Reproductive Barriers in Inter-subspecific Hybrids and Its Relation with Subspecies Differentiation-Related Traits in Rice

Abstract

Asian rice cultivar (Oryza sativa L.) has been classified into two major groups, indica and japonica types. The beneficial traits or heterosis from each type of rice subspecies can be utilized to breed a better variety. However, various reproductive barriers prevent gene exchange or gene flow between two subspecies. Segregation distortion and hybrid sterility have often been encountered in mapping populations generated from crosses involving divergent genotypes. Domestication may often involve gene admixture across isolation barrier as well as intensive selection. The advent of molecular markers, especially in rice where the whole genome is covered, has given the geneticists involved in identifying, characterizing, and cloning the genes related to reproductive barriers. Construction of linkage maps can facilitate further understanding of genetic basis of segregation distortion. QTL analysis by using a set of mapping population with multiple crosses benefits in determination of genetic factors affecting hybrid sterility, such as female and/or male. Moreover it can promisingly enrich diverse sets of QTLs and will also provide a rich source of information about natural genetic variation underlying the evolution, domestication, and breeding of rice. In this study, ten linkage maps were constructed using two reciprocal F2 and eight BC1F1 populations generated from cross between Ilpumbyeo (japonica) and Dasanbyeo (indica). Total length of linkage map and average distance between adjacent markers varied according to different mapping populations. Comparison within map length in BC1F1 populations show that the genetic map length of paternal parent was shorter compared to that of maternal parent. The degree of transmission ratio distortion in BC1F1 populations was low relative to F2 populations. However, Dasanbyeo allele was transmitted in high frequency in both F2 and BC1F1 populations suggesting that indica allele is favored strongly in interspecific cross in rice. The most pronounced reproductive barriers detected in this study were mapped on chromosome 3 (transmitted through male gametophyte), chromosome 6 (transmitted through female gametophyte), and chromosome 5 (both male and female gamete). Although segregation-distorted markers are generally coincided with the previous reports for reproductive barriers, new gametophyte regions responsible for TRD by male and female function were found. A total of 15 main-effect QTLs (M-QTLs) and 17 significant digenic-epistatic interactions (Ep-QTLs) controlling spikelet fertility (SF) was resolved in the entire genome map of F2 and BC1F1 populations. Among detected QTLs responsible for hybrid sterility, four QTLs including qSF5.1 and qSF5.2 on chromosome 5, qSF6.2 on chromosome 6, and qSF12.2 on chromosome 12, were identified as major QTLs since they were located at corresponding positions in at least three mapping populations. Loci qSF5.1, qSF6.1 and qSF6.2 were responsible for both female and male sterility, whereas qSF3.1, qSF7 and qSF12.2 affected the spikelet fertility only through embryo sac factors, and qSF9.1 did through pollen factors. In addition, five new QTLs responsible for hybrid sterility were identified in this study. Despite, several potential epistatic interactions (Ep-QTLs) were identified in six out of the ten mapping populations, their contributions to PVE were relatively low comparing to the main QTLs, suggesting that specific loci are more effective in determining spikelet fertility in hybrid progenies. A total of 26 main-effect QTLs for subspecies-related traits were detected on seven chromosomes including chromosomes 1, 2, 3, 4, 5, 6, and 8. Among the 26, 14 correspond to previously reported QTLs suggesting a functional conservation of QTLs across other rice cultivar. On the other hand, remaining 12 QTLs were novel. In addition, the mendelian factor controlling phenol reaction was mapped on long arm of chromosome 4. We also detected a total of 29 significant epistatic QTLs revealed higher PVE than the main-effect QTLs. These findings demonstrate that epistatic interactions are an important genetic basis for complex subspecies-related traits in rice and should be taken into consideration, especially when a marker-assisted selection strategy is adopted in breeding program.