Genome-Wide Association Study on Berry Traits of Table Grapes

Abstract

Grape (Vitis spp.) is one of the most extensively cultivated fruit plants with table grapes accounting for 36% of the total world production (Office International de la Vigne et du Vin, 2019; Reisch et al., 2012). Breeders in important table grape growing areas have been trying to develop grape cultivars with superior berry traits to those already be on the market. Seedlessness, crisp texture, and suitability for storage are the main goals of table grape breeding (Reisch et al., 2012). Breeding of perennial fruit species is a long-term activity involving a high investment as compared to annual crops due to two challenges: long juvenile periods and the plant size (Zhebentyayeva et al., 2012). To select one with desired traits from the population by introducing molecular breeding technology can be a great benefit in terms of both saving time and cost. Identifying genes involved in target traits for breeding and developing markers are ongoing in various fruit trees. Most studies have been conducted in the form of various mapping approaches including quantitative trait locus (QTL) mapping. QTL mapping has proved a powerful method to identify regions of the target gene either in F2 populations or recombinant inbred line families (Dirlewanger et al., 2009; Kenis et al., 2008; Korte and Farlow, 2013; Martínez-García et al., 2013; Wu et al., 2014). However, QTL mapping has its limitations when it comes to mapping populations (Korte and Farlow, 2013). A new study method was named genome-wide association study (GWAS) investigates the association between genotypes and phenotypes, using the entire genome. This method is based on the hypothesis that there are common genetic variants that affect a trait, independent from the effect of pedigree (Visscher et al., 2012). GWAS can serve as a foundation experiment by providing insights into the genetic architecture of the trait, suggesting an informed choice of parents for QTL analysis. The results of GWAS find out the loci of candidates for mutagenesis and transgenics (Korte and Farlow, 2013). Thus, GWAS are often complementary to QTL mapping, when conducted together (Manenti et al., 2009). GWAS has recently been a common way of studying agriculturally important quantitative traits and natural variations, thanks to the development of next-generation sequencing technologies (Atwell et al., 2010; Guo et al., 2019; Yano et al., 2016). It is also expanding to research related to crops including fruit trees (Cao et al., 2012; Guo et al., 2019; Iwata et al., 2013; Kumar et al., 2013; Lee et al., 2017). There are a few studies on the genetic determination of quantitative traits in fruit trees using GWAS. In apple, GWAS was conducted using 1,200 seedlings of seven full-sib families to reveal significant associations of six fruit traits. Significant associations were found in all six traits, some of which were coincident to known candidate genes (Kumar et al., 2013). In pear, nine agronomic traits were investigated by GWAS, among them harvest time, black spot resistance, and spur number were associated with significant QTLs (Iwata et al., 2013). In peach, GWAS using 104 landrace accessions with 53 simple sequence repeat (SSR) markers detected associated markers for ten traits related to fruit and phenological period (Cao et al., 2012). In grape, several QTL mappings and GWAS of berry related traits have been conducted. Berry size and seedlessness are primary targets of breeding programs for table grapes (Cabezas et al., 2006). To this end, there are several approaches for identifying genes to determine seedlessness. The MADS-box gene VvAGL11 is essential for seed morphogenesis in grapes (Malabarba et al., 2017). The QTL analyses related to seed traits by using seedless grapes as parental resources found QTLs in the region of SDI (seed development inhibitor), VvAGL11 (Cabezas et al., 2006; Doligez et al., 2002; Mejía et al., 2007). Previous studies to identify genes related to soluble solid contents (SSC) and titratable acidity (TA) in grape berries, detected QTLs in various linkage groups (LGs) (Bayo-Canha et al., 2019; Chen et al., 2015; Liu et al., 2007; Viana et al., 2013; Zhao et al., 2015). Several pieces of research regarding berry flesh firmness revealed QTL in LG 18 commonly and other researches detected QTLs in other LGs (Carreño et al., 2014; Correa et al, 2016; Jiang et al., 2020). A GWAS of six table grape berry traits including berry color, berry development period, cluster size, berry weight, and berry flesh texture was conducted by using 179 genotypes comprising a mixture of landraces and cultivars. By the study, QTLs in every trait were founded respectively (Guo et al., 2019). The goals of this study were to determine the significant genetic regions and find the candidate genes governing genetic variations in seven berry related traits of table grapes. A pseudo-F2 population was genotyped by genotyping-by-sequencing (GBS) method and phenotypic variations of the seven traits were measured for two years. QTLs detected by GWAS and linkage analysis by genetic map were compared with QTLs in previous studies and candidate genes underlying the QTLs were predicted. In this study, not only the traits previously studied but the tensile strength, which is advantageous shelf-life characteristic and astringency of berry skin were also investigated. A comprehensive understanding of genetic determinism of the berry related traits will facilitate the breeding of new grape cultivars.

