Advertisement

Molecular Breeding

, 36:81 | Cite as

QTL analysis for the identification of candidate genes controlling phenolic compound accumulation in broccoli (Brassica oleracea L. var. italica)

  • Alicia M. Gardner
  • Allan F. Brown
  • John A. Juvik
Article

Abstract

Quantitative trait loci (QTL) analysis of phenolic compound accumulation was performed in a broccoli mapping population (Brassica oleracea) saturated with single nucleotide polymorphism markers from an Illumina 60K array designed for rapeseed (Brassica napus). In 2 years of analysis in North Carolina and 1 year in Illinois, variation in total phenolic content and antioxidant capacity was associated with 60 QTL. Twenty-three of these loci were identified in at least two analyses (three phenolic/antioxidant assays times 4 years (2009, 2010, 2014, and the mean of the 3 years) making a total of 12 trait-year assays); the two most stable QTL (no. 7 and no. 52) were identified in six and five analyses, respectively. Genome-specific SNP markers were used to identify a priori candidate genes within the QTL marker intervals. Genes involved in the early stages of phenylpropanoid biosynthesis and MYB transcription factors were most heavily represented among the putative candidate genes. The results demonstrate the complexity of the regulatory network involved in phenolic compound accumulation, but highlight potential targets for the development of Brassica vegetables with enhanced phenolic compound profiles.

Keywords

Brassica oleracea Phenolic compounds QTL 

Introduction

Phenolic compounds constitute the largest and most ubiquitous category of secondary metabolites across the plant kingdom, displaying a breadth of structural and functional diversity. Phenolic compounds are characterized by the presence of at least one aromatic ring with at least one hydroxyl group attached. Phenylpropanoid biosynthesis begins with the aromatic amino acids produced by the shikimate pathway. The genes and regulation of phenolic compound biosynthesis have been well characterized, and the reader is directed to two recent reviews for further detail (Cheynier et al. 2013; Saito et al. 2013). A generalized figure of the phenylpropanoid biosynthesis pathway is presented in Online Resource 1. Studies of the phenolic profile of broccoli florets have shown that the primary phenolic compounds present are flavonol and hydroxycinnamic acid derivatives, with the most significant flavonol constituents of broccoli being glycosylated forms of quercetin and kempferol (Price et al. 1997, 1998; Heimler et al. 2006).

Interest in examining the health benefits of phenolic compounds began largely due to epidemiological evidence that diets rich in flavonoids and other phenolic compounds were associated with a lower risk of coronary heart disease (Hertog et al. 1995), breast (Sun et al. 2006), lung (Tang et al. 2009; Knekt et al. 2002), and prostate cancers; type II diabetes; and asthma (Knekt et al. 2002). The role of reactive oxygen species (ROS) in chronic disease (Ma 2014) and the strong ability of phenolic compounds to scavenge free radicals in vitro led to the belief that this was a primary cause for the health benefits associated with their consumption (Masella et al. 2005). However, further investigation of the metabolism and disposition of phenolic compounds in humans revealed modest-to-low absorption and extensive conjugation and breakdown of phenolic metabolites (Clifford 2004; Bergmann et al. 2010). Instead of direct radical scavenging, which would require phenolic metabolites to outcompete the native crop antioxidants, ascorbic acid, and α-tocopherol, interaction with cellular signaling cascades is a much more plausible mode of action (Crozier et al. 2009). There is evidence that supports a role for phenolic compounds in interactions with a number of different signaling molecules, including NF-κB, cyclooxygenase-2, Nrf2, and MAP kinase cascades (Kim et al. 2015; Weng et al. 2011). Of these, the Nrf2/Keap1/ARE cascades are the best characterized. Nrf2 is a transcription factor that activates the expression of phase II detoxification enzymes [e.g., NAD(P)H: quinone reductase and glutathione-S-transferase] by binding to antioxidant response element/electrophile response element (ARE/EpRE) sequences in the promoter region of target genes. Under basal conditions, Nrf2 is bound in the cytoplasm and targeted for ubiquitination by its negative regulator, Keap1 (Osburn and Kensler 2008).

Although direct radical scavenging is not a likely mechanism of cytoprotection in vivo, the reactivity of phenolic compounds with free radicals is not irrelevant. The same structural features that most commonly contribute to in vitro radical scavenging activity—ortho- and para-hydroquinone moieties—also react readily with Keap1 cysteine thiolates. Other polyphenols that initially lack an electrophilic α, β-unsaturated carbonyl can be oxidized by free radicals to generate a quinone capable of reacting with Keap1. This elegantly illustrates how it is possible for “antioxidant” and “pro-oxidant” phenolic compounds and metabolites to positively impact the cellular redox state—by inducing phase II detoxification genes through the Nrf2/Keap1/ARE signaling pathway (Dinkova-Kostova et al. 2001; Forman et al. 2014).

In addition to interacting at Keap1, polyphenols have also been demonstrated to induce MAP kinase activity involved in Nrf2 stabilization, nuclear translocation, and DNA binding. Weng et al. (2011) showed that quercetin treatment of human hepatoma HepG2 cells induced Nrf2 phosphorylation by JNK, ERK, p38, and Akt and promoted greater nuclear translocation and DNA binding. This study also found a significant correlation between the in vitro antioxidant activity of several polyphenols and their ability to promote phase II detoxification gene expression.

Therefore, quantification of phenolic compounds (or antioxidant capacity as a proxy for bioactive phenolic content) and elucidation of the genetic controls involved in their accumulation is an important step in the process of breeding Brassica vegetables for enhanced human health promotion. In this work, three classic phenolic content and antioxidant capacity assays were utilized to phenotype a broccoli [Brassica oleracea L. var. italica (N = 9 CC)] mapping population (VI-158 × BNC) in order to identify quantitative trait loci (QTL) associated with variability in phenolic compound concentration in broccoli florets.

The parents of this population—VI-158, a calabrese-type double haploid of the F1 hybrid “Viking”, and “Broccolette Neri e Cespulgio” (BNC), a broccolette neri (black broccoli) accession (PI 462209)—were selected for their contrasting phytochemical profiles (Kushad et al. 1999; Brown et al. 2002) and crossed to produce an F2:3 mapping population (Brown et al. 2007). This population was recently saturated with single nucleotide polymorphism (SNP) markers from the Illumina 60K iSelect array designed for rapeseed (Brassica napus, N = 19 AACC) and anchored to the TO1000 B. oleracea reference sequence (Parkin et al. 2014) by genome-specific markers. This information was used to conduct QTL mapping of carotenoid and glucosinolate profiles and to identify putative candidate genes co-localizing with the QTL (Brown et al. 2014, 2015). The present effort extends similar analysis to phenolic compound accumulation in the same population.

Materials and methods

Cultivation of plant material

The biparental broccoli population used in this project was developed as described by Brown et al. (2007). The parents were VI-158, a calabrese-type double haploid of the F1 hybrid “Viking” by a broccolette neri (black broccoli) accession BNC, PI 462209. Freeze-dried samples of 92 F2:3 families grown in two field replications in 2009 and 2010 at North Carolina State University as described by Brown et al. (2014) were obtained from the Brown laboratory.

