Mapping Of Kernel Row Numbers Based On Bulked Segregant RNA Sequencing And Meta-analysis In Maize

Maize (Zea mays L.) is the first cereal crop on total yield and sowing area in China. Maize kernel row number (KRN) is one of the most research valuable traits among corn yield components due to the trait stability and high heritability, and easy to have the accurate phenotyping data. Previous studies on genetic dissection on KRN had been carried out based on to the different mapping population along with different types and size of the population. A number of quantitative trait loci (QTL) had been identified and reported. In this study, an F2 segregation population between B73 and FRM (Four-Rowed-Maize) had been constructed to identify the important QTL for KRN in maize through a Bulked Segregant RNA-Seq (BSR-Seq) strategy. In addition, the meta-analysis had been done to integrate the analysis and constitutive QTLs for KRN based on different population across diverse enviornments. The identified QTLs here will found a good basis of data, plant materials and information for subsequent fine mapping and map based clone of the QTL. Meanwhile, the data and results on molecular markers of KRN also provide a good basis for marker assisted selection and breeding for KRN in maize. The key findings of this research were listed as follows:1. The F2 segregation population between parents B73 and FRM containing 948 individuals had been built. The phenotyping data results showed that the distribution of KRN trait was continuous and normal distribution. Sixty-two extreme mutant individuals (5 individuals with 4 rows,57 individuals with 6 rows) and 61 extreme wild-type individuals (57 individuals with 16 rows and 4 individuals with 18 rows) leaves were bulked for RNA extraction. The RNA integrity number (RENT) of each mixing pool RNA samples was 8.0 and 7.6, both of which was higher than sequencing samples required RIN score ob 7.5. RNA samples were 1482ng/ul and 760ng/ul, meeting the quality of the concentration. The OD260/280 was 2.043 and 1.996, meeting the criteria of the range of 1.8～2.2. And,28S RNA:18S RNA was 1.2 and 1.0. The RNA samples were all fitted for the requirements of the RNA-Seq.2. Using Pair-end sequencing technology on Illumina Hiseq2000 to conduct high-throughput sequencing. This study yielded 133942396 101 bp raw reads for wild type pool and 123628466 for mutant pool. In total,26.79 and 24.73 Gb raw data were generated. Among the reads,98% of them had been pass through quality control and clean. 89.1% for the wild type and 88.6% for the mutant reads were aligned to the public reference genome of maize (B73 RefGen.v2). Among them,82.6% and 81.8% trimmed reads were uniquely aligned reads. The results showed that more than 70% reads could be aligned to the reference genome uniquely. And, the mapped read were good for scoring and identifying the Single Nucleotide Polymorphisms (SNPs) and Indels (Insertion/Deletion/polymorphisms) between the pooled samples of wild type and mutant based on the mapped reads on the maize reference genome.3. A total of 202651 polymorphic SNPs were retained between 608608 SNPs discovered from uniquely aligned reads. We performed a modified BSA analysis to identify the regions of genome in which KRN QTLs are located based on the SNPs with high confidence and likely identified five potential KRN QTL regions. Two significant peaks, BSR-QTLl(10-25Mb) and BSR-QTL2(60-150Mb), were observed on chr8. Three other putative QTLs were found on on the chr2 short arm, the chr2 long arms, and one on the long arm of chr4.4. Meta analysis of QTL mapping of KRN had been conducted by using BioMercator2.1 software based on the IBM 2 2008 Neighbors from MaizeGDB as the reference genetic map. The 26 reported KRN QTL studies based on the different bi-parental populations at diverse environments had been projected onto integrated genetic map. One hundred and seventy-six out of 196 reported KRN QTLs showed the common position interval and common markers at a 95% confidence threshold level. Twenty-five "consensus" QTLs with the effect value (R2) higher than 4.5 were found. The consensus QTLs along with their linkage markers might found a good basis for fine mapping, gene cloning and molecular breeding for KRN in maize.5. We combined previous results from different mapping methods and different background by meta analysis with results of KRN QTL by BSR-Seq technology to locate and analysis KRN QTL so that we can screen more reliable and larger effect QTL used for subsequent fine mapping. MQTL4-4 near 295.91cM on chromosome 4 and the KRN QTL near 302cM on chromosome 4 mapped by Peter Bommert using RIL population built with B73 and Mo17 was likely the same QTL. MQTL2-1 from 6 initial QTL and MQTL4-1 from 5 initial QTL had smaller confidence interval and higher genetic contribution. All the knowledge and data of them could be used into the subsequent fine mapping and gene discovery.