Expression of flavonoid 3’-hydroxylase is controlled by P1, the regulator of 3-deoxyflavonoid biosynthesis in maize

Pr1; P1 cob-glumes accumulate luteoforol

Maize plants carrying a functional p1 gene accumulate phlobaphene pigments in kernel pericarps and cob glumes. Although, the Zmf3’h1 has been shown to be required for the formation of purple anthocyanins in kernel aleurones [24], changes in pigment intensity has also been observed in phlobaphene accumulating tissues in the presence of a functional Zmf3’h1 gene [3]. To investigate the role of Zmf3’h1 in the p1 regulated phlobaphene biosynthesis, we developed Pr1 and pr1 near isogenic lines in the genetic background of three p1 alleles: P1rr, P1wr and pdel2 (Figure 2A). Phenotypic characterization of P1wr ears segregating for Pr1 and pr1 showed colorless pericarp and gene dependent cob glumes pigment phenotypes: dark red in Pr1/Pr1; P1wr/P1wr while light red in pr1/pr1; P1wr/P1wr. Moreover, in the presence of P1rr, Pr1 and pr1 plants showed pericarp and cob glumes colour differences: Pr1/Pr1; P1rr/P1rr plants have dark red pericarp and dark red cob glumes as compared to red pericarp and light red cob glumes in pr1/pr1; P1rr/P1rr plants. Importantly, pdel2 plants (lack both p1 and p2), in the presence of Pr1 or pr1 did not show any visible phlobaphene pigmentation in pericarps or cob glumes (see Table 1).

Figure 2

Luteoforol accumulates in Pr1; P1 cob glumes. (A) Ear phenotypes showing cob glumes pigmentation. Cob glumes of pr1 are light red while Pr1 has dark red cob glumes in the presence of P1-wr or P1-rr alleles. Pr1 and pr1 ears carrying p-del2 do not show any pigmentation. (B) Absorption spectra of cob glume extracts. Methanolic extracts from cob glumes of Pr1 ears in genetic background of P1-wr and P1-rr alleles gave maximum absorption at 498 nm. Cob glumes from pr1 ears in P1-wr and P1-rr genetic background gave maximum absorption at 482 nm. These spectra at 498 and 482 correspond to peaks for standard luteolinidin and apigeninidin, respectively (see inset). No absorption peak for flavan 4-ols was observed in p-del2 cob glumes.

Two flavan 4-ols, luteoforol and apiforol have been implicated as precursors of phlobaphene pigments that accumulate in maize genotypes carrying functional p1 or p2 genes. Cob glumes were used to perform biochemical characterization of flavan 4-ols. The dark red cob glumes from Pr1/Pr1; P1wr/P1wr had maximum absorption (λ max) at 552 nm, while light red cob glumes from pr1/pr1; P1wr/P1wr plants had λ max at 535 nm (Figure 2B). These absorption spectra correspond to luteoforol and apiforol, respectively [17]. To further confirm if the Zmf3’h1 plays a role, flavan 4-ols were converted into their corresponding 3-deoxyanthocyanidins by acid treatment of methanolic extracts (Figure 2B). Extracts from Pr1/Pr1; P1wr/P1wr converted to luteolinidin (λ max 498 nm) indicating the presence of luteoforol in the methanolic extracts. Similarly, extracts of pr1/pr1; P1wr/P1wr cob glumes were converted into apigeninidin (λ max 482 nm) indicating presence of apiforol in the extract. Acid treated methanolic extracts from cob glumes of Pr1/Pr1; P1rr/P1rr and pr1/pr1; P1rr/P1rr also had maximum absorption wavelengths corresponding to luteolinidin and apigeninidin, respectively. No detectable flavan 4-ols or corresponding 3-deoxyanthocyanidins accumulated in cob glumes of pr1/pr1; pdel2/pdel2 or Pr1/Pr1; pdel2/pdel2. This was in accordance with p1 gene’s function as a regulator of phlobaphene biosynthesis [12]. Taken together, these results show that cob glumes from pr1; P1 plants accumulate apiforol whereas luteoforol accumulates in cob glumes of Pr1; P1 plants. This result also indicates that the accumulation of apiforol and luteoforol is influenced by a flavonoid 3’-hydroxylase acting in parallel for the conversion of naringenin to eriodictyol.

