Leaf transformation for efficient random integration and targeted genome modification in maize and sorghum

Abstract

Transformation in grass species has traditionally relied on immature embryos and has therefore been limited to a few major Poaceae crops. Other transformation explants, including leaf tissue, have been explored but with low success rates, which is one of the major factors hindering the broad application of genome editing for crop improvement. Recently, leaf transformation using morphogenic genes Wuschel2 (Wus2) and Babyboom (Bbm) has been successfully used for Cas9-mediated mutagenesis, but complex genome editing applications, requiring large numbers of regenerated plants to be screened, remain elusive. Here we demonstrate that enhanced Wus2/Bbm expression substantially improves leaf transformation in maize and sorghum, allowing the recovery of plants with Cas9-mediated gene dropouts and targeted gene insertion. Moreover, using a maize-optimized Wus2/Bbm construct, embryogenic callus and regenerated plantlets were successfully produced in eight species spanning four grass subfamilies, suggesting that this may lead to a universal family-wide method for transformation and genome editing across the Poaceae.

Fig. 1. Schematic depiction of T-DNA constructs used in Agrobacterium-mediated maize leaf transformation experiments.

a, T-DNA vectors used in random integration experiments: PHP86491 is a control vector containing only visible (ZsGreen1) and selectable (Hra) marker genes; PHP96037 and PHP97334 are vectors containing morphogenic genes (Wus2 and Bbm), Cre recombinase, a visible marker gene (ZsGreen1) and either Hra or NptII as a selectable marker gene, respectively. b, T-DNA vector used for Cas9-mediated Wx1 dropout containing morphogenic genes (Wus2 and Bbm), Cre recombinase, Cas9 and two gRNA expression cassettes, visible (ZsGreen1) and selectable (NptII) marker genes. c, Depiction of Cas9-mediated targeted gene insertion experiment, including (i) the T-DNA vector containing the morphogenic genes (Wus2 and Bbm), Cre recombinase, Cas9 and gRNA expression cassettes, a selectable marker gene (Hra), the donor cassette comprising the second selectable marker gene (NptII), homology arms (HR1 and HR2) flanked by Cas9 cut sites (or target sites (TS)) and a visible marker gene (ZsGreen1); (ii) targeted gene insertion via HDR upon Cas9 cleavage of the genomic target site and both target sites within the T-DNA to release the donor sequence and (iii) the targeted gene insertion locus with the positions of PCR primer pairs used for detecting the HR1 (1) and HR2 (2) junctions, and the long-PCR product across the entire insertion (3). RB and LB = the right and left border sequences of the Agrobacterium T-DNA.

Fig. 2. Scoring method for evaluating somatic embryo formation in transformed leaf tissue.

The images show examples of scoring the growth of fluorescent cells and somatic embryos 14 days after Agrobacterium infection. The images are representative of three independent experiments. Scale bar = 1 mm. NA = not applicable since no resulting tissue grew.

Fig. 3. The leaf transformation method used for maize.

a, Surface sterilization of maize seeds. b, Maize seeds on germination medium. c, Fourteen-day-old maize seedlings used as the explant source. d, Preparation of 3 cm segments of the leaf whorl directly above the mesocotyl followed by slicing the tissue longitudinally (not shown). e, The tissue was placed in 100 ml of Agrobacterium suspension in the mini-Cuisinart blender and subjected to ten brief pulses to further slice up the leaf tissue. After a 20-minute incubation in the Agrobacterium suspension, the liquid was poured through a metal sieve to collect the leaf fragments, which were briefly blotted onto dry filter paper to wick away excess Agrobacterium (not shown). f, Leaf fragments distributed onto filter paper on the solid RM. g, Fluorescent somatic embryo growth observed 7–10 days after Agrobacterium infection (the image is representative of over 100 independent experiments; scale bar, 1 mm). h, After culturing on the RM, leaf tissue was transferred to selection medium. i, Following three weeks on filter papers on selection medium, the tissue was subjected to 45 °C/70% RH for two hours (not shown), transferred to maturation medium and cultured for an additional three weeks. j, Formation of T₀ regenerants after two weeks on rooting medium.

Fig. 4. Selected grass species transformed using the maize leaf transformation protocol.

a, The Poaceae family tree indicating the species used in the experiment and the corresponding subfamilies. Species that were successfully transformed and regenerated, producing T₀ plants, are shown in black; species that were transformed and produced callus tissue but did not regenerate are shown in red. b, Results for all successfully transformed grass species. The images on the left show transient expression of ZsGreen1 three to four days after Agrobacterium infection (scale bars, 500 µm), demonstrating the relative T-DNA delivery in various crops; the centre images show examples of green-fluorescent embryogenic callus formation three to four weeks after infection (scale bars, 1 mm); and the images on the right show recovered plantlets with shoots and roots ready for transfer to soil five to eight weeks after infection. The images are representative of three independent experiments.

Extended Data Fig. 1. Early growth of somatic embryos.

a, Fluorescence image of transient ZsGreen1 expression four days after Agrobacterium infection with PHP97334. b, Growing, independent fluorescent foci as somatic embryos begin growing nine days post-infection. c, Higher magnification at nine days post-infection showing fluorescent somatic embryos emerging from the surface of the leaf. d-f, Micrographs of transverse sections of leaf tissue stained with PAS and counter-stained with aniline blue black, showing d, small clusters of dividing cells depicting early developing somatic embryos (arrows) in leaf segments five days after infection, e, cluster of dividing cells ready to emerge through the epidermal layer in leaf segments ten days after infection, and f, larger somatic embryos pushing through the surface of the leaf (arrows) in leaf segments 15 days after infection. See end of Supplementary section for histology methods. Images are representative of over ten independent experiments.

Extended Data Fig. 2. Manual explant preparation and infection.

a, Agrobacterium suspension aliquoted into permeable culture inserts sitting in a six-well culture plate. b, Maize seedling depicting the region harvested for manual explant preparation. c, Manually cut leaf tissue explants in Agrobacterium suspension. d,e, Leaf tissues in a permeable culture insert with excess Agrobacterium suspension being blotted off, and leaf tissues cultured on a filter paper sitting on surface of solid medium.

Extended Data Fig. 3. Transformation of Saccharum officinarum and Triticum aestivum leaf explants.

Transient expression of ZsGreen1 three to four days after Agrobacterium infection (panels on left), demonstrating the relative T-DNA delivery. Green-fluorescent embryogenic callus formation three to four weeks after infection (panels on right). Images are representative of three independent experiments.

Extended Data Fig. 4. Sections of maize leaf base tissue from 14 day-old seedlings grown in the absence (a) or presence (b) of 2 mg/l ancymidol.

a,b, Sections of maize leaf base tissue from 14 day-old seedlings grown in the absence (a) or presence (b) of 2 mg/l ancymidol. Sections are stained with periodic acid-Schiff for polysaccharides and aniline blue-black for protein. Arrow indicates magenta-stained amyloplasts. Note extensive presence of amyloplasts in developing mesophyll of ancymidol-treated leaf tissue and general lack of amyloplasts in tissue grown in the absence of ancymidol. Scale bar = 50 µm.