2. Structural Protein Genes of the Maize Flavonoid Pathway
2.1. Phenylpropanoid Pathway
The first enzymatic steps in the flavonoid pathway are from three genes of the phenylpropanoid pathway (
Table 1). These three enzymes direct the transformation of phenylalanine to coumaroyl-CoA. Those genes are
ZmPAL (phenylalanine ammonium lyase, multiples genes, EC 4.3.1.24) [
16],
ZmC4H (cinnamic acid 4-hydroxylase, Zm00001d009858, EC 1.14.14.91) [
16,
17], and
Zm4CL (4-coumarate CoA ligase,
bm5, EC 6.2.1.12) [
18]. The three genes share a similar expression profile of downstream genes in the flavonoid pathway in anthocyanin-pigmented tissues [
19,
20]. Recent analyses have demonstrated multiple gene families in flavonoid biosynthesis with a tissue-specific expression. In addition, some genes such as
Zm4CL codify various isoforms, each of which has specific functions [
21]. The research on these genes focuses on their roles in lignin biosynthesis [
18]. For example, under sugarcane mosaic virus (SCMV) infection,
ZmPAL and
ZmC4H genes are upregulated, generating the substrate for lignin production [
16]. Meanwhile, studies on a brown
midrib5 maize line demonstrated that a
Zm4CL mutant was responsible for defective lignin biosynthesis [
22]. There is a correlation between anthocyanins and lignin where the fungi
Ustilago maydis activates the anthocyanin but reduces the lignin biosynthesis, thus facilitating its invasion into the maize seed [
23].
Table 1. Summary of genes involved in the early steps of the maize flavonoid pathway.
2.2. Early Biosynthetic Genes of Flavonoids
2.2.1. Chalcone Synthase (ZmCHS, c2, EC 2.3.1.74)
The first crucial step in flavonoid biosynthesis (
Figure 2) is the production of the naringenin chalcone (C6-C3-C6) from the condensation of three molecules of malonyl-CoA (3 × C2) using a 4-coumaroyl-CoA (C6-C3) as substrate [
31]. This gene is also known as polyketide synthase (PKS) type III. The chalcone synthase (CHS) works similarly to other PKS enzymes from the mevalonate/acetate pathway [
4]. The reaction extends the aliphatic chain from the coumaroyl-CoA three times using two carbon units from a malonyl-CoA. Then, an intramolecular Claisen condensation occurs to form the second aromatic ring.
Figure 2. Early genes in the flavonoid pathway. The flavonoid pathway begins with the transformation of phenylalanine to coumaroyl-CoA. The last steps end with the intravacuolar accumulation of acylated anthocyanins. The genes responsible for supplying the coumaroyl-CoA into the flavonoid pathway are phenylalanine ammonium lyase (
ZmPAL, EC 4.3.1.24), cinnamic acid 4-hydroxylase (
ZmC4H, EC 1.14.14.91), and 4-coumarate CoA ligase (
Zm4CL,
bm5, EC 6.2.1.12). The flavonoid genes are divided into early biosynthetic genes (
EBGs) and late biosynthetic genes (
LBGs). EBGs comprise four genes: chalcone synthase (
ZmCHS,
c2, EC 2.3.1.74), chalcone isomerase (
ZmCHI,
chi1, EC 5.5.1.6), flavonoid 3-dioxygenase (
ZmF3H,
fht1, EC 1.14.11.9), and flavonoid 3′-monooxygenase (
ZmF3′
H,
pr1, EC 1.14.14.82). References: [
30,
32].
Genome-wide analysis revealed up to 15
ZmCHS genes in the maize genome (Han et al., 2016). However, the members more consistently studied are the duplicated
c2 (
ZmCHS01) and
whip1 (
ZmCHS02) genes (
Table 1) [
24]. Multiple tissues, including tassels, ear husks, and the aleurone layer of endosperm at different developmental stages, express the genes
c2 and
whip1 [
33,
34]. Indeed, functional alleles for genes
c2 and
whip1 are vital to increasing the biosynthesis of any flavonoids downstream, such as apigenin and tricin, essential for lignin formation, and C-glycosyl flavones [
25,
35]. Meanwhile, members of the chalcone synthase family, such as
ZmCHS013 and
ZmCHS014, compared to
ZmCHS01, had a lower expression in most tissues and different responses under the stimuli of salicylic acid [
36].
