Skip to main content
Download PDF

Genome-wide association study of kernel colour traits and mining of elite alleles from the major loci in maize

BMC Plant Biology volume 24, Article number: 25 (2024) Cite this article

Abstract

Background

Maize kernel colour is an important index for evaluating maize quality and value and mainly entails two natural pigments, carotenoids and anthocyanins. To analyse the genetic mechanism of maize kernel colour and mine single nucleotide polymorphisms (SNPs) related to kernel colour traits, an association panel including 244 superior maize inbred lines was used to measure and analyse the six traits related to kernel colour in two environments and was then combined with the about 3 million SNPs covering the whole maize genome in this study. Two models (Q + K, PCA + K) were used for genome-wide association analysis (GWAS) of kernel colour traits.

Results

We identified 1029QTLs, and two SNPs contained in those QTLs were located in coding regions of Y1 and R1 respectively, two known genes that regulate kernel colour. Fourteen QTLs which contain 19 SNPs were within 200 kb interval of the genes involved in the regulation of kernel colour. 13 high-confidence SNPs repeatedly detected for specific traits, and AA genotypes of rs1_40605594 and rs5_2392770 were the most popular alleles appeared in inbred lines with higher levels. By searching the confident interval of the 13 high-confidence SNPs, a total of 95 candidate genes were identified.

Conclusions

The genetic loci and candidate genes of maize kernel colour provided in this study will be useful for uncovering the genetic mechanism of maize kernel colour, gene cloning in the future. Furthermore, the identified elite alleles can be used to molecular marker-assisted selection of kernel colour traits.

Peer Review reports

Introduction

Maize (Zea mays L.) is one of the world's three major staple foods and is an important feed and industrial crop [1]. Kernel colour is the main index used to evaluate its commodity quality and value. The pigment in maize grain has a high nutritional value, and maize kernels have healthy beneficial functions [2, 3]. Maize kernel pigments mainly include two kinds of natural pigments: carotenoids and anthocyanins [4, 5]. Among them, carotenoids serve a variety of functions, including antioxidative, immune regulation, anticancer and antiaging functions; they are natural antioxidants and colorants and are also the major source of vitamin A in animals [6, 7]. Anthocyanins are flavonoid pigments that also have good antioxidant, free radical scavenging, antitumour, antiaging and skin beautifying effects and have many medical applications [2, 4, 8, 9].

Plants can produce carotenoids, instead of vitamin A, while carotenoids can be converted into vitamin A with physiological activity in the body, a process that can provide necessary vitamin A [6]. Not eating enough vitamin A can result in vitamin A deficiency (VAD), characterized by night blindness, anaemia, impaired immunity, and even death [10]. Instantly supplying vitamin A or improving the diet can reverse the effects of VAD, but chronic lack of vitamin A can lead to irreversible effects. Maize, a major food crop worldwide, is the main focus of carotenoid biofortification [11, 12].

Maize kernel colour is related to the key gene in carotenoid biosynthesis [13]. Y1, encoding phytoene synthase, is the first committed step in carotenoid biosynthesis [14]. Y1 results in an orange kernel colour, and y8 transforms the endosperm of the Y1 background into light yellow. Wc also converts the yellow endosperm caused by Y1 into white [15, 16]. Lycopene is the branch point of carotenoid biosynthesis and is regulated by lycopene β-cyclase (LCYB) and lycopene ε-cyclase (LCYE) [17, 18]. Downregulated expression of LcyE can cause a greater accumulation of β-branch carotenoids than α-branch carotenoids, and LcyE is located on chr8 and affects kernel colour [19, 20]. Ps1 and Vp5 are involved in core carotenoid biosynthesis, and Vp5 is a white kernel mutant that lacks ABA [21]. Vp14 is involved in the cleavage of carotenoids and is also related to the accumulation of kernel colour [16, 22].As inprevious study, two major QTLs were mapped on chr6 and chr9, and one of them was a Y1 gene that controls the coloration of yellow and white kernel [23]. Brenda et al. (2019) identified the known genes Y1 and DXS2 by association analysis and explored the relationship between DXS3, DMES1, LCYE, EP1 and the formation of kernel colour [13]. Chandler performed the visual scored by kernel colour for GWAS, and 11 QTLs were identified, y1, lcyE, zep1, and ccd1 were associated with common QTL [22].

Anthocyanins are water-soluble flavonoids in plants that give plants a wide variety of colours and are robust against adversity experienced by plants [9]. Anthocyanins have an abundance of nutritional and medicinal abilies that enable all kinds of good things for human health [2, 24, 25]. Anthocyanins exist mostly in the aleurone layer of maize kernels, and purple corn is particularly rich in anthocyanins [4, 26]. The anthocyanin biosynthetic pathway involves many structural genes and the regulatory factors of these structural genes [4]. Structural genes encoding chalcone synthase (CHS) and chalcone isomerase (CHI) are key enzymes upstream of the anthocyanin biosynthetic pathway, and their expression levels are positively associated with anthocyanin content [27, 28]. The typical anthocyanin regulatory complex MBW consists of an R2R3-MYB protein, a basic helix-loop-helix (bHLH) protein, and a WD-repeat (WDR) domain protein [29, 30]. Studies of the anthocyanin biosynthetic pathway in maize show that the formation of purple aleurone is controlled by multiple genes, including coloured aleurone 1 (C1, MYB), coloured 1 (R1, bHLH), and pale aleurone colour 1 (Pac1, WDR) [31,32,33,34]. Intensifier1 (In1), with similarity to R1, encodes a bHLH-like recessive intensifier that increases the accumulation of anthocyanins in starch [32]. Booster1 (B1) and plant color1 (Pl1) are regulators of bHLH and MYB, respectively, and are related to the regulation of plant tissues [35].

