. 1). In other cultivated buckwheat species, such as F. tataricum and F. cymosum, these flavones are not present (Margna , Watanabe and Ito 2002).

Plants synthesize C-glycosides via two known pathways (Brazier-Hicks , Du , Yonekura-Sakakibara and Hanada 2011) in which C-glycosylation is performed either before or after the flavone skeleton is completed (Sasaki ). In buckwheat the former pathway is the major method, but it is unclear whether the latter is also utilized.

Flavonols (flavonol glycosides)

Flavonols and their glycosides, as well as other flavonoids, protect plants against biotic and abiotic stresses, such as UV light. They also have important roles in auxin transport (Besseau , Buer and Muday 2004, Peer , Santelia ) and the formation of functional pollen tubes in maize (Mo ) and tobacco (Ylstra ). Flavonols, such as kaempferol, quercetin, and myricetin, and their glycosides are important for coloring of many flowers and fruits via copigmentation with anthocyanins (Baranac ), (Fig. 2C). Rutin (quercetin 3-glucoside-rhamnoside also called quercetin 3-rutinoside) is the main flavonol glycoside in buckwheat and is present in most organs (Fig. 1). It is thought to be a major allelochemical by which buckwheat inhibits the growth of other plants (Golisz ). Several other flavonol glycosides, such as isoquercetin (quercetin 3-glucoside) and quercetin, are found in buckwheat (Fig. 1).

Anthocyanins

Anthocyanins are glycosides of anthocyanidins. Many kinds of anthocyanins are recognized in plants and appear red, purple, or blue depending on structure and pH. Anthocyanins color many organs of buckwheat red. The part excluding sugar and organic acid bound to anthocyanin is also called “aglycones”. By the difference in the number of OH groups in the B ring, three aglycones are recognized: cyanidin, delphinidin, and pelargonidin (Fig. 2D). Buckwheat contains only cyanidin, in the forms of cyanidin-3-rutinoside (cyanidin-3-glucoside-rhamnoside), cyanidin-3-glucoside, cyanidin-3-galactoside, and cyanidin 3-galactopyranosyl-rhamnoside (Kim , Watanabe 2007). This specificity is related to the substrate preference of dihydroflavonol 4-reductase (DFR) of buckwheat (see 3-2.9; Katsu ).

Anthocyanins in buckwheat accumulate mainly at the base of the stem, which fades in color toward the top, and their composition differs depending on the location in the stem (Eguchi ). Some anthocyanins may be transported from roots or the base of the buckwheat stem to the top, since genes involved in anthocyanin synthesis are expressed in roots (Katsu ), but further study is needed to clarify this.

Flavan-3-ol (flavanols)/PAs

Catechins occur as the (+) isomer and epicatechins as the (–) isomer (Fig. 2E). Buckwheat accumulates monomeric (+)-catechin and (–)-epicatechin, as well as epigallocatechin and oligomeric flavan-3-ols (Figs. 1, 2E). The latter are called PAs or condensed tannins, which form by the polymerization of flavan-3-ols and other flavonoids. They are a major group of flavonoids synthesized via the phenylpropanoid biosynthesis pathway, which has not yet been fully characterized in buckwheat or other plants.

3. Synthesis and regulation of flavonoids in buckwheat

Flavonoid biosynthesis pathway in buckwheat

Flavonoids are synthesized via the flavonoid biosynthesis pathway in several steps which is part of the phenylpropanoid biosynthesis pathway, and is similar among plant species, although enzymes and the genes encoding them can differ. Many enzymes in the buckwheat flavonoid biosynthetic pathway remain unidentified and the flow of components through the pathway is unknown (Fig. 3). With the development of the BGDB, the identification of some genes has been accelerated.

Fig. 3.