2.2.2. Chalcone Isomerase (ZmCHI, chi1, EC 5.5.1.6)
This enzyme catalyzes an intramolecular Michael-type addition from the chalcone 2-
O to its α,β-unsaturated carbonyl (
Figure 2). The final product is the typical phenyl-chromanone or flavanone structure [
4]. The first gene sequenced from this family in maize was
ZmCHI (
Table 1) [
26]. Interestingly, mutants have not been reported in maize for this gene, due to the multiple homologous sequences found for
ZmCHI in the maize genome. An experiment designed to find QTLs for resistance to
Fusarium corn fungi detected
ZmCHI3 as a second member of the family [
37]. Indeed, a transformed maize callus with a copy of
ZmCHI3 from a resistant inbred was less susceptible to maize plagues.
2.2.3. Flavonoid 3-Dioxygenase (ZmF3H, fht1, EC 1.14.11.9)
ZmF3H is a Fe
2+ and 2-oxoglutarate-dependent dioxygenase that introduces a hydroxyl group in position 3 of the chalcone structure, generating a dihydroflavonol [
27]. There is just one gene copy known in the maize genome. In a previous report,
ZmF3H was found to be the only gene in the flavonoid pathway in which mRNA expression levels correlate with the synthesis of flavonols in anthers [
38]. Moreover, its expression increases in pigmented kernels compared to white seeds [
19].
2.2.4. Flavonoid 3′-Monooxygenase (ZmF3′H, pr1, EC 1.14.14.82)
This
pr1 or purple aleurone1 gene has been studied in maize because its alleles are responsible for changes in color pigmentation caused by a difference in anthocyanin profile [
39,
40].
ZmF3′
H is monooxygenase hydroxylate in the 3′ position from the phenyl ring B (
Table 1). When the gene is functional, its enzyme can produce blue/violet-colored anthocyanidins (cyanidin and peonidin). If not, it generates a red/orange mono-hydroxylated pelargonidin [
28]. Red kernels are homozygous for the recessive alleles
pr1 that do not produce functional enzymes, resulting in the pelargonidin-base anthocyanins predominating over the anthocyanin profile. The dominant
Pr1 alleles have a gene dose effect in the purple kernel pigmentation, which means that each
Pr1 allele in diploid (vegetative) or triploid (endosperm) tissues increase the cyanidin-base anthocyanins (
Figure 2) in the pigmented tissue [
41].
Moreover,
ZmF3′
H has a role in the biosynthesis of 3-deoxyflavonoids and phlobaphene; as occurs with the anthocyanins, the precursor transforms into a di-hydroxylated phenyl ring B compound [
42]. A
Pr1 allele is essential for the resistance against biotic stress depending on C-glucosyl flavone (maysin) accumulation in salmon-colored silks [
43].
2.3. Late Biosynthetic Genes of Maize Anthocyanins
2.3.1. Dihydroflavonol 4-Reductase (ZmDFR, a1, EC 1.1.1.219)
This enzyme converts the dihydroflavonol (or flavanonol) to a flavan-3,4 diol by reducing the 4-carbonyl (
Figure 3 and
Table 2) [
44]. There is a hypothesis that this enzyme has a role in phlobaphene biosynthesis by transforming the 4-carbonyl into flavanones to produce 4-flavan-4-ol [
45]. The gene locus of
ZmDFR, a1, has been deeply studied for two reasons. The first is its linkage to the
sh2 gene, responsible for the shrunken seed phenotype, that made possible the studies on transposable elements and meiotic recombination hotspots in the a1-sh2 interval [
46,
47]. The second reason is that the gene product is a vital enzyme in the flavonoid pathway, favoring which flavonoid subgroup could be biosynthesized [
48]. If there is a functional allele, it can produce anthocyanidins (
Figure 3) and phlobaphenes (see
Section 2.4.1). However, two copies of a non-functional allele would redirect it to flavanol and flavone biosynthesis [
49].