Mature anthocyanins are transported into vacuoles for storage [4, 36]. According to reports, glutathione S-transferases (GSTs) catalyse conjugation between γ-Glu-Cys-Gly (GSH) and cyanidin-3-glucoside (C3G) or transport anthocyanins into the vacuolar membrane as carriers [37, 38]. Multidrug resistance-associated protein (mrp), an ABC transporter located on vacuolar membranes, can recognize anthocyanins and transport them across the membrane to the vacuole. Multidrug-resistant resistance-associated proteins located on the vacuole membrane recognize anthocyanin glycosides and transport them into the vacuole across the membrane [39, 40]. Chatham and Juvik performed association mapping in purple corn populations, Major QTLs for anthocyanin type were identified: Pr1, R1 and the plant color-associated MYB, Pl1 [4]. The GWAS of colour variation in rice, twenty-six loci were identified, and at least three candidates involved in the flavonoid metabolic pathway [41]. In durum wheat, The genetic mapping identified 4 QTLs disclosed the candidate genes Pp-A3, Pp-B1, R-A1, R-B, bHLH (Myc-1) and MYB (Mpc1, Myb10) [42].

The formation of maize kernel colour is regulated by a series of genes, and it is easily observed and closely related to nutritional quality. In this study, we used natural populations including 244 maize elite inbred lines and used the RGB colour model and visual classification (level) to evaluate the kernel colour, which served as the phenotype. To perform genetic analysis and mine the important genes, we then conducted a genome-wide association study on kernel colour with 3 million SNP markers covering the whole genome of maize. This will help in the deep examination of maize nutritional function and will provide support for the development of the maize industry from the quality of appearance.

Results

Phenotypic data statistical analysis

A statistical analysis was conducted for 6 kernel colour traits of the associated populations in two environments. The results showed that the variation range of the 6 traits was large, and the variation coefficients (CV) were all greater than 10% (Table 1). The absolute values of skewness and kurtosis were lower (Table 1). The maize kernel colour was divided into 7 grades by visual scoring, the darker kernel colour with the higher grades. Kernel colour had a rich genetic diversity in this population (Table 1). Combined with the frequency histogram, these 6 traits satisfy to the heredity of quantitative traits (Figure S1). The correlation coefficient between level 1 and level 2 was 0.56 (P < 0.001) (Fig. 1). There was a negative correlation between level and other colour traits, and a positive correlation between B and R, G, B, and RGB; the correlation coefficient was 0.3–0.53 across the two environments (Fig. 1). The correlation between R, G, B, and RGB was 0.61–0.97, showing a significant positive correlation (Fig. 1). The heritabilities of colour traits were 0.43, 0.59, 0.78, 0.6, 0.43 and 0.8, respectively (Fig. 1), which showed that the colour traits had higher heritability. Correlation analysis showed that the colour traits were positively correlated between the two environments.

Table 1 Descriptive statistics for kernel colour traits in two environments
Full size table
Fig. 1
figure 1

The correlations analysis among the 6 kernel colour traits in two environments. _1: wengyuan experimental station (2020); _2: guangzhou experimental station (2021); The number in the rectangle is the correlation coefficient, Purple indicates negative correlation, Cyan indicates positive correlation, the darker the color the higher the correlation

Full size image

Genome-wide association analysis

GWAS was performed using the Q + K model and PCA + K model in a mixed linear model (MLM), which analysed the 6 colour traits of 244 maize inbred lines across two environments. The QQ plot showed that the model for GWAS was reasonable (Figure S2), and Manhattan plots for each trait are presented in Fig. 2. To combine significant SNPs into QTL intervals The SNPs were in a 50 kb range as a QTL (Table S1, S2, S3 and S4). Under the Q + K model, in total, we identified 877 QTLs significantly associated with kernel colour in the two environments, and 590 QTLs and 440 QTLs were identified in 2020 and 2021, respectively. Among them, 154 QTLs were identified by at least two traits (Fig. 3, Table S4). Under the PCA + K model, we totally identified 475 QTLs significantly associated with kernel colour, 356 QTLs and 304 QTLs in 2020 and 2021, respectively. 163 QTLs of them were identified by at least two traits (Fig. 3, Table S3). A total of 263 QTLs were identified by two models, and 94 were identified by at least two traits. Thirteen QTLs were identified by at least two traits and two environments (Table S5). These loci have important research value and were distributed across the 10 chromosomes.

Fig. 2
figure 2

Manhattan-plots for GWAS of 6 kernel colour traits in maize.Two GWAS models for the control of false positive (Q-Q plots). The manhattan plots of two models include MLM_PCA + K (left) and MLM_Q + K (right); E1: wengyuan experimental station (2020); E2: guangzhou experimental station (2021)

Full size image
Fig. 3
figure 3

Number of significant QTLs and stable QTLs for the concentration of the 6 kernel colour traits in two environments and GWAS models. A, B, C, D, E and F is R, B, RGB, level, G and Gray, respectively. E1: wengyuan experimental station (2020); E2: guangzhou experimental station (2021). Horizontal bars show the number of QTLs for different environments and methods. The colours of circles corresponding to Horizontal bars indicate the environment in which QTLs was detected and the method applied

Full size image

Two known genes involved in maize carotenoid biosynthesis were detected. A single SNP rs6_85061523 that significant associations with B was detected in the coding region of Y1, with a small MAF (0.06) (Table 2), and the R2 was 0.29. Y1 encodes phytoene synthase, which is the key enzyme in the first step of carotenoid biosynthesis and is a typical yellow-and-white gene. The rs3_219867520 that also significant association with B was in the region of A1 gene, with a small MAF (0.08), and the R2 was 0.1 (Table 2). which encodes bifunctional dihydroflavonol 4-reductase (DFR).

