Application of CRISPR/Cas9 for targeted mutagenesis in teosinte Zea mays ssp. parviglumis

To Editor,

Maize (Zea mays ssp. mays L.) is one of the major cereal crops grown for food, fibre and bioenergy production. Modern maize was domesticated ~9000 years ago from teosinte with the closest progenitor being Zea mays ssp. parviglumis. Teosinte contains a greater amount of genetic diversity compared to modern maize (Hufford et al., 2012) and may serve as a model system to understand maize evolution and provide genetic resources to further improve maize (Barnes et al., 2022). However, the genetic differences between maize and teosinte present challenges in incorporating teosinte traits into cultivated maize (Doebley et al., 2006).

Recently, clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR‐associated protein (Cas) systems have been successfully applied to generate new tomato and rice varieties from wild species through de novo domestication (Yu et al., 2021; Zsögön et al., 2018). De novo domestication requires genomic sequence data and a method for transforming the regeneratable explants with CRISPR reagents to create desired allelic changes and sequence modifications. We recently established the first genetic transformation method for teosinte (Zea mays ssp. parviglumis) using a biolistic delivery method and leaf‐derived callus explants (Zobrist et al., 2021). In this work, we report the first successful use of CRISPR/Cas9 in teosinte to generate targeted gene knockouts via short indel mutations using a single guide RNA (sgRNA) and precise sequence deletion using two sgRNAs.

We utilised our T‐DNA binary vector system (Figure 1a) to generate CRISPR/Cas9 constructs (Figure 1b). First, we tested if CRISPR/Cas9 can induce targeted mutagenesis in teosinte using a proven construct pKL2013 (Figure 1b, top) that targets Glossy2 (Gl2) gene (Lee et al., 2019; McCaw et al., 2021). The target sequence, including protospacer adjacent motif (PAM) and protospacer, was conserved among the Zea mays subspecies (Figure 1c; Methods S1). We performed 10 bombardment experiments with calli derived from four different leaf whorl segments (WS) using the pKL2013 DNA (Figure 1d). A total of 89 T0 plants were regenerated from Cas9‐positive callus events that were derived from two different WS. Inference of CRISPR Edits (ICE) analysis (Conant et al., 2022) revealed that 37 T0 plants were biallelic mutants (Figure 1e; Table S1). Because there were only two types of indel mutations (Figure 1e; 1‐bp deletion and 1‐bp insertion, 27 plants; 1‐bp insertion and 2‐bp insertion, 10 plants), it is likely that many of those T0 plants were clonal. Gl2 is responsible for the biosynthesis and deposition of waxy cuticle layer on the juvenile leaf tissues and water adheres to the surface of null mutant plants when misted. Indeed, biallelic mutant T1 plants exhibited gl2 knockout phenotype (Figure 1f, bottom), confirming highly efficient CRISPR/Cas9‐mediated targeted mutagenesis in teosinte.

Figure 1.

CRISPR/Cas9‐mediated targeted mutagenesis in teosinte. (a) Gateway destination vector pKL2393. (b) T‐DNA regions of pKL2013 (Gl2) and pJZ113 (KRN2). (c) Sequence conservation in Gl2 target region among Zea mays subspecies. B73, maize inbred B73; RS, parviglumis from Restoration Seeds (https://www.restorationseeds.com/products/teosinte); parvi, parviglumis; (d) teosinte transformation. (d‐1 & d‐2) Teosinte seeds before and after removal of fruit case. (d‐3) Germinating seedlings. (d‐4) Leaf‐derived embryogenic callus. (d‐5) Transgenic callus expressing red fluorescent protein (RFP, mCherry). (d‐6) Regenerating transgenic shoots. (d‐7) T1 seeds expressing RFP. (e) Summary of targeted mutagenesis of Gl2. (f) Phenotype of gl2 mutant on a young leaf. Scale bar = 1 cm. (g) Targeted mutagenesis of KRN2. Deleted sequences are shaded in grey.