During the 2014 growing season, 115 F2:3 families from the VI-158 × BNC mapping population were grown at University of Illinois at Urbana–Champaign (UIUC). Seeds were planted on May 28 in 48-cell flats filled with Sunshine® LC1 (Sun Gro Horticulture, Vancouver, BC, Canada) professional potting mix. Seedlings were germinated in the plant science laboratory greenhouse under a 14/10 h and 25/15 °C day/night temperature regime for 3 weeks and then were hardened off in an outdoor ground bed for 2 weeks prior to transplanting. The plants were transplanted on July 7 to the University of Illinois South Farm (40°04′38.89″N, 88°14′26.18″W) and irrigated during establishment.

The field design consisted of a randomized complete block design with three replications. Guard rows were planted around the experimental plot to avoid border effects. Within each replicate, rows of ten plants spaced approximately 0.3 m apart were included for each family, with 0.6 m between rows. Broccoli heads were harvested between August 3 and October 23, with at least five heads collected from each replicate of each family. Heads were cut to similar size and stalk proportions, flash-frozen immediately after harvest, and stored at −20 °C. The florets were lyophilized, ground to a fine powder with a coffee grinder, and stored at −20 °C prior to analysis. Because the families were segregating for maturity, multiple harvests were conducted to obtain heads at uniform maturity. In instances where multiple harvests were required for a given replicate of a family, proportional dry weight per heads harvested on each date were pooled to create a bulk sample for the family replicate. Days to harvest (DTH) were calculated using the transplant date as day 0, with a family DTH value represented by the mean harvest date weighted by heads harvested per day.

Aqueous phenolic extraction

Freeze-dried broccoli powder samples of 75 mg were extracted with 1.5 mL ddH2O in 2-mL microcentrifuge tubes (Fischer Scientific). Tubes were vortexed and stored at room temperature in the dark for 24 h. After incubation, the tubes were centrifuged for 4 min at 12,000 rpm, and the supernatants transferred to new 2-mL tubes. The supernatants were spun for another 2 min, and 200 µL aliquots were pipetted into 96-well plates.

Total phenolic content

The phenolic content of aqueous broccoli extracts was quantified using the Folin–Ciocalteu reagent (FCR) method as described by Ku et al. (2010), with minor modifications. Briefly, 10 µL of sample aqueous extract was pipetted into a 96-well plate, followed by 100 µL of 0.2 M FCR. After 3 min, 90 µL of saturated sodium carbonate solution was added to each well. The samples were incubated for 1 h at room temperature, after which the optical density was measured at 630 nm on a BioTek EL 808 microplate reader (Biotek Instruments Inc., Power Wave XS, Winooski, VT, USA). The total phenolic content was calculated using a gallic acid standard curve, with concentrations ranging from 31.25 to 500 µg/mL and was reported as milligrams of gallic acid equivalents (GAE) per 100 g dry weight. Three analytical replications were measured for each biological replicate (field replicate).

ABTS radical scavenging assay

The antioxidant capacity of aqueous broccoli extracts was quantified by the 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical assay according to Ku et al. (2010). The ABTS radical solution was generated by the treatment of 7 mM ABTS (aqueous) with 2.45 mM potassium persulfate. The mixture was then allowed to stand for 12–16 h for full color development (dark blue-green). The solution was diluted with phosphate-buffered saline until the absorbance (measured at 630 nm) reached 1.0 ± 0.02 for use in the assay reaction. Ten microliter of the aqueous sample extract was treated with 190 µL of 7 mM ABTS. The samples were incubated for 6 min at room temperature and then optical density was read at 630 nm on a microplate reader. The antioxidant capacity was calculated as millimolar 6-hydroxy-2,5,7,8-tetramethylchroman-2-carbonsäure (Trolox) equivalents per 100 g sample dry weight (TE), based upon a Trolox standard curve with concentrations ranging from 0.08 to 2.5 mM. All tests were performed in triplicate.

DPPH radical scavenging assay

A second measure of the antioxidant capacity of aqueous broccoli extracts was taken using the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging assay, according to Ku et al. (2010). Briefly, 0.5 mM DPPH in 9:1 methanol/water was diluted with 90 % methanol (100 % methanol was used for 2009 and 2010) to an optical density of 1.0 ± 0.02 at 490 nm. Then, 190 µL of DPPH methanol solution was added to 10 µL of aqueous broccoli extract in a 96-well plate and incubated for 30 min at room temperature. The optical density was read at 490 nm and the antioxidant capacity calculated as mM Trolox equivalents per 100 g sample dry weight (TE), based on the same Trolox standard curve used in the ABTS assay. All tests were performed in triplicate.

Statistical analysis

Statistical analyses were performed using JMP 12 software (SAS Institute, Cary, NC, USA). Analysis of variance (ANOVA) was performed for all traits with all factors considered fixed. The general linear model was \(y_{ijklm} = \mu + G_{i} + Y_{j} + R(Y)_{jk} + E(Y)_{jl} + \varepsilon_{ijklm}\) where y = the trait measurement associated with the individual ijklm, µ = overall population mean, G = genotype (family), Y = year (macro environment), R(Y) = replication (block) within year, E(Y) = within year environment (DTH), and ε = experimental random error. The interaction between genotype and year was not fitted because there were insufficient degrees of freedom to estimate the effect. F tests of effect significance were obtained using the standard least squares (SLS) method. Pearson’s correlation coefficients were generated for all trait-by-year combinations using the restricted maximum likelihood (REML) method.

Identification of QTL

The linkage map used in this analysis was developed by Brown et al. (2014), with minor updates. It consists of 553 SNP markers from the Brassica napus (AACC) Illumina 60K iSelect array that are named according to the progenitor genome (“A” = B. rapa or “C” = B. oleracea) followed by the position of the SNP as referenced by the “Chiifu-401” or “TO1000” genome sequence, respectively. The map covers 435,229,440 bp of the 446,905,700-bp TO1000 reference assembly (97 %). MapQTL® 5 (Van Ooijen 2004) was used to identify QTL associated with TPC, ABTS, and DPPH assay results from 2009, 2010, and 2014, and the mean of the 3 years. Analysis was performed on the Z-transformation of the average F2:3 family values for each assay from each year. Initially, nonparametric single-factor analysis was performed using the Kruskal–Wallis test to select a group of markers that were subsequently evaluated using the program’s default settings for automated backward-elimination cofactor selection. The resulting model was tested by nonrestricted multiple QTL mapping (MQM) with the default settings adjusted to a scan distance of 0.5 cM. The process was iterated several times to produce an optimal set of cofactors. The genome-wide LOD score threshold value for declaring the presence of a QTL (P < 0.05) was estimated by 1000 permutations of bootstrapping for each phenotypic trait. Confidence intervals were set using a two-LOD drop-off on either side of the maximum score. A QTL was considered common between analyses if the confidence intervals overlapped.