Putative flavonoid biosynthesis pathway in buckwheat. Enzymes are indicated in upper-case letters. Identified genes are written under the enzymes. Postulated genes are indicated with [ ]. Dashed arrows indicate uncertain steps. White arrows indicate putative transportation pathways. CHS, chalcone synthase; CHI, chalcone isomerase; F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3′-hydroxylase; F3′5′H, flavonoid 3′5′-hydroxylase; FLS, flavonol synthase; FNS, flavone synthase, GT, flavonol glycosyltransferase; DFR, dihydroflavonol 4-reductase; ANS/LDOX, anthocyanidin synthase/leucoanthocyanidin dioxygenase; UFGT, UDP glucose-flavonoid 3-O-glycosyltransferase; ANR, anthocyanidin reductase; LAR, leucoanthocyanidin reductase.

The flavonoid biosynthetic pathway in Arabidopsis is broadly divided into two phases, early and late (Gonzalez , Pelletier , Quattrocchio ). Early flavonoid biosynthesis involves a major group of flavones and flavonols and is regulated by transcription factors (TFs) of the R2R3 MYB family. Late flavonoid biosynthesis includes anthocyanins and PAs and is regulated mainly by TFs of the R2R3 MYB, basic helix–loop–helix (bHLH), and WD40 repeat (WDR) protein families in many species. This model for the pathway could be a reasonable model for the pathway in buckwheat (Fig. 3). However, there seem to be some differences. In this section, we discuss key genes encoding enzymes and transcription factors in the flavonoid biosynthesis pathway in buckwheat and also discuss the differences to other plant species. We also added information of the results of the search with BGDB in each key enzyme. However, the draft genome of buckwheat in the BGDB still includes gaps between contigs because short reads were used (Yasui ), so confirmation of the sequence of these genes will be required.

Genes of key enzymes in flavonoid biosynthesis pathway

CHS, chalcone synthase

CHS catalyzes the conversion of 4-coumaroyl-CoA and malonyl-CoA to naringenin chalcone. Two full length genes encoding CHS are registered in the database of the International Nucleotide Sequence Database Collaboration (INSDC; DDBJ/EMBL-EBI/NCBI) (HM149787 and GU172166), but it is not yet known whether these genes occur at the same locus or not. A key word search for “chalcone synthase” in the BGDB (v. 1.1) returned nine scaffolds.

CHI, chalcone isomerase

CHI catalyzes the conversion from chalcones to flavanones. A locus (HM149788) encoding buckwheat CHI is registered in the INSDC database. A search for “chalcone isomerase” in the BGDB identified six scaffolds with full or partial sequence similarity to CHI.

FNS, flavone synthase

FNS catalyzes the conversion from flavanones to flavones. No buckwheat FNS gene is registered in the INSDC database. Two types of FNS occur in plants: FNS I and FNS II; FNSI class belongs to the superfamily of soluble Fe2+/2-oxoglutarate-dependent dioxygenases (2-ODDs) and FNSII correspond to oxygen- and NADPH-dependent cytochrome P450 (CYPs) membrane-bound monooxygenases (Jiang et al. 2016, Martens and Mithofer 2005). Flavone formation in many plant species is catalyzed by FNS II (Martens and Mithofer 2005) but recently FNSI enzymes have been characterized from maize and Arabidopsis (Jiang et al. 2016). A key word search for “flavone synthase” with the BGDB returned two scaffolds with full or partial sequence similarity to FNS II. Flavones of buckwheat are all C-glycosylated: orientin, isoorientin, vitexin, and isovitexin (Nagatomo , Watanabe and Ito 2002). In buckwheat, these flavone glycosides are C-glycosylated before the skeleton of apigenin or luteolin flavones is completed (Nagatomo ) indicating FNS is not required for the synthesis of flavones, so whether these genes encode actual FNS enzymes involved in flavone synthesis will need to be confirmed.

CGT, UDP-glucose: 2-hydroxyflavanone C-glucosyltransferase: FeCGTa and FeCGTb

C-glucosyltransferase (CGT) of buckwheat was first partially purified by Kerscher and Franz (1988). FeCGTa (UGT708C1, AB909375) and FeCGTb (UGT708C2, AB909376) were identified by Nagatomo and thought to be allelic due to the high similarity of deduced amino acid sequence and observation of hybrid genes arising through a possible recombination event between FeCGTa and FeCGTb in the cDNA library (Nagatomo ).

