Dissection The Genetic Architecture Of Drought Stress In Maize By Genotyping-by-Sequencing SNPs

Maize(Zea mays)is an important and multi-functional crop,the production of which is critical to the world food security and economic development.With the rapid increasing of worldwide population,the demands of maize production are dramatically increasing.However,the environment is worse and worse,and drought stress(DS)is much more severely and occurs more often,to be a main abiotic factor strongly restricting the production of maize.Thus,it is very meaningful for reduction the production loss due to DS that dissection the genetic structure of drought tolerance and developing drought tolerant germplasm in maize.In this study,>3000maize germplasms collected from tropical and sub-tropical were subjected to revel the genetic basis of drought tolerance,and to explore the genomic selection scheme in order to supply efficient methods for marker assisted selection(MAS)breeding in maize.All the lines were genotyped using genotyping-by-sequencing(GBS),a genome-wide scale and high-throughput genotyping method.With the disadvantage of low-coverage on genome of GBS SNPs,we developed two methods to build genomic Block markers for genetic analysis.The main results are as follows:(1)SL(Similarity/Linkage)and FPP(Fixed Physical Position)methods were developed to build genomic Block markers merged from GBS SNPs,respectively.They were applied to build Block markers for six different bi-parental populations(DH,RIL,IBM,F₂,BC₁ and BC₁F₂).The missing rate(MR)and the genotyping error of heterozygous populations was higher than that in the homologous populations at the same sequencing multiplex level.Based on the Block markers via SL and FPP methods,the heterozygosity of heterozygous populations was reached to the expected values,and the other properties of genotypes were improved significantly;the recombination landscape identified in this study was consistent with the previously reported,average crossovers for each population was also matched to the theoretical values;the length of the genetic maps were in reasonable range(IBM:?5K c M,others:1-2K c M),of which the SNP densities were high(such as 35.43 SNP/c M for F₂).The genetic maps were validated by QTL mapping for TSC(tar spot complex)and Trp(tryptophan)in maize endosperm.These results indicate that the genetic maps were in high quality constructed using the Blocks via SL and FPP methods,and the SL does better for homologous populations and FPP does well in heterozygous populations.(2)Grain yield(GY)and the related secondary traits of 15 bi-parental maize populations(named as WEMA)were evaluated under wel-watered(WW)and drought stress(DS)conditions.The FPP method were used to build linkage maps and 265 QTL were detected in linkage mapping analysis.The joint genome-wide association study(GWAS)was carried out as well,and 2979 candidate genes significantly associated with the target traits under different conditions were overlapped between linkage association mapping totally.In the expression and gene ontology enrichment analysis,354 out of 2886 genes were differentially expressed genes(DEGs)under DS vs.WW conditions in maize seedlings,and the DEGs were enriched to catalytic activity in molecular function and abiotic stimulus and oxidation reduction.In addition,in the GS analysis of the target traits under WW and DS conditions between half-sib populations(HSP)using equivalent,tow-fold and three-fold prediction schemes and three-fold cross-validation scheme,the prediction accuracy of three-fold scheme between HSP was the optimal,only slightly lower than the equivalent scheme under WW condition and the cross-validation scheme under DS condition,respectively.These results are meaningful for developing drought tolerance maize using MAS method.(3)The DTMA association panel,consisted of 300 inbred lines,was collected and evaluated under WW,DS,heat stress(HS)and the combined DS and HS(DHS)conditions for the target traits of GY and the related secondary traits(AD and ASI).In total,1549 SNP were identified significantly associated with the 12 trait-environment combinations in the SNP based GWAS.Only 589 out of them have PVE>5%,indicating that the flowering time related traits are complex under WW and stress conditions,and are controlled by minor effect genes distributed on the whole genome.Haplotype-based GWAS have higher PVE(average value=10.56%),indicating that the higher mapping power of haplotype-based GWAS.In total,690candidate genes were identified in the SNP based and haplotype based GWAS,46 out of which were DEGs in DS condition vs.WW condition in maize seedlings.The 46 genes could be key candidate genes for further analysis.In the GS,the less complex trait AD has higher prediction accuracy compared with the more complex traits of ASI and GY.Besides,the introduction of the trait associated SNPs into the GS model could improve the prediction accuracy for the complex traits,but the quality control is necessary to avoid the background noise,which has practical guiding significance for MAS breeding and stress tolerance improving in maize.This study also highlighted the importance and paved the way of multi-stress tolerant maize germplasm breeding.In the whole study,SL and FPP methods were newly developed and were assessed and validated in different maize populations.Based on the Block markers in the WEMA population using FPP method,many candidate genes were detected,the overlapped genes with DTMA and the DEGs,which were annotated involving ABA and Jasmonic Acid signaling pathway or unknown function,could be the key candidate genes in the further study.This could be the complementation of the regulation network in response to DS.We will expand the genetic application of FPP method in the joint analysis in different populations and the multi-parental populations.In addition,the DS is always accompanied with the co-occurrence of HS,the genetic basis of the genes responding to HS and DHS are meaningful for developing heat and combined stress tolerant maize germplasm.Taking the findings in this study together,the locus associated with GY,as well AD,ASI,in combination with high-quality genome-wide SNPs could efficiently accelerate the efforts on rapid development of the stress-tolerant maize germplasm through marker-assisted selection and/or genomic selection.