P1 regulates the transcription of Zmf3'h1

The Zmf3’h1 gene appears to affect the composition of flavan 4-ols in floral tissues of P1 alleles. Thus, it is possible that the p1 gene regulates the Zmf3’h1 transcription. To test this hypothesis, we examined steady state transcript levels of Zmf3’h1 in pericarps, cob-glumes, and silks of Pr1 plants carrying functional and null p1 alleles (Figure 3). Increased levels of Zmf3’h1 transcripts were present in pericarps of Pr1/Pr1; P1rr/P1rr plants, while these transcripts were not detected in pericarps of Pr1/Pr1; P1wr/P1wr and Pr1/Pr1; P1ww/P1ww plants (Figure 3B). In cob glumes of Pr1/Pr1; P1rr/P1rr and Pr1/Pr1; P1wr/P1wr plants there were appreciable amounts of Zmf3’h1 transcripts, while Pr1/Pr1; P1ww/P1ww cob glumes did not show Zmf3’h1 transcripts (Figure 3B). Transcript levels of p1, c2, chi1, a1, and Zmf3’h1 were also compared in silks (Figure 3C). Similar to the p1 expression, Zmf3’h1 was highly expressed in silks of Pr1/Pr1; P1rr/P1rr and Pr1/Pr1; P1wr/P1wr. Although, both P1ww [4Co63] and P1ww-1112 alleles have non-functional p1 gene, detectable levels of p2 expression was present. The the p gene probe used in this study can detect both p1 and p2 specific transcripts (see Methods). Thus the Zmf3’h1 expression detected in P1ww silks is because of a functional p2 gene [13]. Further, the p-specific transcripts were not detected in pdel2 silks because of deletions within p1 and p2 genes. In addition, other biosynthetic genes c2, chi1, and a1 showed expression pattern similar to that of Zmf3’h1 indicating a commonality in the regulation of these genes. Overall, these results show that the p1 and p2 genes regulate the transcriptional expression of Zmf3’h1 in maize pericarp, cob glume, and silk.

Figure 3

Expression of Zmf3’h1 requires P1 and/or P2 genes. (A) Ear phenotypes of different p1 alleles: P1-rr, red pericarp, red cob glumes; P1-wr, white pericarp, red cob glumes; P1-ww [4Co63] and P1-ww-1112, white pericarp and white cob glumes. (B) Gel blot of RNA extracted from pericarps and cob glumes of different p1 alleles was probed with Zmf3’h1 and c2 probes. P1-rr pericarps and cob-glumes and P1-wr cob-glumes show Zmf3’h1 expression. (C) RNA gel blot of silks from different p1 alleles was hybridized with various flavonoid gene fragments as probes. rRNA bands showing equal loading of RNA in different lanes.

P1 binds to two sites in the Zmf3’h1 promoter

The p1 gene regulates expression of Zmf3’h1 during accumulation of luteoforol in maize floral tissues. To further confirm the mechanism by which p1 controls Zmf3’h1 expression, we found that the Zmf3’h1 promoter contains the cis-regulatory elements CCTACC (-614 ~ -553) and CCAACC (-83 ~ -78), that resemble P1 DNA-binding sites [12]. To test this possibility, electrophoretic mobility shift assays (EMSA) were performed using as probes two DNA fragments spanning regions containing the putative P1 binding sites from -649 to -553 and from -128 to -41 (Figure 4). Our results show that P1 binds both the sites, which is in agreement to consensus sequences CC^T/_AACC derived from SELEX [12].

