Cloning And Function Validation Of A Major QTL,QPH3.1, For Plant Height In Maize

Maize is an important crop for food, feed, raw materials of industry and biofuels. Grain yield, biomass and quality are all the target traits for maize breeding. The plant height (PH) is an important yield-associated trait that highly correlated with the grain yield and biomass of maize. Firstly, PH is related to the plant lodging of maize. Previous work suggested that increase the planting density is an effective way to enhance the grain yield, but high-density planting also involves the risk of lodging, which can reduce grain yield. To avoid lodging, one breeding strategy is to moderately decrease plant height. Secondly, the modification of plant architecture has the potential to enhance biomass production, and plant height is an important component of architecture that is highly correlated with biomass yield. Therefore, plant height is a very agronomic importantly trait in maize genetics and breeding and the cloning of the genes for PH can guide the genetic improvement of crops. In a previous study from our lab, a PH QTL, qPH3.1which located in chromosome3was detected using chromosome segment substitution line (CSSL) population. In present study, qPH3.1was finely mapped using map-based cloning strategy and candidate gene for qPH3.1was successfully cloned and the evidences from mutant, expression, association mapping, cytology, and physiological analysis supported the candidate gene. The main research results are as follows:1. SL15is one line from the Z3HBILs CSSL population, which was developed by crossing a recipient line (Zong3) with a donor line (HB522) through four cycles of advanced backcrosses. The introgressed segment in SL15is flanked by ND1and ND75on chromosome3. A total of245genome-wide SSR markers showing polymorphism between Zong3and HB522were used to screen the SL15. Except target segment, only one donor segment was detected, near dupssr23on bin3.06. The genetics difference between SL15and Zong3based on the SSR analysis was about2%.2. The mean PH of SL15was significantly taller than that of Zong3by increase the cell length of internode. The comparison of PH, internode number and internode length between SL15and Zong3showed that both SL15and Zong3have the same number of elongated internodes, indicating that the observed difference of PH between the two lines was caused by difference in the length of the internodes. Measurement of length of each internode and tassel showed that all of the internodes of SL15were significantly longer than those of the Zong3, while the length of the tassel of SL15had no significant difference from that of Zong3. Cytological observation revealed that internodes cells in SL15were significantly longer than that in Zong3(p<0.01), thus indicating that longer internodes in SL15relative to Zong3were due to longitudinally increased cell lengths but not to an increase number of cell in SL15internodes.3. The introgressed segment of SL15harbored the gene which increase the PH. a NIL-F2population containing161individuals was developed by crossing SL15with Zong3. QTL for PH and yield-associated traits, including100-kernel weight, number of rows of kernels, kernels per row, ear length, and ear diameter, were detected. QTL mapping revealed a QTL for PH (qPH3.1) in the target region, accounting for32.3%of the phenotypic variance, and the additive and dominant effects of the SL15allele were9.6cm and4.5cm, respectively. No QTL for yield-associated traits were detected indicated that qPH3.1controlled only the PH phenotype and not any other yield-associated trait. Linkage analysis indicated that the introgression segment near dupssr23on bin3.06was not associated with PH and yield-associated traits.4. Fine mapping of qPH3.1and cloning of candidate gene. In2008, a NIL-F2populations containing617individuals was developed, and the effect of qPH3.1on yield-associated traits (number of rows of kernels, kernels per row, ear length, and ear diameter) and flowering traits (days to tassel, and days to silk) were evaluated and results indicated that qPH3.1has no effect on the yield and flowering. Those individuals containing recombination breakpoints in the QTL interval were selected to self-pollinate and three sub-NILs were developed. In2009, another five sub-NILs were developed from a larger NIL-F2population containing2153individuals. Substitution mapping with eight sub-NILs represent the probability of single crossover among markers in the QTL interval narrowed down qPH3.1within a12.6kb interval between marker ND87—ND88. One ortholog of OsGA3ox2, designated as ZmGA3ox2, encoding a GA3β-hydroxylase was successfully cloned. RLM-RACE (RNA Ligase Mediated Rapid Amplification of cDNA Ends) were performed to amplify full-length cDNA of ZmGA3ox2, and the results showed that the length of5’-UTR and3’-UTR of transcript were238bp and505bp, respectively. The coding region and promoter region were amplified and ZmGA3ox2was found to contain three exons and two introns. The length of ORF was1149bp and the deduced protein is expected to contain382amino acids.27polymorphisms were found between SL15and Zong3in the coding region and promoter region. 5. Expression difference was detected between SL15and Zong3. BLAST analysis using sequence of ZmGA3ox2revealed that the maize genome has only two GA3ox genes, ZmGA3ox1and ZmGA3ox2. RT-PCR analysis with total RNA extracted from root, stem, leaf, tassel, and ear from B73revealed that ZmGA3ox2was expressed in all the tissues and organs tested, whereas ZmGA3ox1expression was observed only in tassel. The results indicated that ZmGA3ox2is the only GA3ox gene that functions for stem elongation during the vegetative stage. Our qRT-PCR results showed that significantly higher levels of ZmGA3ox2were expressed in the stem apices in four stages in SL15than in Zong3.6. Genetics analysis of maize dwarf mutant dwarf-1(dl) indicated that the variation of ZmGA3ox2could affect PH. Sequence analysis revealed a2304bp deletion of ZmGA3ox2in d1-6016allele, including734bp upstream sequence of the start codon and1570bp of the coding sequence. It loses promoter core region and almost whole coding region with only64bp of last exon remaining. The ZmGA3oxl was expressed in tassel of d1-6016indicated that ZmGA3ox1can not compensate the deficiency of ZmGA3ox2and the ZmGA3ox2play a decisive role for stem elongation during the vegetative stage. In particular, allelic variation determined by gene-specific marker (ND88) in ZmGA3ox2locus co-segregated with the plant height phenotype in a backcross population containing406individuals. These evidences from cosegregation and sequence variation in the dl-6016allele experimentally support that ZmGA3ox2is the candidate gene for d1-6016.7. PH difference between SL15and Zong3could be removed by exogenous GA3treatment. The results described above indicated that ZmGA3ox2is the candidate gene for qPH3.1and the phenotypic difference between Zong3and SL15was ascribed to the difference of accumulation of GA. Exogenous GA treatment for SL15and Zong3showed that the PH of treated SL15and Zong3were significantly taller than that of untreated SL15and Zong3, respectively. Treated SL15becomed very similar to treated Zong3for PH. The evidence from physiological analysis indicated that the difference between SL15and Zong3was related to GA level and the difference was caused by the variation of GA-biosynthesis gene. Significantly greater endogenous GA1levels in SL15relative to Zong3was measured in the stem apices in four stages after jointing.8. Candidate gene association mapping showed that natural variation in promoter region affect the PH of maize. Association mapping was carried out to test the association between the polymorphisms in the candidate gene and PH using a population with244inbreds. The results revealed that two sites (S-575and S-566) in the promoter were associated with PH variation across two field experiments at the P<0.01level. The SL15allele of the two association loci results in an increased PH. The loci of S-575and S-566both have three alleles. These two sites were in strong LD (r2=0.53), and formed four main haplotypes in the association panel. The PH of haplotype4(SL15type) was significantly taller than that of haplotype3(Zong3type).