Microbe-associated molecular pattern (MAMP) signatures, synergy, size and charge: influences on perception or mobility and host defence responses

Abstract

Triggering of defences by microbes has mainly been investigated using single elicitors or microbe-associated molecular patterns (MAMPs), but MAMPs are released in planta as complex mixtures together with endogenous oligogalacturonan (OGA) elicitor. We investigated the early responses in Arabidopsis of calcium influx and oxidative burst induced by non-saturating concentrations of bacterial MAMPs, used singly and in combination: flagellin peptide (flg22), elongation factor peptide (elf18), peptidoglycan (PGN) and component muropeptides, lipo-oligosaccharide (LOS) and core oligosaccharides. This revealed that some MAMPs have additive (e.g. flg22 with elf18) and even synergistic (flg22 and LOS) effects, whereas others mutually interfere (flg22 with OGA). OGA suppression of flg22-induced defences was not a result of the interference with the binding of flg22 to its receptor flagellin-sensitive 2 (FLS2). MAMPs induce different calcium influx signatures, but these are concentration dependent and unlikely to explain the differential induction of defence genes [pathogenesis-related gene 1 (PR1), plant defensin gene 1.2 (PDF1.2) and phenylalanine ammonia lyase gene 1 (PAL1)] by flg22, elf18 and OGA. The peptide MAMPs are potent elicitors at subnanomolar levels, whereas PGN and LOS at high concentrations induce low and late host responses. This difference might be a result of the restricted access by plant cell walls of MAMPs to their putative cellular receptors. flg22 is restricted by ionic effects, yet rapidly permeates a cell wall matrix, whereas LOS, which forms supramolecular aggregates, is severely constrained, presumably by molecular sieving. Thus, MAMPs can interact with each other, whether directly or indirectly, and with the host wall matrix. These phenomena, which have not been considered in detail previously, are likely to influence the speed, magnitude, versatility and composition of plant defences.

Figure 1

Effect of concentrations of three microbe‐associated molecular patterns (MAMPs) on the induction of calcium ion influx in Arabidopsis leaves. (A) Elongation factor peptide elf18. (B) Flagellin peptide flg22. (C) Oligogalacturonan (OGA). All showed similar trends of slower and lower induction of calcium influx with decreasing concentrations. All data are means of at least three replicates. Significant differences (different letters) between treatments were detected by the Tukey–Kramer honestly significant difference (HSD) test.

Figure 2

Microbe‐associated molecular pattern (MAMP) combinations show additive effects in the induction of calcium ion influx or reactive oxygen species (ROS) generation in Arabidopsis leaves. (A) Flagellin peptide flg22 and elongation factor peptide elf18 in combination enhance induced calcium influx. (B) flg22 and elf18 in combination enhance the generation of ROS as peroxide (RLU, relative light units). (C) flg22 and lipo‐oligosaccharide (LOS) in combination result in higher calcium influx. (D) flg22 and core oligosaccharides in combination induce a higher calcium influx than individual MAMPs. (E) LOS and core oligosaccharides show additive effects in calcium influx induction. All data are means of at least three replicates. Significant differences (different letters) between treatments were detected by Tukey–Kramer honestly significant difference (HSD) test.

Figure 3

Microbe‐associated molecular pattern (MAMP) combinations show interference in the induction of calcium influx or reactive oxygen species (ROS) generation in Arabidopsis. (A) Flagellin peptide flg22 and peptidoglycan (PGN) in combination reduce the calcium influx triggered by flg22. (B) flg22 and hydrolysed PGN in combination reduce flg22‐induced calcium influx. (C) Oligogalacturonan (OGA) and flg22 in combination reduce flg22‐induced calcium influx. (D) OGA and flg22 in combination reduce the flg22‐induced oxidative burst (RLU, relative light units). Note that, in this example, interference occurred between MAMPs, even when both were used at higher concentrations than in other experiments. (E) OGA and elongation factor peptide elf18 in combination reduce the calcium influx induced by elf18. (F) In bak1– plants, the flg22‐induced oxidative burst is reduced by OGA. Note that the level of peroxide generated in response to flg22 is markedly lower than in the Arabidopsis Col‐0 plants shown in (D). All data are means of at least three replicates. Significant differences (different letters) between treatments were detected by Tukey–Kramer honestly significant difference (HSD) test.

Figure 4

Oligogalacturonan (OGA) does not reduce flagellin peptide flg22 binding to Arabidopsis cells. (A) Pretreatment of Arabidopsis cells for 30 min with 50 µg/mL OGA, followed by exposure to 0.4 nm ¹²⁵I‐Tyr‐flg22. (B) Exposure of cells to a mixture of 0.4 nm ¹²⁵I‐Tyr‐flg22 and 0.5, 1, 2.5 and 5 mg/mL OGA. There were no significant differences with OGA concentration, and therefore only 500 µg/mL OGA is shown here. Arabidopsis cells were incubated with ¹²⁵I‐Tyr‐flg22 either alone (total binding; grey columns) or with an excess of 10 µm unlabelled flg22 (non‐specific binding; open columns). To determine specific binding, non‐specific binding is subtracted from total binding (mean of three replicates ± standard deviation).

Figure 5

Calcium signatures: patterns of cytoplasmic calcium ion influx induced in Arabidopsis by five bacterial microbe‐associated molecular patterns (MAMPs). These concentrations were chosen to give maximum or near‐maximum responses and to reveal the most characteristic signatures. (See Fig. 1, which shows the influence of concentration on elf18, flg22 and OGA signatures.) All data are means of at least three replicates. Significant differences (different letters) between treatments were detected by Tukey–Kramer honestly significant difference (HSD) test. flg22, flagellin peptide; elf18, elongation factor peptide; LOS, lipo‐oligosaccharide; OGA, oligogalacturonan.

Figure 6

Expression of three defence‐related genes in Arabidopsis leaves challenged by three bacterial microbe‐associated molecular patterns (MAMPs), obtained by real‐time reverse transcriptase‐polymerase chain reaction (RT‐PCR). Data are the mean of two replicates ± standard deviation. Data significant at P < 0.001 unless marked as NS (not significant). Three independent biological repetitions of each experiment were performed with similar results. The data shown from one experiment are therefore representative. elf18, elongation factor peptide; flg22, flagellin peptide; OGA, oligogalacturonan; PAL1, phenylalanine ammonia lyase gene 1; PDF1.2, plant defensin gene 1.2; PR1, pathogenesis‐related gene 1.

Figure 7

Influence of a plant cell wall matrix on the movement of two bacterial microbe‐associated molecular patterns (MAMPs). (A) A cell wall (ex. Lycopersicon esculentum) column (0.5 × 15 cm) shows molecular sieving properties using two size markers (FMN, flavin mononucleotide). Markers and MAMPs were eluted with 10 mm phosphate buffer (pH 6) containing 3 mm sodium azide. (B) Elution of flagellin peptide flg22 (¹²⁵I‐Tyr‐ flg22) through the cell wall column in the presence and absence of 0.5 m NaCl. Radiolabelled flg22 was determined by γ‐counting. (C) Elution of lipo‐oligosaccharide (LOS) through the cell wall column. LOS was assayed by Limulus Amoebocyte Chromogenic Endpoint (LAL) assay.