Genome-wide pedigree analysis of elite rice Shuhui 527 reveals key regions for breeding

REN Yun ,CHEN DanLI Wen-jieTAO LuoYUAN Guo-qiangCAO YeLI Xue-meiDENG QimingWANG Shi-quanZHENG Ai-pingZHU JunLIU Huai-nianWANG Ling-xiaLI PingLI Shuang-cheng

1 State Key Laboratory of Hybrid Rice/Rice Research Institute,Sichuan Agricultural University,Chengdu 611130,P.R.China

2 College of Landscape Architecture and Life Science/Institute of Special Plants,Chongqing University of Arts and Sciences,Yongchuan 402160,P.R.China

Abstract Hybrid rice significantly contributes to the food supply worldwide. Backbone parents play important roles in elite hybrid rice breeding systems. In this study,we performed pedigree-based analysis of the elite backbone parent rice variety,namely,Shuhui 527 (SH527,Oryza sativa),to exploit key genome regions during breeding. Twenty-four cultivars (including SH527,its six progenitors and 17 derived cultivars) were collected and analyzed with high-density single nucleotide polymorphism(SNP) array. Scanning all these cultivars with genome-wide SNP data indicated the unique contributions of progenitors to the SH527 genome and identified the key genomic regions of SH527 conserved within all its derivatives. These findings were further supported by known rice yield-related genes or unknown QTLs identified by genome-wide association study.This study reveals several key regions for SH527 and provides insights into hybrid rice breeding.

Keywords:backbone parent,Shuhui 527,SNPs,genome-wide association study

1.Introduction

Rice is an important cereal crop cultivated worldwide and serves as a carbohydrate source for more than half of the world population. Improving rice productivity by optimizing grain size and plant architecture is the major method employed in rice breeding programs. The rice yield has been increased through two breakthroughs:the utilization of semi-dwarf genes that improve plant architecture and harvest index and the development of hybrid heterosis (Yu and Zhu 2009). Scientists have bred thousands of rice varieties (Lin and Min 1991; Wan 2010) and found that backbone parents play an important role in breeding new cultivars (Tanget al.2012). Backbone parents of hybrid rice contain at least two important characteristics:1) common parent of several hybrid rice varieties released publicly,and 2) major contributors of important breeding lines widely used. Thus,backbone parents always exhibit high yield,good quality,multiresistance,and wide adaptability (Wan 2006; Sunet al.2014).

Shuhui 527 (SH527,Oryza sativa),developed at the Rice Research Institute,Sichuan Agricultural University in 2000,is a famous restorer line characterized by heavy panicles,good leaf shape,strong resilience,high combining ability,excellent quality,and resistance to rice blast (Fig.1-A).The female parent R1318 is a heavy-panicle material for blast resistance developed by the foreign-introduced strong restorer materials G630,GU154 and rice blast resistant material IR1544-28-2-3; the father parent 88-R3360 is a high-quality strong-resilience material developed by the hybrid of F36-2 and IR24 (Fig.1-B). SH527 is one of the main backbone parents in modern hybrid rice breeding in China contributing to more than 30 hybrid rice varieties approved by the national or provincial level,such as Gangyou 527,Dyou527,Cyou527,Byou827,Xieyou 527,and Mian5you527 (Wang Y Pet al.2004; Wang Zet al.2004;Chenet al.2005),which have been widely cultivated due to high quality and high yield. More than 30 elite restorer lines,such as Nanhui 511 (NH511; Jieet al.2005),3301 (Chenet al.2013),and Shuhui 238 (SH238; Xiao and Zhang 2010)have also been derived from SH527; these lines are widely used in hybrid rice breeding (Chenet al.2008; Cuiet al.2008;Qianet al.2009). Although SH527 plays an important role in modern rice breeding in China,the genetic mechanism underlying its excellent performance remains unknown.

