Modification of the PthA4 effector binding elements in Type I CsLOB1 promoter using Cas9/sgRNA to produce transgenic Duncan grapefruit alleviating XccΔpthA4:dCsLOB1.3 infection

Abstract

Citrus canker caused by Xanthomonas citri subspecies citri (Xcc) is a severe disease for most commercial citrus cultivars and responsible for significant economic losses worldwide. Generating canker-resistant citrus varieties will provide an efficient and sustainable solution to control citrus canker. Here, we report our progress in generating canker-resistant grapefruit by modifying the PthA4 effector binding elements (EBEs) in the CsLOB1 Promoter (EBEPthA4 -CsLOBP) of the CsLOB1 (Citrus sinensis Lateral Organ Boundaries) gene. CsLOB1 is a susceptibility gene for citrus canker and is induced by the pathogenicity factor PthA4, which binds to the EBEPthA4 -CsLOBP to induce CsLOB1 gene expression. There are two alleles, Type I and Type II, of CsLOB1 in Duncan grapefruit. Here, a binary vector was designed to disrupt the PthA4 EBEs in Type I CsLOB1 Promoter (TI CsLOBP) via epicotyl transformation of Duncan grapefruit. Four transgenic Duncan plants with targeted modification of EBEPthA4 -T1 CsLOBP were successfully created. As for Type I CsLOB1 promoter, the mutation rate was 15.63% (#D13), 14.29% (#D17), 54.54% (#D18) and 81.25% (#D22). In the presence of wild-type Xcc, transgenic Duncan grapefruit developed canker symptoms similarly as wild type. An artificially designed dTALE dCsLOB1.3, which specifically recognizes Type I CsLOBP, but not the mutated Type I CsLOBP or Type II CsLOBP, was developed to infect Duncan transformants. Consequently, #D18 had weakened canker symptoms and #D22 had no visible canker symptoms in the presence of XccΔpthA4:dCsLOB1.3. Our data suggest that activation of a single allele of susceptibility gene CsLOB1 by PthA4 is sufficient to induce citrus canker disease, and mutation in the promoters of both alleles of CsLOB1 is probably required to generate citrus canker-resistant plants. This work lays the groundwork to generate canker-resistant citrus varieties via Cas9/sgRNA in the future.

Keywords: Cas9/sgRNA; CsLOB1; PthA; Xanthomonas; genome editing.

Figure 1

CsLOBP of Duncan grapefruit and binary vectors. (a) Two kinds of CsLOBP, Type I and Type II, in Duncan grapefruit, and part of their sequences and chromatograms are shown, in which the difference (1 bp) was indicated by arrows, and the PthA4 EBEs were highlighted by blue. Among 23 colonies sequenced, 13 belong to Type I CsLOBP, and 10 belong to Type II CsLOBP. (b) A sgRNA (sgRNA:CsLOBP1) was designed to target EBE _pthA4‐TI CsLOBP, which was indicated by blue; sgRNA:CsLOBP1‐targeting site was underlined by a red line. (c) Schematic diagram of p1380N‐Cas9/sgRNA:CsLOBP1 CaMV 35S and 35T, the cauliflower mosaic virus 35S promoter and its terminator; NosP and NosT, the nopaline synthase gene promoter and its terminator; LB and RB, the left and right borders of the T‐DNA region; Flag‐Cas9‐NLS, the Cas9 endonuclease containing Flag tag at its N‐terminal and nuclear location signal at its C‐terminal; target, the 20 nucleotides of EBE _pthA4‐TI CsLOBP highlighted by red, was located upstream of protospacer‐adjacent motif (PAM); sgRNA scaffold, a synthetic single‐guide RNA composed of a fusion of CRISPR RNA and trans‐activating CRISPR RNA; NptII, the coding sequence of neomycin phosphotransferase II. EBE, effector binding element.

Figure 2

Cas9/sgRNA:CsLOBP1‐mediated EBE_{P thA4}‐TI CsLOBP modification via Xcc‐facilitated agroinfiltration in Duncan grapefruit leaves. (a) Targeted mutations induced by Cas9/sgRNA:CsLOBP1 to grapefruit EBE _pthA4‐TI CsLOBP. The p1380N‐Cas9/sgRNA:CsLOBP1‐targeted sequence in Type I CsLOBP was shown in red, and the mutations were shown in purple. Among 100 colonies sequenced, there were 55 Type I CsLOBP, 43 Type II CsLOBP and two mutant Type I CsLOBP. The results verified that only Type I CsLOBP could be targeted by Cas9/sgRNA:CsLOBP1. (b) The representative chromatograms of Type I CsLOBP and its mutations. The targeted sequence within Type I CsLOBP was shown by black lines, and the mutant site was pointed out by an arrow. EBE, effector binding element.

