High-Throughput CRISPR/Cas9 Mutagenesis Streamlines Trait Gene Identification in Maize.

Haijun Liu ,Yunxiu Ding,Maolin Zhang,Liumei Jian,Wenhao Song,Jiali Yan,Jieting Xu,Xiaofeng Yang,Baozhu Han,Fan Gao,David Jackson,Qinghua Zhang,Yuming Lu,Minliang Jin,Ya Liu,Jie Liu,Yangwen Qian,Jiuran Zhao,Alisdair R Fernie,Hanmo Chen,Jiamin Sun,Minji Bai,Wenqiang Li,Lei Huang,Mingliang Zhang,Xiangguo Liu,Yong Peng,Wenjie Wei,Zhenxian Li,Xiang Dong Sun

doi:10.1105/tpc.19.00934

Abstract

Maize (Zea mays) is one of the most important crops in the world. However, few agronomically important maize genes have been cloned and used for trait improvement, due to its complex genome and genetic architecture. Here, we integrated multiplexed CRISPR/Cas9-based high-throughput targeted mutagenesis with genetic mapping and genomic approaches to successfully target 743 candidate genes corresponding to traits relevant for agronomy and nutrition. After low-cost barcode-based deep sequencing, 412 edited sequences covering 118 genes were precisely identified from individuals showing clear phenotypic changes. The profiles of the associated gene-editing events were similar to those identified in human cell lines and consequently are predictable using an existing algorithm originally designed for human studies. We observed unexpected but frequent homology-directed repair through endogenous templates that was likely caused by spatial contact between distinct chromosomes. Based on the characterization and interpretation of gene function from several examples, we demonstrate that the integration of forward and reverse genetics via a targeted mutagenesis library promises rapid validation of important agronomic genes for crops with complex genomes. Beyond specific findings, this study also guides further optimization of high-throughput CRISPR experiments in plants.

Highlights

Global crop production will need to double by 2050 in order to feed the increasing world population
(A) Candidates selected from quantitative trait loci (QTL) fine mapping, genome-wide association mapping studies (GWAS), and comparative genomics. (B) Line-specific sgRNA filtering based on assembled pseudo-genome of the receptor line KN5585. (C) Different vector construction approaches of double sgRNA pool (DSP) and single sgRNA pool (SSP). (D) Measuring the coverage and uniformity during plasmid pool by deep-sequencing. (E) to (G) Transformation and assignment of targets to each T0 individual by barcode-based sequencing. (H) to (J) Identification of mutant sequences by Sanger sequencing. (K) and (L) Identification of mutant sequences by Capture-based deep-sequencing. (M) Measuring phenotype changes and identification of functional genes
This high-targeting frequency is consistent with a previous study (51 to 91%; Li et al, 2017) and may be a consequence of using a maize endogenous RNA polymerase III promoter to drive the expression of the sgRNA (Qi et al, 2018)

Summary

Introduction

Global crop production will need to double by 2050 in order to feed the increasing world population. Transposon tagging and mutagenesis by the Activator (Ac) and Dissociation (Ds) transposable elements (Cowperthwaite et al, 2002; Vollbrecht et al, 2010) and UniformMu (May et al, 2003; McCarty et al, 2005; Settles et al, 2007) or chemical mutagens such as ethyl-methanesulfonate (Lu et al, 2018) have all been used in maize, the exact identification of causal gene(s) among the tens or even hundreds of loci within a line that might have been mutated but are not responsible for the phenotype under question is still costly due to the complexity of the maize genome. The laborious and low-throughput nature of classical forward genetics approaches that rely on the segregation of the causal mutation(s) in a mapping population hinders the successful and rapid application of these resources in many plant species

Methods

Results

Conclusion