Rice is a model plant for crop genomics research (Izawa and Shimamoto 1996). The genome sequences of the two major rice subspeciesindicaandjaponicawere released over a decade (IRGSP 2005; Yuet al.2006). Based on these genome sequences,re-sequencing allows the redesigning of genotyping strategies for effective genetic and genomic analysis in rice (Huanget al.2009,2013; Guoet al.2014).Several genomics-based genotyping platforms have been developed using the large amount of genome re-sequencing data and high-throughput genotyping technologies available;these platforms include Illumina SNP (single nucleotide polymorphism) Chip Platform (http://www.illumina.com),Affymetrix SNP Array Platform (http://www.affymetrix.com),Genome Lab SNP Stream Genotyping System (Beckman Coulter,https://www.beckmancoulter.com),andTaq ManOpen Array Genotyping System (Applied Biosystems,http://www.appliedbiosystems.com). These platforms enable low-cost,fast and high-throughput genotyping of rice cultivars and are widely used in background analysis,genetic mapping and genome-wide association studies(GWAS) (Huanget al.2010; Keyanet al.2010; Luoet al.2011; Zhaoet al.2011; Takagiet al.2013)

Considering the rapidly accumulating rice genomic sequencing data and the robustness and convenience of Illumina Infinium Platform,Chenet al.(2014) selected 51 478 SNPs by using the public and their own rice resequencing data to develop a large-scale genotyping array on the platform RiceSNP50. SNP probes on the array were selected and designed based on the analysis of over 10 000 000 SNPs extracted from the sequence data of 801 rice accessions. These SNPs were preferentially located in annotated genes and evenly distributed across the genome.Thus,this high-resolution SNP array can be used for rice genomic investigations,especially for developing molecular markers for important rice traits.

Fig.1 Pedigree of backbone parent Shuhui 527 (SH527).A,plant of backbone parent SH527. B,breeding pedigree of SH527.

The clear breeding history with a carefully recorded pedigree of SH257 makes it an ideal candidate for dissecting the genetic basis of modern hybrid rice breeding (Fig.1-B).In this study,the RiceSNP50 array was used to genotype 24 cultivars comprised by SH527,its six progenitors and 17 derivative lines. This analysis uncovered the specific contributions of progenitors to the SH527 genome and the key genomic regions of SH527 conserved in all derivatives.Further analysis of known rice yield-related genes or unknown QTLs identified by GWAS analysis revealed the basis for the excellent performance of SH527 and all its derivatives.Our study provides a foundation for further hybrid rice breeding.

2.Materials and methods

2.1.Plant materials

Materials used in this study include SH527,six progenitors of SH527 (88-R3360 was not obtained) and 17 SH527 derivatives (Appendices A and B). All materials were planted in the experimental field of the Rice Research Institute,Sichuan Agricultural University,Wenjiang,China in 2012,2013 and 2014. Each material is randomly planted in three plots,each plot is planted with 6 to 8 rows,and 12 plants are planted in each row. The spacing between plants is 20 cm×20 cm.

2.2.Agronomic trait investigation

Ten traits including heading date (HD),plant height (PH),tiller number (TN),effective panicle number (EPN),panicle seed-setting rate (PSSR),thousand-kernel weight (TKW),spikelet number per panicle (SpNPP),seed number per panicle (SeNPP),seed weight per plant (WPPL),and seed weight per panicle (WPPA) of all pedigree materials of SH527 were investigated during the summer of 2012,2013 and 2014 in Wenjiang. Remove marginal plants and randomly select 20-30 plants from three plots for investigation. The investigation methods were as follows:HD,the time when 50% of the plants were heading; PH,the average height from the soil surface to the highest panicle(not to the awn) at the maturity stage; TN,the total number of tillers in the wax maturity period; EPN,except the panicle with less than five grains,all other panicles are counted as effective panicles; PSSR (%),average number of fulfilling grains per panicle/average total number of grains×100(grains less than two-thirds filling were count as empty granules); TKW,200 fulfilling grains were weighed for TKW measurement with two replications; SpNPP,SeNPP and WPPA,three main panicles of each plant were selected for measurement; WPPL=WPPA×EPN.

2.3.Rice SNP microarray scanning and data analysis

All the cultivars in the pedigree of SH527 (indica),together with Minghui 63 (M63,indica) and Taipei 309 (TP309,japonica),served as the inner and outer group controls,respectively,were hybridized with the RiceSNP50 microarray by using the methods described by Chenet al.(2014).

Based on the genome-wide SNP microarray data of SH527 progenitors and all derivatives,the typicaljaponicarice TP309 as the outer group control,a phylogenetic tree of the whole pedigree material was constructed using MEGA5 Software.

Contributions of progenitors to the SH527 genomeAccording to the similarities and differences of genetic markers of SH527 and its six progenitors,the contribution is analyzed. For example,the fragments same as G630 were recorded as G630 contributions. In addition,if the fragments were same as G630 and same as other progenitors simultaneously,it was considered to be originated from two or more progenitors; if the fragments were same as G630 but different from other progenitors,it was considered to be derived from G630-specific contributions. During the analysis,the TP309 SNP set information was always used as the control and the same marker fragment as TP309 was knocked out (the same fragment as TP309 was recorded as a conserved fragment in rice instead of a progenitor contribution).