Figure 4

P1 binds to Zmf3’h1 promoter. The locations of two probes are indicated with reference of the first nucleotide of Transcription Start Site (TSS) as +1. +: P1 + probe; -: free probe.

P1 binds the Zmf3^′h1 promoter in vivo

In vitro binding assays indicate that P1 directly binds the Zmf3’h1 promoter. In order to investigate whether P1 binds the Pr1 promoter in vivo, chromatin immunoprecipitation (ChIP) experiments were performed. Pericarps from P1ww and P1rr were collected at 15 days after pollination (DAP), and subjected to ChIP assays using an anti-P1 polyclonal antibody that was successful in a1 gene ChIP experiments previously [32]. Figure 5 clearly shows that the anti-P1 antibody precipitate complexes containing the Zmf3’h1 promoter fragment with the presence of P1 protein whereas the antibody did not enrich copia or actin fragments. These results suggest that P1 directly binds Zmf3’h1 promoter in vitro and in vivo.

Figure 5

ChIP assay. Relative enrichment was measured by a ratio of P1-rr to P1-ww of % input, which was calculated as a percentage of DNA amount in a fraction of chromatin immuno-precipitation by anti-P1 antibody to total DNA amount in an input fraction. Error bar indicates standard error taken from three independent ChIP experiments. Asterisk shows P < 0.05 (t-test).

pr1; P1 plants do not accumulate luteolinidin in silks

In maize, p1 is known to regulate 3-deoxyanthocyanidins in silks, but their biosynthesis is poorly understood [16,17]. Several lines of evidence showed that flavan 4-ols are the precursors of 3-deoxyanthocyanidins [16,17,33-35]. Since, Zmf3’h1 plays a role in the differential accumulation of p1 regulated flavan 4-ols, we tested if Zmf3’h1 also influences the accumulation of 3-deoxyanthocyanidins. Silk extracts from Pr1 and pr1 plants carrying different p1 alleles were analysed by reverse-phase high performance liquid chromatography (HPLC) (Figure 6). Extracts from Pr1/Pr1; P1rr/P1rr and Pr1/Pr1; P1wr/P1wr plants showed four major peaks at 480 nm: three for luteolinidin glycosides labelled as a, b, and c with retention times of 13.6 min, 14.1 min, and 14.9 min, respectively and the fourth peak corresponds to anthocyanins, eluting at 17.9 min. (Figure 6A). The peaks a, b, and c had the luteolinidin aglycone spectra but eluted considerably before the aglycone suggesting that they are more polar and therefore most likely correspond to glycosylated flavonoids. Based on their different elution times, these may be mono-glucosyl or di-glucosyl luteolinidins. Interestingly, silks from pr1/pr1; P1rr/P1rr and pr1/pr1; P1wr/P1wr plants, showed much smaller peaks for all luteolinidin glycosides and few additional peaks of unknown compounds. Pr1 and pr1 plants carrying P1ww or pdel2 alleles did not produce any detectable levels of these compounds. This latter result is consistent with the fact that both of these p alleles lack a functional p1 gene to induce 3-deoxyanthocyanidins accumulation in silks. Quantitative HPLC data (Figure 6B) showed that Pr1/Pr1; P1rr/P1rr and Pr1/Pr1; P1wr/P1wr silks accumulated significantly higher levels of total luteolinidin glycosides as compared to pr1/pr1; P1rr/P1rr and pr1/pr1; P1wr/P1wr. None or very little amount of 3-deoxyanthocyanidins accumulated in pdel2 or P1ww plants. In summary, HPLC analysis of 3-deoxyanthocyanidins in silks of Pr1; P1, and pr1; P1 plants show that the Zmf3’h1 participates in the synthesis of luteolinidin.

