Genetic Analysis And QTL Mapping For Characteristic Parameters Related To Kernel Filling And Dehydration In Maize

The kernel filling and dehydration are complex and dynamic process in maize.The features of kernel filling and dehydration directly affect kernel field and quality in harvest.At present,the genetic analysis and QTL mapping for characteristic parameters based on whole filling and dehydration process in maize have not been reported.In this study,208 recombinant inbred lines derived from the maize inbred lines DH1M and T877 were evaluated in 2015 Nantong,2016 Yangzhou and 2017 Sanya,respectively.Two similar ears of each recombinant inbred lines were sampled at 10,15,20,25,30,35,40,43,46,49,52,55,58 and 61 days after pollination,of which 50 kernels dry weight and moisture content at middle were measured.Based on Logistic function,the change process of kernel dry weight and moisture content of each recombinant inbred line were fitted to achieve the filling curve and dehydration curve.On this basis,12 filling characteristic parameters were estimated,including kernel final dry weight,relative growth rate of dry weight,maximum filling rate and so on.Eight dehydration characteristic parameters were estimated,including dehydration duration,average dehydration rate,final moisture content and so on.Some analysis related to characteristic parameters of kernel filling and dehydration were performed using major genes plus polygene mixed genetic model,QTL mapping?BSR-seq and association analysis of candidate genes.The major results are as follow:1.The kernel filling features of two parents had obvious difference.DH1M had smaller final dry weight,shorter filling duration and larger maximum filling rate than T877.The coefficient of variation of filling characteristic parameters in three environments was from 9.53%to 45.11%,and broad heritability was from 69.26%to 79.33%.Maximum filling rate had significantly positive correlation with parameters related to kernel dry weight and significantly negative correlation with filling duration and active filling period,which was a noteworthy parameter.The final dry weight of kernel,time for median final growth amount,filling duration,accumulation during gradual filling period and accumulation during fast filling period were controlled by two major genes with complement effect in three environments.Relative growth rate of dry weight,time for first inflexion of growth rate,filling duration,accumulation during slow growth period,active filling period and maximum filling rate were controlled by two major genes with complement effect in two environments.Other filling characteristic parameters were controlled by three or four major genes,of which genetic model were different in three environments due to environmental impact.2.The kernel filling features of parents had obvious difference.DH1M had longer dehydration duration,earlier dehydration initial time and later dehydration termination time than T877.The coefficient of variation of dehydration characteristic parameters in three environments was from 10.9%to 234.7%,and broad heritability was from 27.9%to 61.4%.The correlation between most dehydration characteristic parameters were relatively stable in different environments.However,the correlation between different environments for same dehydration characteristic parameter was lower,indicating that dehydration characteristic parameters were easily impacted by environments.Dehydration termination time was controlled by two major genes in three environments,but the interaction between major genes was different.The change of moisture content in dehydration were controlled by four major genes with partial equal additive effects in three environments.Dehydration duration,dehydration initial time and final moisture content were controlled by two major genes in two environment.Average dehydration rate were controlled by four major genes with additive and epistasis effect in two environments.Initial dehydration rate were controlled by four major genes with partial equal additive effect in two environments.There was no major gene related to initial moisture content in 2016 Yangzhou and 2017 Sanya.However,in 2015 Nantong,two gene with complement were related to initial moisture content.3.We constructed a high resolution genetic linkage map with 3227 bin markers,of which overall length was 2450.31cM and range of chromosome length was from 102.29 to 373.06cM.Chromosome 1 had the largest number of markers,503.Chromosome 2 had the least number of markers,111 and marker density was the lowest.Chromosome 5 had the highest marker density.The average distance of adjacent marker is 0.76cM,and maximum distance was 25.39cM.4.A total of 156 QTL related to filling characteristic parameters were detected,where 2015 Nantong had 42 QTL,2016 Yangzhou had 51 QTL,2017 Sanya had 33 QTL and 30 QTL were detected using average value.Kernel final dry weight,relative growth rate of dry weight,and accumulation during fast filling period detected 10 QTL,respectively.Accumulation