Quantitative Trait Locus Analysis For Seed Size Related Traits And Gene Expression In A RIL Population In Peanut

Peanut（Arachis hypogaea L.）is an important oil crop and cash crop,being a good source of vegetative oil and proteins and playing an important role in national economy and society development.To meet the growing demand of vegetative oil,it is important to improve peanut yield.Generally,seed size and weight significantly effects peanut yield and is important to measure potential of yield.In our previous study,we developed a recombinant inbred line（RIL）population derived from crossing Zhonghua 10 and ICG 12625 and constructed genetic map for F₂population and F₆ population,respectively,based on simple sequence repeat（SSR）markers.However,it was unknown about the physical distance of QTL region for traits in peanut genome because of lack of physical location of SSR markers.In this study,we sequenced the whole transcriptome of peanut immature seeds after flowering 30 days sampled in the RIL population,constructed a genetic map based on single nucleotide polymorphism（SNP）markers and performed QTL analysis for seed length,seed width and hundred seed weight in three environments.Meanwhile,gene expressions,which also could be considered as phenotypic traits,were performed genetic variation analysis and genome wide QTL analysis.A major QTL of testa color as an example was analyzed to illustrate the application of population transcriptome data in identification of candidate genes for traits.The main results were described as follows:1. Phenotypic variation of seed size and weight traits in RIL population.The seed length,seed width and hundred seed weight were evaluated in the RIL population and the parents,Zhonghua 10 and ICG 12625 in three years（2015-2017）.Each of the three traits,which had significantly difference between parents,had large variation in the RIL population and distributed normally in three environments.Each pair of three traits had significantly correlation.Two-way analysis of variance revealed that genetic,environmental effects and genotype by environment interaction significantly influenced seed length.However,only genetic and environmental effects significantly influenced seed width and hundred seed weight.2. Gene expression variations in peanut immature seed.There were 49,691 genes that expressed in both parents and more than 90% of RILs,which were employed in subsequent analyses.In the RIL population,over 99%genes showed much broader variation range in their expression in RIL population.The coefficient of variation for expression in the population strongly depended on the initial parents’expression difference.There were 28,392（57.1%）genes expressed in a bimodal distribution,while the expression of 15,928（32.1%）genes exhibited a normal distribution,leaving 5,371（10.8%）genes as unclassified distributions.For the majority of 49,691 genes,the population mean expression showed roughly approximate to the parents mean expression.Interestingly,however,we found that the expression of 92 genes in the population apparently departed from the parents mean expression,following a paramutation-like expression pattern.3. Genetic map construction based on SNP markers.A total of 1,285 SNPs（aa*bb）were clearly called to be the homozygous genotype for both parents.Considering the genomic complexity of tetra-polyploid peanut,we used these SNPs to construct genetic map,which had 20 linkage groups 1301 loci,1911.57c M of length and 1.47c M of density.Synteny analysis revealed high co-linearity between genetic map and the physical map（reference genome）of two wild diploid ancestors,albeit a small fraction of inverted segments existed.The genetic map covered approximately 96.4%of diploid peanut reference genomes.4. QTL analysis for seed size and weight traits.QTL analysis were performed for seed length,seed width and hundred seed weight in three environments.We identified 7,8and 10 QTLs,respectively for seed length,seed width and hundred seed weight,with phenotypic variance explained 5.24%-25.55%（LOD>3）.QTL q SLA2.2 and q SLB5 for seed length,QTLq SWA3,q SWA7.1 and q SWB6.2 for seed width,QTLq HSWA5,q HSWA7.2,q HSWA9.1 and q HSWB6.1 for hundred seed weight were identified in each environment.There were 5 QTL hotspots regulating more than 2 traits.Condition QTL analysis indicated that QTL q SLA2.2 and QTLq SWA3 not only directly regulated seed length and seed width,respectively,but also indirectly regulated hundred seed weight.These two QTLs were novel QTLs for seed size and weight.5. Genome-wide e QTL analysis in peanut.By treating gene expression as a quantitative trait,a global e QTL analysis was performed for 49,691 genes in RIL population.Totally,17,044 e QTLs were detected to regulate the expression variation of 11,268 genes in the RIL population（LOD>4.19）.The e QTLs for bimodal-expression genes were found to explain significantly higher than the e QTL for normal-expression genes.On the basis of whether an e QTL regulate the gene expression nearby or far away,all e QTLs were designated into 1,207 local e QTLs and 15,837 distant e QTLs.On average,the local e QTLs explained higher gene expression variation than distant e QTLs.We detected a total of 94e QTL hotspots across the whole genome,which were lack on chromosome A08.Gene ontology analysis was performed for genes in the e QTL hotspot and found that these genes mainly participated in metabolic process.6. Application of population transcriptome data in identification of candidate genes for purple testa in peanut.QTL analysis was performed for purple testa in the RIL population and identified a major QTL in region of 97,001,012-102,338,287bp on chromosome A10,which contained 196 genes.Integration of correlation analysis between population gene expression and testa color,gene expression difference analysis between parents and gene annotation analysis rapidly identified a novel candidate gene,Aradu.10025440.To identify the DNA sequence variation of candidate genes,the whole genome resequencing data was generated for parents and In Del02 marker（CGCCTCG/C）linked the candidate gene was developed.The amplification in two parents,4 white testa germplasms,10 pink testa germplasms,4 red testa germplasms and 10 purple testa germplasms indicated that In Del02 marker had highest detection accuracy of purple testa in different peanut accessions.Purple testa as a good case illustrated the powerful potential in application of population transcriptome data for trait dissection.