Figure 6

Characterization of the 3-deoxyanthocyanidins in Pr1 and pr1 silks. Compounds were identified based on their retention time to known standards. (A) HPLC chromatograms of silk methanolic extracts at 480 nm. Luteolinidin glycosides a, b, and c were eluted at approximately 13.6 min, 14.1 min, and 14.9 min, respectively. Anthocyanidins elute at approximately 17.9 min. (B) Total luteolinidin and total anthocyanidin levels in silk tissue of homozygous pr1; P1-wr, Pr1; P1-wr, pr1; P1-rr, Pr1; P1-rr, pr1; p-del2, Pr1; p-del2, pr1; P1-ww, and Pr1; P1-ww were determined by HPLC analysis at 495 nm. All data are presented as mean of six replicates.

Silks of pr1; P1 plants have reduced maysin accumulation

Our results thus far suggest that, the Zmf3’h1 is under the regulatory control of p1 and participates in the biosynthesis of both phlobaphenes and 3-deoxyanthocyanidins. To test if Zmf3’h1 also has a role in p-regulated C-glycosyl flavones biosynthesis, flavone levels in silks of Pr1 and pr1 plants in various p1 allelic backgrounds were measured (Figure 7). Maysin and apimaysin were identified by comparing their retention times of 22.0 min and 23.5 min, respectively and UV absorption spectrum with authentic standards. In Pr1/Pr1; P1-wr/P1-wr silks, maysin was the major compound while smaller peaks for apimaysin and rhamnosylisoorientin were detected at 340 nm (rhamnosylisoorientin retention time was 16.0 min). Silk extracts from pr1/pr1; P1-wr/P1-wr showed a single dominant peak for apimaysin, a smaller peak for rhamnosylisoorientin and no peak for maysin. Peaks for maysin and apimaysin were not observed at an appreciable level in Pr1/Pr1; p-del2/p-del2 silks. Quantitative data measured from HPLC showed that Pr1/Pr1; P1-rr/P1-rr and Pr1/Pr1; P1-wr/P1-wr plants have significantly higher levels of maysin and substantially lower levels of apimaysin (Figure 7B, panels I and II). In contrast, pr1/pr1; P1-wr/P1-wr silks produced significantly high levels of apimaysin and very little amount of maysin. In p-del2 silk extracts, no significant levels of C-glycosyl flavones were detected irrespective of the type of pr1 allele present. However, in Pr1 and pr1 lines carrying P1-ww [4Co63] allele, very low levels of maysin and apimaysin were detected; this can be attributed to the functional p2 gene present in this allele. Interestingly, Pr1/Pr1; P1-rr/P1-rr plants accumulated significantly higher amount of rhamnosylisoorientin in silks as compared to pr1/pr1; P1-rr/P1-rr (see Figure 7B, panel III). Overall, these results suggest that the Zmf3’h1 plays a role in the accumulation of C-glycosyl flavones in silks in the presence of functional p1 alleles.

Figure 7

Characterization of the C-glycosyl flavones in Pr1 and pr1 silks. (A) HPLC chromatograms of silk methanolic extracts from Pr1 and pr1 silks measured at 340 nm. Maysin, apimaysin, and rhamnosylisoorientin elute at approximately 22 min, 23.5 min, and 16 min, respectively. (B) Maysin, apimaysin, and rhamnosylisoorientin levels in silks of homozygous pr1; P1-wr, Pr1; P1-wr, pr1; P1-rr, Pr1; P1-rr, pr1; p-del2, Pr1; p-del2, pr1; P1-ww, and Pr1; P1-ww were determined by HPLC analysis. All data are presented as mean of ten biological replicates. (C) Mean corn earworm larval weight and days to pupation on fresh silk from pr1; P1-wr, Pr1; P1-wr, pr1; P1-rr, Pr1; P1-rr, pr1; p-del2, Pr1; p-del2, pr1; P1-ww, and Pr1; P1-ww lines.

