CONSTRUCTION OF A MICROSATELLITE LINKAGE MAP IN A TWO-LINE RICE VARIETY PEIAI64s/E32 AND ITS APPLICATION

A doubled haploid population (DH) consisting of 86 lines derived by anther culture of Peiai64s/E32, a two-line hybrid rice variety with high heterosis, was used to construct a microsatellite or SSR linkage map of six rice chromosomes (chromosomes 1, 3, 5, 7, 9 and 11). A total of 163 PCR primers for SSR analysis on these chromosomes were chosen from a map published by Cornell University (designated CUMAP) and 71(43.6%) of them were found polymorphic between the two parents. Those polymorphic PCR primers were used for population genotyping. The software Mapmaker/Exp 3.0 was used for the map construction. The map (designated PEMAP) comprises 68 microsatellite maker loci, covering a total length of 414.3 cM, with an average distance of 6.1 cM between adjacent markers and an average marker number of 11.3 on each chromosome. The PEMAP is highly comparable with the CUMAP. Most of the markers were mapped onto the same chromosomes and aligned in the same order. Only two markers (RM269 and RM236) were mapped onto different chromosomes and two pairs of markers (RM86 and RM84, RM243 and RM259) on chromosome 1 were mapped in reversed order. The PEMAP is 47% shorter than the CUMAP in the regions covered by the same set markers. Serious segregation distortion was observed in this DH population, with 23 (33.8%) markers showing significant deviation. It is noted that all markers on chromosomes 1, 3 and 11 were biased to Peiai64s, while those on chromosomes 7 and 9 were opposite. Mapping of QTLs underlying grain length (GL), grain breadth (GB) and grain weight (GW) was performed, using the complete SSR maker linkage map (the other six chromosomes were constructed by another student) and the statistical method of composite interval mapping (CLM). A total of 13 QTLs were detected, 5 for GL, 5 for GB and 3 for GW QTLs g13 and gw5 showed the largest effects on GL and GW respectively, explaining 17.36% and 21.33% of phenotypic variances of the traits. Previous studies also detected QTLs on these regions. Therefore, the two QTLs are credible. So, in marker-assisted breeding, these two regions should be firstly considered. In addition, maker intervals showing pleiotropy were found in this study. For example, QTL gl5b controlling GL, gb5b controlling GB and gw5 controlling GW were all located in the same marker interval on chromosome 5.