Genes involved in flavonoid biosynthesis and regulation were identified by Blast searches in the TO1000 B. oleracea reference genome (personal communication Yu Chun Chiu and Allan Brown). The results of this analysis are provided in Online Resource 2. Genes from this pool of a priori candidates that co-localized within or adjacent to a significant QTL interval were declared putative candidates genes contributing to the QTL effect.

Results

Total phenolic content and radical scavenging assays

Means, standard deviations, and ranges of TPC, ABTS, and DPPH assay results for the broccoli F2:3 families from 3 years of analysis are presented in Table 1. Frequency distributions of phenolic compound accumulation (from all three assays) approximated normality (Online Resource 3). The population response for the ABTS assay was consistent across the 2009 and 2010 growing seasons, which were both conducted in North Carolina. The 2014 population from Illinois showed ABTS assay results with a tighter distribution and a lower mean (4628.0 ± 486.0 TE) than the 2009 and 2010 populations (5838.7 ± 944.8 TE and 5870.7 ± 1015.8, respectively). DPPH assay results were similar across all 3 years, with the exception that in 2010 the minimum of values extends lower (271–2317 TE) than in 2009 and 2014 (403–2566 TE and 491–2328 TE, respectively). The results of the total phenolic content assay are not informative for absolute quantification because a fresh solution of 7 % sodium carbonate was prepared in 2014 that was more concentrated than the solution used for analysis in 2009 and 2010, resulting in a scalar difference in the response of samples to the assay. However, the relative response of samples in a given year was similar for all 3 years, and rankings of high and low TPC among samples are meaningful.
Table 1

Means, standard deviations, and ranges of the total phenolic content (TPC), ABTS radical scavenging capacity (ABTS), and DPPH radical scavenging capacity (DPPH) from floret tissue of the F2:3 VI-158 × BNC broccoli population over 3 years of analysis

Assay

Mean ± SD

Range

TPC

 2009

366.5 ± 26.6a

298–437

 2010

370.9 ± 26.4

313–463

 2014

857.1 ± 83.4

665–1185

ABTS

 2009

5838.7 ± 944.8b

3610–9359

 2010

5870.7 ± 1015.8

3732–9608

 2014

4628.0 ± 486.0

3689–6245

DPPH

 2009

1256.0 ± 376.6

403–2566

 2010

1107.6 ± 417.3

271–2317

 2014

1204.3 ± 336.2

491–2328

aTPC expressed as gallic acid equivalents/100 g DW

bABTS and DPPH expressed as mM Trolox equivalents/100 g DW

Considerable variation was demonstrated for all traits and in all years, with significant genetic and environmental effects detected by ANOVA (Table 2). Twenty-eight out of thirty-six trait-year combinations were significantly correlated, with the highest correlation seen between ABTS 2014 and DPPH 2014 (r = 0.7632) (Table 3). DPPH 2009 did not correlate with any of the 2014 traits, and TPC 2010 and DPPH 2010 did not correlate with TPC 2014 and DPPH 2014. The final nonsignificant relationship was between TPC 2009 and DPPH 2014.
Table 2

Analysis of variance of total phenolic content (TPC), ABTS radical scavenging capacity (ABTS), and DPPH radical scavenging capacity (DPPH) from floret tissue of the F2:3 VI-158 × BNC broccoli population evaluated in 2009 and 2010 in North Carolina and 2014 in Illinois

Source

TPC

ABTS

DPPH

Family

9.94e5*

1.73e8**

3.81e7**

Rep (year)

1.74e4

3.43e6

2.69e6**

DTH (year)

1.25e5**

2.81e7**

2.77e6**

Year

5.59e6**

8.35e7**

5.23e5

Error

2.77e6

2.37e8

7.27e7

R 2

0.94

0.70

0.43

Values presented are sums of squares

*, ** Significant at P < 0.05 and P < 0.01, respectively

Table 3

Pearson’s correlation coefficients (r) between individual trait-years measured in the F2:3 VI-158 × BNC broccoli population

 

ABTS 09

DPPH 09

TPC 10

ABTS 10

DPPH 10

TPC 14

ABTS 14

DPPH 14

TPC 09

0.535**

0.499**

0.559**

0.226**

0.260**

0.147*

0.169*

0.094

ABTS 09

 

0.525**

0.323**

0.584**

0.250**

0.231**

0.263**

0.302**

DPPH 09

  

0.342**

0.234**

0.345**

0.034

0.110

0.039

TPC 10

   

0.560**

0.410**

0.126

0.182**

0.101

ABTS 10

    

0.500**

0.273**

0.313**

0.292**

DPPH 10

     

0.102

0.196**

0.082

TPC 14

      

0.506**

0.550**

ABTS 14

       

0.763**

*, ** Significant at P < 0.05 and P < 0.01, respectively

QTL associated with phenolic compound accumulation

Analysis identified 60 QTL associated with phenolic compound accumulation at or above the genome-wide LOD threshold (3.7–3.9) across all nine B. oleracea chromosomes (Table 4). Of these, 23 QTL were detected in multiple analyses [three phenolic/antioxidant assays times 4 years (2009, 2010, 2014, and the mean of the 3 years) making a total of 12 trait-year assays]; QTL no. 7 was the most consistent, identified as significant in six of the twelve analyses. It was located on chromosome 1 (C01) between markers Bn-C1-p40578360 and Bn-C1-p41497525 and accounted for 9.6–18.6 % of the phenotypic variation for the respective traits. The putative candidate gene MYB65 co-localized with QTL no. 7.
Table 4

QTL linkage map locations, flanking SNP markers, allele effects, phenotypic variation explained by the QTL, and candidate genes located within or adjacent to significant QTL intervals

QTL

Trait

Chra

Posb

M1c

M2d

LODe

μ_Af

μ_Hg

μ_Bh

%P i

Putative candidates

1

ABTS 09j

C01

2.6

Bn-C1-p00029741

Bn-C1-p00370644

5.11

−0.11

0.75

0.14

15.2

MYB4

ABTS m k

C01

2.6

Bn-C1-p00029741

Bn-C1-p00370644

7.59

1.02

1.11

1.01

17.7

 

2

DPPH 09

C01

13.7

Bn-C1-p01703986

Bn-C1-p01789416

4.8

0.06

0.73

−1.07

12.1

 

3

ABTS 10l

C01

15.5

Bn-C1-p01923307

Bn-C1-p02090899

4.45

0.02

0.65

0.08

8.6

MYB7, HCT

DPPH 09

C01

16.5

Bn_A01_01422384

Bn-C1-p02340807

3.83

−1.53

−1.42

0.56

8.8

 

4

DPPH m

C01

36.2

Bn-C1-p04396925

Bn-C1-p06708408

5.59

0.97

0.99

1.19

12.3

 

5

ABTS 10

C01

40.0

Bn_A01_04948316

Bn-C1-p07392382

4.61

−0.47

−0.15

0.57

8.8

 