Table 2 the carotenoid-related and anthocyanins-related genes within 200 kb of most significant SNP for each trait
Full size table

The significant SNP rs9_20232174 was near the previously identified Dxs3 gene, was approximately 7.4 kb, and had a small MAF (0.06) (Table 2). Dxs3 encodes a 1-deoxy-D-xylulose 5-phosphate synthase, which catalyses the first and committed step of the MEP pathway [43]. Cgt1 was located 5.7 kb downstream of rs6_123785816 and MAF (0.07) (Table 2), which encodes c-glucosyl transferase and is a structural gene in the anthocyanin biosynthetic pathway [44]. A SNP located on chromosome 9 (15,687,532) was located 25 kb upstream of hyd5 and MAF (0.12) (Table 2), which encodes an enzyme with hydroxylase domains and plastid-targeting signals and is involved in carotenoid degradation. Psy2 was approximately 26 kb away from SNP rs8_173494185 and MAF (0.12) (Table 2), which encodes phytoene synthase and is involved in the carotenoid biosynthesis pathway [45].

In addition, the physical distance between at least 8 QTLs and kernel colour regulation genes is less than 200 kb (Table 2). Three of them are involved in carotenoid biosynthesis: whitecap1 (Wc1) carotenoid cleavage dioxygenase1, which catalyses the cleavage of carotenoids to their corresponding apo-carotenoid products [46]; Dxs1, which catalyses the first and committed step of the MEP pathway [47]; and Crti3, which encodes carotenoid isomerase 3 [20]. Five of the genes are involved in the anthocyanin biosynthetic pathway, 4 of which are structural genes: colored1 (R1); anthocyaninless 2 (A2); chalcone isomerase 1 (Chi1); chalcone isomerase 3 (Chi3) and Bronze 1 (Bz1), which encodes UDP-glucose flavonol glycosyltransferase [4].

Candidate genes

Based on the B73 reference genome (B73 ref_V4), we obtained 136 candidate genes within 200 kb upstream and downstream of 13 high confidence SNPs, and 95 of them had function annotation (Table S6). Three key candidate genes were selected based on the gene annotation. Zm00001d048621 encodes an ABC transporter involved in anthocyanin transport; Zm00001d048626 encodes a cytochrome P450 enzyme; Zm00001d048623 encodes the MYB transcription factor MYB59.

The effect of allelic variation

The R2 of 13 SNPs with high credibility ranged from 0.6%-23.2%, and the analysis of the phenotypic data showed that there were significant correlations between the phenotypic data for the 6 kernel colours and each dominant SNP among different allelic variation inbred lines (Fig. 4). For example, AA genotypes at rs1_40605594 sites and rs5_2392770 were largely detected in inbred lines with higher levels, such as yellow or purple kernels rich in anthocyanins and carotenoids. AA genotypes at rs2_231499616 and rs7_22639260 sites were largely concentrated in inbred lines with higher B. Therefore, the SNPs mined in this study have significant effects on maize kernel colour and are important targets for genetic improvement of maize kernel colour.

Fig. 4
figure 4

The superior and alternative alleles. _1: wengyuan experimental station (2020); _2: guangzhou experimental station (2021). P < 0.05: differences, P < 0.01: significant differences, P < 0.001: highly significant differences

Full size image

Discussion

Phenotypic analysis of kernel colour

In this study, we performed statistical analysis of the kernel colour phenotype data in two environments and the variation rates of phenotypes were more than 10% higher (Table 1), indicating that the kernel colour of these associated populations was diversity. We also found a certain correlation between two environmental factors (Fig. 1). However, the correlation coefficient was small, which may be due to environmental differences and other factors. A heritability analysis showed that kernel colour had higher heritability (Table 1). This result indicates that kernel colour is mainly regulated by genetic factors, and also influenced by environmental factors [48].

GWAS model selection analysis

With the rapid development of plant genomics, the development and application of sequencing technology and cost reduction, quantitative trait loci (QTL) and GWAS have been widely used to analyse the genetic basis of plant traits [49]. GWAS is a way to mine genetic variation based on linkage disequilibrium, and there have been many GWAS statistical models [13, 20]. The research shows that the maize GWAS is affected by the community structure and kinship, so choosing the best statistical model to study the relationship between genotypes and traits increases the statistical effect of GWAS [16]. In this study, we used two statistical models, Q + K and PCA + K, and found that the two models could control false positives well (Fig. 2, Figure S2). But because the algorithms vary, Phenotype G and Gary's data in 2020 did not get a reasonable result under the Q + K model, so we analysed the GWAS results of these two models simultaneously.

Comparative analysis of kernel colour location results

Maize kernel colour is a quantitative trait controlled by multiple genes and has stable heritability. By dividing 2448 inbred lines into 12 levels according to visual kernel colour, and 11 QTLS were identified through linkage analysis, half of which were related to carotenoid biosynthesis genes. Research findings showed that the visual score could be applied to studies of kernel colour [22]. With the same method. Lin et al. (2021) identified a major QTL on chromosome 6 and chromosome 9, and one QTL was Y1, which controls yellow and white kernels [23]. Owens identified Y1 and Dxs2 by GWAS and explored the relationship between Dxs3, Dmes1, LcyE and EP1 and kernel colour formation [13]. In this study, both visual scoring and the RGB system were used to evaluate kernel colour, which was taken as phenotypic data, a GWAS was performed for kernel colour-related traits (Fig. 2, Figure S2), and multiple known genes related to kernel colour were identified, such as Y1.The rs6_85061523 was in exon 4 of Y1 and significantly associated with B_1 (Table 2). The Y1 gene dose effect on endosperm carotenoids was identified in 1940. Sequencing analysis later confirmed that Y1 encodes phytene synthetase 1 (PSY1), which plays a key role in the formation of phytoene from two molecules of geranylgeranyl pyrophosphate (GGPP) [50]. PSY1 is involved in carotenoid biosynthesis in leaves and endosperm, and its allelic variation to a large extent determines the variation in kernel colour from white to orange [15, 16]. Overexpression of Y1 can change the colour of the kernel from white to yellow. In addition, Psy2 is 26 kb downstream of the significant SNP rs8_173494185 (Table 2). Crti3 encodes a carotenoid isomerase, and the distance from the rs5_1569528 is 201 kb (Table 2). These are all key enzymes in the process by which GGPP produces lycopene [45].