Figure 3. Biosynthetic genes for maize anthocyanin pathway. After the formation of the dihydroflavonol, five enzymatic steps catalyze its biotransformation into acylated maize anthocyanins. Those genes are the following: dihydroflavonol 4-reductase (
ZmDFR,
a1, EC 1.1.1.219), anthocyanidin synthase (
ZmANS,
a2, EC 1.14.20.4), anthocyanidin 3-
O-glucosyltransferase (
ZmAGT,
bz1, EC 2.4.1.115), malonyl-CoA: anthocyanin 3-
O-glucoside-6′′-
O-malonyltransferase (
Zm3MAT,
aat1, EC 2.3.1.171), and flavonoid 3′,5′-
O-methyltransferase (
ZmAOMT, EC 2.1.1.267). The glutathione S-transferase (
ZmGST,
bz2, EC 2.5.1.18) and multidrug resistance protein (
ZmABCC3 and
ZmABCC4, MRP3 and MRP 4, EC 7.6.2.2) are required to deliver them inside the vacuole. References: [
30,
32].
Table 2. Summary of anthocyanin genes in the maize flavonoid pathway.
The role of
ZmDFR in the diversification of flavonoids is further exemplified by its interaction with multiple transcription factors [
19,
20]. The identity and function of those transcription factors are discussed in
Section 3.1.
ZmDFR1 has a gene duplication in the maize genome, known as
a4. Nevertheless, it is not clear if there is an active protein in the tissue from the genomic sequences alone [
44]. Both genes have a higher expression in pigmented kernels than in anthocyanin-less seeds [
19].
2.3.2. Anthocyanidin Synthase (ZmANS, a2, EC 1.14.20.4)
The dioxygenase
ZmANS oxidizes at the C-3 position of a flavan-3,4 diol, generating a flavan-3,3,4 triol (
Figure 3) [
27]. After oxidation, two water molecules are removed, producing an anthocyanidin molecule [
57]. Moreover, the
ZmANS gene expression is upregulated in pigmented kernels compared to white seeds through elements that conserve the promoter region for the MBW complex [
19,
50]. The
a2 is the unique copy known in the maize genome.
2.3.3. Anthocyanidin 3-O-Glucosyltransferase (ZmAGT, bz1, EC 2.4.1.115)
This enzyme is also known as UDP-flavonoid glucosyltransferase (
ZmUFGT). It catalyzes the transference of glucose to the C-3 position of anthocyanidins (
Figure 3) [
58,
59]. This locus is named bronze1 since
bz1 alleles cannot produce a functional gene product and are responsible for the bronze-colored aleurone [
51,
60]. Glycosylated anthocyanidins (anthocyanins) accumulate in a vacuole only when the
ZmAGT is functional. If not, the anthocyanidins are prone to oxidation, turning into brown pigments in the cell wall [
61]. The expression occurs in all anthocyanin pigmented tissue because it contains conserved elements in its promoter, as other genes are upregulated simultaneously by the MYB-bHLH-WD40 (MBW) complex [
60,
62].
The locus
bz1 is located in the intergenic region
bz1-stc1, known for the varying copies of transposable elements [
63,
64]. A relevant study included the first discovery of the first DNA transposable element, the Ac/Ds transposon, that resulted in a Nobel Prize being awarded to Dr. McClintock [
65]. The Ds activation by marker Ac produces a chromosome rupture of chromosome 9 short arm region, which was recognized phenotypically by the apparition of bronze-colored spots in the kernel [
66].
2.3.4. Malonyl-CoA: Anthocyanin 3-O-Glucoside-6′′-O-Malonyltransferase (Zm3MAT, aat1, EC 2.3.1.171)
Two types of acyl moieties can modify the glycosidic part of the anthocyanins in the
Plantae kingdom, aromatic and aliphatic dicarboxylic acids.
