Agrobacterium-produced and exogenous cytokinin-modulated Agrobacterium-mediated plant transformation

Abstract

Agrobacterium tumefaciens is a plant pathogenic bacterium that causes neoplastic growths, called 'crown gall', via the transfer and integration of transferred DNA (T-DNA) from the bacterium into the plant genome. We characterized an acetosyringone (AS)-induced tumour-inducing (Ti) plasmid gene, tzs (trans-zeatin synthesizing), that is responsible for the synthesis of the plant hormone cytokinin in nopaline-type A. tumefaciens strains. The loss of Tzs protein expression and trans-zeatin secretions by the tzs frameshift (tzs-fs) mutant is associated with reduced tumorigenesis efficiency on white radish stems and reduced transformation efficiencies on Arabidopsis roots. Complementation of the tzs-fs mutant with a wild-type tzs gene restored wild-type levels of trans-zeatin secretions and transformation efficiencies. Exogenous application of cytokinin during infection increased the transient transformation efficiency of Arabidopsis roots infected by strains lacking Tzs, which suggests that the lower transformation efficiency resulted from the lack of Agrobacterium-produced cytokinin. Interestingly, although the tzs-fs mutant displayed reduced tumorigenesis efficiency on several tested plants, the loss of Tzs enhanced tumorigenesis efficiencies on green pepper and cowpea. These data strongly suggest that Tzs, by synthesizing trans-zeatin at early stage(s) of the infection process, modulates plant transformation efficiency by A. tumefaciens.

Figure 1

Tzs is responsible for trans‐zeatin secretion and promotes Agrobacterium tumefaciens virulence on white radish stems. (A) Immunoblot analyses of Tzs and neomycin phosphotransferase II (NptII) in the wild‐type (WT) (NT1RE[pJK270]), tzs‐fs mutant (tzs‐fs) and tzs‐complemented (tzs‐fs+tzs) strains grown in AB‐MES in the presence of 200 µm acetosyringone (AS) at 19 °C for 40 h. Equal amounts of protein were resolved by Tricine‐sodium dodecylsulphate‐polyacrylamide gel electrophoresis (SDS‐PAGE) and analysed by immunoblotting with antibodies against Tzs and NptII, which served as a loading control. The molecular weight markers are indicated on the left in kilodaltons. (B) Concentration of trans‐zeatin in the culture medium of the wild‐type, tzs‐fs mutant and tzs‐complemented strains by ultra‐performance liquid chromatography (UPLC). Culture medium from each bacterial strain grown in AB‐MES containing AS was subjected to UPLC analysis, and trans‐zeatin was eluted by methanol at a specific retention time (min, x‐axis) and detected by absorbance at 265 nm wavelength (y‐axis). The mean concentration of trans‐zeatin (µg/L) detected in each sample from three independent experiments is shown with standard deviations. N.D., not detected. The lowest concentration of trans‐zeatin detected with our UPLC system was 100 µg/L, whereas no signal could be detected when 10 µg/L of trans‐zeatin was analysed (data not shown). (C) Virulence of the wild‐type, tzs‐fs mutant and tzs‐complemented strains on white radish stems. The tumorigenesis efficiency of white radish stems infected with the wild‐type strain NT1RE(pJK270), with approximately 80%–100% of plants forming tumours in each experiment, was set at 100%, and that of the mutant is shown relative to that of the wild‐type strain. Mean values for relative tumorigenesis frequencies from at least three independent experiments are shown with standard errors. An asterisk above each mean represents a statistically significant difference between wild‐type (NTIRE[pJK270]) and tzs‐fs mutant strains based on analysis of variance (anova) and pairwise Student's t‐test. P < 0.05 for all significant differences between strain combinations.

