Obtaining genetic variation information from indica rice hybrid parents and identification of loci associated with heterosis are important for hybrid rice breeding. Here, we resequence 1,143 indica accessions mostly selected from the parents of superior hybrid rice cultivars of China, identify genetic variations, and perform kinship analysis. We find different hybrid rice crossing patterns between 3- and 2-line superior hybrid lines. By calculating frequencies of parental variation differences (FPVDs), a more direct approach for studying rice heterosis, we identify loci that are linked to heterosis, which include 98 in superior 3-line hybrids and 36 in superior 2-line hybrids. As a proof of concept, we find two accessions harboring a deletion in OsNramp5, a previously reported gene functioning in cadmium absorption, which can be used to mitigate rice grain cadmium levels through hybrid breeding. Resource of indica rice genetic variation reported in this study will be valuable to geneticists and breeders.

BackgroundWith an increasing world population and a gradual decline in the amount of arable land, food security remains a global challenge. Continued increases in rice yield will be required to break through the barriers to grain output. In order to transition from hybrid rice to super-hybrid rice, breeding demands cannot be addressed through traditional heterosis. Therefore, it is necessary to incorporate high yield loci from other rice genetic groups and to scientifically utilize intersubspecific heterosis in breeding lines. In this study, 781 lines from a segregating F2 population constructed by crossing the indica variety, “Giant Spike Rice” R1128 as trait donor with the japonica cultivar ‘Nipponbare’, were re-sequenced using high-throughout multiplexed shotgun genotyping (MSG) technology. In combination with high-density single nucleotide polymorphisms, quantitative trait locus (QTL) mapping and genetic effect analysis were performed for five yield factors (spikelet number per panicle, primary branches per panicle, secondary branches per panicle, plant height, and panicle length) to explore the genetic mechanisms underlying the formation of the giant panicle of R1128. Also, they were preformed to locate new high-yielding rice genetic intervals, providing data for super-high-yielding rice breeding.ResultsQTL mapping and genetic effect analysis for five yield factors in the population gave the following results: 49 QTLs for the five yield factors were distributed on 11 of 12 chromosomes. The super-hybrid line R1128 carries multiple major genes for good traits, including Sd1 for plant height, Hd1 and Ehd1 for heading date, Gn1a for spikelet number and IPA1 for ideal plant shape. These genes accounted for 44.3%, 21.9%, 6.2%, 12.9% and 10.6% of the phenotypic variation in the individual traits. Six novel QTLs, qph1-2, qph9-1, qpl12-1, qgn3-1, qgn11-1 and qsbn11-1 are reported here for the first time.ConclusionsHigh-throughout sequencing technology makes it convenient to study rice genomics and makes the QTL/gene mapping direct, efficient, and more reliable. The genetic regions discovered in this study will be valuable for breeding in rice varieties because of the diverse genetic backgrounds of the rice.Electronic supplementary materialThe online version of this article (doi:10.1186/1939-8433-6-21) contains supplementary material, which is available to authorized users.

This paper presents investigations of segregation distortion of six rice F2 populations generated from reciprocal F1 hybrids grown at three locations varied at altitudes from 400 to 2200 m. The F1s were derived from reciprocal crosses between cv. XMG, which is a japonica landrace traditionally grown at 2650 m altitude, and cv. N34, which is a japonica restorer possessing a fertility restoring (Rf) gene and cytoplasm of male sterility (CMS) donated by an indica cultivar. Among nine morphological traits of the F2 populations, only one was in normal distribution, eight were distorted in all or at least one population. Out of 16 polymorphic PCR markers, 10 markers distributed on 7 chromosomes were significantly distorted. Among these markers, RMAN7 and RM257 were distorted in both of the reciprocal populations, which suggested that nuclear genes had strong effects on segregation distortion. The other makers were distorted only in the populations with cytoplasm donated by XMG or N34. The results indicated that segregation of DNA markers was affected by cytoplasm background. Segregation distribution was also affected by altitude, since segregation distortions of most of the markers were detected not in all the three populations generated from F1 grown at the three altitudes, but only in one population from F1 grown at one altitude. Marker M45461, which is located within Rf-1 locus, was severely distorted towards N34 in all the populations with cytoplasm donated by N34, but not in the populations with cytoplasm provided by XMG. The results indicated that interaction between CMS and Rf gene had strong effects on distortion. Results of this study indicated that japonica cytoplasm did not cause distortion favouring a special parent, but indica cytoplasm made distortion favouring a maternal parent. The results suggested that indica cytoplasm was not well compatible with japonica nuclear background, while japonica cytoplasm did not have such trouble with indica nuclei. This study also found that the six F2 populations were divergent into two groups due to difference of cytoplasm background.