6

ABTS 14m

C01

42.7

Bn-C1-p08800375

Bn-C1-p09517731

7.65

−0.65

−0.09

0.67

16.9

MYB92, LDOX, AT4g22870

ABTS m

C01

43.0

Bn-C1-p08918925

Bn-C1-p10181336

9.16

0.93

1.00

1.12

22.1

 

DPPH 14

C01

44.2

Bn-C1-p09517731

Bn-C1-p11794504

14.27

−0.47

0.10

1.26

23.4

 

TPC 09

C01

44.7

Bn_A01_06679669

Bn-C1-p12003811

12.25

0.37

−0.58

1.08

28.2

 

7

TPC 14

C01

71.9

Bn-C1-p40578360

Bn-C1-p41034424

7.26

−0.42

−0.42

0.28

9.6

MYB65

ABTS 09

C01

72.3

Bn-C1-p40875666

Bn-C1-p41170721

4.96

−0.49

0.28

0.57

13.6

 

TPC m

C01

72.3

Bn-C1-p40875666

Bn-C1-p41170721

14.16

0.96

0.99

1.04

20.2

 

DPPH 10

C01

74.7

Bn-C1-p41034424

Bn-C1-p41497525

7.11

−0.92

0.08

−0.42

12.1

 

TPC 09

C01

74.7

Bn-C1-p41034424

Bn-C1-p41497525

8.63

0.60

1.80

0.96

18.6

 

DPPH m

C01

74.7

Bn-C1-p40875666

Bn-C1-p41497525

6.53

1.00

1.25

1.16

16.5

 

8

TPC m

C02

3.0

Bn-C2-p00685796

Bn-C2-p00961123

11.26

0.90

0.99

1.10

15.9

 

9

TPC m

C02

7.3

Bn-C2-p01274486

Bn-C2-p01401759

8.07

1.09

1.02

0.91

10.8

 

10

ABTS 10

C02

23.5

Bn-C2-p02890787

Bn-C2-p03659632

6.03

0.61

0.22

−0.51

12.1

TT4

TPC 10

C02

23.5

Bn-C2-p02868435

Bn-C2-p02971211

11.85

0.13

−0.14

−1.38

19.1

 

11

DPPH 10

C02

55.4

Bn-C2-p42194925

Bn-C2-p43425491

4.97

−1.12

−0.73

−0.22

9

 

12

TPC m

C02

69.5

Bn-C2-p48137237

Bn-C2-p48985239

7.52

0.99

1.03

1.00

8.9

MYB50

13

DPPH m

C03

7.8

Bn-C3-p00240810

Bn-C3-p02003089

4.19

1.02

0.97

1.16

10.1

MYB92

14

ABTS 14

C03

23.4

Bn-C3-p04651804

Bn_A03_04180932

3.97

−0.39

0.06

0.41

8.1

 

15

TPC m

C03

27.9

Bn_A03_04180932

Bn-C3-p06146358

7.08

0.97

1.00

1.03

8.7

MYB48/59

ABTS 10

C03

29.5

Bn-C3-p05308401

Bn-C3-p06280470

4.61

−0.38

0.22

0.48

8.8

 

ABTS m

C03

29.5

Bn-C3-p05308401

Bn-C3-p06280470

4.04

0.97

1.02

1.08

8.2

 

DPPH 10

C03

30.5

Bn-C3-p05308401

Bn-C3-p06280470

6.05

−1.06

−1.03

−0.16

10.8

 

16

DPPH 14

C03

37.7

Bn_A03_05866896

Bn-C3-p10393377

13.03

−0.20

0.67

0.99

20.9

MYB49/101, OMT1, C4H

17

TPC 14

C03

41.8

Bn-C3-p10393377

Bn-C3-p12420975

6.23

−0.74

−0.13

0.64

8.7

UGT73C6, TTG2, CAC3

18

TPC 14

C03

48.4

Bn-C3-p13064413

Bn_A03_10081359

4.24

0.56

0.26

−0.69

5.2

CYP98A3

19

TPC 14

C03

58.6

Bn_A03_11772387

Bn_A03_12737336

6.79

−0.76

−0.49

0.62

9.1

MYB6

20

TPC 14

C03

81.3

Bn-C3-p28076430

Bn-C3-p32296708

5.22

0.16

0.50

−0.32

6.8

4CL5, PAP1, MYB7/113/114

ABTS 14

C03

82.1

Bn-C3-p28076430

Bn-C3-p32296708

4.32

−0.38

0.32

0.36

9.4

 

TPC 09

C03

85.5

Bn-C3-p28076430

Bn-C3-p33785454

4.77

0.35

1.01

1.25

8.8

 

DPPH 14

C03

86.2

Bn-C3-p32296708

Bn_A03_20584351

5.81

−0.19

0.08

0.96

7.9

 

21

TPC 10

C03

114.1

Bn-C3-p57296637

Bn-C3-p60769039

4.87

−0.19

−0.56

−1.05

7.3

MYB48/59

22

TPC m

C03

133.8

Bn-C3-p64658821

Bn-C3-p64972155

7.22

0.97

0.98

1.03

9.3

 

23

TPC 14

C04

9.7

Bn-C4-p01549566

Bn-C4-p02381936

4.51

0.20

0.32

−0.35

5.3

 

ABTS 14

C04

10.2

Bn-C4-p01549566

Bn-C4-p02381936

4.5

0.44

0.35

−0.40

9.4

 

24

TPC 10

C04

13.9

Bn-C4-p02440944

Bn-C4-p02943481

4.12

−0.15

−0.43

−1.09

5.7

 

25

TPC m

C04

31.6

Bn-C4-p06534838

Bn-C4-p08271696

11.57

1.01

1.06

0.99

15.7

TGG2, PAL1/2, UGT73C6

26

DPPH 10

C04

40.1

Bn-C4-p10349525

Bn-C4-p11015238

4.14

−0.64

−0.08

−0.69

7.2

MYB101

27

TPC 09

C04

49.9

Bn-C4-p37595206

Bn-C4-p38747703

6.85

0.68

2.30

0.89

14.7

PAL3

28

TPC 14

C04

54.1

Bn-C4-p45767999

Bn-C4-p43279394

7.72

0.36

−0.44

−0.50

10.5

UF3GT, PAL4

TPC 09

C04

54.3

Bn-C4-p42892908

Bn_A04_13116453

6.74

−0.29

−0.52

1.86

12.3

 

29

DPPH 10

C04

58.1

Bn-C4-p45821281

Bn-C4-p46609817

4.43

−0.96

−1.25

−0.38

7.6

C4H, MYB101

30

TPC 09

C04

66.7

Bn-C4-p49547471

Bn-C4-p50153979

5.87

1.64

0.58

−0.06

10.7

PAL1/2

31

TPC 10

C04

77.8

Bn-C4-p51056930

Bn-C4-p51701964

5.76

−1.13

−0.91

−0.11

8.1

CYP98A3

32

DPPH 14

C04

83.9

Bn-C4-p52177780

Bn-C4-p52368561

10.75

0.92

0.79

−0.11

16.1

 