Key genomic regions of SH527Fragment of the derivatives that is identical to SH527 and different from TP309 is recorded as key region of SH527 that is inherited to the next generation.

2.4.Construction of a SH527 genomic structure map

A SH527 genomic structure map was constructed by integrating data,including the unique contribution regions of the progenitor,the key regions,the known genes,and the SNP loci associated with important agronomic traits,into the rice genome. Information on the known genes/QTLs was obtained from the National Rice Data Center (http://www.ricedata.cn/index.htm) and the Rice Genome Annotation Project (http://rice.plantbiology.msu.edu/).

2.5.GWAS analysis

To determine putative traits related to SNPs,we tested the association between SNPs and the mean data of agronomic traits collected from 2012,2013 and 2014 in Wenjiang using compressed general linear model (GLM) and mixed linear model (MLM).

3.Results

3.1.Agronomic trait investigation

Evaluation of 10 agronomic traits was conducted for all the cultivars (Appendix C). It can be seen from six progenitor traits survey in Fig.2-A that the agronomic traits of SH527 were excellent,especially in the TKW and WPPL. The TKW of SH527 is higher than that of all other progenitors,and the WPPL is higher than that of all progenitors except G630;in addition,the PH of SH527 was also higher than that of other progenitors,and the EPN was similar to IR24. It can be seen that the TKW of G630,R1318 and SH527 is 31.03,30.95 and 35.67 g,respectively,and the other progenitors of SH527 do not have the traits of large TKW (＞30 g).Tracing the transmission law of SH527 yield-related traits,we preliminarily inferred that the TKW of SH527 is mainly inherited from the G630-R1318-SH527 path,whereas the EPN trait is mainly inherited from the IR24-SH527 path.Therefore,the yield performance of SH527 is mainly associated with the coordination of the large grain weight of G630 and the excellent plant architecture phenotype of IR24.

It can be seen from the investigation of the traits of 17 SH527 derivatives that SH527 derivatives are similar to SH527 in agronomic traits (Appendix C; Fig.2-B). The PH of the derivatives and SH527 was basically the same except for the shorter SH238. In terms of WPPL,only two derivatives,LH11 and NH511,exceeded SH527,and the WPPL of other derivatives was lower than or close to SH527.Among them,MH2088 had the lowest WPPL,which may be related to its earlier maturity.In addition,most derivatives have the important characteristics of “the heavy panicle type rice” and with large TKW (＞30 g),especially NH511 (37.04 g) and SH288 (35.75 g). On the whole,these derivatives all reflect the characteristics of SH527 to a certain extent,which indicates that SH527 provides a large number of genes or fragments related to its own superior traits for the newly derived restorer lines in the genetic process.

Fig.2 Agronomic characters of all the cultivars in the pedigree of Shuhui 527 (SH527). A,agronomic characters of SH527 and its progenitors. B,agronomic characters of SH527 and its derivatives. PH,plant height; EPN,effective panicle number; TKW,thousand-kernel weight; WPPL,seed weight per plant. Bars are SD (n=20).

3.2.Phylogenetic analysis

A total of 44 800 high-quality SNP data were obtained by hybridizing the genomic DNA of the cultivars in the SH527 pedigree with the RiceSNP50 array. Based on the 44 800 high-quality SNP data for all SH527 pedigree cultivars and the two controls,inner control M63 (indica) and outer control TP309 (japonica),a neighbor-joining tree was constructed by using MEGA5 Software (Fig.3). It can be seen that the SH527 progenitors and its derivatives show great differences with TP309,which is consistent with the fact that these materials are mainlyindicabackground. Comparative analysis of 17 derivatives with SH527 showed that all the derivatives had higher similarity with SH527. Among which the highest two were LH5240 and MH2088,and the genome similarity to SH527 was 98.53 and 98.47%,respectively. The lowest two were SH238 and B4114,and the genome similarity was 68.96 and 65.50%,respectively,indicating that the SH527’s derivative was highly similar to SH527 as a whole. The progenitors of SH527,IR24,F36-2,and GU154,are far away from SH527 with low genome similarity.