The carotenoid precursor substance GGPP is synthesized by the methylerythritol phosphate (MEP) pathway in higher plants. The key enzyme in the first step of the MEP pathway is 1-deoxy-D-xylulose 5-phosphate synthase (DXS), which is the enzyme with the highest control coefficient in this pathway [51]. In this study, the rs9_20232174 was 7.4 kb away from Dxs3, and rs6_150537590 was 121 kb away from Dxs1 (Table 2). In addition, hyd5 (crtRB5), approximately 25 kb away from rs9_156874532 (Table 2), is involved in hydroxylation reactions downstream of the carotenoid biosynthetic pathway. Carotenoid cleavage dioxygenase 1 (CCD1) is involved in carotenoid degradation. The rs9_155118340 is 25 kb away from Wc1 (Ccd1) [46]. The above findings indicated that the results of this study are highly valuable as a reference.

The anthocyanin synthesis pathway is divided into three stages: the initial reaction of flavonoid metabolism; important reactions of flavonoids; and anthocyanin synthesis [52]. The anthocyanin synthesis pathway is catalysed by a series of enzymes encoded by structural genes, for example, phenyl alanine ammonialyase (PAL) in the first stage; chalcone synthase (CHS), chalcone isomerase (CHI) and flavonoid 3’—hydroxylase (F3’H) in the second stage; and dihydroflavonol4—reductase (DFR) and anthocyanidinaynthase (ANS) in the third stage [27, 28]. In this study, significant signals were detected near Chi1, Chi3 and A1 (DFR); the rs1_298633704 was located 52 kb downstream of chi1, rs5_2392770 was located 189 kb downstream of Chi3, and rs5_68147228 was located 122 kb downstream of A1 (Table 2).

Anthocyanin skeleton modification is necessary for its maturation, and the most common method of anthocyanin modification is glycosylation, which can enhance the stability and water solubility of anthocyanins. The key enzyme that catalyses this process is UDP-glucose flavonol glycosyl transferase (UFGT) [4, 53]. In this study, a significant SNP rs7_19965244 was found near Bz1 at a distance of 138 kb (Table 2). In maize, Bz2 encodes a GST, which helps transport anthocyanins and prevent anthocyanin oxidation, resulting in the bronze colour of kernels [37].

Anthocyanin synthesis structural genes are directly involved in the formation of anthocyanins and their regulation by transcription factors [54]. In this study, the rs3_219867520 was located at the first exon of R1, which can activate the expression of A1 and cause anthocyanin accumulation [55]. The rs10_139859410 located 80 kb downstream of In1 encodes a bHLH-like inhibitor that increases anthocyanin accumulation in starch [32].

At present, research on the biosynthesis of carotenoids and anthocyanins is fairly clear [7, 8]; however, the mechanism of their regulation of kernel colour formation needs to be studied and explored further. Three key candidate genes were identified in this study. Zm00001d048623 encodes the MYB transcription factor MYB59. MYB transcription factors are important regulatory factors for the structural genes of the anthocyanin synthesis pathway and are the largest gene family in higher plants [56]. Therefore, Myb59 may be a key gene that modulates maize kernel colour by regulating anthocyanin synthesis. Zm00001d048621 encodes an ABC transporter. In maize, Mrp3 encodes a multidrug resistance-associated protein, an ABC transporter that transports anthocyanins into the vacuole [39]. Thus, we conclude that Zm00001d048621 is a key gene for anthocyanin transport in maize kernels, which affects kernel colour. Zm00001d048626 encodes a cytochrome P450 enzyme. In maize, lut1 encodes CYP97C, and lut5 encodes CYP97A, which are cytochrome P450-type monooxygenases. LUT1 catalyses the conversion of α-carotene to zeinoxanthin and hydroxylation of zeinoxanthin to yield lutein [20, 22]. CYP97A is an ε-ring carotenoid hydroxylase. Therefore, it is speculated that Zm00001d048626 encodes a cytochrome P450 enzyme that is involved in the biosynthesis of xanthophylls and regulates kernel colour.

Identification of superior allelic variation of important loci

There are many superior allelic variations in crop germplasm, such as wild versions or related species, and superior allelic variations of important genetic loci were mined and developed, and new cultivars were bred by molecular assistant selection (MAS) [57]. For example, the diversity of alleles of LcyE in maize demonstrates that the favourable allele is more common in tropical lines [19], and the favourable allele for CrtRB1 is more common in temperate germplasm [58]. In this study, the phenotypic effects of the identified new and pleiotropic loci were analysed, and it was found that the inbred lines carrying different allelic variations had significant differences in phenotype. Moreover, the superior allelic variations of the corresponding loci were identified. rs1_40605594 and rs5_2392770 were significantly associated with the kernel colour level, and selecting A/A superior allelic variation was expected to improve the kernel colour trait (Fig. 4). These results indicate that the superior allelic variations of important loci identified in this study can be used in marker-assisted selection breeding of maize kernel traits for further genetic improvement of crops.