Pr1; P1 silks is detrimental to corn earworm larvae development

To determine the biological relevance of differential accumulation of maysin and apimaysin in silks of Pr1 and pr1 plants, we performed insect silk feeding bioassays. Corn earworm (Helicoverpa zea Boddie) larvae (neonate stage) were fed on fresh silks (Additional file 1: Figure S1) collected from the same Pr1 and pr1 plants that were used for HPLC analysis. Larvae fed on Pr1 silks had lower body weight and took longer time to pupate as compared to those fed on pr1 silks (Figure 7C). These results are in agreement with the accumulation of higher amounts of maysin in Pr1 silks. Interestingly, larvae fed on pr1/pr1; P1wr/P1wr silks showed lower weight and longer time to pupate as compared to those fed on Pr1/Pr1; P1wr/P1wr silks, even though the Pr1 silks had higher levels of maysin. This anomaly could possibly be because of the accumulation of exceptionally higher level of apimaysin in pr1/pr1; P1wr/P1wr (see Figure 7B, panel II). As shown in previous studies, apimaysin has insecticidal activity against lepidopteron insects, although, apimaysin’s activity levels were lower than maysin [36]. Overall, combined data from HPLC and larvae feeding bioassay indicated that a functional Zmf3’h1 participates in the accumulation of 3’-hydroxylated C-glycosyl flavones that affect the growth of corn earworm larvae.

Maize genetic stocks

Standard maize genetic nomenclature is used in the current study [56]. Alleles of the maize (Zea mays) p1 have been identified based on their expression in the floral organs and are named according to their pericarp and cob-glumes pigmentation: P1wr (white pericarp, red cob), P1rr (red pericarp, red cob), and P1ww (white pericarp, white cob) (Figure 3A) [57-60]. The maize inbred lines W23 (genotype P1wr Pr1 c1 rg), W22 (P1wr Pr1 C1 R1), and other genetic stocks MGS 14273 (P1wr pr1 C1 R1) and MGS 14284 (P1ww pr1 C1 R1) were kindly provided by the Maize Genetics Co-operation Stock Centre (USDA-ARS, University of Illinois, Urbana, IL). The P1ww [4Co63] inbred line was obtained from the National Seed Storage Laboratory (Fort Collins, CO), while P1rr 4B2, P1ww1112 and pdel2 genetic stocks were obtained from Dr. Thomas Peterson, Iowa State University, Ames, IA [61,62]. The pdel2 deletion mutant was derived from P1vv9D9A and has a deletion encompassing both p1 and p2[14,63]. All p1 alleles except pdel2 and P1ww1112 are in 4Co63 genetic background. Our genetic tests have shown that all these p stocks carry a functional Pr1 allele and their pigmentation phenotypes are presented in Table 1. To develop F₂ populations, pr1MGS14273 plants were crossed with P1wr, P1rr4B2, P1ww, and pdel2 and progenies were grown from selfed F₁ plants. These F₂ populations showed a 3:1 segregation for purple to red aleurones. To develop homozygous Pr1 and pr1 stocks containing different p alleles, plants from F₂ ears showing desirable pericarp, cob-glumes, and kernel aleurone pigmentation phenotypes (see Table 1) were subjected to six subsequent cycles of self-pollination and selection. To confirm the presence of Pr1 or pr1, PCR based genotyping was done using primers in the promoter region [24].