3.3.Contributions of progenitors to the SH527 genome

By comparing SNPs between SH527 and its progenitors,we found that the SH527 genome can be divided into traceable fragments to elucidate the contributions of each progenitor to SH527. In all fragments,14.20% of the SH527 genome was unambiguously derived from G630,3.58%from IR24,1.18% from F36-2,0.98% from GU154,and 0.07% from IR1544 (Appendix D). This result suggests that the SH527 genome contains significant contributions from the progenitors G630 and IR24. This is consistent with the breeding objective of fully utilizing “the large grain and heavy panicle” characteristic of the restorer line G630 and “the compact plant architecture” of the restorer line IR24.IR1544 may only improve the rice blast resistance of SH527,so relatively fewer regions were identified as uniquely inherited from IR1544.

Numerous traceable fragments possibly originated from two or more progenitors. A total of 515 fragments (18.09% of the entire genome) may originate from two progenitors,619 fragments (15.42%) from three progenitors,1 383 fragments(24.74%) from four progenitors and 3 911 fragments(31.91%) from five progenitors (no polymorphism between all progenitors) (Appendix D). Interestingly,we also detected 60 SH527 unique fragments,covering 716 412 bp and accounting for 0.19% of the entire genome (Appendix D).Thus,although SH527 inherited lager number of fragments from its progenitors,it possesses substantial variations.

We also constructed a haplotype map for SH527 by using continuous specific SNP markers from each progenitor (Fig.4; Appendix E). The map shows the specific contributions of each progenitor maintained in SH527. G630 exhibited the highest specific contribution to SH527 by providing 366 unique fragments (5 348 693 bp),accounting for 1.44% of the entire genome. The remaining progenitors were ranked according to the degree of contribution:F36-2＞GU154＞IR24＞IR1544,and the genomic contribution rate was between 0.02-0.27% (Appendix E).

Fig.3 Neighbor-joining tree of all Shuhui 527 (SH527) pedigree cultivars. SH527 is represented in red; upstream progenitors are represented in orange; derivatives are represented in green; and the control,M63 (indica) and TP309 (japonica), are represented in gray; similarity of each cultivar to the SH527 genome is shown in the right side.

In order to distinguish the yield-related genes on SH527 derived from progenitors,we collected information of 78 known yield-related genes (http://www.ricedata.cn/index.htm,Appendix F),including those related to grain weight,grain shape,grain number,plant height,tiller number,and seed setting rate. The results showed that a total of 13 genes were found located in G630-specific contributions(Appendix G),including genes related to grain weightGW8(Wanget al.2012),GS5(Liet al.2011),GW2(Songet al.2007),OsPPKL3(Zhanget al.2012),andSRS3(Kitagawaet al.2010),genes related to plant architecture and yieldOsCD1(Luanet al.2010),EP3(Piaoet al.2009) andSGL1(Nakagawaet al.2012),as well as genes related to stress and yieldOsFAD8(Liuet al.2012),OsNAC10(Jin and Kim 2010) andOsJMT1(Kimet al.2009). These genes probably lead to the large grain,heavy panicle,high yield,and strong stress resistance of SH527. Meanwhile,four genes were located in IR24 originated fragments,of which,D14(Ariteet al.2009),d-27(Linet al.2009) andSP1(Liet al.2009) are mainly associated with the plant architectural traits,such as tiller and panicle type. This suggests that IR24 may play a role in optimizing SH527 plant architecture.

3.4.Key genomic regions of SH527

Fig.4 Unique contribution of progenitor to Shuhui 527 (SH527). Each unique contribution region of the progenitor to SH527 is represented in different colors. The progenitor name is placed in the left side,and the percentage of these regions in the SH527 is shown in the right side.

As an important restorer line,SH527 directly contributed to the development of many new restorer lines. For identifying the key genomic regions of SH527 in these derivatives,we compared SNPs between SH527 and its derivatives and found that 3 147 fragments (accounting for 65 305 677 bp,about 17.54% of the SH527 genome) are stably maintained in these derivatives (Fig.5). These fragments contain 10 fragments longer than 200 kb (Appendix H). The largest fragment is 597 148 bp in length and located on Chr06. We believe that these fragments may play an important role in the excellent performance of SH527 and its derivatives are considered the key genomic regions of SH527.