Conclusions

In summary, we identified 1029 QTLs associated with maize kernel colour by GWAS. Key candidate genes were predicted through functional gene annotation and previous reports, laying the foundation for subsequent gene function verification and providing a reference for analysing the genetic basis of kernel colour and improving the nutritional quality of maize.

Methods

Plant materials and field experiments

An association panel was constructed by 244 inbred lines from the laboratory of professor Jinsheng Lai of China Agricultural University containing 3 million SNP markers [59]. These inbred lines were planted in Wengyuan County, Shaoguan City, Guangdong Province (24.35°N, 114.13°E) in 2020 and Haizhu District, Guangzhou City, Guangdong Province (23.10°N, 113.26°E) in 2021. Single row plantings, the row spacing was 65 cm, and intra-row spacing of 25 cm, with the conventional field management and artificial self-pollination. Harvest and dry at post maturity, and then select the consistent maize ears for the further experiments.

Kernel colour determination

Thirty mature and dry maize kernels with a consistent appearance were selected, and original kernels’ images were captured using EPSON EU-88 scanning devices and EPSON Scan software. The colour values of the top of the endosperm near the style vestige were extracted and calculated based on the image in the RGB colour model, which could get the R, G, B and Gary Value, and RGB = 2562*R + 256*G + B (Fig. 5). In addition, the kernel colour of the inbred lines was graded by visual scoring [22], with the scoring divided into 7 grades, which were used as the visual grade phenotype data for kernel colour (Fig. 6).

Fig. 5
figure 5

Location diagram of extracting colour values near the maize kernel style. The red circle is where the colour is extracted

Full size image
Fig. 6
figure 6

Standardized colour scale representative kernels from the association mapping families. The ordinal colour scale ranges from A (lightest) to F (darkest), 6 levels

Full size image

The kernel colour data were organized and averaged. The descriptive statistics analysis and data visualization were conducted using IBM SPSS Statistics 25 and R (4.2.2). The Pearson correlation matrix was drawn using the corrplot function and pheatmap package in R statistical analysis program. The Broad-sense heritability (h2) was calculated for kernel colour traits according Nyquist as: h2 = δ2G / (δ2G + δ2E / r) where δ2G and δ2E is genetic variance and residual variance, respectively [60].

Genome-wide association study

Association analysis for the 6 indexes of colour was conducted via emmax software with a mixed linear model (MLM), taking both the K and Q matrices into account to avoid spurious associations. PLINK was used to calculate the R2 of adjacent windows with the parameters of R2 > 0.2. A total of 198,910 independent SNPs were ultimately obtained. Then P value ≤ 1/198910 (P ≤ 1 × 10–6) was used as the GWAS significance threshold. Q-Q plot was used to estimate the difference between the observed and predicted P values.

To combine significant SNPs into QTL intervals, we combine SNPs within the range of 50 kb as a QTL. If there is only one SNP in the range, we use that as a new starting point and searched forward another 50 kb, the search ends until the distance between two SNPS is larger than 50 kb, SNPs in this range are combined into a QTL, and then the search is repeated with SNPs with distances larger than 50 kb apart as the starting point [61, 62].

Identification and annotation of candidate genes

According to the linkage disequilibrium analysis of the natural populations, 100 kb was taken as the LD decay distance [63]. All potential candidate genes within 200 kb (100 kb upstream and 100 kb downstream of the lead SNP) of the detected loci were identified. The candidate genes were obtained from the B73 genome reference (version 4) in the MaizeGDB genome browser (https://www.maizegdb.org/). Complementary information was collected from the U.S. National Center for Biotechnology Information (http://0-www-ncbi-nlm-nih-gov.brum.beds.ac.uk/) and MaizeGDB.

Analysis of superior allelic variations

On the basis of the results of the GWAS, the most significant SNPs were selected, and the allelic variation effects of these major SNPs were analysed by the R package ofggplot2, ggsignif and ggpubr.

Availability of data and materials

All data generated or analyzed during this study are included in this article and its supplementary information files or are available from the corresponding author on reasonable request.

References

  1. Nuss ET, Tanumihardjo SA. Maize: a paramount staple crop in the context of global nutrition. Compr Rev Food Sci Food Saf. 2010;9(4):417–36.

    Article  PubMed  CAS  Google Scholar 

  2. Petroni K, Pilu R, Tonelli C. Anthocyanins in corn: a wealth of genes for human health. Planta. 2014;240(5):901–11.

    Article  PubMed  CAS  Google Scholar 

  3. Jaradat AA, Goldstein W. Diversity of maize kernels from a breeding program for protein quality: I. Physical, biochemical, nutrient, and color traits. Crop Sci. 2013;53(3):956–76.

    Article  CAS  Google Scholar 

  4. Chatham LA, Juvik JA. Linking anthocyanin diversity, hue, and genetics in purple corn[J]. G3 (Bethesda). 2021;11(2):jkaa062.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  5. Zilic S, Serpen A, Akillioglu G, et al. Phenolic compounds, carotenoids, anthocyanins, and antioxidant capacity of colored maize (Zea mays L.) kernels. J Agric Food Chem. 2012;60(5):1224–31.

    Article  PubMed  CAS  Google Scholar 

  6. DellaPenna D, Pogson BJ. Vitamin synthesis in plants: tocopherols and carotenoids. Annu Rev Plant Biol. 2006;57:711–38.

    Article  PubMed  CAS  Google Scholar 

  7. Walter MH, Strack D. Carotenoids and their cleavage products: biosynthesis and functions. Nat Prod Rep. 2011;28(4):663–92.

