以高通量 SNP 基因型分析臺灣芒果種原之遺傳歧異度

Abstract

芒果 (Mangifera indica L.) 為世界第六大水果作物，廣泛栽種於熱帶及亞熱帶地區，在臺灣為重要經濟作物，產值逾 76 億新臺幣。芒果在臺灣已歷經超過 300 年的栽培，過去數十年來從國外引進許多栽培種，並選育許多知名且受到歡迎的芒果品種。調查種原的遺傳歧異度對於作物改良及重要性狀的研究皆能提供助益，本研究旨在了解臺灣芒果種原以及本地種土芒果－柴檨族群之遺傳歧異度。自試驗單位之芒果種原圃及田野採集共得到 476 個收集系，透過 genotyping-by-sequencing (GBS) 方法，以次世代定序探勘了超過 19 萬個 single nucleotide polymorphism (SNP) 分子標誌，經由定序深度、缺值率及次等位基因頻率 (minor allele frequency, MAF) 之篩選，與連鎖失衡 (linkage disequilibrium) 位點之刪節，得到 2,356 個分布於 20 條染色體之 SNP。透過收集系之間的 identity by descent (IBD) 在種原辨識出 14 個同物異名及 9 個異物同名組合，去除重複基因型的個體，臺灣芒果核心種原達 168 個。經由族群結構分析，將種原收集系劃分為兩個次族群，並於主成分分析 (principal component analysis, PCA) 中分布於第一主成分軸兩邊，是以，臺灣芒果核心種原中存在兩個不同遺傳背景。Pop I 中主要為來自美國的商業品種及許多臺灣品種，顯示美國芒果在台灣受到歡迎並常作為育種親本；Pop II 來源廣泛，主要為亞洲（尤其東南亞地區）品種組成，具有較高遺傳歧異度並可再被分為四個次族群，其中本地種柴檨具獨特之遺傳背景，並在親緣樹上與臺灣多個品種形成一群集，說明臺灣芒果中有一群品種應和柴檨有較接近之親緣。然而在全臺採集得到之 142 個柴檨收集系族群中，發現此地方種即使已在臺灣超過 300 年之久，收集系之間遺傳相似度極高，且無地區性的差異，整體歧異度極低，此現象可能是當初引進臺灣時受到創始者效應 (founder effect) 與強烈選拔所致，再者與柴檨之無融合生殖特性 (apomixis) 有極大相關。本研究建立臺灣芒果種原之全基因體 SNP 分子標誌，將作為未來重要性狀基因座探勘的基礎，也可用以設計一套具辨識度之 SNP 套組，應用於芒果品種鑑別與品種保護。族群結構與遺傳歧異度評估，將提供未來育種工作親本選擇之參考，使芒果種原獲得更有效率之利用。

Mango (Mangifera indica L.) ranks sixth in the global fruit production and is widely cultivated among tropical and subtropical regions. In Taiwan, mango is an important economic crop with a production value reached NT$ 7.6 billion. Mango has been cultivated for more than 300 years in Taiwan. Cultivars had been introduced into the country and many popular varieties were selected over the past decades. Investigation of genetic diversity in the germplasm is necessary for crop improvement and research of important traits, so this study aimed to evaluate the population structure and genetic diversity of the mango germplasm and the population of the landrace, ‘Tsar-Swain’, in Taiwan. For this purpose, 476 mango accessions were obtained from the germplasm of three national experimental institutes and through field collection. 198,602 single nucleotide polymorphism (SNP) sites were discovered among accessions by the genotyping-by-sequencing (GBS) method. A subset of 2,356 SNPs distributed on 20 chromosomes were retained after the filtering and pruning by sequencing depth, missing rate, minor allele frequency, and linkage disequilibrium. Pairwise identity by descent (IBD) between accessions was used to identify fourteen synonyms and nine homonyms from the mango germplasm. 168 core accessions were retained after removing duplicates in the germplasm. We defined two distinct genetic backgrounds in the mango core accessions by population structure analysis with similar clustering in the principal component analysis (PCA). Accessions in Pop I were primarily commercial varieties from the US and Taiwanese varieties, indicating the US cultivars were popular in Taiwan and usually served as parents in breeding works. Pop II was mainly composed of accessions from Asia, with higher genetic diversity that it could be further divided into another four subgroups. ‘Tsar-swain’ showed a unique genetic background in this study and formed a clade with many Taiwanese varieties on the phylogenetic tree, suggesting a group of Taiwanese mango that has a close relationship with ‘Tsar-Swain’. However, in the population of 142 ‘Tsar-swain’ accessions collected from the field, the high similarity between individuals and low diversity within the population was observed. This finding might be linked with the apomixis characteristics of this variety, selection, and the probable founder effect at the time it was introduced. This study established genome-wide SNP markers of the mango germplasm in Taiwan. It would be useful in the discovery of locus associated with important horticultural traits and the development of SNP set for mango variety identification and protection. The assessment of population structure and genetic diversity of mango germplasm also provides information in the future breeding work to make better utilization of the genetic resource.