Buckwheat synthesizes vitexin (apigenin-8-C-glucoside), isovitexin (apigenin-6-C-glucoside), orientin (luteolin 8-C-glucoside), and isoorientin (luteolin 6-C-glucoside). FeCGTs do not perform C-glycosylation for apigenin and luteorin, but 2-hydroxyflavanones and 2-phenyl-2′,4′,6′-trihydroxyacetophenone (Nagatomo ). Thus, flavone glycoside is C-glycosylated before the flavone skeleton is completed (Nagatomo ). Nagatomo reported differences in molecular mass, peak activity, and sugar donor between the C-glucosyltransferase reported by Kerscher and Franz (1988) and the FeCGTs that they identified. Some species, such as Gentiana, have a CGT which can catalyze direct C-glycosylation of the flavone skeleton with a C-linkage (Sasaki ). The presence of such CGTs in buckwheat will need to be investigated.

F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3′-hydroxylase and F3′5′H, flavonoid 3′5′-hydroxylase

Flavanone 3-hydroxylase (HM149789) catalyzes the conversion of flavanones to dihydroflavonols. F3′H (HM149790, partial) catalyzes hydroxylation at the 3′ position of the B-ring of the flavonoid while F3′5′H catalyzes hydroxylation at the 3′ and 5′ positions of the B-ring of flavonoids. No buckwheat F3′5′H gene is registered in the INSDC database. Key word searches for these genes in the BGDB identified sixteen, two and three scaffolds with full or partial sequences respectively.

FLS, flavonol synthase: FeFLS FeFLS1 and FeFLS2

Flavonol synthase (FLS) converts dihydroflavonols to flavonol aglycones (kaempferol, quercetin, and myricetin) (Fig. 3). Three buckwheat loci encoding FLS (JF274261), are registered in the INSDC database. The full-length sequences of FeFLS1 (HM357804), and FeFLS2 (HM357805) are still to be identified (Li ). FeFLS1 is expressed mainly in buds and flowers whereas FeFLS2 is expressed in roots, leaves, and stems of seedlings, and in seeds (Li , Matsui ). FeFLS is strongly expressed in flowers (Matsui ).

DFR, dihydroflavonol 4-reductase: FeDFR (ACZ48698), FeDFR1a (BBC48377), and FeDFR2 (BBC48378)

DFR is the key enzyme committed to the formation of anthocyanins and PAs in the flavonoid biosynthesis pathway (Fig. 3). It catalyzes the NADPH-dependent reduction of dihydroflavonols to leucoanthocyanidins, including dihydrokaempferol, dihydroquercetin, and dihydromyricetin to leucopelargonidin, leucocyanidin, and leucodelphinidin, respectively, in many species (Fig. 1). Although DFR proteins in many plants can catalyze these three substrates, some species cannot produce some leucoanthocyanidins, so some forms of DFR may have substrate preferences (Forkmann and Ruhnau 1987, Gerats , Johnson , 2001). The relationships between substrate preference and amino acids in the region responsible for substrate specificity have been investigated in several plant species (Johnson , Leonard , Shimada , Xie ).

In buckwheat, anthocyanins accumulate in several organs, notably stems and petioles, coloring them red. On the other hand, PAs accumulate in most organs. In buckwheat seeds, cyanidin and pelargonidin were identified by acid treatment of PAs (Durkee 1977, Ölschläger ). Furthermore, (–)-epigallocatechin, which is synthesized by anthocyanidin reductase from delphinidin, was detected in stems, leaves, and flowers (Li ). However, only cyanidin is recognized as an aglycone of buckwheat anthocyanins (Kim , Troyer 1964).

Two gene loci, FeDFR1 and FeDFR2, and one allele of FeDFR1, FeDFR1a, have been reported (Katsu ). FeDFR2 has sequence similarity to FeDFR1a but with a different exon–intron structure. The two loci are genetically linked but not allelic. Unlike common DFR proteins in other plant species, FeDFR2 has a valine instead of the typical asparagine at the position 3 and an extra glycine between position 6 and 7 in the region that determines substrate specificity (Johnson ), and has less activity using dihydrokaempferol as a substrate than FeDFR1a, which has an asparagine at the position 3 (Katsu ). The substrate preferences of buckwheat DFR would be determined not only by the amino acid at the third position of the 26-amino-acid region identified to determine substrate specificity, but also by its neighboring residues that are involved in the formation of the substrate-binding pocket (Katsu ).