Analysis of flavan 4-ols

To detect the presence of flavan 4-ols, 500 mg of cob-glumes were macerated with a plastic grinder in an Eppendorf tube containing 1 mL of 30% HCl/70% butanol (v/v) and incubated for 60 min at 37°C [64]. Samples were spun for 10 min at 14,000 rpm and the absorption spectra of the supernatants were determined using a Shimadzu UV-mini 1240 spectrophotometer (Shimadzu Corporation, Columbia, MD) [65,66]. Apiforol and luteoforol are flavan 4-ols previously described from maize and sorghum that give flavylium ions in acidic butanol with a λ max of 535 and 552 nm, respectively [64]. To confirm the identity of the major flavan 4-ols in cob glumes of Pr1 and pr1 alleles in the genetic background of different p alleles, methanol extract were treated with aqueous HCl. This converts flavan 4-ols such as apiforol and luteoforol to their corresponding 3-deoxyanthocyanidins (i.e. apigeninidin and luteolinidin). The treated Pr1 extracts had λ max of 498 nm that shifted in alcoholic AlCl₃ to a shoulder at 546 nm. The addition of HCl restored its absorption to 498 nm. The results of our samples were verified using commercial standards for apigeninidin and luteolinidin (Extrasynthese, Genay Cedex, France). The commercial sample of apigeninidin had a λ max of 475 nm and did not respond to AlCl₃, whereas luteolinidin had a λ max of 495 nm that shifted in AlCl₃ to 546 nm and reverted to 498 nm upon re-addition of HCl.

RNA gel blot analysis

Silks were collected 2 d after emergence, and pericarps and cob glumes were dissected 20 DAP. To isolate total RNA, tissues were ground in liquid nitrogen and then extracted using Tri-Reagent (Molecular Research Centre Inc., Cincinnati, OH). RNA gel blot hybridizations were performed as described previously [24]. Probe fragments used for RNA gel blot analysis were: plasmid pC2 containing a maize c2 cDNA [67], pCHI1 containing a maize chi1 cDNA [68], pA1 with a maize a1 cDNA [69], pF3’H1 containing Zmf3’h1 cDNA, and pP1 containing full length p1 cDNA [70]. The p1 probe used here can recognize both p1 and p2 transcripts [71]. Filters were stripped by washing thrice in a boiling solution of 0.1% (w/v) SDS before re-hybridization.

Protein expression and purification

NHis₆-P1^MYB used for EMSA was expressed in Escherichia coli and affinity purified using Ni-NTA beads under natural condition as described previously [72]. Briefly, IPTG induction of 1-liter culture, the cells were harvested by centrifugation and re-suspended in 20 ml of SB buffer (50 mM sodium phosphate, pH 8.0, 100 mM NaCl, and 100 μg/ml phenylmethylsulfonyl fluoride) and passed twice through a French press. The cell lysate was centrifuged and the supernatant was filtered through Mira-cloth (Calbiochem). One ml of 50% slurry Ni-NTA beads (Qiagen) was incubated with the cell lysate supernatant for 2 h with gentle rocking at 4°C. The beads were gently harvested by centrifugation, re-suspended in five ml of SB, and loaded onto a column. The column was washed with SB five times and WB (50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 1% Tween 20, 5 mM 2-mercaptoethanol, 10 mM EDTA, and 10% glycerol) three times. The protein was eluted with five washes of five column volumes of WB containing 50 mM imidazole. The elutions were then dialyzed against A-0 buffer (10 mM Tris pH 7.5, 50 mM NaCl, 1 mM DTT, 1 mM EDTA, and 5% glycerol) and stored at -80°C until further use.

Electrophoretic mobility shift assay (EMSA)

EMSA was performed as previously described [12]. The two Zmf3’h1 promoter fragments used as probes for EMSA were generated by PCR amplification using the following primer pairs: Pr1-1, 5′- GAGTGGGTTGTGGGATTGTT-3′ and 5′- ACCGTAAGGCCAACTCCAAC-3′; Pr1-2, 5′- GCCCGCGAAGAAAAATATAA-3′ and 5′- CCACTTGCGTGCTTCATCTA-3′, in which one of the primer was radioactively labeled with [γ-³²P]ATP by using T4-polynucleotide kinase. The radioactively labeled DNA fragments were purified by PAGE and quantified by scintillation counter. Ten ng of purified P1^MYB was incubated with an equal molar amount of probes (with radioactivity ~10⁵ CPM) for 30 min and the P1^MYB-probe complex was separated by PAGE. After PAGE, the gel was dried onto Whatman paper and then subjected to autoradiography at -70°C.