We also detected 25 yield-related genes located in the key regions stably inherited in the derivatives of SH527(Appendix G). These genes includeTAD1(Linet al.2012),qGL3.1(Qiet al.2012),TGW6(Ishimaruet al.2013),GS5(Liet al.2011),andTH1(Yanet al.2013),which are related to grain weight and/or grain shape; this finding may explain why most of the derivative restorer lines possess large grains. In addition,some genes related to controlling grain number were found,which includesDST(Liet al.2013),APO1(Teraoet al.2010) andGn1a(Ashikariet al.2005). This finding confirms that most of these derivative restorer lines possess heavy panicles.We also found genes related to some plant architecture,panicle architceture and tiller number,such asOsTB1/FC1(Takedaet al.2003),DEP2(Zhuet al.2010),d-27(Linet al.2009),OsPIN2(Chenet al.2011),OsPIN1/REH1(Xuet al.2005),EUI1(Zhanget al.2008),andSGL1(Nakagawaet al.2012). Moreover,we found many genes associated with stress response,likeSQS(Manavalanet al.2012),OsSDR(Kimet al.2009),OsFAD8(Liuet al.2012),OsJMT1(Kimet al.2009),andOsPIP1(Mosaet al.2012); this explains the good tolerance of these derivative restorer lines to stresses.Ghd7is a key gene that synchronously regulates the growth period,grain number,plant height,and yield (Wenget al.2008),which was found stably inherited in all derivative restorer lines and thus explains their excellent yield performance.

3.5.ldentification of yield-related sites using GWAS

We performed GWAS to identify trait-associated sites by using agronomic trait data collected from three independent surveys in Wenjiang during 2012-2014 and the entire genome SNP data. A total of 261 significant association sites were detected using GLM and MLM models (Appendix I).Among these sites,146 SNPs were found to be associated with TN,37 SNPS with WPPA,31 SNPS with PSSR,20 SNPS with PH,11 SNPS with EPN,five SNPS with SeNPP and WPPL,only three SNPs with HD and TKW,and no SNP was associated with SpNPP. Interestingly,132 associated sites were located in regions uniquely associated with specific progenitors and key fragments of SH527(Appendix J). Eleven sites were located in G630 contributed regions,33 sites were located in IR24 contributed regions,and the other 88 sites were located in the key fragments of the SH527 genome. In addition,we identified seven possible candidate genes at/or adjacent to the peak SNP sites of three traits,one in PSSR,five in TN and one in WPPL(Fig.6; Appendix K). These GWAS results can be used for genetic dissecting of yield-related genes on the important genomic fragments of SH527.

3.6.Construction of a SH527genomic structure map

The genomic structure map of SH527 was constructed by integrating all the above-mentioned data,including the contribution regions of the progenitors,the key regions maintained in derivatives,the known genes,and the traitassociated SNPs within the SH527 genome (Fig.7).The map suggests that G630 and IR24 are the most important progenitors of SH527. Some known genes inherited from G630 (genes on yellow bar) and IR24 (genes on red bar)could explain the large grain and ideal plant and panicle architecture phenotype of SH527. Moreover,several genes associated with stress responses were located within the key regions,suggesting their roles in the improvement of stress tolerance of the SH527 derivatives. In addition,the map lists known yield-related genes located in the SH527 key regions (genes on blue and purple bar). The map also lists the SNPs associated with important agronomic traits within the SH527 genome (10 different color bars on the left side of the chromosome).

Fig.5 Key genomic regions of Shuhui 527 (SH527). The key genomic regions of SH527 stably inherited by its elite derivatives are represented in orange.The derivative name is shown in the left side,and the similarity of each derivative to the SH527 genome is shown in the right side.

4.Discussion

4.1.Contributions of progenitors to the SH527 genome

In the late 1980s,breeders first used G630,a strong restorer line with large grains,GU154 and IR1544,a rice blast resistant variety,breed “heavy panicle rice” variety,R1318,by polymerization hybridization (Wanget al.2004).Simultaneously,IR24 and F36-2 were crossed to improve the plant architecture for meeting the ecological conditions of Southwest China. These efforts finally leaded to the development of SH527,a strong restorer line with large grain and well blast resistance. Our analysis of the genetic structure of SH527 suggests that G630 and IR24 are major contributors of SH527 genome,which is consistent with the pedigree of SH527. The location of known important yield-associated genes in these regions further supported the important roles of the two progenitors.