Figure 2

The tzs‐fs mutant shows lower tumorigenesis and transient transformation efficiency in Arabidopsis roots. Root segments from 3–4‐week‐old Arabidopsis plants were infected with the wild‐type NT1RE(pJK270), tzs‐fs mutant and tzs‐complemented Agrobacterium tumefaciens strains at two concentrations [10⁸ and 10⁷ colony‐forming units (cfu)/mL] for both tumorigenesis (A) and transient transformation (B) assays. The transformation efficiency of root segments infected with the wild‐type strain NT1RE(pJK270), with approximately 40%–70% of roots forming tumours or β‐glucuronidase (GUS) activity in each experiment, was set at 100% and that of the mutant is shown relative to that of the wild‐type strain. Mean values of virulence from at least three independent experiments are shown with standard errors. (C) Representative plates of Arabidopsis roots with tumours infected at a bacterial concentration of 10⁸ cfu/mL.

Figure 3

Cytokinin supplementation during infection increases the transient transformation efficiencies of strains lacking Tzs. Arabidopsis root segments were infected with a bacterial concentration of 10⁷ colony‐forming units (cfu)/mL of various Agrobacterium tumefaciens strains, the wild‐type (NT1RE[pJK270]), the tzs‐fs mutant, A208, LBA4404 and EHA101, on Murashige and Skoog (MS) medium with or without different kinetin concentrations for 2 days. Seven days after infection, root segments were stained with 5‐bromo‐4‐chloro‐3‐indolyl‐β‐d‐glucuronic acid (X‐gluc) staining solution. Mean values of the transformation frequencies from at least three independent experiments are shown with standard errors. An asterisk above each mean represents a statistically significant difference between the transformation frequencies of the wild‐type and various tested strains in the absence of exogenous cytokinins during infection based on analysis of variance (anova) and pairwise Student's t‐test. P < 0.05 for all significant differences between strains. The letters ‘a’, ‘b’ and ‘c’ above each mean represent statistically significant differences with and without exogenous cytokinins using the same strain during infection based on anova and pairwise Student's t‐test. GUS, β‐glucuronidase.

Figure 4

Addition of cytokinin enhances transient transformation efficiencies on Arabidopsis roots infected with NT1RE(pJK270) at lower bacterial concentrations. Arabidopsis root segments were infected with different concentrations [10⁵, 10⁶, 10⁷ and 10⁸ colony‐forming units (cfu)/mL] of the wild‐type Agrobacterium tumefaciens strain NT1RE(pJK270) on Murashige and Skoog (MS) medium with or without kinetin for 2 days. Root segments were stained with 5‐bromo‐4‐chloro‐3‐indolyl‐β‐d‐glucuronic acid (X‐gluc) staining solution 7 days after infection. The mean values of the transformation frequencies from at least three independent experiments are shown with standard errors. Letters ‘a’ or ‘b’ above each mean represent statistically significant differences between the presence and absence of exogenous cytokinins with the same wild‐type bacterial concentration during infection based on analysis of variance (anova) and pairwise Student's t‐test. GUS, β‐glucuronidase.

Figure 5

The tzs‐fs mutant decreased virulence on potato tubers, pai‐tsai, loose leaf lettuce and carnation and caused more plants to produce tumours on cowpea and green pepper. Potato tuber discs (A), stems of pai‐tsai, loose leaf lettuce and carnation (B), stems of cowpea and green pepper (C) were infected with the wild‐type NT1RE(pJK270) and tzs‐fs mutant as indicated. Tumours were scored 5 weeks after infection. The tumorigenesis efficiency of potato tuber discs (A) infected with the wild type was approximately 20–30 tumours per disc, and those of pai‐tsai, loose leaf lettuce and carnation (B) were approximately 40%–90% of stems forming tumours in each experiment. The tumorigenesis efficiency of the wild‐type shown in (A) and (B) was set at 100% and that of the mutant is shown relative to that of the wild‐type strain. The transformation efficiency of stems of cowpea and green pepper (C) infected with the wild‐type Agrobacterium tumefaciens strain NT1RE(pJK270), with approximately 10%–30% of stems forming tumours in each experiment, was set at 10% and that of the mutant is shown relative to that of the wild‐type strain. Mean values of the tumorigenesis frequencies from at least three independent experiments are shown with standard errors. An asterisk above each mean represents a statistically significant difference between the wild‐type and various tested strains based on analysis of variance (anova) and pairwise Student's t‐test. P < 0.05 for all significant differences between strains.