Capping protein integrates multiple MAMP signalling pathways to modulate actin dynamics during plant innate immunity

Plants and animals perceive diverse microbe-associated molecular patterns (MAMPs) via pattern recognition receptors and activate innate immune signalling. The actin cytoskeleton has been suggested as a target for innate immune signalling and a key transducer of cellular responses. However, the molecular mechanisms underlying actin remodelling and the precise functions of these rearrangements during innate immunity remain largely unknown. Here we demonstrate rapid actin remodelling in response to several distinct MAMP signalling pathways in plant epidermal cells. The regulation of actin dynamics is a convergence point for basal defence machinery, such as cell wall fortification and transcriptional reprogramming. Our quantitative analyses of actin dynamics and genetic studies reveal that MAMP-stimulated actin remodelling is due to the inhibition of capping protein (CP) by the signalling lipid, phosphatidic acid. In addition, CP promotes resistance against bacterial and fungal phytopathogens. These findings demonstrate that CP is a central target for the plant innate immune response.

. LYK1 and LYK4 are expressed in dark-grown Arabidopsis seedlings. The transcript levels for LYK1 and LYK4 were quantified by RT-qPCR analysis on dark-grown seedlings of WT, lyk1 and lyk4 single mutants, as well as a lyk1 lyk4 double mutant. The expression of LYK1 was more abundant than LYK4 in etiolated seedlings and was completely absent in lyk1 mutant, whereas LYK4 transcripts remained similar to WT seedlings. No LYK4 transcripts were detected in lyk4 mutant, whereas LYK1 was expressed normally. The transcripts for both genes were absent in the lyk1 lyk4 double mutant. Mean values from triplicate biological samples and technical replications are plotted ± s.e.m., normalized to GAPD expression. Figure 3. LYK1 plays a major role in translating the perception of chitin into cytoskeleton remodeling. Actin architecture analysis was performed on hypocotyl epidermal cells of lyk1 and lyk4 single mutant seedlings following 5 min treatment with mock, 1 µM elf26 and 1 µM chitin, the wild-type siblings of lyk1 mutant were used as positive controls. The density of actin filament arrays in WT cells was significantly increased in response to both MAMPs. When compared with mock control, only a minor increase in filament density was induced in the lyk1 single mutant treated with chitin. The density of actin filament arrays was significantly increased following chitin treatment in the lyk4 single mutant, but not as strongly as in chitin-treated WT cells. However, elf26 treatment stimulated an increase in the density of actin filament arrays that was comparable to WT in both lyk mutants (a). No significant change in filament bundling was observed in any genotype or treatment tested (b). Values given are means ± s.e.m. (n ≥ 200 cells from 10 hypocotyls for each treatment and genotype; *P <0.05; ***P < 0.001 compared to mock control of the same genotype; Student's t test).

Supplementary Figure 4. Actin architecture in cpb-3 mutant cells fails to respond to both chitin and elf26 treatments.
Actin architecture analysis was performed on epidermal cells of WT and cpb-3 mutant treated with mock and MAMPs for 5 min. The cortical actin array in mocktreated cpb-3 mutant cells was more dense compared to WT cells, as shown previously 1,2 . Following MAMP treatment, the actin filament abundance in WT cells was significantly enhanced, whereas no significant changes in actin filament abundance were detected in cpb-3 mutant cells treated with elf26 or chitin compared to mock control (a). The extent of filament bundling was not altered in WT and cpb-3 mutant cells after MAMP treatments compared with their respective mock controls (b). Values given are means ± s.e.m. (n ≥ 200 cells from 10 hypocotyls for each treatment and genotype; ***P < 0.001; nd, no significant difference compared to mock control of the same genotype. Student's t test).

Supplementary Figure 5. Reduction of CP blocks the PA-induced increase in filament abundance.
Actin architecture analysis was performed on epidermal cells of WT and cpb-1 mutant treated with 0, 50 µM PA or 100 µM PS for 30 min, as described previously 1 . WT cells treated with PA had a significant increase in filament abundance, whereas PA treatment of the cpb-1 mutant did not increase filament density. Treatments with PS had no measurable effects on actin filament levels in either WT or cpb-1 seedlings (a). No significant differences in the extent of filament bundling were observed in WT and cpb-1 mutant cells among treatments (b). Values given are means ± s.e.m. (n ≥ 200 cells from 10 hypocotyls for each treatment and genotype; ***P < 0.001; nd, no significant difference compared to mock control of the same genotype. Student's t test).

Supplementary Figure 6. FIPI treatment blocks MAMP-triggered actin remodeling.
Percentage of occupancy or density was measured on epidermal cells of WT (a,c,e) or cpb-1 mutant (e) treated with 750 nM FIPI for different time periods (30 min in [e]), and followed by 5-min treatment with 1 µM elf26 or 1 µM chitin. Pretreatments with the PLD inhibitor 5-fluoro-2-indolyl des-chlorohalopemide (FIPI) in the absence of MAMP stimulation had no significant effect on the density of actin arrays in WT cells when compared with mock control. However, FIPI inhibited the increase in actin filament abundance following MAMP treatments, in a time-dependent manner (a,c). By contrast, actin filament abundance in the cpb-1 mutant was not altered by either treatment (e). No significant differences in the extent of filament bundling were observed in WT and cpb-1 mutant cells among the various treatments (b,d,f). Values given are means ± s.e.m. (n ≥ 200 cells from 20 hypocotyls for each treatment and genotype; ***P < 0.001; nd, no significant difference. Student's t test).

Supplementary Figure 7. Basal expression levels of defense responsive genes in WT and cpb-1 mutant.
The basal expression level of defense responsive genes was monitored by real-time quantitative PCR (RT-qPCR) of untreated, dark-grown WT and cpb-1 mutant seedlings. The transcript levels of MAPK-specific reporter gene, FRK1, as well as MAPK dominant-pathway genes CYP81F2 and WRKY33; A CDPK-specific response gene, PHI1 and the CDPKsynergistic pathway gene, NHL10 were tested. Mean values from triplicate biological samples and technical replications were normalized to GAPD expression in the same genotype and plotted ± s.e.m., (**P < 0.01; Student's t test).

Supplementary Figure 8. Transcriptional activation of CDPK and MAPK pathways is altered in cpb-1 mutant following MAMP treatments.
Quantitative analysis of marker genes for initial defense signaling. In WT plants, the transcript levels of defense responsive genes: FRK1 (a), CYP81F2 (b), WRKY33 (c), PHI1 (d) and NHL10 (e) were activated by elf26 or chitin treatments. Except for FRK1, the expression of these defense-induced genes was not altered in cpb-1 mutant treated with elf26. The induction of all the MAMP-responsive genes tested was impaired in cpb-1 mutant following chitin treatment. Mean values from triplicate biological samples and technical replications are plotted ± s.e.m., normalized to GAPD expression and presented as fold induction from mock (**P < 0.01; ***P < 0.001; nd, no significant difference compared to mock control of the same genotype; Student's t test).

Supplementary Table 1. Actin dynamic parameters from MAMP-treated WT and cpb-1 mutant epidermal cells
Values given are mean ± s.e.m., with n > 50 filaments from n > 10 epidermal cells and at least 10 hypocotyls per treatment. For filament origin: n > 30 cells from at least 10 hypocotyls per treatment.