Liuet al.(2012) preliminarily investigated the genetic composition of SH527 and traced the contribution fragments of some progenitors by screening SSR makers. The result indicated that the contribution of G630 and IR24 to SH527 is higher than that of the other progenitors. However,for identifying the precise blocks contributing to SH527,the low density of SSR markers (1 000 markers around) is apparently not enough. In contrast,SNP scanning based on gene chip has high density and good polymorphism. In this study,we analyzed the SH527 pedigree with a high-density rice SNP array containing 51 478 evenly distributed markers. SNP probes were designed by screening more than 10 000 000 SNP loci extracted from the re-sequencing data of 801 rice varieties. Through SNP chip typing technology,we realized the genetic analysis of SH527 and its progenitors and derivatives,and clarified the genetic makeup of SH527,and made a more reliable and detailed explanation for the source and inheritance of SH527 excellent traits.

4.2.Key genomic regions of SH527 and its derivatives

Fig.6 Manhattan plot of genome-wide association studies (GWAS) analysis for three traits. PSSR,panicle seed-setting rate; TN,tiller number; WPPL,seed weight per plant. Dashed line represents a 0.2 false discovery rate (-log10P=4.00).

Several elite new restorer lines have been bred using SH527 as parent. Thus,SH527 genomic regions stably transmitted to its derivatives are important for the excellent performance of SH527 and its derivatives. By comparing the SNPs between SH527 and its derivatives,we found that 17.54% of the SH527 genome is stably inherited by its derivative restorer lines. We also detected 25 yield-related genes located in the key regions and stably inherited in the derivative restorer lines of SH527. These genes,to some extent,explain the large grains and heavy panicles,good plant architectures,and strong stress tolerance of the derivative restorer lines. A total of 261 significant association sites were detected,among which 132 sites were located on specific fragments. This might be helpful for identifying the genes in these important genomic regions of SH527.However,most key regions do not contain known genes,and most of GWAS SNPs are currently unknown function,indicating the complexity of the genetic basis of SH527 key genomic regions.As an important parent of hybrid rice,SH527 is elite not only because of its outstanding performance and traits but also due to its strong combining ability. Zhanget al.(2008) identified several sites associated with the combining ability of SH527. This result and the present findings provide insights into dissecting of the genetic basis for the combining ability exhibited by SH527.

4.3.Narrow genetic background of the derivatives

A neighbor-joining tree was constructed based on the SNP data for all SH527 pedigree cultivars (Fig.3). SH527 pedigree cultivars were clustered into one group,which has a huge genetic distance compared with TP309,thejaponicaouter-group control. This clustering tree clearly shows that all derivatives are similar to their parent SH527,which suggests that SH527 greatly contributes to its derivatives.However,the tree reveals an extraordinary narrow genetic background for the restorer lines,which is frequently observed in hybrid breeding practice and leads to some risks in practice due to narrow genetic diversity. In contrast to the SH527 derivatives,our results showed that the progenitors of SH527 have a relatively distant genetic relationship,indicating that the different genetic backgrounds of the parent materials are necessary for successful breeding of a new and elite variety. Thus,the genetic diversity of parental materials must be expanded for future hybrid rice breeding.

Fig.7 Genomic structure map of Shuhui 527 (SH527). Blocks on chromosomes represent the unique regions of the progenitor,and the round dots on the left of the chromosomes represent the markers supporting these regions; blocks on the right of the chromosomes represent the key genomic regions; horizontal lines on the left of the chromosomes represent the genome-wide association studies (GWAS) association sites. Known genes are directly labeled on the map. HD,heading date; SeNPP,seed number per panicle; SpNPP,spikelet number per panicle; EPN,effective panicle number; PSSR,panicle seed-setting rate; TN,tiller number; PH,plant height; TKW,thousand-kernel weight; WPPA,seed weight per panicle; WPPL,seed weight per plant.

5.Conclusion

We have deeply analyzed the genetic composition of backbone parent SH527 through SNP chip data,and at the same time determined important candidate genes on various types of genomic regions through GWAS,which basically clarified the possible molecule basis of SH527’s excellent yield traits and key region of stable inheritance.However,these studies are relatively preliminary as a whole.Subsequent studies need to design practical and feasible genetic experiments to genetically verify these key regions and important candidate genes.

Acknowledgements

We thank Dr.Chen Haodong in School of Life Sciences,Peking University for the assistance of microarray analysis.This study was supported by the Sichuan Science and Technology Support Project,China (2016NZ0103),the National Natural Science Foundation of China (91435102 and 31570004),the Sichuan Provincial Founding for Distinguished Young Scholars,China (2015JQ0048),and the Open Research Fund of State Key Laboratory of Hybrid Rice,Hunan Hybrid Rice Research Center,China(2016KF10).

Appendicesassociated with this paper can be available on http://www.ChinaAgriSci.com/V2/En/appendix.htm