Zm3MAT (
Figure 3) was the first anthocyanin acyltransferase (AAT) discovered not only in maize but also in monocots [
52,
67].
Zm3MAT is necessary to produce mono-malonylated anthocyanins, the most common type of anthocyanins in the aleurone layer [
68].
Zm3MAT was selected as a QTL for the reduced acylation phenotype and then corroborated through a knockout by Mu transposon insertion [
52]. Further research showed that
Zm3MAT exerts a dimalonyl transferase activity and can utilize both acyl moieties malonyl-CoA and succinyl-CoA, but it is more specific for malonyl-CoA [
67]. The spectrum of anthocyanin selectivity ranges from the most preferable to the least preferable as follows: cyanidin-3-
O-glucoside, pelargonidin-3-
O-glucoside, peonidin-3-
O-glucoside, and delphinidin-3-
O-glucoside.
2.3.5. Flavonoid 3′,5′-O-Methyltransferase, or Anthocyanin S-Adenosyl-l-Methionine-Dependent O-Methyltransferase (ZmFOMT or ZmAOMT, EC 2.1.1.267)
This enzyme catalyzes the methylation of a hydroxyl group in the -3′ or -5′ position of the 3-hydroxyflavonoid’s phenyl B-ring (
Figure 3) [
53]. The enzyme uses several flavonoids as substrates, not just anthocyanins. These include aglycone and glycosylated forms of flavonols or anthocyanidins. However, every member has a specific affinity that favors some substrate above others [
68]. Unfortunately, in maize, this enzyme has not been characterized yet. However, Chapman and collaborators mentioned two candidate genes, namely
omt1 (Zm00001d052841) and
omt4 (Zm00001d05284), for anthocyanin
O-methyltransferases related to QTLs for peonidin-base anthocyanins [
54].
2.3.6. Glutathione-S-Transferase (ZmGST, bz2, EC 2.5.1.18)
The glutathione S-transferase (GST) family in maize includes more than 40 GST gene sequences [
69]. This family of enzymes detoxifies cells affected by xenobiotics, such as herbicides, by conjugating a glutathione (GSH) molecule [
70,
71,
72]. After being labeled with glutathione, these molecules are sent out of the cell by an ATP-dependent glutathione conjugate export pump [
73]. However, the
bz2 gene, a GST type III, is supposed to label the anthocyanin to be recognized by a vacuolar glutathione pump, and then the labeled anthocyanin is transported into the vacuolar lumen [
55,
73]. Until now, there is no evidence that shows that anthocyanins are conjugated with GSH. However, the role of
bz2 in the accumulation of anthocyanins is accepted. Other authors suggested that this enzyme may function as a carrier protein for vacuolar anthocyanin sequestration [
74].
When
ZmGST is not functional, the anthocyanins are not transported to the vacuole interior. Then, the intravacuolar pH and environment contribute to maintaining these molecules without degradation [
75]. As described for
bz1, a maize plant without functional alleles will develop a bronze-colored kernel [
61].
ZmGST is upregulated in pigmented tissue because it shares conserved binding sites in the promoter region for the MBW complex interaction, a characteristic shared with other upstream genes in the flavonoid pathway [
19,
76].
2.3.7. Multidrug Resistance Protein (ZmABCC3 and -4, mrpa3, EC 7.6.2.2)
ZmABCC3 is part of a broader ATP-binding cassette (ABC) superfamily protein containing up to 130 open reading frames [
72]. In maize, this superfamily of transmembrane proteins anchored to the cell membrane is highly specialized in expelling xenobiotics from the intracellular environment [
77]. However,
ZmABCC3 and
ZmABCC4 are present in the tonoplast of vegetative tissues and in the aleurone layer, respectively [
56].
This protein follows a similar expression profile to other genes related to anthocyanin biosynthesis [
19,
78]. Recent research in species such as
Vitis vinifera and
Arabidopsis thaliana shows that their homologous sequences to
ZmABBC3 are GSH/anthocyanin co-transporters [
79,
80].