    Article  PubMed  CAS  Google Scholar 

  8. Gould K, Davies KM, Winefield C. Anthocyanins: biosynthesis, functions, and applications: Springer; 2008.

  9. de Pascual-Teresa S, Sanchez-Ballesta MT. Anthocyanins: from plant to health. Phytochem Rev. 2008;7:281–99.

    Article  Google Scholar 

  10. Rice A, West K, Black R. Vitamin A deficiency. Comparative quantification of health risk. Global and regional burden of disease attributable to selected major risk factors. 2005;5(1):211–56.

    Google Scholar 

  11. Pixley K, Rojas NP, Babu R, et al. Biofortification of maize with provitamin a carotenoids. Carotenoids Hum Health. 2013;271–92.

  12. Maqbool MA, Aslam M, Beshir A, et al. Breeding for provitamin A biofortification of maize (Zea mays L.). Plant Breeding. 2018;137(4):451–69.

    Article  CAS  Google Scholar 

  13. Owens BF, Mathew D, Diepenbrock CH, et al. Genome-wide association study and pathway-level analysis of kernel color in maize. G3 (Bethesda). 2019;9(6):1945–55.

    Article  PubMed  CAS  Google Scholar 

  14. Buckner B, Kelson TL, Robertson DS. Cloning of the y1 locus of maize, a gene involved in the biosynthesis of carotenoids. Plant Cell. 1990;2(9):867–76.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  15. Orlovskaya OA, Vakula SI, Khotyleva LV, et al. Association between total carotenoid content of maize kernels (Zea mays L.) and polymorphic site INDEL1 in PSY1 gene. Russian J Genet: Appl Res. 2018;8:74–9.

    Article  CAS  Google Scholar 

  16. LaPorte MF, Vachev M, Fenn M, et al. Simultaneous dissection of grain carotenoid levels and kernel color in biparental maize populations with yellow-to-orange grain. G3 (Bethesda). 2022;12(3):jkac006.

    Article  PubMed  CAS  Google Scholar 

  17. Singh M, Lewis PE, Hardeman K, et al. Activator mutagenesis of the pink scutellum1/viviparous7 locus of maize. Plant Cell. 2003;15(4):874–84.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  18. Lu S, Li L. Carotenoid metabolism: biosynthesis, regulation, and beyond. J Integr Plant Biol. 2008;50(7):778–85.

    Article  PubMed  CAS  Google Scholar 

  19. Harjes CE, Rocheford TR, Bai L, et al. Natural genetic variation in lycopene epsilon cyclase tapped for maize biofortification[J]. Science. 2008;319(5861):330–3.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  20. Owens BF, Lipka AE, Magallanes-Lundback M, et al. A foundation for provitamin A biofortification of maize: genome-wide association and genomic prediction models of carotenoid levels. Genetics. 2014;198(4):1699–716.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Hable WE, Oishi KK, Schumaker KS. Viviparous-5 encodes phytoene desaturase, an enzyme essential for abscisic acid (ABA) accumulation and seed development in maize. Mol Gen Genet. 1998;257(2):167–76.

    Article  PubMed  CAS  Google Scholar 

  22. Chandler K, Lipka AE, Owens BF, et al. Genetic analysis of visually scored orange kernel color in maize. Crop Sci. 2013;53(1):189–200.

    Article  CAS  Google Scholar 

  23. Lin G, He C, Zheng J, et al. Chromosome-level genome assembly of a regenerable maize inbred line A188. Genome Biol. 2021;22(1):175.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  24. Riaz M, Zia-Ul-Haq M, Saad B. Anthocyanins effects on carcinogenesis, immune system and the central nervous system. In: Anthocyanins and Human Health: Biomolecular and therapeutic aspects: Springer International Publishing; 2016.

  25. Tena N, Martin J, Asuero AG. State of the art of anthocyanins: antioxidant activity, sources, bioavailability, and therapeutic effect in human health. Antioxidants (Basel). 2020;9(5):451.

    Article  PubMed  CAS  Google Scholar 

  26. Cone KC. Anthocyanin synthesis in maize aleurone tissue[M]//endosperm: developmental and molecular biology. Springer. 2007;15:121–39.

    Google Scholar 

  27. Shih CH, Chu H, Tang LK, et al. Functional characterization of key structural genes in rice flavonoid biosynthesis. Planta. 2008;228(6):1043–54.

    Article  PubMed  CAS  Google Scholar 

  28. Ferrer JL, Austin MB, Stewart CJ, et al. Structure and function of enzymes involved in the biosynthesis of phenylpropanoids. Plant Physiol Biochem. 2008;46(3):356–70.

    Article  PubMed  CAS  Google Scholar 

  29. Gonzalez A, Zhao M, Leavitt JM, et al. Regulation of the anthocyanin biosynthetic pathway by the TTG1/bHLH/Myb transcriptional complex in Arabidopsis seedlings. Plant J. 2008;53(5):814–27.

    Article  PubMed  CAS  Google Scholar 

  30. Lloyd A, Brockman A, Aguirre L, et al. Advances in the MYB-bHLH-WD repeat (MBW) Pigment regulatory model: addition of a WRKY factor and co-option of an anthocyanin MYB for Betalain regulation. Plant Cell Physiol. 2017;58(9):1431–41.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  31. Cone KC, Cocciolone SM, Burr FA, et al. Maize anthocyanin regulatory gene pl is a duplicate of c1 that functions in the plant. Plant Cell. 1993;5(12):1795–805.

    PubMed  PubMed Central  CAS  Google Scholar 

  32. Burr FA, Burr B, Scheffler BE, et al. The maize repressor-like gene intensifier1 shares homology with the r1/b1 multigene family of transcription factors and exhibits missplicing. Plant Cell. 1996;8(8):1249–59.