ANS, anthocyanidin synthase/leucoanthocyanidin dioxygenase

ANS catalyzes the conversion of leucoanthocyanidins to anthocyanidins. Buckwheat ANS (ADT63066) was first reported by Li and is the only the locus encoding buckwheat ANS that is registered in the INSDC database. A search for “anthocyanidin synthase” in the BGDB also returned one scaffold validating the hypothesis of a single gene.

ANR, anthocyanidin reductase, FeANR1

ANR catalyzes the conversion of anthocyanidins to (–)-epicatechins. A buckwheat ANR gene (BAV56574) was identified by degenerate PCR and RACE (Matsui ). Only the locus encoding buckwheat ANR is registered in the INSDC database and a search for “anthocyanidin reductase” in the BGDB identified one scaffold. Matsui reported that the expression of FeANR1 was high in roots and cotyledons and resembled the pattern of PA accumulation.

LAR, leucoanthocyanidin reductase: FeLAR1 FeLAR2 FeLAR3

LAR can convert leucocyanidins to (+)-catechins. In buckwheat, three loci, FeLAR1–FeLAR3, (BAV56575, BAV56576, BAV56577), have been characterized (Matsui ). These genes were identified by sequence homology with other plant species and searches in the BGDB. FeLAR1–FeLAR3 are expressed in most organs but have different expression patterns in different organs at the flowering stage and in seeds at different maturation stages. The expression of FeLAR1 is high in cotyledons and low in roots and hypocotyls, and FeLAR2 and FeLAR3 are expressed mainly in roots. The unique expression patterns of FeLAR1–FeLAR3 suggest that all three genes play a role in PA synthesis in a distinct spatiotemporal manner (Matsui ).

Regulation of flavonoid biosynthesis by transcription factors

The expression of genes encoding enzymes of the flavonoid biosynthesis pathway is regulated by many TFs, several of which have been well studied in various plant species (e.g. Bogs , Czemmel , Kobayashi , Lloyd , Mehrtens , Takos , Walker ). Most of them belong to the two largest families of TFs in plants: the R2R3 MYB family and the bHLH family. As yet there is no report of a bHLH related to flavonoid biosynthesis in buckwheat.

MYB transcription factors

MYB proteins are believed to be key factors because of the specific expression patterns of their encoding genes (Stracke ). The MBW complex (MYB factor– bHLH factor–WD40 protein) is required for regulation of the expression of genes encoding enzymes in the late flavonoid biosynthesis pathway (Fig. 3) in many species (e.g. Baudry , Gonzalez , Quattrocchio , Walker ), but not those in the early pathway (flavonol synthesis) in Arabidopsis (Mehrtens , Stracke , Zimmermann ).

Yasui identified 274 potential MYB TFs in the BGDB using “MYB” in a key word search. After exclusion of partial sequences, pseudogenes, and genes that did not contain fully conserved R2R3 regions, 71 putative R2R3 MYB TFs remained. Six putative R2R3 MYBs were assigned within known functional groups consisting of representatives from other plant species (Yasui ; Fig. 4). Matsui reported a flavonol-specific MYB TF, FeMYBF1 (see 3-3.2), but a gene encoding FeMYBF1 was not found among the 71 candidate genes. The buckwheat genome has 16 chromosomes (2n = 2x = 16), but the draft genome in the BGDB (FES_r1.0) includes many gaps between contigs due to the use of short reads (Yasui ), and still includes 387 594 scaffolds. Improvement of BGDB and functional studies would be needed to identify more MYB TFs in buckwheat.

Fig. 4.

A phylogenetic tree for six putative R2R3 MYB transcription factors with representatives from other plants. Bootstrap values (500 replicates) are shown next to the branches. The scale bar corresponds to 0.05 substitutions per site. The root was assumed at the midpoint of the tree. This figure was derived from Supplementary Figure S2 of Yasui .