DPPH 10

C04

85.1

Bn-C4-p52225633

Bn-C4-p52368561

4.56

−1.13

−0.42

−0.04

7.5

 

33

DPPH 09

C05

32.1

Bn-C5-p02588197

Bn-C5-p04170795

5.05

−1.24

0.11

0.27

15.3

CAC2

TPC 10

C05

37.2

Bn-C5-p04170795

Bn-C5-p04997700

3.92

−0.96

−0.28

−0.28

5.1

 

34

DPPH 10

C05

40.7

Bn-C5-p05223886

Bn-C5-p07277640

4.28

−0.42

−1.15

−0.92

7.1

MYB63

DPPH 09

C05

44.1

Bn-C5-p07277640

Bn-C5-p08047780

4.32

0.38

−0.66

−1.35

13

 

35

TPC 10

C05

62.7

Bn-C5-p40304729

Bn-C5-p41959864

4.79

−0.36

−0.09

−0.89

6

BGLU10

TPC 14

C05

62.7

Bn-C5-p40304729

Bn-C5-p41959864

6.85

0.23

−0.49

−0.37

9.1

 

36

TPC 09

C06

3.7

Bn-C6-p00315140

Bn-C6-p01409673

4.4

1.17

0.31

0.48

9.1

MYB63

37

TPC 09

C06

29.5

Bn-C6-p03032332

Bn-C6-p03981788

7.43

0.08

0.48

1.48

14.7

MYB95, MYBL2

TPC m

C06

29.5

Bn-C6-p03032332

Bn-C6-p03981788

11.5

0.97

0.97

1.03

13.6

 

38

ABTS 09

C06

44.6

Bn-C6-p06726844

Bn-C6-p07241341

5.37

−0.33

−0.62

0.48

17.1

AT1g24735, CCOAMT

39

TPC 10

C06

47.3

Bn-C6-p07628656

Bn_A07_14975152

9.08

−0.89

−1.40

−0.29

13.6

4CL3, MYB62

40

DPPH 14

C06

62.6

Bn-C6-p17633267

Bn-C6-p20200556

4.25

0.07

0.25

0.72

5.3

PAL1/2, MYB93, TT12/1/5, 4CL1/2

ABTS 10

C06

63.0

Bn-C6-p19506663

Bn-C6-p31037793

9.00

−0.67

−0.23

0.77

19.5

 

ABTS m

C06

63.0

Bn-C6-p19506663

Bn-C6-p31037793

9.06

0.93

0.99

1.12

21.9

 

ABTS 14

C06

63.5

Bn-C6-p19506663

Bn-C6-p31037793

3.93

−0.38

−0.25

0.45

8.5

 

41

DPPH m

C06

68.8

Bn-C6-p34088688

Bn-C6-p36713260

8.58

0.88

1.11

1.28

22.6

 

42

TPC 14

C07

3.4

Bn-C7-p13115130

Bn-C7-p14997616

6.96

−0.29

−0.71

0.15

8.9

AT1g24735

43

DPPH m

C07

21.9

Bn-C7-p30487360

Bn_A06_22938934

4.79

1.16

0.99

1.01

10.7

 

TPC m

C07

22.4

Bn-C7-p32304272

Bn-C6-p39049912

3.87

0.97

1.00

1.03

4.4

 

DPPH 10

C07

24.6

Bn_A06_22938934

Bn-C7-p33366127

7.09

−1.25

−0.25

−0.08

12.2

 

44

TPC m

C07

34.6

Bn-C7-p35347848

Bn_A06_18388778

6.58

1.04

0.99

0.96

7.6

MYB111

DPPH 10

C07

35.9

Bn-C7-p35347848

Bn-C7-p36343153

9.69

0.49

−0.31

−1.83

18.4

 

45

DPPH 14

C07

46.1

Bn-C7-p39438047

Bn-C7-p40890433

6.95

0.86

0.77

−0.08

10.3

F3H

46

DPPH 10

C07

51.4

Bn-C7-p41241233

Bn-C7-p43333164

10.48

−1.67

−1.35

0.33

20.5

LDOX, AT4g22870

TPC 14

C07

54.7

Bn-C7-p43333164

Bn-C7-p44016812

4.78

0.15

0.32

−0.30

5.7

 

47

DPPH 09

C07

59.1

Bn-C7-p44043402

Bn-C7-p44512932

4.11

−1.06

−0.70

0.04

11.4

OMT1

DPPH m

C07

59.1

Bn-C7-p44043402

Bn-C7-p44512932

4.45

1.00

0.98

1.17

8.6

 

48

TPC 10

C07

62.3

Bn-C7-p44529761

Bn-C7-p45406477

5.15

−0.81

−0.10

−0.43

6.8

 

49

TPC 10

C08

6.3

Bn-C8-p15189574

Bn-C8-p19415560

6.55

−0.41

−1.53

−0.83

9.5

 

50

DPPH 10

C08

13.6

Bn_A08_19541990

Bn-C8-p21602272

4.26

−0.85

−0.04

−0.49

6.5

AHA10

TPC m

C08

14.6

Bn-C8-p21451543

Bn-C8-p21833129

8.25

0.97

1.02

1.03

9.4

 

51

DPPH 09

C08

23.0

Bn-C8-p25706196

Bn-C8-p27802012

4.18

−0.79

0.11

−0.32

11.9

MYB84/77

TPC 10

C08

24.0

Bn-C8-p25706196

Bn-C8-p27802012

13.07

−1.36

0.31

0.06

21.5

 

52

DPPH 14

C08

28.9

Bn-C8-p29068278

Bn-C8-p30785808

7.44

−0.03

0.03

0.82

10.4

TT5

TPC 09

C08

31.4

Bn-C8-p29395056

Bn-C8-p31688245

8.7

0.33

1.44

1.24

17.6

 

TPC 14

C08

33.2

Bn-C8-p30785808

Bn-C8-p31898562

7.32

−0.53

−0.09

0.39

9.9

 

ABTS m

C08

33.2

Bn-C8-p30785808

Bn-C8-p31898562

6.22

0.96

1.01

1.09

13.5

 

ABTS 09

C08

34.2

Bn-C8-p31688245

Bn-C8-p32125534

4.93

−0.46

0.26

0.54

13.9

 

53

ABTS 10

C08

40.3

Bn-C8-p35741949

Bn-C8-p35982715

8.34

−0.61

−0.08

0.71

17.7

F3H

54

ABTS 14

C08

53.6

Bn-C8-p37703534

Bn-C8-p38577059

4.03

−0.48

0.01

0.48

8.8

TTG1

55

TPC 09

C08

60.6

Bn-C8-p39998098

Bn-C8-p40317098

5.56

0.70

−0.01

0.84

9.1

MYB50, AT1g06000

56

TPC 14

C09

8.45

Bn-C9-p02086982

Bn-C9-p02689607

7.19

−0.29

0.49

0.15

9.5

AT5MAT

TPC 10

C09

11.5

Bn-C9-p02309674

Bn-C9-p02919300

4.51

−0.11

−0.42

−1.13

6.1

 

