Study On The Genetic Basis Of Maize Photoperiod Response Traits

Maize is the most important crops for in the world.The flowering time and photoperiod response traits are not only important for maize reproductive development,but also play a key role in the processes of maize domestication and adaption.The utilization of elite hybrid varieties is a crucial factor that promotes the increase of maize production.However,the narrowed genetic diversity of germplasm has severely restricted the sustainable development of modern maize breeding program.To solve this problem,introducing new favourable allele to broaden the germplasm basis is one of the most imminent tasks.Tropical maize which has abundant genetic diversity is a preferred candidate for germplasm improvement.Unfortunately,most tropical maize lines are not well suited to temperate breeding program because they demonstrate severe photoperiod response such as increased lodging,excessive grain moisture,excessive plant height,and delayed flowering time when grown in long-day environments.Therefore,dissecting the genetic basis of maize flowering time and photoperiod response traits by genetics and genomics tool will greatly facilitate the utilization of tropical germplasm.In this study,we employed linkage and association analysis to identify flowering time and photoperiod response QTL.What's more,the important role of gene ZmCCT on maize photoperiod response and post-domestication was studied by candidate gene association mapping,linkage population validation,haplotype and evolution analysis.1.QTL mapping for photoperiod response(PR)was performed using the BLUP values of 9 long-day and 2 short-day environments in 10 RIL populations with high density genetic maps which containing 1,979-3,071 genetic bins.A total of 51 QTL were mapped in 10 RIL populations which including 25 and 26 QTL for anthesis photoperiod response(APR)and silk photoperiod response(SPR)respectively.For APR,the QTL detected were distributed in all chromosomes except Chr.4 and Chr.10,with QTL numbers in each RIL population ranging from 1 to 5,and phenotypic variation explained(PVE)by single QTL ranging from 4.81%to18.62%;Among them,9 QTL have PVE larger than 10%(43%).The confidence interval for each QTL ranged from 3.87 to 50.9Mb and with an average of 9.78Mb.For SPR,the QTL detected were distributed in all chromosomes except Chr.10,with QTL numbers in each RIL population ranging from 1 to 4,and PVE by single QTL ranging from 5.25%to 20.82%;Among them,6 QTL have PVE larger than 10%(23%).The confidence interval for each QTL ranged from 1.3 to 72.2Mb and with an average of 11.3Mb.Combining 10 populations,an integrated genetic map with 14,613 genetic bins was constructed to perform joint linkage mapping(JLM).In total,72 QTL were detected for maize photoperiod response traits with 32 and 40 QTL for APR and SPR respectively.For APR,the QTL detected by JLM were distributed in all chromosomes except Chr.5.The physical length of each QTL peak bin ranged from 16.9kb to 1.49Mb and with an average of 235.8kb;For SPR,the QTL detected by JLM were distributed in all ten maize chromosomes.The physical length of each QTL peak bin ranged from 7kb to 3.29Mb and with an average of 247.6kb.In addition,103 million high quality SNPs obtained by RNA-seq of 14 founder lines were projected to all the RIL family lines.RIL based genome wide association study(GWAS)was conducted for PR based on a stepwise regression model with resampling procedure and a total of 64 significant SNPs were detected.Among the 64 significant SNPs,26 of them were retained in the final model by backward selection of the liner model which were defined as candidate SNP.For APR,the significant and candidate SNP were 36 and 25 respectively,and the PVE for the full model and SNP effects were 34.7%and 22.7%respectively.For SPR,the significant and candidate SNP were 28 and 11 respectively,and the PVE for the full model and SNP effects were 31.9%and 13.5%respectively.Among the 25 QTL for APR detected by SLM,10(40.0%)QTL were overlapped with 18(56.3%)QTL identified by JLM;10(40.0%)QTL of SLM covered the genomic regions of 16(44.4%)significant GWAS SNPs;Among the 36 significant SNPs detected by RIL based GWAS,8(22.2%)of them localized within 6(18.8%)JLM-indentified QTL;Regarding to all the loci detected by three methods for APR,there were 8 significant SNPs of RIL based GWAS localized within the physical interval of 5 JLM-indentified QTL,which corresponding to the confidence intervals of 7 SLM-indentified QTL.For SPR,Among the 26 QTL detected by SLM,10(38.5%)QTL were overlapped with 17(42.5%)QTL identified by JLM;10(38.5%)QTL of SLM covered the genomic regions of 12(42.9%)significant GWAS SNPs;Among the 28 significant SNPs detected by RIL based GWAS,13(46.2%)of them localized within 7(17.5%)JLM-indentified QTL;Regarding to all the loci detected by three methods for SPR,there were 4 significant SNPs of RIL based GWAS localized within the physical interval of 2 JLM-indentified QTL,which corresponding to the confidence intervals of 4 SLM-indentified QTL.Comparison of the co-localized loci of APR and SPR indicates that the genetic basis for photoperiod response of the maize male and female flower organ was quite different.For the three analysis methods,the confidence interval of SLM-indentified QTL were very large which make it difficult to anchor the candidate gene.However,by JLM method,we could narrow the QTL to a relatively small interval.What's more,by GWAS,it is possible for us to indentify candidate genes nearby the significant SNPs.Each of the three methods has their own advantages for digging unique loci,while they can also co-localize common chromosome segments that responsible for maize photoperiod response,which is a strong support for the corresponding loci.By employing all the three methods,we could gain a comprehensive perspective for the genetic basis of maize photoperiod response.2.We performed a GWAS using the BLUP values of 6 long-day and 2 short-day environments across China in a panel of 368 maize inbred lines to identify loci associated with flowering time andphotoperiod response traits.For each line,557,955 high quality polymorphic sites developed by RNA-seq,Illumina 56K chip and resequencing of candidate gene(ZmCCT)were used for the GWAS.A total of 48 polymorphicsites,including 41 SNPs,and seven insertions or deletions(InDels)were significantly associated with flowering time traits(P?1.8×104)and an additional 20 SNPs and one InDel were with marginal significance(1.8×10-6<P?1.8×10-5).Of the 48 highly significant sites,42 were located with in the ZmCCT promoter region,which indicated the important effect of gene ZmCCT for maize flowering time and photoperiod response traits.3.The ZmCCT genomic locus within a diverse group of 180 lines of maize was resequenced.A total of 136 variants with MAF of?0.05 across the ZmCCT locus were discovered,with eight in the coding region.Of these 136 variants,29 weresignificantly associated with photoperiod sensitivity(P?2.8×10-4,n=180).A CACTA transposon element(TE)in the promoter region of ZmCCTshowed the strongest association.The genetic effect of TE was attenuated with the decreasing of latitude in the association panel and four F2 populations,and the TE effect in tropical germplasm was larger than that of temperate germplasm.Combining TE with variants in the promoter region to conduct haplotype analysis,we found the TE-related haplotypes displayed variable photoperiod sensitivities.The ZmCCT genomic locus of 32 teosintes and the ZmCCT promoter region of 61 teosintes and 481 maize lines were also resequenced.Evolution analysis found that maize lines with the TE insertionexhibited dramatically reduced diversity in the promoter region which indicated that this region experienced a strong selective sweep and a history of positive selection.Furthermore,a minimum-spanning tree of the haplotypes constructed by variants of maize lines and teosintes of the ZmCCT promoter region,revealed two distinct clusters-a maize haplotype cluster and a teosinte haplotype cluster.The maize haplotype cluster could be further divided into two distinct subgroups.These results suggest that the TE insertion occurred after domestication and accumulated during the adaptation of maize from tropical to temperate regions.