QTL Identification For Rice Yield-related Traits Using Chromosome Segment Substitution Lines And Genetic Interaction Analysis Among Four Heading Date Genes

Part I:Yield per plant in rice is determined by three component traits:number of panicles per plant,number of grains per panicle,and grain weight.Mining and cloning QTLs related to yield traits are essential for marker assisted selection breeding.Chromosome segment substitution lines?CSSLs?are excellent materials for QTL mapping due to the advantages of simple genetic composition and strong ability to QTL detection.In this study,a set of high-quality CSSLs were constructed using two landraces,Chuan 7?C7?and Haobuka?HBK?,which have the significant differences in agronomic traits.QTLs for yield related traits were detected in these CSSLs.The main results are as follows:1. Using C7 as the donor parent and HBK as the recurrent parent,a set of CSSLs?BC₄F₄?with 123 families was constructed,of which 69 lines were chromosome single segment substitution lines.The length of single introgression segment ranged from 0.3Mb to 25.8Mb,with an average length of 7.9Mb,which covered 95.7%of the entire genome of the donor parent.2. Five QTLs for heading date?HD?,7 QTLs for plant height?PH?,6 QTLs for panicle length?PL?and 1 QTL for tiller number?TN?were identified in CSSLs,respectively.Among them,q HD10/Ehd1 was the major QTL for HD,and its additive effect was 5.1 days.qPH1.2/SD1 and qPH8 were major QTLs for PH,of which additive effects were 12.4 cm and-15.2 cm,respectively?negative values indicating that C7 allele reduced plant height?.q PL1 and q PL10 were major QTLs for PL,of which additive effects were 1.8 cm and 2.1cm,respectively.q PL1 and q PL10 were co-located with qPH1.2 and q HD10,respectively.In addition,3 QTLs for heading date,3 QTLs for plant height,3 QTLs for panicle length and 1 QTL for tiller number were detected at the 99%confidence interval level utilizing the BSA-seq strategy in the F₂ population?940 plants?,and all these QTLs were verified in either F₂?200 or BC₄F₂ populations.3.Four QTLs for grain length?GL?,3 QTLs for grain width?GW?,5 QTLs for thousand-grain weight?TGW?and 5 QTLs for spikelets per plant?SPP?were detected in CSSLs,respectively.Except for q GL5,the C7 allele of the other QTLs for GL,GW and TGW decreased the phenotypic value.q GL3.1 and q GW5 were the major QTLs for GL and GW,corresponding to GS3 and GW5,respectively.The additive effects of these two QTLs were-1.25 mm and-0.30 mm,respectively.q SPP3.2 and q SPP10 were the major QTLs for SPP,with additive effects of 16.3 and 14.6,respectively.q SPP10 was co-located with q HD10,but q SPP3.2 was a new QTL.4.qPH8 was mapped to an interval between the markers M15 and B4 as a new QTL for PH,which mainly controlled the basal internodes elongation.C7 allele of qPH8significantly reduced the plant height,improved the lodging resistance,but didn't significantly reduce grain yield,which contributed to genetic improvement of rice breeding.In addition,qPH8 is a semi-dominant gene,and only one parent is needed to introduce the C7 allele of qPH8 to achieve semi-dwarf breeding in hybrid rice breeding.Part II:Heading date determines the ecogeographical adaptation and yield of rice.Cultivars with an appropriate heading date will be conductive to high grain yield by fully utilizing the light and temperature resources.The heading date of rice is controlled by multiple genes.Ghd7,Ghd8,Ghd7.1/Os PRR37?PRR37?and Hd1 are important heading date genes.Their combinations determine the heading date variation and cultivation region of rice,but the genetic interactions among them are still unclear.Here,we performed a genetic interaction analysis among these four genes in the near-isogenic background under both natural long-day?NLD?and natural short-day?NSD?conditions.The main results are as follows:1.The 4-gene segregating population of Ghd7,Ghd8,PRR37 and Hd1 was constructed under ZS97 background,and 16 homozygous genotype combinations of these four genes were obtained.This population exhibited a large heading date variation with more than 95days under NLD and 42 days under NSD conditions,respectively.2.Digenic,trigenic and tetragenic interactions among these four genes were observed under both conditions but more significant under NLD conditions.In the functional Hd1backgrounds,the strongest digenic interaction was Ghd7 by Ghd8 under NLD but was Ghd7 by PRR37 under NSD conditions.3. PRR37 acted as a flowering suppressor under NLD conditions.However,PRR37 promoted heading date at 4.8 days,18.0 days,and 5.3 days in Ghd7ghd8Hd1,Ghd7Ghd8Hd1,and Ghd7Ghd8hd1 background,respectively,under NSD conditions.The function conversion?from suppression of heading to promotion of heading?of PRR37under NSD conditions depended on different combinations of the other three genes.4. Hd1 acted as an activator of heading date under both conditions,but it exhibited a function conversion under NLD conditions.Hd1 promoted rice heading for 9.5 days,4.9days,and 3.8 days in ghd7ghd8prr37,ghd7ghd8PRR37,and ghd7Ghd8prr37 background,respectively,under NLD conditions,but delayed rice heading for 2.5 days in Ghd7ghd8prr37 background.It indicated that PRR37 or Ghd8 alone couldn't trigger Hd1function conversion,but Ghd7 did.In addition,the interactions among Ghd8,PRR37 and Ghd7 enhanced the function conversion of Hd1 under NLD conditions.5. We identified the relationship between heading date and spikelets per panicle on the basis of performance of 16 homozygous four-gene combinations.A positive correlation between heading date and spikelets per panicle was found under NSD conditions,but changed to a negative correlation when heading date was over 90 days under NLD conditions.