Genetic analysis of maize kernel color using segregation, bulked segregant, and nested association mapping analysis

Carotenoids are phytopigments that give many fruits, vegetables and grains their yellow and orange color. Carotenoids are very important nutritionally, as sources of provitamin A for the developing world, and as antioxidants important for eye health for both developing and developed country populations. Maize is an optimal target crop for carotenoid biofortification due to broad phenotypic variation for this trait. Simple visual selection for darker orange color has been associated with an increase in total carotenoids. Visual scoring is much less expensive and time-consuming compared to other quantitative carotenoid measurements such as HPLC and NIR. Segregation analysis using visual scoring of S1 kernels from a four-parent synthetic revealed that colors segregate in a relatively simple mendelian manner and that very few genes are involved in the conversion from yellow to orange kernel color. With this knowledge, we became confident that visual scoring of kernel color can be a reliable method of phenotyping. Bulked Segregant Analysis (BSA) was performed on bulks of light orange and dark orange kernels from an F2:3 mapping population. The kernels making up the bulks were scored visually and the genotyping was done using an Illumina MaizeSNP50 chip. Paired t-tests were performed to detect genetic differences between the bulks. Several loci were considered statistically different between bulks and some mapped to regions of the genome near known carotenoid biosynthesis genes. QTL analysis was performed on 10 Nested Association Mapping (NAM) families segregating for yellow and orange kernel color. The results of both individual family and joint family analysis further indicated that only a few genes are largely involved in the conversion from yellow to orange kernel color. Markers near two logical candidate genes, phytoene synthase 1 (psy1) and lycopene epsilon-cyclase (lcyϵ), were significantly associated with the trait in several families as well as in the joint analyses. Other QTL having smaller effects on kernel color are located near known carotenoid biosynthesis genes including whitecap 1 (wc1), beta-carotene hydroxylase 1 (crtRB1), zeta-carotene desaturase 1 (zds1), and yellow 8 (y8). The combined information from these studies provides confirmation that few genes are involved in the conversion from yellow to orange kernel color, and that phenotypes associated with genetic differences at these loci can be detected visually. This proven ability to select for higher total carotenoids visually will be especially useful in the developing world as local breeders and farmers can make selections for darker orange maize kernels which are much higher in provitamin A than the traditional white maize presently consumed. In the developed world, lines containing favorable alleles of the loci detected in these studies will be used in Marker-Assisted Selection (MAS) breeding programs to biofortify maize for higher antioxidant levels. Selection for darker orange ears can be paired with MAS for alleles of genes associated with increased provitamin A to increase flux into the carotenoid biosynthetic pathway and to improve provitamin A content