ABTS 09

C09

13.6

Bn-C9-p02689607

Bn-C9-p03558943

4.55

−1.90

−0.45

2.15

13.4

 

57

TPC 09

C09

27.4

Bn_A09_02730673

Bn-C9-p05218392

4.13

1.25

0.53

0.37

7.5

MYB96/94, HCS1

58

TPC 10

C09

41.1

Bn-C9-p05218392

Bn-C9-p08945466

3.87

−0.90

−0.17

−0.52

6.2

MYB84

59

TPC m

C09

81.7

Bn_A10_14394392

Bn-C9-p51214841

5.79

0.98

0.97

1.02

7.5

MYB92

60

TPC 10

C09

93.1

Bn-C9-p51460548

Bn-C9-p52270996

10.19

−0.20

−1.38

−0.90

16.9

TT5

aChromosome number

bMost likely position for QTL on chromosome (cM)

cFront flanking SNP marker

dBack flanking SNP marker

eLogarithm of the odds

fMean for homozygous QTL marker allele from parent 1 (VI-158)

gMean for heterozygous QTL marker allele

hMean for homozygous QTL marker allele from parent 2 (BNC)

iPercent of the phenotypic variation explained by the QTL (overestimated for linked QTL)

j2009

kMean

l2010

m2014

The second most consistent QTL, no. 52, was identified in five analyses. QTL no. 52 was located on chromosome 8 (C08) between markers Bn-C8-p29068278 and Bn-C8-p32125534. It explained 9.9–17.6 % of the phenotypic variation in TPC 09, ABTS 09, TPC 14, DPPH 14, and the ABTS mean. QTL no. 52 functioned primarily in an additive manner, with the positive allele contributed by the parent BNC. The candidate gene chalcone-flavanone isomerase 1 [CHI, aka TRANSPARENT TESTA 5 (TT5)] was identified within this interval at 29,634,423 bp.

Four QTL were consistent across four analyses: QTL no. 6, no. 15, no. 20, and no. 40. QTL no. 6 was located on C01 between markers Bn-C1-p08800375 and Bn-C1-p12003811. This locus explained up to 28.2 % of the phenotypic variation in TPC 09, ABTS 14, DPPH 14, and the ABTS mean, the greatest contribution of any individual QTL. BNC was the parent that contributed the positive allele, with the gene effect functioning in a primarily additive manner. Three candidate genes were identified within QTL no. 6: leucoanthocyanidin dioxygenase [LDOX, aka anthocyanidin synthase (ANS)], AT4g22870 (an LDOX-like protein), and MYB92.

On chromosome 3 (C03), QTL no. 15 was flanked by markers Bn_A03_04180932 and Bn-C3-p06280470. 8.2–10.8 % of the phenotypic variation was explained by this QTL in ABTS 10, DPPH 10, the TPC mean, and the ABTS mean. The positive additive gene effect was contributed by the BNC allele. Two putative candidate genes, MYB48 and MYB59, co-localized with QTL no. 15.

Also on C03, Bn-C3-p28076430 and Bn_A03_20584351 flanked QTL no. 20. Associated with all three traits from 2014 and TPC 09, it accounted for 6.8–9.4 % of the phenotypic variation. Five candidate genes were identified within the QTL no. 20 interval, specifically 4-coumarate: CoA ligase 5 (4CL5), production of anthocyanin pigment 1 (PAP1, aka MYB75), MYB7, MYB113, and MYB114.

QTL no. 40 was located on chromosome 6 (C06) between markers Bn-C6-p17633267 and Bn-C6-p31037793, explaining up to 21.9 % of the variation in ABTS 10, ABTS 14, DPPH 14, and the ABTS mean. BNC contributed the positive allele at this locus, demonstrating an additive effect. The no. 40 interval contained eight putative candidate genes, the most of any QTL. These genes included two copies of phenylalanine ammonia-lyase (PAL), two copies of 4-coumarate: CoA ligase (4CL), MYB93, CHI, TRANSPARENT TESTA 1 (TT1), and TT12.

Two QTL, no. 43 on chromosome 7 and no. 56 on chromosome 9, were identified three times. The remaining 15 QTL that were identified in multiple analyses were each identified twice. Their chromosomal distribution was as follows: 2 on C01, 1 on C02, 3 on C04, 3 on C05, 1 on C06, 3 on C07, and 2 on C08.

Of the 60 total QTL, only 15 lacked an a priori candidate gene. These loci could contain novel regulators of phenolic compound accumulation that have not yet been reported. The most frequently identified a priori candidate genes were PAL, CHI, and 4CL. PAL was identified in five QTL regions and nine individual analyses (trait-year combinations). CHI and 4CL were each identified in three QTL regions associated with ten and nine trait-years, respectively. As a gene family, the MYB transcription factors were most heavily represented among the identified candidate genes. Twenty-four different MYB genes were identified within significant QTL intervals.

Discussion

Characterization of the variation in total phenolic content and antioxidant capacity in the VI-158 × BNC broccoli mapping population corroborated the expectation that both genetic and environmental factors play a significant role in determining phenolic compound accumulation. The genetic factor (family) was highly significant (P < 0.001) in all three assays, which allowed for successful QTL mapping of this genetic effect.

The DPPH assay appears to have been more prone to experimental error or was influenced by uncharacterized factors, as its R 2 value in the ANOVA was only 0.42871. The most likely source of interference is the presence of soluble protein in the aqueous broccoli extracts that may be precipitated out in the methanol solution used in the DPPH assay. This issue was largely mitigated in the 2014 DPPH analysis by altering the solvent system from 100 % methanol to a 9:1 methanol/ water solution. This improvement is reflected in the highly significant correlation between the ABTS and DPPH results in 2014 (r = 0.7632). However, the paired correlation between the ABTS and DPPH results in each year of analysis suggests that similar effects were captured by both assays, albeit with less power by DPPH.

For TPC and ABTS, both in-season environmental effects captured by the factor DTH and the specific year/location environmental effect (the factor termed “year”) significantly contributed to the variation in ANOVA. The DTH factor also made a significant contribution to the variation in DPPH. This highlights the importance of measuring phenolic content over multiple environments and years in order to capture genes that control phenolic accumulation in response to different environmental stimuli. To our knowledge, this is the first QTL mapping study of phenolic compound accumulation and antioxidant capacity conducted in a broccoli population over multiple environments. Analysis of the VI-158 × BNC mapping population is particularly powerful due to its dense genetic linkage map composed SNP markers anchored to the TO1000 rapid cycling B. oleracea reference sequence.