    PubMed  PubMed Central  CAS  Google Scholar 

  33. Selinger DA, Chandler VL. A mutation in the pale aleurone color1 gene identifies a novel regulator of the maize anthocyanin pathway. Plant Cell. 1999;11(1):5–14.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  34. Piazza P, Procissi A, Jenkins GI, et al. Members of the c1/pl1 regulatory gene family mediate the response of maize aleurone and mesocotyl to different light qualities and cytokinins. Plant Physiol. 2002;128(3):1077–86.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  35. Paulsmeyer MN, Juvik JA. R3-MYB repressor Mybr97 is a candidate gene associated with the Anthocyanin3 locus and enhanced anthocyanin accumulation in maize. Theor Appl Genet. 2023;136(3):55.

    Article  PubMed  CAS  Google Scholar 

  36. Poustka F, Irani NG, Feller A, et al. A trafficking pathway for anthocyanins overlaps with the endoplasmic reticulum-to-vacuole protein-sorting route in Arabidopsis and contributes to the formation of vacuolar inclusions. Plant Physiol. 2007;145(4):1323–35.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  37. Marrs KA, Alfenito MR, Lloyd AM, et al. A glutathione S-transferase involved in vacuolar transfer encoded by the maize gene Bronze-2. Nature. 1995;375(6530):397–400.

    Article  PubMed  CAS  Google Scholar 

  38. Gomez C, Conejero G, Torregrosa L, et al. In vivo grapevine anthocyanin transport involves vesicle-mediated trafficking and the contribution of anthoMATE transporters and GST. Plant J. 2011;67(6):960–70.

    Article  PubMed  CAS  Google Scholar 

  39. Goodman CD, Casati P, Walbot V. A multidrug resistance-associated protein involved in anthocyanin transport in Zea mays. Plant Cell. 2004;16(7):1812–26.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  40. Francisco RM, Regalado A, Ageorges A, et al. ABCC1, an ATP binding cassette protein from grape berry, transports anthocyanidin 3-O-Glucosides. Plant Cell. 2013;25(5):1840–54.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  41. Wang W, Qiu X, Wang Z, et al. Deciphering the Genetic Architecture of Color Variation in Whole Grain Rice by genome-wide association. Plants. 2023;12(4):927.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Sgaramella N, Nigro D, Pasqualone A, et al. Genetic mapping of flavonoid grain pigments in durum wheat. Plants. 2023;12(8):1674.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  43. Cordoba E, Porta H, Arroyo A, et al. Functional characterization of the three genes encoding 1-deoxy-D-xylulose 5-phosphate synthase in maize. J Exp Bot. 2011;62(6):2023–38.

    Article  PubMed  CAS  Google Scholar 

  44. Falcone FM, Rodriguez E, Casas MI, et al. Identification of a bifunctional maize C- and O-glucosyltransferase. J Biol Chem. 2013;288(44):31678–88.

    Article  Google Scholar 

  45. Bartley GE, Scolnik PA. cDNA cloning, expression during development, and genome mapping of PSY2, a second tomato gene encoding phytoene synthase. J Biol Chem. 1993;268(34):25718–21.

    Article  PubMed  CAS  Google Scholar 

  46. Sun Z, Hans J, Walter MH, et al. Cloning and characterisation of a maize carotenoid cleavage dioxygenase (ZmCCD1) and its involvement in the biosynthesis of apocarotenoids with various roles in mutualistic and parasitic interactions. Planta. 2008;228(5):789–801.

    Article  PubMed  CAS  Google Scholar 

  47. Zhang M, Li K, Zhang C, et al. Identification and characterization of class 1 DXS gene encoding 1-deoxy-D-xylulose-5-phosphate synthase, the first committed enzyme of the MEP pathway from soybean. Mol Biol Rep. 2009;36(5):879–87.

    Article  PubMed  CAS  Google Scholar 

  48. Murcray CE, Lewinger JP, Gauderman WJ. Gene-environment interaction in genome-wide association studies. Am J Epidemiol. 2009;169(2):219–26.

    Article  PubMed  Google Scholar 

  49. Doerge RW. Mapping and analysis of quantitative trait loci in experimental populations. Nat Rev Genet. 2002;3(1):43–52.

    Article  PubMed  CAS  Google Scholar 

  50. Rodriguez-Concepcion M, Boronat A. Elucidation of the methylerythritol phosphate pathway for isoprenoid biosynthesis in bacteria and plastids. A metabolic milestone achieved through genomics. Plant Physiol. 2002;130(3):1079–89.

    Article  PubMed  CAS  Google Scholar 

  51. Wright LP, Rohwer JM, Ghirardo A, et al. Deoxyxylulose 5-Phosphate synthase controls flux through the methylerythritol 4-Phosphate pathway in Arabidopsis. Plant Physiol. 2014;165(4):1488–504.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  52. Holton TA, Cornish EC. Genetics and biochemistry of anthocyanin biosynthesis. Plant Cell. 1995;7(7):1071.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  53. Vogt T, Grimm R, Strack D. Cloning and expression of a cDNA encoding betanidin 5-O-glucosyltransferase, a betanidin- and flavonoid-specific enzyme with high homology to inducible glucosyltransferases from the Solanaceae. Plant J. 1999;19(5):509–19.

    Article  PubMed  CAS  Google Scholar 

  54. Xu W, Dubos C, Lepiniec L. Transcriptional control of flavonoid biosynthesis by MYB-bHLH-WDR complexes. Trends Plant Sci. 2015;20(3):176–85.