Flavonol-specific R2R3 MYB transcription factor in buckwheat

As rutin is a major flavonoid in buckwheat, it will be important to clarify the mechanisms of rutin synthesis and regulation. FeMYBF1, an R2R3 MYB TF, was identified as a regulator of flavonol synthesis (Matsui ). It was isolated using degenerate primers and its full-length sequence was determined by RACE and genome walking. By phylogenetic analysis with the deduced amino acid sequences of the R2R3 MYB domains including Arabidopsis regulators of flavonol synthesis, AtMYB11, AtMYB12 and AtMYB111, FeMYBF1 was inferred to be a MYB protein with the ability to regulate flavonol synthesis (Matsui ).

The function of FeMYBF1 was demonstrated by introducing FeMYBF1 under the control of the AtMYB111 promoter into the flavonol-deficient Arabidopsis triple mutant myb11 myb12 myb111 where it was shown to complement these mutations. Matsui investigated whether FeMYBF1 can complement PA synthesis by introducing FeMYBF1 under the control of the AtTT2 promoter into a PA-deficient Arabidopsis tt2 mutant, but the failure to restore PA synthesis suggests that FeMYBF1 does not regulate this process.

The WDR protein complex regulates late flavonoid biosynthesis in many species. But FeMYBF1 activated the promoter of one of the DFRs committed to anthocyanin and PA in the flavonoid biosynthesis pathway without the WDR complex (Matsui ). Maize DFR, which functions in the synthesis of reddish or brownish water-insoluble pigments called phlobaphenes, is activated by a MYB TF, P, alone (Grotewold , Pooma ). In buckwheat, FeDFR1a could probably be activated by FeMYBF1 and may contributes to flavonol synthesis through an alternate pathway, and FeDFR2 could be activated by different TFs that may participate in the formation of the MBW complex and contribute to anthocyanin synthesis (Matsui ). Furthermore, FeMYBF1 activated the promoter of ANS/LDOX, which is involved in catalyzing the conversion of leucoanthocyanidins to anthocyanidins. Arabidopsis LDOX has FLS-like side activity (Stracke ), so buckwheat LDOX may contribute to the synthesis of flavonols in addition to anthocyanidins (Matsui ).

4. Transport of flavonoids

Transport of flavonoids to vacuole

Several mechanisms for flavonoid transport have been proposed and demonstrated, such as vesicle trafficking, involvement of membrane transporters, and glutathione S-transferase (GST)-mediated transport (Zhao ). In buckwheat, a gene related to transport of anthocyanins has been reported (Matsui ).

The basal stem color in wild-type buckwheat is red owing to the accumulation of anthocyanins. Matsui reported a green-stem mutant line that cannot accumulate anthocyanins but does synthesise rutin and PAs. The glucosyltransferase (GT) activity in the green stem was similar to that in red-stem plants, indicating that the green-stem phenotype is not caused by non-functional GT, but instead seems to be caused by dysfunction of the system that transports anthocyanins to the vacuole.

In phylogenetic analysis of the 56 buckwheat candidate GST genes from the BGDB with six GST genes known to play a role in anthocyanin accumulation, five of the genes (the exception was Bz2) were classified together with a candidate gene, Fes_sc0002828.1.g000005.aua.1 (named FeGST1) (Matsui ). FeGST1 transcripts were present in the stem of wild-type cultivars but not in that of the green-stem lines. Interestingly, the genomic region around the FeGST1 locus was deleted in the mutant, indicating that the null allele of FeGST1 is caused by deletion of the FeGST1 locus. Linkage analysis with a self-compatible line showed that the green-stem locus is homozygous for the deletion of FeGST1 (Matsui ).

Movement of flavonoids between organs

Reports that the expression of genes related to flavonoid synthesis in buckwheat is not correlated with the amount of flavonoids (Li , Matsui ) suggest the involvement of other, non-cell-autonomous mechanisms for the accumulation of flavonoids, such as transport from other tissues. Buer reported that exogenously applied naringenin, dihydrokaempferol, and dihydroquercetin were taken up at the root tip, midroot, or cotyledons and traveled long distances via cell-to-cell movement to distal tissues. However, quercetin and kaempferol were taken up only at the root tip and were not transported. Some flavonoids are released from roots (Hungria , Pueppke , Yoo ), but there are no reports that flavonols move between cells and travel long distances. Although buckwheat uses rutin as an allelochemical (Golisz , 2008), it is still unclear whether rutin can be directly released from the roots.

