Agromonas: a rapid disease assay for Pseudomonas syringae growth in agroinfiltrated leaves

Abstract

The lengthy process to generate transformed plants is a limitation in current research on the interactions of the model plant pathogen Pseudomonas syringae with plant hosts. Here we present an easy method called agromonas, where we quantify P. syringae growth in agroinfiltrated leaves of Nicotiana benthamiana using a cocktail of antibiotics to select P. syringae on plates. As a proof of concept, we demonstrate that transient expression of PAMP receptors reduces bacterial growth, and that transient depletion of a host immune gene and transient expression of a type-III effector increase P. syringae growth in agromonas assays. We show that we can rapidly achieve structure-function analysis of immune components and test the function of immune hydrolases. The agromonas method is easy, fast and robust for routine disease assays with various Pseudomonas strains without transforming plants or bacteria. The agromonas assay offers a reliable approach for further comprehensive analysis of plant immunity.

Keywords: Agrobacterium; Nicotiana benthamiana; Pseudomonas syringae; disease assay; plant immunity; technical advance.

Figure 1

Selection against Agrobacterium tumefaciens. (a) Chemical structures of cephaloridine, fucidin and cetrimide (CFC). (b) Pseudomonas syringae grows on CFC selection, Agrobacterium does not. Nicotiana benthamiana leaves were infiltrated with 1 × 10⁶ CFU ml⁻¹ P. syringae or 1 × 10⁸ CFU ml⁻¹ AtumGV3101, and bacterial populations were determined 3 days later using colony count method using Luria−Bertani (LB) plates containing CFC or not. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics (***P < 0.001). CFU, colony‐forming units.

Figure 2

Concept of agromonas assay. (a) Experimental procedure for agromonas assay. Two days after agroinfiltration, agroinfiltrated leaves are infiltrated with Pseudomonas syringae bacteria. Bacterial growth is measured 3 days later (3 days post‐infiltration, 3 dpi) by a classic colony count on Luria−Bertani (LB) agar plates containing cephaloridine, fucidin and cetrimide (CFC). (b) CFC selects P. syringae from agroinfiltrated leaves. Agroinfiltrated leaves were infiltrated with 1 × 10⁶ CFU ml⁻¹ PtoDC3000(∆hQ) and 3 days later leaf extracts were diluted, and each dilution was plated onto medium supplemented with or without CFC. Pictures were taken 48 h later. (c) Selective isolation of Pseudomonas spp. from agroinfiltrated leaves. Agroinfiltrated leaves were infiltrated with 1 × 10⁶ CFU ml⁻¹ P. syringae and, at 0 and 3 dpi, leaf extracts were plated on medium containing CFC or gentamicin to select P. syringae or Agrobacterium, respectively. Error bars represent SE of n = 3 biological replicates.

Figure 3

PAMP receptors reduce bacterial growth in the agromonas assay. (a) Transient expression of tomato FLS3 and Arabidopsis EFR in Nicotiana benthamiana confers flgII‐28 and elf18 responsiveness, respectively. Leaf discs from agroinfiltrated leaves expressing FLS3 (blue), EFR (red) or empty vector (EV; grey) were treated with 100 nm flgII‐28 or elf18, and reactive oxygen species (ROS) was measured in relative light units (RLU). Error intervals (shaded regions) represent SE of n = 12 biological replicates. (b) Transient expression of FLS3 or EFR reduces Pseudomonas syringae growth. Two days after agroinfiltration, agroinfiltrated leaves expressing FLS3 (blue), EFR (red) or EV (grey) were spray‐inoculated with the indicated strains of P. syringae (at 1 × 10⁸ CFU ml⁻¹), and bacterial growth was measured 3 days later using cephaloridine, fucidin and cetrimide (CFC) selection. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics (*P < 0.05). (c) Transient expression of EFR, but not FLS3, affect Agrobacterium growth. Bacterial growth of AtumGV3101 was measured by plating the leaf extracts described in (b) on medium containing gentamicin. Error bars represent SE of n = 3 replicates. Student’s t‐test statistics (^* P < 0.05).