    Article  PubMed  CAS  Google Scholar 

  55. Carey CC, Strahle JT, Selinger DA, et al. Mutations in the pale aleurone color1 regulatory gene of the Zea mays anthocyanin pathway have distinct phenotypes relative to the functionally similar TRANSPARENT TESTA GLABRA1 gene in Arabidopsis thaliana. Plant Cell. 2004;16(2):450–64.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  56. Feng K, Xu ZS, Que F, et al. An R2R3-MYB transcription factor, OjMYB1, functions in anthocyanin biosynthesis in Oenanthe javanica. Planta. 2018;247(2):301–15.

    Article  PubMed  CAS  Google Scholar 

  57. Varshney RK, Hoisington DA, Tyagi AK. Advances in cereal genomics and applications in crop breeding. Trends Biotechnol. 2006;24(11):490–9.

    Article  PubMed  CAS  Google Scholar 

  58. Yang X, Yan J, Shah T, et al. Genetic analysis and characterization of a new maize association mapping panel for quantitative trait loci dissection. Theor Appl Genet. 2010;121(3):417–31.

    Article  PubMed  Google Scholar 

  59. Liu H, Shi J, Sun C, et al. Gene duplication confers enhanced expression of 27-kDa gamma-zein for endosperm modification in quality protein maize. Proc Natl Acad Sci U S A. 2016;113(18):4964–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  60. Nyquist WE, Baker RJ. Estimation of heritability and prediction of selection response in plant populations. CRC Crit Rev Plant Sci. 1991;10(3):235–322.

    Article  Google Scholar 

  61. Chen S, Liu F, Wu W, et al. A SNP-based GWAS and functional haplotype-based GWAS of flag leaf-related traits and their influence on the yield of bread wheat (Triticum aestivum L.) [J]. Theor Appl Genet. 2021;134(12):3895–909.

    Article  PubMed  CAS  Google Scholar 

  62. Li X, Lu S, Chen W, et al. Genome-wide association study of root hair length in maize. Trop Plant Biol. 2023;1–8.

  63. Barrett JC, Fry B, Maller J, et al. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21(2):263–5.

    Article  PubMed  CAS  Google Scholar 

Download references

Acknowledgements

We thank Dr. Jinsheng Lai from China Agricultural University for generously providing the maize inbred lines and their genotypes for our research.

Funding

This work was supported by the GDAS’ Project of Science and Technology Development (2022GDASZH-2022010102), and the National Natural Science Foundation of China (32072027), the Guangdong Province special projects in key fields of ordinary colleges and universities, the Guangdong Province key construction discipline research ability enhancement project (2022ZDJS023), and the Special Project for Rural Revitalization Strategy in Guangdong Province (2022NPY000235), and the Basic and Applied Basic Research Fund of Guangdong Province (2021A1515110745), the stable support project of Guangdong Academy of Sciences, and germplasm innovation and new variety breeding of heat resistant fresh-eating maize.

Author information

Author notes

  1. Weiwei Chen, Fangqing Cui and Hang Zhu are co-first author.

Authors and Affiliations

  1. Institute of Nanfan & Seed Industry, Guangdong Academy of Science, Guangzhou, 510316, Guangdong, China

    Weiwei Chen, Fangqing Cui, Hang Zhu, Xiangbo Zhang, Siqi Lu, Chuanli Lu, Hailong Chang, Lina Fan, Huanzhang Lin, Junteng Fang, Yuxing An, Xuhui Li & Yongwen Qi

  2. College of Agriculture and Biology, Zhongkai University of Agriculture and Engineering, Guangzhou, 510325, Guangdong, China

    Fangqing Cui, Hang Zhu, Siqi Lu, Lina Fan, Huanzhang Lin, Junteng Fang & Yongwen Qi

  3. College of Agriculture, Yangtze University, Jingzhou, 434025, Hubei, China

    Hang Zhu

Authors
  1. Weiwei Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Fangqing Cui
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Hang Zhu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Xiangbo Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Siqi Lu
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Chuanli Lu
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Hailong Chang
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Lina Fan
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Huanzhang Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Junteng Fang
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Yuxing An
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Xuhui Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Yongwen Qi
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Y.Q and X.L designed the research; W.C, F.C, H.Z and X.L analyzed data; W.C wrote the paper; W.C, F.C, H.Z, X.Z, S.L, C.L, H.C, L.F, H.L, J.F, Y.A, X.L, and Y.Q carried out the experiments. All authors have read and approved the manuscript.

Corresponding authors

Correspondence to Yuxing An, Xuhui Li or Yongwen Qi.

Ethics declarations

Ethics approval and consent to participate

All experimental studies on plants were complied with relevant institutional, national, and international guidelines and legislation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Supplemental figure 1.

Frequency distribution of 6 kernel colour traits under two environments. _1: wengyuan experimental station (2020); _2: guangzhou experimental station (2021). Supplemental figure 2. Two GWAS models for the control of false positive (Q-Q plots). The X-axis and Y-axis is expected -log10(p) and observed -log10(p) of the 6 kernel colour traits in maize; The Q-Q plots of two models include MLM_PCA+K (above) and MLM_Q+K (below); E1: wengyuan experimental station (2020); E2: guangzhou experimental station (2021).

Additional file 2: Table S1.

SNPs identified by PCA+K models in two environments. Table S2. SNPs identified by Q+K models in two environments. Table S3. QTLs identified by PCA+K models in two environments. Table S4. QTLs identified by Q+K models in two environments. Table S5. The list of significant SNPs identified by at least two traits and two environments. Table S6. The information of candidate genes.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chen, W., Cui, F., Zhu, H. et al. Genome-wide association study of kernel colour traits and mining of elite alleles from the major loci in maize. BMC Plant Biol 24, 25 (2024). https://0-doi-org.brum.beds.ac.uk/10.1186/s12870-023-04662-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://0-doi-org.brum.beds.ac.uk/10.1186/s12870-023-04662-5

Keywords

BMC Plant Biology

ISSN: 1471-2229

Contact us