One possible mechanism buckwheat might use to transport flavonoids long distances via cell-to-cell movement relies on aluminum. Unlike other crops, buckwheat is resistant to aluminum: it takes up aluminum in the roots and transfers it to leaves and stems where it accumulates in vacuoles (Ma ). Aluminum can bind with flavonoids, including flavonols, altering flower color (Tanikawa ). Because there is no report of the relationship between the movement of flavonols and aluminum, further study is desirable.

5. Breeding for high-value buckwheat cultivars using information on flavonoid biosynthesis pathway and key genes

Development of non-discoloring line by stopping biosynthesis of PAs

PAs have important roles in the mouth feel and astringency of many fruits and in the quality of wine and tea (Aron and Kennedy 2008). They can also cause undesirable colloidal haze in beer (Gramshaw 1970) and discoloration in cooked barley grains (Tonooka ). Buckwheat kernels also turn brown by oxidation of PAs after the hull is shed.

PA-free mutants of barley have been detected (von Wettstein ), and a PA-free barley cultivar has been developed from a PA-free line recessive-homozygous for ant28 (Himi , Tonooka ). To produce a PA-free buckwheat cultivar would be valuable, because Japanese consumers prefer pale buckwheat.

The first step toward this goal is to identify genes related to PA synthesis and regulation in buckwheat. Many genes related to PA synthesis have been reported, but not all can be used for this purpose: because consumers expect to get functional substances, especially rutin, in buckwheat, it will be necessary to breed lines that can synthesize high levels of rutin but not PAs. To produce such lines, genes for DFRs and TFs which regulate PA synthesis would be candidate targets for manipulating flavonoid composition. Two genes for DFR have been identified and could be used in breeding if mutant versions can be discovered. On the other hand, the TFs that regulate PA synthesis in buckwheat are not yet known. Yasui reported a gene which was clustered with known regulators of PAs, such as the grapevine VvMYBPA1 (Fig. 4), but its function will need to be determined if it is to be used as a target gene.

Finding lines with nonfunctional genes caused by mutations would be the next step. The Targeting Induced Local Lesions In Genomes (TILLING) method, which is based on reverse genetics, is a powerful method of producing lines with mutated genes of interest. Using TILLING, valuable lines such as waxy wheat have been produced (Slade ). But genes which govern traits of interest must first be found.

Development of a high-antioxidant line by increasing flavonoids

Matsui demonstrated the ability of FeMYBF1 to regulate flavonol synthesis in Arabidopsis, and to efficiently upregulate it in transgenic Arabidopsis plants. By altering the flux through the flavonoid biosynthesis pathway, it might be possible to produce cultivars with higher level of flavonoids, especially rutin. Unlike Tartary buckwheat, common buckwheat has a low level of rutinosidase, which catalyzes rutin to quercetin (Suzuki ), so the bitterness which is thought to be due to quercetin would not be present in common buckwheat with its high rutin content.

However, FeMYBF1 may regulate genes in other parts of the phenylpropanoid biosynthetic pathway (Matsui ), because Arabidopsis MYB TFs AtMYB11 and AtMYB12, which regulate flavonol synthesis, also regulate the expression of genes related to caffeoylquinic acid in tomato (Li , Luo ) and tobacco (Li ). To use FeMYBF1 in breeding, further study would be needed to clarify whether it can regulate the expression of other genes involved in other metabolic pathways.

FeMYBF1 could be used to increase flavonoid content and to improve flavonoid composition by using plant transformation techniques, including genome editing or a cisgenic approach, although it is still difficult to transform buckwheat. Molecular techniques such as marker-assisted selection and TILLING combined with conventional breeding would also accelerate the breeding of high-antioxidant lines by increasing the content of rutin in buckwheat.

Author Contribution Statement

K.M. and M.W. wrote the manuscript. Both authors have read and approved the final version of the manuscript.