Figure 3

Cas9/sgRNA:CsLOBP1‐directed EBE_PthA4‐TI CsLOBP modification in transgenic Duncan grapefruit. (a) Four Cas9‐sgRNA:CsLOBP1‐transformed Duncan grapefruit plants (#D13, #D17, #D18 and #D22) were established, which were PCR‐positive using primers Npt‐5 and 35T‐3. Plasmid p1380N‐Cas9/sgRNA:CsLOBP1 was used as a positive control. M, 1‐kb DNA ladder; WT, wild type. (b) Cas9/sgRNA:CsLOBP1‐induced mutations in grapefruit EBE_PthA4‐TI CsLOBP. The Type I CsLOBP sequence targeted by Cas9/sgRNA:CsLOBP1 was shown in red, and the mutations were shown in purple. Among 124 colonies sequenced, there were 58 Type I CsLOBP, 33 Type II CsLOBP and 33 mutant Type I CsLOBP. (c) The representative chromatograms of Type I CsLOBP and its mutations. The targeted sequence within the Type I CsLOBP was indicted by black lines, and the mutant sites were highlighted by arrows. (d) The representative chromatograms of PCR product direct sequencing. The PCR products were amplified from wild‐type Duncan and four transgenic plants, and primer CsLOB4 was used for direct sequencing. The beginning sites of double peak/multiple peaks were highlighted by arrows. EBE, effector binding element.

Figure 4

Pathogenicity assay for Cas9/sgRNA:CsLOBP1‐transformed Duncan grapefruit and GUS assay for CsLOBP. (a) Five days postinoculation, similar canker symptoms were readily detected on wild‐type and transgenic Duncan grapefruit. (b) Schematic diagram of three binary plasmids (p1380‐TI CsLOBP‐GUSin, p1380‐TII CsLOBP‐GUSin and p1380‐MTI CsLOBP‐GUSin) developed in this work. TI CsLOBP, Type I CsLOBP of Duncan grapefruit; TII CsLOBP, Type II CsLOBP; MTI CsLOBP, mutant Type I CsLOBP; GUSin, the intron‐containing β‐glucuronidase; HptII, the coding sequence of hygromycin phosphotransferase II. (c) Via Xcc‐facilitated agroinfiltration, quantitative GUS assay and GUS histochemical staining were used to evaluate the effects of Xcc‐derived PthA4 on CsLOBPs. The results indicated that PthA4 could activate GUS expression under the control of Type I CsLOBP, Type II CsLOBP or mutant Type I CsLOBP, but not affect GUS expression directed by AtHSP70B. SD values were calculated from three replicates of one experiment. The experiments were repeated twice with similar results. (d) Quantitative RT‐PCR analyses of CsLOB1 expression in Duncan plants. The expression level of CsLOB1 was analysed at 48 h postinoculation of Xcc or water on Duncan grapefruit. The expression was normalized to housekeeping gene CsEF1α. Data bars represent the mean ± SD with three technical replicates of one experiment.

Figure 5

Artificial dTALE dCsLOB1.3 for Type I CsLOBP activation. (a) Repeat variable diresidues (RVDs) of artificial dTALE dCsLOB1.3 and the corresponding effector binding element (EBE) sequence in host genome. EBE _pthA4‐TI CsLOBP was underlined, red font represents putative TATAA box. Artificial dCsLOB1.3, binding to a sequence 7 bp downstream of EBE _pthA4‐TI CsLOBP, was designed to specifically activate Type I CsLOBP, but not Type II CsLOBP or mutant Type I CsLOBP. (b) Via Xcc306ΔpthA4:dCsLOB1.3‐facilitated agroinfiltration, quantitative GUS assay and GUS histochemical staining were performed to study the effects of Xcc‐derived dCsLOB1.3 on CsLOBPs. As expected, GUS expression, only under the control of Type I CsLOBP, could be activated. The experiments were repeated twice with similar results. (c) In the presence of Xcc, citrus canker symptoms were observed on Duncan (containing Type I CsLOBP and Type II CsLOBP), Valencia (containing Type I CsLOBP and Type II CsLOBP) and pummelo (containing Type II CsLOBP) at 4 DPI, because PthA4 derived from Xcc could activate Type I CsLOBP and Type II CsLOBP. In the presence of Xcc306ΔpthA4:dCsLOB1.3, citrus canker symptoms were not observed on pummelo, because dCsLOB1.3 could not activate Type II CsLOBP, which is present in pummelo.

Figure 6

#D18 and #D22 resistant against Xcc306ΔpthA4:dCsLOB1.3. (a) At 3 days postinoculation (DPI) with Xcc306ΔpthA4:dCsLOB1.3, there was no citrus canker symptoms on #D18 and #D22, whose EBE _pthA4‐TI CsLOBP mutation rates were higher, whereas typical canker symptoms were observed on #D13 and #D17, which had lower mutant rates. Wild‐type Duncan grapefruit was used as a control. (b) At 7 DPI with Xcc306ΔpthA4:dCsLOB1.3, weak canker symptom was detected on #D18, which was pointed out by an arrow. There was no citrus canker symptom on #D22, whose EBE _pthA4‐TI CsLOBP mutant rate was highest among four transgenic plants. The results indicated that #D18 and #D22 could resist against Xcc306ΔpthA4:dCsLOB1.3. EBE, effector binding element.