Plant chemical genetics reveals colistin sulphate as a SA and NPR1-independent PR1 inducer functioning via a p38-like kinase pathway

Abstract

In plants, low-dose of exogenous bacterial cyclic lipopeptides (CLPs) trigger transient membrane changes leading to activation of early and late defence responses. Here, a forward chemical genetics approach identifies colistin sulphate (CS) CLP as a novel plant defence inducer. CS uniquely triggers activation of the PATHOGENESIS-RELATED 1 (PR1) gene and resistance against Pseudomonas syringae pv. tomato DC3000 (Pst DC3000) in Arabidopsis thaliana (Arabidopsis) independently of the PR1 classical inducer, salicylic acid (SA) and the key SA-signalling protein, NON-EXPRESSOR OF PR1 (NPR1). Low bioactive concentration of CS does not trigger activation of early defence markers such as reactive oxygen species (ROS) and mitogen activated protein kinase (MAPK). However, it strongly suppresses primary root length elongation. Structure activity relationship (SAR) assays and mode-of-action (MoA) studies show the acyl chain and activation of a ∼46 kDa p38-like kinase pathway to be crucial for CS' bioactivity. Selective pharmacological inhibition of the active p38-like kinase pathway by SB203580 reverses CS' effects on PR1 activation and root length suppression. Our results with CS as a chemical probe highlight the existence of a novel SA- and NPR1-independent branch of PR1 activation functioning via a membrane-sensitive p38-like kinase pathway.

Figure 1

CS is a novel activator of the PR1 defence gene in Arabidopsis. (A) Hydroponically grown (96-well microtiter plate format) 14-day-old PR1p::GUS Arabidopsis reporter seedlings were treated in duplicate with the compounds from the chemical library (20 µM) for 24 h. GUS activity was recorded and converted to Z-score values. Compounds above a Z-score of 4 (indicated by a magenta dotted line) were re-screened leading to the identification of CS (green diamond). (B) Chemical structure of CS and two related polymyxins: PmB and PmBN. (C) 14-day-old hydroponically grown PR1p::GUS reporter seedlings were treated with increasing concentrations of CS for 24 h before measuring GUS activity. (D) Leaves of 28-day-old and 14-day-old Col-0 plants were treated with DMSO (1%) or CS (5 µM) for 24 h. PR1 transcript was quantified by qRT-PCR and normalised to the reference gene At4g26410 (Expressed Protein, EXPR). The values represent the mean (±SD) of at least three biological replicates. All experiments were repeated at least twice with similar outputs. Asterisks indicate significant differences from control values (*P < 0.05, **P < 0.01, and ***P < 0.001, two-tailed Student’s t-test).

Figure 2

CS does not stress plants at its EC₅₀ but affects seed germination and root growth. (A) 14-day-old Col-0 seedlings, treated with increasing concentrations of CS for 24 h were vacuum-infiltrated with Evans blue (0.1%, w/v) for 30 min and then kept for at least 6 h in the vacuum chamber. Dye bound to dead cells was quantified at A₆₀₀/A₆₈₀. (B) Properly sterilised and stratified Col-0 seeds were hydroponically grown for a week with the indicated CS concentration under long day conditions. (C) Properly sterilised Col-0 seeds were sown on solid half-MS phytagel media plates containing indicated concentrations of CS. Seeds on plates were stratified for 48 h and then kept vertically for 14 days in a long day growth chamber. (D) Root length of plants in (C) were quantified using the ImageJ software. At least 10 to 15 seedlings were used for root length calculation. All experiments were repeated at least twice with similar outputs. Asterisks indicate significant differences from respective controls (**P < 0.01 and ***P < 0.001, two-tailed Student’s t-test).