Chromatin immuno-precipitation assay (ChIP)

ChIP experiments using pericarps were performed as previously described [32]. Real-time PCR was used to detect enriched DNA fragments after ChIP experiments with a minor modification. To adjust for different PCR efficiencies as a consequence of the presence of inhibitory compounds in chromatin obtained from P1rr pericarps, equal amounts of the pPHP611 plasmid [73] were spiked in the real-time PCR reaction buffer (Epicentre). Copy number of pPHP611 plasmid [17] in each reaction was estimated by using the normalization primer sets directed to the β-lactamase gene responsible for Amp^r in the plasmid. The primer sets used for ChIP-qPCR were the following: qChIP-ZmCopia-F, 5′-CGATGTGAAGACAGCATTCCT-3′, qChIP-ZmCopia-R, 5′-CTCAAGTGACATCCCATGTGT-3′, qChIP-ZmAct1-5UTR-F, 5′-TTTAAGGCTGCTGTACTGCTGTAGA-3′, qChIP-ZmAct1-5UTR-R, 5′-CACTTTCTGCTCATGGTTTAAGG-3′, qrt Zmf3′h1Prom-A1, 5′-AGATCGCGGGTAGGTAGGAG-3′, and qrt Zmf3’h1Prom-B1, 5′-ACTGGTGGCGAGGGTGTAGT-3′. The following primer sets were used to detect Amp^r gene: ChIP-Amp-F, 5′-GTAGTTATCTACACGACGGGGAGT-3′, and ChIP-Amp-R, 5′-ATCAGTGAGGCACCTATCTCAGC-3′.

Analysis of C-glycosyl flavones and 3-deoxyanthocyanidins

Primary ear shoots were covered prior to silk emergence to prevent random pollination. Silks were collected on ice 2 d after emergence from the ear and subsequently freeze dried. Silk samples were then shipped on dry ice to the Richard B. Russell Research Centre (USDA-ARS, Athens, Georgia) for biochemical analysis. Flavones were extracted with 125 mL methanol at -20°C for 14 d. Concentration of flavones were determined by reversed-phase HPLC [16,74] and expressed as percent dry silk weight. For 3-deoxyanthocyanidins, silks were extracted for 24 h at -20°C with 10 mL of 1% HCl-Methanol (v/v). Their levels were detected at 495 nm by HPLC, with the same column and solvent program used for flavone analysis. Commercial standard of luteolinidin hydrochloride (different than the luteolinidin standard used for spectrophotometric analysis of cob-glumes) was used for quantification (Roth-Atomergic Chemicals Corp., Farmingdale, N.Y.). Chrysin was used as an internal reference standard. Total 3-deoxyanthocyanidin concentration was calculated as the sum of three distinct luteolinidin glycoside peaks a, b, and c [16].

Insect bioassay

Corn earworm (Helicoverpa zea Boddie) eggs were obtained from Benzon Research Company, Carlisle, PA. Eggs were incubated at 28°C. They hatched after 48 h to produce neonate larvae. Maize lines with Pr1 or pr1 alleles in genetic backgrounds of four different p1 alleles; P1wr, P1rr, P1ww, and pdel2 were grown during the summer of 2007. Silks collected 2 to 3 d after emergence were pooled from 20 field grown plants per genotype. The experiment was conducted as a randomized complete block design with 30 replications and two cups per replicate. Freshly collected silks were filled into 1 oz. plastic diet cups containing 10 mL of 2.5% (w/v) agar to prevent silk drying. Instead of adding silk extracts to the artificial insect diet, fresh silk tissues were used to maximize the resemblance to natural larval feeding conditions. One neonate larvae was introduced into each cup and larvae were allowed to feed in a controlled environment maintained at 28°C, 75% RH, and a photoperiod of 14/10 h (light/dark). Larval weights were recorded after eight days of feeding and larvae were subsequently transferred to artificial diet [75] until pupation.