The identification of PAL, 4CL, and CHI as the most consistently occurring candidate genes is encouraging because it suggests that the influence of these early phenylpropanoid biosynthesis genes is stable across environments. PAL, the gene responsible for the first committed step in phenylpropanoid biosynthesis, was a putative candidate within five different loci. Of these, one was consistent across four analyses and another was identified in two analyses. CHI was the gene candidate associated with the second most consistent QTL, no. 52, which was associated with five trait-years, as well as QTL no. 40, which was consistent across four trait-years. 4CL also co-localized with QTL no. 40, and with another locus consistent across four analyses, QTL no. 20. The remaining three of the first six enzymes involved in flavonoid biosynthesis were also identified as putative candidates within at least one QTL. Cinnamic acid 4-hyrdoxylase (C4H) was identified within no. 16 and no. 29, chalcone synthase (CHS aka TT4) co-localized with no. 10, and flavanone 3-hydroxylase (F3H) was identified within no. 45 and no. 53. These results are consistent with another recent QTL mapping study of antioxidant activity in a rapid cycling B. oleracea × B. oleracea var. italica population that identified 4CL and other phenylpropanoid biosynthesis genes as QTL candidates (Sotelo et al. 2014). The frequency of co-localization of significant QTL with core phenylpropanoid biosynthesis genes, especially PAL, 4CL, and CHI, suggests the presence of useful genetic variation within these genes that can be leveraged to make substantial breeding progress.

While the biosynthetic genes are clearly of importance, this analysis presents stark evidence that an array of transcription factors play a major role in determining phenolic compound accumulation. 24 MYB transcription factors, a WRKY transcription factor, and a WD40 repeat protein, known to complex with MYBs, were all identified as putative candidates. This complexity is not surprising given the diverse roles of phenolic compounds in plant survival. The regulatory effect of PAP1 (MYB75) is a good illustration of the integration of environmental signals in phenolic compound biosynthesis. Induction of PAP1 expression has been demonstrated in response to high light conditions and nitrogen and phosphorous depletion. Downstream targets activated by PAP1 include 20 flavonoid biosynthesis genes, resulting in accumulation of flavonoids in response to abiotic stress (Lotkowska et al. 2015). PAP1 was identified as a candidate gene within QTL no. 20 (consistent across four analyses), along with three other known activators of phenylpropanoid biosynthesis (MYB7, MYB113, and MYB114).

One intriguing example that emerged from this research was the identification of MYB65 as the putative candidate gene underlying QTL no. 7, which was the most consistent QTL. It was identified in six analyses that included all 3 years (TPC 09, ABTS 09, DPPH 10, TPC 14, the TPC mean, and the DPPH mean). MYB65 has close homology to the barley transcription factor GAMYB, which is known to transduce gibberellin signaling. In Arabidopsis, MYB65 has been shown to play a role in anther development, with its primary expression localized to the floral tissue (Millar and Gubler 2005). This is interesting because phenolic compounds are known to be important constituents of pollen grain walls (Quilichini et al. 2014) and because floral tissue is the major consumed portion of broccoli. The role of phenolic compounds in plant development logically necessitates the regulation of their biosynthesis in coordination with developmental programming.

With an eye toward crop improvement, the involvement of such a complex regulatory network may at first glance appear to hinder the ability of breeders or producers to enhance phenolic compound accumulation. However, overexpression or repression of MYB activity through the use of transgenes and miRNA systems has the potential to profoundly change phenolic profiles in Brassica vegetables. Proof of concept has been demonstrated in tomato, where overexpression of AtMYB12 has been shown to up-regulate the biosynthesis of caffeoylquinic acids and flavonols (Luo et al. 2008).

While substantial environmental effects on phenolic compound accumulation are a challenge for consistency in the performance of field-grown crops, it also provides the opportunity for cultural practices to impact phenolic compound accumulation. An even greater impact may be seen in controlled environment systems such as greenhouse production, where the manipulation of light could be used to enhance phenolic compound accumulation. This is particularly relevant for specialty market production, as for broccoli sprouts, which may be marketed and sold at a premium for enhancing health-promoting phytochemical composition.

The results of this study highlight a number of targets for the enhancement of phenolic compound accumulation in Brassica cultivars. While SNP markers from the QTL could be used for immediately marker-assisted selection, fine mapping of these QTL is needed to provide greater functional characterization and validate putative candidates.

Notes

Acknowledgments

This research was supported in part by the Hatch Multistate Project NC-7 (ILLU-802-931). We kindly acknowledge Professor Ray Ming for the use of his QTL mapping software, Dr. Kang Mo Ku for his help with the antioxidant assays, and Dr. Talon Becker for assistance with field planting, harvesting, experimental design.

Supplementary material

11032_2016_497_MOESM1_ESM.pdf (255 kb)
Supplementary material 1 (PDF 255 kb)
11032_2016_497_MOESM2_ESM.pdf (158 kb)
Supplementary material 2 (PDF 157 kb)
11032_2016_497_MOESM3_ESM.pdf (777 kb)
Supplementary material 3 (PDF 777 kb)