Figure 3

CS’ unique MoA induces PR1 independent of SA and SA-signalling components. (A) 14-day-old GUS reporter seedlings, inducible by selective phytohormones, were treated either with CS (5 µM) or with the respective inducers IAA (5 µM), ABA (100 µM) or MeJA (100 µM)) for 24 h in at least 3 biological replicates. GUS activity was normalised to the respective DMSO control of each reporter line. (B) 14-day-old Col-0 seedlings were treated with DMSO (1%), CS (5 µM), and flg22 (0.1 µM) followed by immediate addition of luminol-peroxidase mix to detect chemiluminescence [(reported as relative light units (RLU)] by a luminometer for the indicated time course. The values represent the mean (±SD.) of at least 5–10 biological replicates. (C) Hydroponically grown Col-0 seedlings were treated with bacterial elicitor Flg22 (1 µM), CS (5 µM) and DMSO (1%) for indicated time durations before harvesting and detecting phosphorylated MPK6 and MPK3 bands by immunoblotting using α-phospho-p44/42. For the uncropped blot, see Supplemental Fig. S7. (D) Hydroponically grown, 14-day-old seedlings of indicated genotypes were treated with DMSO (1%), CS (5 µM) or SA (100 µM) for 24 h. PR1 gene levels were plotted as change fold to the respective DMSO controls of each genotype. The values represent the mean (±SD) of at least three biological replicates. (E) Four-week-old leaves of different Arabidopsis genotypes were infiltrated with DMSO (1%) or CS (5 µM) for 24 h before spray-inoculating with a virulent luminescent luxCDABE-tagged Pst DC3000 strain (Fan et al., 2008). Pathogen growth was plotted as log₁₀ CPS per cm² relative to the DMSO control of each genotype. (F) 14-day-old Col-0 wild-type and fmo1-1 mutant seedlings were treated with DMSO (1%) and CS (5 µM) for 24 h. PR1 transcripts were quantified via qRT-PCR and are reported relative to EXPR. The values represent the mean (±SD) of at least three biological replicates. All experiments were repeated at least twice with similar outputs. Asterisks indicate significant differences from respective DMSO controls (*P < 0.05, **P < 0.01, and ***P < 0.001, two-tailed Student’s t-test).

Figure 4

Structure-activity-relationship assays with CS. (A) 14-day-old hydroponically grown PR1p::GUS reporter seedlings were treated with increasing concentrations of PmB for 24 h before measuring GUS activity. (B) 14-day-old Col-0 seedlings were treated with DMSO (1%), PmB (5 µM), and flg22 (0.1 µM) followed by immediate addition of luminol-peroxidase mix to detect chemiluminescence (reported as relative light units (RLU)) by a luminometer for the indicated time course. The values represent the mean (±SD) of at least 5–10 biological replicates. (C) 14-day-old PR1p::GUS seedlings were treated with CS (5 µM), PmB (5 µM), PmBN (5 µM) and surfactin (5 µM) for 24 h before quantifying GUS activity. The values represent the mean (±SD) of at least 3 biological replicates. All experiments were repeated at least twice with similar outputs. Asterisks indicate significant differences from respective controls (***P < 0.001, two-tailed Student’s t-test).

Figure 5

CS functions via a stress-sensitive p38-like protein kinase pathway. (A) 14-day-old PR1p::GUS reporter seedlings were pre-treated for 1 h with increasing concentrations of SB203580, followed by application of CS (5 µM) for 24 h before measuring resulting GUS activity. (B) 14-day-old PR1p::GUS reporter seedlings were pre-treated for 1 h with either SB203580 (10 µM) or SB202474 (10 µM) and were supplemented with CS (5 µM) and SA (100 µM) as indicated. (C) Western blot analysis of protein extracts from 14-day-old Col-0 wild-type seedlings treated with surfactin (5 µM), PmBN (5 µM), PmB (5 µM), CS (5 µM), or DMSO (1%) and the p38-like kinases inhibitor SB203580 (10 µM) either alone or together with CS (5 µM). Total protein was extracted and after gel electrophoresis and western blotting the membrane was probed with a commercially available α-phospho p38 antibody. Equal protein loading is demonstrated by Ponceau S staining. For the uncropped blot, see Supplemental Fig. S8. (D) Properly sterilised Col-0 seeds were sown on solid half-MS phytagel media plates containing indicated concentrations of CS and 10 µM SB203580. Root length were quantified using the ImageJ software. All experiments were repeated at least twice with similar outputs. Asterisks indicate significant differences from respective controls (***P < 0.001, **P < 0.01, two-tailed Student’s t-test).