2.3.8. Flavanol-Anthocyanin Condensed Forms
The flavanol-anthocyanin condensed forms are compounds found in maize; however, there is still no description of a known enzyme producing them [
81]. Their biosynthesis starts with the generation of the flavan-3-ol unit (
Figure 3). The leucoanthocyanidin reductase (E.C. 1.17.1.3) participates in a reduction reaction in the C-3 position of the leucoanthocyanidin [
54,
82]. This enzyme is yet unidentified in maize. Then, a linkage occurs between the anthocyanin and the flavan-3-ol, but there is no recognized enzyme for this process (
Figure 3). However, it is known that a QTL for the flavanol-anthocyanin condensed form was mapped near the
p1 locus [
54].
In wine, the presence of flavanol-anthocyanin condensed forms is related to aging. However, in maize, there is evidence of natural formation [
81]. The production of flavanol-anthocyanin condensed forms consumes monomeric anthocyanin, therefore reducing the total concentration [
67].
2.4. Biosynthesis of Flavonols, Flavones C-Glycosides, and Phlobaphenes in Maize
2.4.1. Flavonol Synthase (ZmFLS1, fls1, EC 1.14.20.6)
The flavonols are important in maize due to their effects on male fertility and UV-B protection [
83]. Flavonol synthesis depends on flavanone 3-dioxygenase and flavonol synthase, a Fe
2+/2-oxoglutarate dependent dioxygenase (
Figure 4 and
Table 3). The transcription factors that regulate the expression of anthocyanins and C-flavone glycosylated biosynthetic genes can also upregulate the expression of
ZmFLS1 [
1,
27]. In the maize genome are two copies (
ZmFLS1 and
ZmFLS2) in tandem in the long arm of chromosome 5. The expression of both enzymes was augmented under UV-B light and in high-altitude landraces compared to the inbred lines through an increased
p1 expression [
1,
3].
Figure 4. The biosynthetic genes of flavonol and phlobaphenes. The flavanones naringenin and eriodyctiol are the starting substrates for the other flavonoid subgroups. Flavonol synthesis depends on flavanone 3-dioxygenase (
ZmF3H,
fht1, EC 1.14.11.9) and flavonol synthase (
ZmFNS1,
fns1, EC 1.14.20.5). Phlobaphene synthesis begins with the action of dihydroflavonol 4-reductase (
ZmDFR,
a1, EC 1.1.1.219) on flavanones, generating flavan-4-ol molecules that undergo a non-enzymatic polymerization into phlobaphenes. References: [
8,
30,
32].
Table 3. Summary of flavonol and flavone C-glycoside genes in the maize flavonoid pathway.
2.4.2. Flavone Synthase I (ZmFNSI1-2, fnsi1, EC 1.14.20.5) and Flavone Synthase II (ZmFNSII-1, fnsii1, EC 1.14.19.76)
Maize possesses three enzymes that can synthesize flavones from a flavanone, flavone synthases I and II, and flavone 2-hydroxylase (
Figure 5) [
2,
86]. The flavone synthase produces a desaturation in the C2–C3 bond in the flavanone through an oxidation reaction. The oxidative mechanism in
ZmFNSI is a Fe
2+/2-oxoglutarate-dependent dioxygenase, like in
ZmFLS1, whereas that in
ZmFNSII is CYP450 [
2]. In addition,
ZmFNSI1 is upregulated more in tassels than in silks compared to
ZmF2H [
88]. The
p1 transcription factor regulates the expression of
ZmFNSI. Meanwhile, the anthocyanin MBW complex regulates the expression of
ZmFNSII. Both types of flavone synthases generate apigenin, which defends the plant against UV-B radiation-induced damage [
2].
2.4.3. Flavanone 2-Hydroxylase (ZmF2H1, fns1, EC 1.14.14.162)
In maize, this is the third known enzyme that can produce the flavone backbone of the flavone C-glycosides in the salmon-colored silks [
85]. This enzyme is phylogenetically closer to FNS type II, both being CYP proteins [
86,
89]. Flavanone-2-hydroxylase adds a hydroxyl group into the flavanone C-2, producing the opening of the C-ring and finally generating the 3-oxo-dihydrochalcone (
Figure 5). After this opening, it can be glycosylated in either of the two positions of the A-ring, closing the C-ring, eliminating water (spontaneous or not), and then generating in vitro a mixture of C-6 or C-8 glycosylated flavones [
86].