과수 육종의 주된 목표는 고품질의 과실을 생산하는 품종을 선발하는 것이다. 과실의 품질을 결정하는 대부분의 형질들은 양적 형질이다. 이러한 형질들에 관여하는 다수의 유전자를 탐색하고 그 역할을 밝히는 것은 육종을 효율적으로 수행하기 위해 필요한 사항이다. 본 연구에서는 육종의 목표에 해당하는 7가지 형질에 관여하는 유전자를 탐색하기 위하여 유전체 전장 연관 분석을 수행하였다. 타노레드와 루비씨들리스 그리고 그 둘을 교배하여 만든 269개체의 자손을 이용하여 실험을 진행하였다. 실험에 이용한 형질은 총 7개로 과립 무게, 종자 무게, 인장 강도, 당도, 산 함량, 과육 경도, 과피 내 고분자 타닌 함량이다. 표현형 조사는 2018년과 2019년 두 해 동안 실시하였는데 성숙기에 해당하는 8월부터 10월에 과실 시료를 무작위로 채취하여 위의 형질들을 조사하였다. 두 해 동안 조사한 표현형 데이터의 상관 계수는 0.34-0.81에 해당하여 넓게 분포하였으나, P-값은 1.126e-07부터 2.2e-16에 해당하여 각각의 형질은 연차 간 상관 관계가 있는 것으로 나타났다. Genotyping-by-sequencing 방법을 통해 유전체 DNA의 염기 서열 분석을 수행하였다. 총 271개체로부터 얻은 148.69Gb의 유전 정보를 포도 표준 유전체인 PN40024 12X.v2에 정렬하였다. 총 400,648개의 단일 염기 다형성을 얻었으며 이후 조건에 맞춰 25,421개의 단일 염기 다형성을 선발하여 유전체 전장 연관 분석에 이용하였다. 또한 그 중 2,243개의 단일 염기 다형성을 이용하여 연관 지도를 작성하고 양적 형질 유전자좌를 탐색하여 본 연구의 결과를 검토하였다. 과립 무게, 종자 무게, 인장 강도와 관련된 유의한 단일 염기 다형성은 18번 염색체 상에 있는 무핵 연관 주동 유전자로 알려진 VvAGL11의 주변에 위치하였으나 당도, 산도, 과육 경도의 경우에는 두 해 모두 공통되는 유의한 결과를 찾을 수 없었다. 포도 과피의 떫은맛을 결정하는 과피 내 고분자 타닌 함량의 경우 연관된 단일 염기 다형성이 11번째 염색체에서 발견되었으며 그 주변에서 발견된 VvMybPA2이 후보 유전자로 선발되었다. 따라서 본 연구 결과, 유전체 전장 연관 분석을 통해 생식용 포도 과실의 품질을 결정하는 형질과 관련된 후보 유전자들을 선발할 수 있었으며 해당 분석 방법의 적용 가능성을 확인할 수 있었다.