We then tested if another target gene KRN2 (kernel row number2) can be precisely edited using two sgRNAs. Recently, it was reported that KRN2, a WD40 protein‐coding gene, had been selected during maize domestication and plays an important role in maize kernel raw number regulation (Chen et al., 2022). In their study, null mutant lines generated by transposon insertion and CRISPR/Cas9‐mediated mutagenesis produced more kernel row numbers per ear compared to the wild‐type segregants (16 vs. 14), and increased yield by about 10% (Chen et al., 2022). In addition, CRISPR/Cas9‐mediated mutagenesis of KRN2 ortholog in rice also increased yield by about 8%, suggesting that KRN2 would be an agronomically important target gene in maize and other relevant crop species and can contribute to yield increase.

For precise sequence deletion, we designed two sgRNAs to target the second exon of KRN2 gene with 37‐bp distance (Figure 1g). Precise deletion would produce a loss‐of‐function mutant by frameshift. Callus pieces derived from 20 independent WS were bombarded with pJZ113 DNA (Figure 1b, bottom) and a total of 18 T0 plants were regenerated from callus events derived from four independent WS. We analysed 15 T0 events using ICE analysis (Conant et al., 2022) and found four wild‐type (27%), nine homozygous mutants (60%), and two mosaic mutants (13%). Among the nine homozygous mutants (Figure 1g), eight had precise 37‐bp deletion (lines 171, 177, 178, 181, 182, 184, 185, 186) while one event (line 179) had a 38‐bp deletion including additional 1‐bp deletion at the second sgRNA target site. The two mosaic mutant events (lines 172 and 176) had a precise 37‐bp deletion on one allele but had additional mutations on the other allele. These results suggest that precise sequence deletion by two sgRNAs can be an efficient approach to generate gene knockouts or sequence modifications in teosinte.

Both mutant and wild‐type transgenic plants exhibited normal development under the environmental growth chamber conditions and produced viable T1 seeds. Due to limited resources and a lack of suitable growth spaces, we were unable to perform a quantitative seed production analysis demonstrated by Chen et al. (2022), thus further studies are necessary to determine if the krn2 gene knockout by CRISPR/Cas9 alters seed production or causes other phenotypes in teosinte.

The lack of genetically homogenous teosinte populations or inbred lines remains a major hurdle for both genetic transformation and genome engineering in teosinte. In addition to the development of genomic tools, we recognise that it is equally important to have appropriate growth infrastructure for transgenic and CRISPR‐edited teosinte plants. Growing transgenic teosinte to maturity for adequate trait analysis is limited by teosinte's photoperiodic requirements, growth habit and ability to cross‐pollinate with maize. A controlled pollination process including ear‐covering, detasseling and hand‐pollination is effectively used to prevent cross‐contamination and ensure seed purity for maize. However, this practice is not applicable to growing transgenic teosinte plants. Unlike maize, which typically has one or two large ears and one large tassel on top of the plant, a teosinte plant produces multiple inflorescences and tassels distributed across multiple tillers. In practice, transgenic teosinte plants require individual growth chambers to avoid any cross‐contamination and to achieve maximum seed production. We are confident that the robust transformation and genome‐editing protocols described here, when paired with adequate growth facilities (such as inexpensive, compact and portable growth systems) will pave the way for our understanding of the maize domestication process, deciphering teosinte's functional genomics and moving a step forward to utilising the rich genetic resources in teosinte for maize breeding and trait improvements.

Author contributions

KW conceived the project idea, and JDZ and KL designed the project and conducted molecular analysis. KL and JDZ designed and built the plasmid vectors. JDZ performed the teosinte transformation. KL, JDZ and KW wrote the manuscript.

Supporting information

Method S1 Molecular analysis of target genes, Glossy2 and KRN2.

Table S1 List of primers used in this study.

DATA AVAILABILITY STATEMENT

The data that supports the findings of this study are available in the supplementary material of this article.