Figure 5. Biosynthetic genes of flavone C-glycosides. The flavanones naringenin and eriodictyol are the initial substrates for the other flavonoid subgroups. There are two possible ways to generate C-glycosyl flavones, indirectly or directly, from any flavanone. The indirect pathway begins through flavanone-2-hydroxylase (
ZmF2H,
fnsii1, EC 1.14.14.162) opening the C-ring, producing a 3-oxo-dihydrochalcone. Then, UDP C-glycosyl transferase (
ZmCGT,
cgt1, EC 2.4.1.360) generates a glycosidic bond in the A-ring. Then, there is a dehydration reaction (spontaneous or enzymatic) that produces the C6-flavone glycoside. The direct pathway firstly involves flavone synthase I (
ZmFNSII-2, fnsii2, EC 1.14.20.5) and flavone synthase II (
ZmFNSII-1,
fnsi2, EC 1.14.19.76) producing the same reaction by the addition of a double bond between C2 and C3 in the flavanone. Then, a flavone functions as a substrate for the UDP C-glycosyl transferase (
ZmCGT,
cgt1, EC 2.4.1.360). The enzymatic action of UDP-rhamnosyl transferase (
ZmCGT,
sm2, EC 2.4.1.159) and glucose 4,6 dehydratase (
sm1, EC 4.2.1.76) produces either apimaysin or maysin. References: [
30,
43,
90].
2.4.4. UDP-Glucose:2-Hydroxyflavanone C-Glucosyltransferase (ZmCGT, cgt1, EC 2.4.1.360)
UDP-glucose:2-hydroxyflavanone C-glucosyltransferase generates a glycosidic bond in the A-ring from the C-1 of the glucose to the C-6 in the C-glycosyl flavones (
Figure 5) [
43]. In vitro and in vivo experimental evidence has demonstrated that the
ZmCGT enzyme has a bifunctional capacity to form glycosidic bonds with C or O atoms. On the contrary, there is only in vitro evidence for C-8 flavone glycosides [
85]. The likely reason for that is the possibility of an enzyme that only selects C-6 glycosylated 2-hydroxyflavanone for dehydration into C-6 glycosyl flavones [
91].
2.4.5. UDP-Rhamnosyl Transferase (sm2, UGT91L1, EC 2.4.1.159)
The UDP-rhamnosyl transferase enzyme forms the glycosidic bond between the glucose C-2 and the rhamnose C-1 (
Figure 5) [
87]. Functional alleles confer a characteristic salmon color to the silks due to the accumulation of maysin/apimaysin in the silks. This is due to
p1 upregulating
sm2 and is expressed principally in silks [
43] but also in non-vegetative tissues such as pollen, tassels, and seeds [
88].
2.4.6. Glucose-4,6 Dehydratase (ZmRHS1, sm1, EC 4.2.1.76)
The biosynthesis of C-flavones glycosides in maize ends with a modification to the glucose structure of the rhamnosylisoorientin (or rhamnosylisovitexin) to produce maysin/apimaysin (
Figure 5) [
92]. These metabolites give the ear of maize the ability to deter the herbivore
Helicoverpa zea, commonly known as corn earworm [
25,
43]. This locus was found to be responsible for producing the last step in the maize flavone pathway and found to be a putative UDP-rhamnose synthase (
ZmRHS1) [
43,
87]. The gene has two putative domains; the first domain is a UDP-glucose dehydratase, and the second domain corresponds to UDP glucose 4-keto-6-deoxyglucose epimerase/reductase. The former domain is the exclusive one catalyzing maysin or apimaysin biosynthesis [
43]. Its gene expression pattern in the tissues is similar to the
sm1 profile [
2].