Allelic variation in the autotetraploid potato: genes involved in starch and steroidal glycoalkaloid metabolism as a case study

Abstract

Background: Tuber starch and steroidal glycoalkaloid (SGA)-related traits have been consistently prioritized in potato breeding, while allelic variation pattern of genes that underlie these traits is less explored.

Results: Here, we focused on the genes involved in two important metabolic pathways in the potato: starch metabolism and SGA biosynthesis. We identified 119 genes consisting of 81 involved in starch metabolism and 38 in the biosynthesis of steroidal glycoalkaloids, and discovered 96,166 allelic variants among 2,169 gene haplotypes in six autotetraploid potato genomes. Comparative analyses revealed an uneven distribution of allelic variants among gene haplotypes and that the vast majority of deleterious mutations in these genes are retained in heterozygous state in the autotetraploid potato genomes. Leveraging full-length cDNA sequencing data, we find that approximately 70% of haplotypes of the 119 genes are transcribable. Population genetic analyses identify starch and SGA biosynthetic genes that are potentially conserved or diverged between potato varieties with varying starch or SGA content.

Conclusions: These results deepen the understanding of haplotypic diversity within functionally important genes in autotetraploid genomes and may facilitate functional characterization of genes or haplotypes contributing to traits related to starch and SGA in potato.

Keywords: Allelic variation; Potato; Starch; Steroidal glycoalkaloid.

Fig. 1

The landscape of haplotype-based allelic variation in starch and SGA genes among autotetraploid potato. a, Number of SNPs, InDels and SVs identified in the six tetraploid potato varieties. b, Number of variants containing different counts of alleles. A bi-allelic variation denotes that it only possesses one alternative allele and the reference allele among the six potato cultivars. c, Functional annotation of the identified variants as shown in the pie chart. d, Uneven distribution of allelic variation identified in 38 SGA genes among their haplotypes in the Colomba genome. Number of allelic variations identified in each haplotype is displayed in heat maps. Grey boxes indicate missing data (This gene does not contain this haplotype number). e, Violin plots depict number and distribution of genetic variants in genes carrying different numbers of haplotypes in each of the six potato genomes. The blue dashed lines indicate 75%, median and 25% quartiles. Multiple comparisons are performed using Kruskal-Wallis test with α = 0.001 f, Number of deleterious mutations in genes with one to four haplotypes in the six potato genomes. g, Number of types of haplotypes that can be defined in the six potato cultivars for starch and SGA genes. h, Domestication targeted on a gene encoding a squalene synthase (Soltu.DM.01G050130.3) may lead to a conserved amino acid haplotype structure among tetraploid potato cultivars, which is possibly essential in significant reduction of tuber SGA content in cultivated potato. The number of regulation haplotypes for Soltu.DM.01G050130.3 varies from 2 to 4 across the six potato genomes (left panel), while only one amino-acid haplotype was identified (right panel). A high level of sequence divergence was observed around the predicted transmembrane region of this gene when comparing a wild potato species S. chacoense with cultivated potato. For f and g, data are presented as mean ± SD. One-way ANOVA and Turkey’s multiple comparisons with α = 0.01 are applied

Fig. 2

Patterns of transcribable and un-transcribable haplotypes unraveled in genes involved in starch and SGA metabolism among autotetraploid potato. a, Composition of transcribable and un-transcribable haplotypes in 36 SGA-related genes among six tetraploid potato cultivars illustrated by a 36 × 6 matrix of pie charts. b, Percentage of transcribable haplotypes in starch and SGA genes carrying one to four haplotypes in each of the six cultivars. One-way ANOVA and Turkey’s multiple comparisons with α = 0.01 are applied. c, Density of genetic variations (per one kilo base pairs) identified in un-transcribable haplotypes (UNTR hap) and transcribable haplotypes (TR hap). P value is calculated using two-tailed t-test. d, Number of deleterious mutations predicted in un-transcribable haplotypes (UNTR hap) and transcribable haplotypes (TR hap). **** p < 0.0001 in two-tailed Wilcoxon rank sum test. e, Three types of transcripts derived from a single locus carrying three haplotypes as exemplified by Soltu.DM.09G030970.1, a phosphoglucan, water dikinase. The reference haplotype is Hap2 with a transcript whose translated peptides containing two predicted functional domains CMB_2 and PPDK_N. A large deletion present in Hap3 probably leads to the non-transcript outcome. The complete transcript is divided into two independent properly aligned full-length transcripts in Hap 1 possibly due to a substitution of a divergent region. Data are presented as mean ± SD in b-d

Fig. 3

Potentially functionally important genes involved in SGA and starch metabolism. a, Five putatively conserved genes (red color) and five diverged (orange color) genes between cultivars with high and low total SGA content in the proposed SGA biosynthesis pathway. C5-SD, delta(7)-sterol-c5(6)-desaturase; CAS1, cycloartenol synthase; GAME9, GLYCOALKALOID METABOLISM 9; HMGR, 3-hydroxy-3-methylglutaryl coenzyme-A reductase; HMG-CoA, 3-hydroxy-3-methylglutaryl-coenzyme A; LAS1, lanostetrol synthase; PGA1, POTATO GLYCOALKALOID BIOSYNTHESIS 1; PGA2, POTATO GLYCOALKALOID BIOSYNTHESIS 2; SMT1, sterol C24-methyltransferase; SGT1, galactosyltransferase; SGT2, glucosyltransferase; SGT3, glycosyltransferase; SQE, squalene epoxide; SQS, squalene synthase; SSR2, sterol side chain reductase 2. b, Potentially conserved (red color) and diverged (orange color) genes between potato accessions bred for starch industry and other usages. The proposed starch biosynthesis and degradation pathway in potato tubers are depicted. ADP-Glc, ADP-glucose; AGPase, ADP-glucose pyrophosphorylase; AM, amylose; Amy, α-amylase; AP, amylopectin; BAM, β-amylase; DBE, starch branching enzyme; DPE, disproportionating enzyme; G1P, glucose 1-phosphate; G6P, glucose 6-phosphate; GBSS, granule-bound starch synthase; Glc, glucose; GLT, glucose transporter; GPT, glucose 6-phosphate/phosphate translocator; GWD, α-glucan, water dikinase; LSF, Like starch-excess Four; MEX, maltose transporter; NTT, nucleotide translocator; PGM, phosphoglucomutase; PHO, α-glucan phosphorylase; PWD, phosphoglucan, water dikinase; P_i, inorganic pyrophosphate; SuSy, sucrose synthase; SS, starch synthase; SBE, starch branching enzymes; SEX4, starch excess 4; UGPase, UDP-glucose pyrophosphorylase; UDP-Glc, UDP-glucose; VGT, vacuolar glucose transporter. c, Expression pattern of Soltu.DM.05G018750.1 (GPT2.1) on a log₂ scale in seven tissues of DM. In a and b, dashed arrows indicate steps containing multiple catalytic reactions