during gradual filling period and accumulation during slow filling detected 11 QTL,respectively.Filling duration and time for first inflexion of growth rate detected 12 and 13 QTL,respectively.Time for second inflexion of growth rate,active filling period,average filling rate and maximum filling rate detected 15 QTL,respectively.Time for median final growth amount detected 19 QTL.The phenotypic contribution rate of each QTL was from 1.09%to 18.55%.For relative growth rate of dry weight,average filling rate and maximum filling rate,their synergistic site came from T877.A pleiotropic locus controlled average filling rate and maximum filling rate had an interaction with environments,which located on chromosome 10.Based on distribution of BLUP values of maximum filling rate,13 low valued inbred lines and 13 high valued inbred lines were sampled to construct two extreme bulks.There was an obvious difference in filling curves and all filling characteristic parameters except for time for median final growth amount had significantly difference between two extreme bulks.Further,eight candidate genes were identified from 512 differentially expressed gene,which were related to synthesis and metabolism of glucide and carbohydrates.Through transcriptional analysis in different filling stages,we revealed the expression model of these candidate genes in kernel filling process.Association analysis for two candidate genes located on key node of metabolic pathway,GRMZM2G391936 and GRMZM2G008263,were performed in 352 maize inbred lines.Two SNP in intron 9 of GRMZM2G391936 were significantly associated with 100 kernel weight,grain weight per ear and ear weight.One indel in 5'-UTR region of GRMZM2G008263 was significantly associated with grain weight per ear and ear weight.Two SNP in intron 7 and intron 10 were significantly associated with starch content and hundred-kernel weight,respectively.5.A total of 73 QTL related to dehydration characteristic parameters were detected in three environments,where 2015 Nantong had 31 QTL,2016 Yangzhou had 25 QTL and 2017 Sanya had 20 QTL.Initial dehydration rate,dehydration duration and final moisture content detected 12 QTL,respectively.Dehydration initial time detected 9 QTL.Dehydration initial moisture content detected 10 QTL.Change of moisture content in dehydration detected 8 QTL.Dehydration termination time detected 7 QTL.Average dehydration rate detected 6 QTL.The phenotypic contribution rate of each QTL was from 1.03%to 15.24%.For dehydration duration,dehydration termination time and average dehydration rate,their synergistic site came from T877.Through multi-environment QTL analysis,21 QTL were detected,of which phenotypic contribution rate was from 0.50%to 38.15%.Dehydration duration detected a QTL with significant major effect.Dehydration termination time detected 4 QTL,including a QTL with significant major effect and environmental interaction effect and 3 QTL with significant environmental interaction effect.Dehydration initial time detected 4 QTL,including 2 QTL with significant major effect and environmental interaction effect and 2 QTL with environmental interaction effect.Initial dehydration rate detected a significant major effect QTL and an environmental interaction QTL.Change of moisture content in dehydration detected a QTL.Final moisture content detected 7 QTL,including a QTL with significant major effect,a QTL with significant major effect and environmental interaction effect and 5 QTL with significant environmental interaction effect.Through multi-trait QTL analysis,58 QTL was detected,where 51 QTL had pleiotropic effects.22 QTL controlled 2 dehydration characteristic parameters simultaneously.16 QTL controlled 3 parameters.Eight QTL controlled four parameters simultaneously.Four QTL controlled 5 parameters simultaneously.A QTL controlled 6 parameters simultaneously.Based these results,12 co-located QTL were identified,where chromosome 2,5 and 8 had one respectively;chromosome3,4 and 6 had 2 co-located QTL respectively and chromosome 7 had 3 co-located QTL.Using 12 co-located QTL,the kernel dehydration was divided into 3 dehydration models.The dehydration curves under different model showed obvious difference,of which dehydration characteristic parameters had significant difference except for dehydration duration.This study analyzed the features of kernel filling and dehydration from many aspects and revealed the genetic mechanism related to kernel filling and dehydration in DH1M and T877 hybredized combination.These findings not only laid a foundation for deep understanding of genetic rule of kernel filling and dehydration but also provided a theory reference for genetic improvement of maize with high efficiency filling and dehydration.These research results using Logistic function-based method was based overall dynamic process and revealed genetic structure of developmental traits and regulatory relationship of related loci at overall level,which provided a new insight to research other complex dynamic traits.