The genetic architecture of amylose biosynthesis in maize kernel

Abstract

Starch is the most abundant storage carbohydrate in maize kernel. The content of amylose and amylopectin confers unique properties in food processing and industrial application. Thus, the resurgent interest has been switched to the study of individual amylose or amylopectin rather than total starch, whereas the enzymatic machinery for amylose synthesis remains elusive. We took advantage of the phenotype of amylose content and the genotype of 9,007,194 single nucleotide polymorphisms from 464 inbred maize lines. The genome-wide association study identified 27 associated loci involving 39 candidate genes that were linked to amylose content including transcription factors, glycosyltransferases, glycosidases, as well as hydrolases. Except the waxy gene that encodes the granule-bound starch synthase, the remaining candidate genes were located in the upstream pathway of amylose synthesis, while the downstream members were already known from prior studies. The linked candidate genes could be transferred to manipulate amylose content and thus add value to maize kernel in the breeding programme.

Keywords: SNP; amylose; genome-wide association study; kernel; maize; starch.

Figure 1

The elucidated pathway of starch biosynthesis in maize. The model summarizes the elucidated key enzymes involved in starch synthesis. Basically, ADP‐glucose as the glucosyl donor for starch biosynthesis is activated by ADP‐glucose pyrophosphorylase (AGPase), which is composed of two large subunits (AGP LSU) and two small subunits (AGP SSU). In amyloplast, the amylose is synthesized by granule‐bound starch synthase (GBSS), while amylopectin biosynthesis requires three more coordinated enzymes of soluble starch synthase (SS), starch‐branching enzyme (BE) and starch‐debranching enzyme (DBE).

Figure 2

The frequency distribution of amylose content. x‐axis shows the amylose content, and y‐axis shows the population frequency.

Figure 3

Linkage disequilibrium decay on maize chromosomes and whole genome. The red dashed horizontal line shows the LD threshold for the association panel (r ² = 0.1).

Figure 4

Quantile–quantile and Manhattan plots for the association study of amylose content in maize kernels. (a) Quantile–quantile for amylose content. (b) Manhattan plot of amylose content. The dashed line indicates the significance threshold of P‐value 5 × 10–8. 27 unique SNPs are labelled with red dots, and the corresponding genes in the amylose/amylopectin of carbohydrate metabolic pathway are highlighted.

Figure 5

Candidate causative genes and variants underlying amylose content in maize kernel. (a) and (b) Regional Manhattan plot of the mfs‐like genomic region on chromosome 9 and the ppr‐like genomic region on chromosome 4. The 250‐kb genomic region on either side of the most significant SNP is shown. The lead SNP is shown with a largest red diamond. The red broken circle indicates the SNPs in coding region for the corresponding genes. (c) and (d) Candidate causative variants in mfs‐like and ppr‐like. Red nabla shows the locations of SNPs. The green colour showed the predicted coding region. (e) and (f) Allele effects for corresponding SNPs in mfs‐like and ppr‐like genes.

Figure 6

A model of amylose/amylopectin biosynthesis pathway. The model is summarized based on the previous knowledge and our GWAS analysis. In leaves, sucrose is produced from the Calvin cycle by carbon fixation with the activity of mitochondrial NAD‐dependent malic enzyme (ME encoded by GRMZM2G085747), phosphoribulokinase (PRK encoded by GRMZM2G026024 and GRMZM2G143804) and other enzymes. Then, sucrose is transported through phloem to the storage organ of endosperm by sucrose transporter (SUT6 encoded by GRMZM2G106741), where it is imported into the cytosolic compartment of each cell. In the cytosol, the sucrose synthase (SUS) or invertase (INV encoded by GRMZM2G089836) cleaves sucrose into fructose and UDP‐glucose. After glucose 6‐phosphate is transported into amyloplasts, phosphoglucomutase (PGM encoded by GRMZM2G173674) catalyses glucose 6‐phosphate into glucose 1‐phosphate that is activated into ADP‐glucose as the substrate for starch synthesis. The content of ADP‐glucose is also negatively regulated by Nudix hydrolases (NUDT encoded by GRMZM5G809417 and GRMZM2G069008) that break down ADP‐glucose linked to starch biosynthesis. Granule‐bound starch synthase (GBSS encoded by GRMZM2G024993) cooperates with other enzymes to catalyse and elongate sugar chains. All enzymes encoded by candidate genes found from GWAS are highlighted in red colour.