References

  1. Bergmann H, Triebel S, Kahle K, Richling E (2010) The metabolic fate of apple polyphenols in humans. Curr Nut Food Sci 6(1):19–35CrossRefGoogle Scholar
  2. Brown AF, Yousef GG, Jeffery EH, Klein BP, Wallig MA, Kushad MM, Juvik JA (2002) Glucosinolate profiles in broccoli: variation in levels and implications in breeding for cancer chemoprotection. J Am Soc Hortic Sci 127(5):807–813Google Scholar
  3. Brown AF, Jeffery EH, Juvik JA (2007) A polymerase chain reaction-based linkage map of broccoli and identification of quantitative trait loci associated with harvest date and head weight. J Am Soc Hortic Sci 132(4):507–513Google Scholar
  4. Brown AF, Yousef GG, Chebrolu KK, Byrd RW, Everhart KW, Thomas A, Reid RW, Parkin IA, Sharpe AG, Oliver R (2014) High-density single nucleotide polymorphism (SNP) array mapping in Brassica oleracea: identification of QTL associated with carotenoid variation in broccoli florets. Theor Appl Genet 127(9):2051–2064CrossRefPubMedGoogle Scholar
  5. Brown AF, Yousef GG, Reid RW, Chebrolu KK, Thomas A, Krueger C, Jeffery E, Jackson E, Juvik JA (2015) Genetic analysis of glucosinolate variability in broccoli florets using genome-anchored single nucleotide polymorphisms. Theor Appl Genet 128(7):1431–1447Google Scholar
  6. Cheynier V, Comte G, Davies KM, Lattanzio V, Martens S (2013) Plant phenolics: recent advances on their biosynthesis, genetics, and ecophysiology. Plant Physiol Biochem 72:1–20CrossRefPubMedGoogle Scholar
  7. Clifford MN (2004) Diet-derived phenols in plasma and tissues and their implications for health. Planta Med 70(12):1103–1114CrossRefPubMedGoogle Scholar
  8. Crozier A, Jaganath IB, Clifford MN (2009) Dietary phenolics: chemistry, bioavailability and effects on health. Nat Prod Rep 26(8):1001–1043CrossRefPubMedGoogle Scholar
  9. Dinkova-Kostova AT, Massiah MA, Bozak RE, Hicks RJ, Talalay P (2001) Potency of michael reaction acceptors as inducers of enzymes that protect against carcinogenesis depends on their reactivity with sulfhydryl groups. Proc Natl Acad Sci USA 98(6):3404–3409CrossRefPubMedPubMedCentralGoogle Scholar
  10. Forman HJ, Davies KJA, Ursini F (2014) How do nutritional antioxidants really work: nucleophilic tone and para-hormesis versus free radical scavenging in vivo. Free Radic Biol Med 66:24–35CrossRefPubMedGoogle Scholar
  11. Heimler D, Vignolini P, Dini MG, Vincieri FF, Romani A (2006) Antiradical activity and polyphenol composition of local Brassicaceae edible varieties. Food Chem 99(3):464–469CrossRefGoogle Scholar
  12. Hertog ML, Kromhout D, Aravanis C et al (1995) Flavonoid intake and long-term risk of coronary heart disease and cancer in the seven countries study. Arch Intern Med 155(4):381–386CrossRefPubMedGoogle Scholar
  13. Kim M, Choi S-, Lee P, Hur J (2015) Neochlorogenic acid inhibits lipopolysaccharide-induced activation and pro-inflammatory responses in BV2 microglial cells. Neurochem Res 40(9):1792–1798CrossRefPubMedGoogle Scholar
  14. Knekt P, Kumpulainen J, Jarvinen R, Rissanen H, Heliovaara M, Reunanen A, Hakulinen T, Aromaa A (2002) Flavonoid intake and risk of chronic diseases. Am J Clin Nutr 76(3):560–568PubMedGoogle Scholar
  15. Ku KM, Choi JN, Kim J, Kim JK, Yoo LG, Lee SJ, Hong Y, Lee CH (2010) Metabolomics analysis reveals the compositional differences of shade grown tea (Camellia sinensis L.). J Agric Food Chem 58(1):418–426CrossRefPubMedGoogle Scholar
  16. Kushad MM, Brown AF, Kurilich AC, Juvik JA, Klein BP, Wallig MA, Jeffery EH (1999) Variation of glucosinolates in vegetable crops of Brassica oleracea. J Agric Food Chem 47(4):1541–1548CrossRefPubMedGoogle Scholar
  17. Lotkowska ME, Tohge T, Fernie AR, Xue G, Balazadeh S, Mueller-Roeber B (2015) The arabidopsis transcription factor MYB112 promotes anthocyanin formation during salinity and under high light stress. Plant Physiol 169(3):1862–1880PubMedPubMedCentralGoogle Scholar
  18. Luo J, Butelli E, Hill L, Parr A, Niggeweg R, Bailey P, Weisshaar B, Martin C (2008) AtMYB12 regulates caffeoyl quinic acid and flavonol synthesis in tomato: expression in fruit results in very high levels of both types of polyphenol. Plant J 56(2):316–326CrossRefPubMedGoogle Scholar
  19. Ma Q (2014) Advances in mechanisms of anti-oxidation. Discovery Medicine 17(93):121–130PubMedPubMedCentralGoogle Scholar
  20. Masella R, Di Benedetto R, Varì R, Filesi C, Giovannini C (2005) Novel mechanisms of natural antioxidant compounds in biological systems: involvement of glutathione and glutathione-related enzymes. J Nutr Biochem 16(10):577–586CrossRefPubMedGoogle Scholar
  21. Millar AA, Gubler F (2005) The arabidopsis GAMYB-like genes, MYB33 and MYB65, are microRNA-regulated genes that redundantly facilitate anther development. Plant Cell 17(3):705–721CrossRefPubMedPubMedCentralGoogle Scholar
  22. Osburn WO, Kensler TW (2008) Nrf2 signaling: an adaptive response pathway for protection against environmental toxic insults. Mutat Res Rev Mutat Res 659(1–2):31–39CrossRefGoogle Scholar
  23. Parkin IAP, Koh C, Tang H, Robinson SJ, Kagale S, Clarke WE, Town CD, Nixon J, Krishnakumar V, Bidwell SL et al (2014) Transcriptome and methylome profiling reveals relics of genome dominance in the mesopolyploid Brassica oleracea. Genome Biol 15(6):R77CrossRefPubMedPubMedCentralGoogle Scholar
  24. Price KR, Casuscelli F, Colquhoun IJ, Rhodes MJC (1997) Hydroxycinnamic acid esters from broccoli florets. Phytochemistry 45(8):1683–1687CrossRefGoogle Scholar
  25. Price KR, Casuscelli F, Colquhoun IJ, Rhodes MJC (1998) Composition and content of flavonol glycosides in broccoli florets (Brassica olearacea) and their fate during cooking. J Sci Food Agric 77(4):468–472CrossRefGoogle Scholar
  26. Quilichini TD, Samuels AL, Douglas CJ (2014) ABCG26-mediated polyketide trafficking and hydroxycinnamoyl spermidines contribute to pollen wall exine formation in Arabidopsis. Plant Cell 26(11):4483–4498. doi: 10.1105/tpc.114.130484 CrossRefPubMedPubMedCentralGoogle Scholar
  27. Saito K, Yonekura-Sakakibara K, Nakabayashi R, Higashi Y, Yamazaki M, Tohge T, Fernie AR (2013) The flavonoid biosynthetic pathway in arabidopsis: structural and genetic diversity. Plant Physiol Biochem 72:21–34CrossRefPubMedGoogle Scholar
  28. Sotelo T, Cartea ME, Velasco P, Soengas P (2014) Identification of antioxidant capacity-related QTLs in Brassica olearacea. PLoS One 9(9):e107290CrossRefPubMedPubMedCentralGoogle Scholar
  29. Sun CL, Yuan JM, Koh WP, Yu MC (2006) Green tea, black tea and breast cancer risk: a meta-analysis of epidemiological studies. Carcinogenesis 27(7):1310–1315CrossRefPubMedGoogle Scholar
  30. Tang N, Wu Y, Zhou B, Wang B, Yu R (2009) Green tea, black tea consumption and risk of lung cancer: a meta-analysis. Lung Cancer 65(3):274–283CrossRefPubMedGoogle Scholar
  31. Van Ooijen J (2004) MapQTL® 5. Software for the mapping of quantitative trait loci in experimental populations. Kayazma B.V., WageningenGoogle Scholar
  32. Weng C, Chen M, Yeh C, Yen G (2011) Hepatoprotection of quercetin against oxidative stress by induction of metallothionein expression through activating MAPK and PI3K pathways and enhancing Nrf2 DNA-binding activity. New Biotechnol 28(6):767–777CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2016

Authors and Affiliations

  • Alicia M. Gardner
    • 1
  • Allan F. Brown
    • 2
  • John A. Juvik
    • 1
  1. 1.Department of Crop SciencesUniversity of IllinoisUrbanaUSA
  2. 2.International Institute of Tropical AgricultureArushaTanzania

Personalised recommendations