Figure 4

Depletion of host immune gene increases Pseudomonas syringae growth in the agromonas assay. (a) Experimental procedure for studying the role of endogenous immune components in agromonas assay. Ten days after agroinfiltration, agroinfiltrated leaves are spray‐inoculated with P. syringae bacteria. Bacterial growth is measured 3 days later by a classic colony count on Luria−Bertani (LB) agar plates containing cephaloridine, fucidin and cetrimide (CFC). (b) FLS2 depletion reduces reactive oxygen species (ROS) production upon flg22 treatment. Leaves were agroinfiltrated (OD₆₀₀ = 0.2) with hpGFP (grey) or hpFLS2 (blue) and, at 3 and 10 dpi, leaf discs were treated with 100 nm flg22. Error intervals represent SE of n = 12 replicates. (c) FLS2 depletion increases P. syringae growth. Agroinfiltrated leaves expressing hpGFP (grey) or hpFLS2 (blue) were spray‐inoculated at 10 dpi with 1 × 10⁸ CFU ml⁻¹ Pta6605 and bacterial growth was measured 3 days later using CFC selection. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics (**P < 0.01). (d) FLS2 depletion does not affect Agrobacterium growth. Bacterial growth of AtumGV3101 was measured by plating the leaf extracts described in (c) on medium containing gentamicin. Error bars represent SE of n = 3 replicates. Student’s t‐test statistics.

Figure 5

Rapid functional analysis of immune components. (a) Time scale for functional analysis by generation transgenic plants and by agroinfiltration (agromonas assay). (b) Phosphomutant EFR^Y836F is unable to trigger reactive oxygen species (ROS) burst upon elf18 treatment. Leaves were agroinfiltrated with EFR (red), EFR^Y836F (light red) or empty vector (EV; grey), and the ROS burst was measured at 3 dpi in leaf discs treated with 100 nm elf18. Error intervals represent SE of n = 12 replicates. (c) Phosphomutant EFR^Y836F is blocked in elf18‐triggered immunity. Two days after agroinfiltration, agroinfiltrated leaves expressing EFR, EFR^Y836F or EV were spray‐inoculated with 1 × 10⁸ CFU ml⁻¹ PtoDC3000(∆hQ) and bacterial growth was measured 3 days later using cephaloridine, fucidin and cetrimide (CFC) selection. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics (*P < 0.05). (d) Agroinfiltration of EFR, but not EFR^Y836F, reduces Agrobacterium growth. Bacterial growth of AtumGV3101 was measured by plating the leaf extracts described in (c) on medium containing gentamicin. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics (^** P < 0.01).

Figure 6

T3 effector suppresses immunity in the agromonas assay. (a,b) Expression of AvrPto blocks reactive oxygen species (ROS) production upon flgII‐28 and flg22 treatment. Leaf discs from agroinfiltrated leaves expressing FLS3 with empty vector (EV; blue), FLS3 with AvrPto (red) or EV alone (grey) were treated with 100 nm flgII‐28 (a) or flg22 (b) and the ROS burst was measured in RLU. Error intervals represent SE of n = 12 replicates. (c) Agroinfiltration of AvrPto increases Pseudomonas syringae growth in Nicotiana benthamiana and suppresses FLS3‐mediated immunity. Two days after agroinfiltration, agroinfiltrated leaves expressing FLS3 in combination with either AvrPto (red) or EV (blue) were spray‐inoculated with 1 × 10⁸ CFU ml⁻¹ PtoDC3000(∆hQ) and bacterial growth was measured 3 days later using cephaloridine, fucidin and cetrimide (CFC) selection. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics (**P < 0.01). (d) AvrPto does not affect growth of AtumGV3101. Bacterial growth of AtumGV3101 was measured by plating the leaf extracts described in (c) on medium containing gentamicin. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics.

Figure 7

β‐Galactosidases reduce bacterial growth of BGAL‐sensitive strains in agromonas assay. (a) NbBGAL1 and AtBGAL8 have β‐galactosidase activity. FDG‐hydrolysing activity was measured in apoplastic fluids isolated from bgal1 mutant leaves transiently expressing NbBGAL1 or AtBGAL8. Error bars represent SE of n = 3 biological replicates. (b) Agroinfiltration of NbBGAL1 and AtBGAL8 reduce Pta6605 growth. Two days after agroinfiltration, agroinfiltrated leaves expressing NbBGAL1 or AtBGAL8 were spray‐inoculated with 1 × 10⁸ CFU ml⁻¹ Pta6605 and bacterial growth was measured 3 days later using cephaloridine, fucidin and cetrimide (CFC) selection. Error bars represent SE of n = 6 biological replicates; t‐test P‐values (^* P < 0.05). (c) NbBGAL1 or AtBGAL8 do not reduce PsyB728a growth. Two days after agroinfiltration, agroinfiltrated leaves expressing NbBGAL1 or AtBGAL8 were spray‐inoculated with 1 × 10⁸ CFU ml⁻¹ PsyB728a and bacterial growth was measured 3 days later using CFC selection. Error bars represent SE of n = 3 biological replicates; t‐test P‐values. (d) NbBGAL1 and AtBGAL8 do not affect Agrobacterium growth. Bacterial growth of AtumGV3101 was measured by plating the leaf extracts described in (c) on medium containing gentamicin. Error bars represent SE of n = 3 biological replicates. Student’s t‐test statistics.