Morphogene-assisted transformation of Sorghum bicolor allows more efficient genome editing

Abstract

Sorghum bicolor (L.) Moench, the fifth most important cereal worldwide, is a multi-use crop for feed, food, forage and fuel. To enhance the sorghum and other important crop plants, establishing gene function is essential for their improvement. For sorghum, identifying genes associated with its notable abiotic stress tolerances requires a detailed molecular understanding of the genes associated with those traits. The limits of this knowledge became evident from our earlier in-depth sorghum transcriptome study showing that over 40% of its transcriptome had not been annotated. Here, we describe a full spectrum of tools to engineer, edit, annotate and characterize sorghum's genes. Efforts to develop those tools began with a morphogene-assisted transformation (MAT) method that led to accelerated transformation times, nearly half the time required with classical callus-based, non-MAT approaches. These efforts also led to expanded numbers of amenable genotypes, including several not previously transformed or historically recalcitrant. Another transformation advance, termed altruistic, involved introducing a gene of interest in a separate Agrobacterium strain from the one with morphogenes, leading to plants with the gene of interest but without morphogenes. The MAT approach was also successfully used to edit a target exemplary gene, phytoene desaturase. To identify single-copy transformed plants, we adapted a high-throughput technique and also developed a novel method to determine transgene independent integration. These efforts led to an efficient method to determine gene function, expediting research in numerous genotypes of this widely grown, multi-use crop.

Keywords: Agrobacterium; CRISPR/Cas9 editing; Sorghum; altruistic morphogene-assisted transformation; engineering; morphogene-assisted transformation.

Figure 1

RTx430 tissues from morphogene‐assisted transformation (MAT) and altruistic MAT. RTx430 IEs, 1.5‐2mm, were transformed with Agrobacterium LBA4404 Thy‐ strain, containing appropriate constructs depending on the experiment. IEs were cultured for one week each on co‐cultivation then resting medium. For MAT, using pPHP81814 (ZSG), images by brightfield (a) and by fluorescence (b) were taken while tissues were on EMM. For MAT RTx430 tissues were moved to EMM with 0.05 mg/L imazapyr (IMZ) (c, d, e). For altruistic MAT 20 mg/L hygromycin was used as a selection agent for RTx430 (f, g, h). Images were taken using brightfield (i, k) and fluorescence microscopy (j, l). For altruistic MAT, using pGL190 (ZSG) and pANIC10A (RFP), IEs expressed ZSG and RFP (j). For comparison a brightfield image of the same tissue is shown (i). As a control, pGL190 was used to transform RTx430 IEs and imaged using brightfield (k) and fluorescence (l); ZSG expression in transformed tissue is shown (l).

Figure 2

Transgene independent integration (TII) was determined with adaptor‐ligation‐mediated PCR. Examples of T‐DNA/gDNA junctions in the sorghum genome from the left border (LB, a‐f). Transgenic lines derived from the same immature embryo are labelled with alphabetic letters at the bottom; wild‐type genomic DNA (WT) is the negative control. Right border (RB) assay was performed if left border assay failed to amplify bands (e). DNA ladder (DL) is on the right. Dots indicate bands from T‐DNA/gDNA junctions. Plants from the same immature embryo in a and b are independent; plants in c, d and e are the same event.

Figure 3

Construct maps for transformation. (a) pGL190 for altruistic morphogene‐assisted transformation (MAT). (b) pGL196 for MAT‐mediated editing of phytoene desaturase genes. The SpCas9 is driven by Sb‐Ubi1pro. Black rectangles indicate terminator regions. (c) pGL199 for MAT‐mediated editing of phytoene desaturase genes. The maize codon‐optimized SpCas9 is driven by Zm‐Ubi1pro. Triangles indicate T‐DNA borders; rectangles with arrows indicate promoters; blank rectangles indicate genes or buffer regions; Black rectangles indicate terminator regions. All three constructs were created from pPHP85425, as a Gateway‐compatible destination vector.

Figure 4

Editing of pds genes using CRISPR/Cas9 system. (a) Four guide RNAs (gRNAs) targeting pds. Dark blue rectangles indicate untranslated regions; yellow rectangles indicate exons; light grey lines indicate intron regions. gRNA1.1 is located in the third exon of pds. gRNA1.2 is located in the fifth exon of pds. These two gRNA are in pGL196. gRNA2.1 is located in the second exon of pds. gRNA2.2 is located in the fourth exon of pds. These two gRNAs are in pGL199. (b) Sanger sequencing of edited pds genes are shown with the main editing versions predicted by Synthego. gRNAs are underlined and PAM sites for gRNAs are in bold black font. Deletions are shown with dashes. (c) Summary of transformation efficiency and CRISPR/Cas9‐induced genome‐editing efficiency in T₀ generation from MAT‐mediated editing in RTx430 using pGL196 and pGL199. (d) Albino phenotype from pds knockout, as frameshift mutations, in CRISPR/Cas9 edited plants. Phenotypic characterization of chimeric (i, ii) and fully albino (iii) mutants in T₀ generation. Scale bar, 1 cm.