RIN4 recruits the exocyst subunit EXO70B1 to the plasma membrane

Abstract

The exocyst is a conserved vesicle-tethering complex with principal roles in cell polarity and morphogenesis. Several studies point to its involvement in polarized secretion during microbial pathogen defense. In this context, we have found an interaction between the Arabidopsis EXO70B1 exocyst subunit, a protein which was previously associated with both the defense response and autophagy, and RPM1 INTERACTING PROTEIN 4 (RIN4), the best studied member of the NOI protein family and a known regulator of plant defense pathways. Interestingly, fragments of RIN4 mimicking the cleavage caused by the Pseudomonas syringae effector protease, AvrRpt2, fail to interact strongly with EXO70B1. We observed that transiently expressed RIN4, but not the plasma membrane (PM) protein aquaporin PIP2, recruits EXO70B1 to the PM. Unlike EXO70B1, RIN4 does not recruit the core exocyst subunit SEC6 to the PM under these conditions. Furthermore, the AvrRpt2 effector protease delivered by P. syringae is able to release both RIN4 and EXO70B1 to the cytoplasm. We present a model for how RIN4 might regulate the localization and putative function of EXO70B1 and speculate on the role the AvrRpt2 protease might have in the regulation of this defense response.

Keywords: Autophagy; EXO70B1; EXO70B2; RIN4; exocyst; plant immunity; secretion.

Fig. 1.

Yeast-two hybrid assays showing interactions of exocyst subunits with RIN4. (a) Full-length RIN4 protein (on the left) and the C-terminal membrane-anchored RIN4 fragment (on the right). (b) An N-terminal fragment and a C-terminal fragment without palmitoylation sites. SEC3A yeasts autoactivate expression of selection markers (as shown by Hála et al., 2008). (c) The overall structure of RIN4. The AvrRpt2 cleavage consensus sequences (RCS) within NOI domains are shown above the bar scheme, with actual cleavage sites indicated by triangles. C-terminal cysteine residues responsible for membrane anchoring are shown as CCC above the bar scheme. The range of N- (N-RCS2) and C- (RCS2-C) terminal RIN4 fragments used in Y2H experiments is also depicted above the bar scheme of the protein. The model of RIN4 was drawn with CSS-Palm software (Ren et al., 2008). (This figure is available in colour at JXB online.)

Fig. 2.

Co-immunoprecipitation assay (co-IP) showing an interaction between HA-EXO70B1 and GFP–RIN4 proteins transiently expressed in N. benthamiana. Proteins were immunoprecipitated with anti-GFP magnetic beads, and EXO70B1 was detected with anti-HA antibody. Experiments were performed three times with similar results. Marker sizes are shown in kDa next to the blots. Left: detection of HA-EXO70B1 in total cell extracts (CE) from plants co-expressing free YFP; in total cell extracts from plants co-expressing GFP–RIN4, in eluate (E) after co-IP with free YFP, and in eluate after co-IP with GFP–RIN4. Right: detection of GFP/YFP in plants expressing HA-EXO70B1 and co-expressing YFP in total cell extract (CE), GFP–RIN4 in total cell extract, YFP in eluate (E) after co-IP, and GFP–RIN4 in eluate after co-IP. The arrow marks bands corresponding to free YFP, and the asterisk marks the position of GFP–RIN4. (This figure is available in colour at JXB online.)

Fig. 3.

RIN4 recruits EXO70B1 to the plasma membrane. Shown are the confocal microscopy images of fluorescently labeled constructs transiently expressed in N. benthamiana leaves. While RFP–EXO70B1 alone (upper two panels) localized mainly to the cytoplasm and nucleus, when co-expressed with GFP–RIN4, RFP–EXO70B1 was almost exclusively localized to the plasma membrane. In contrast to RIN4, aquaporin PIP2;1–GFP did not change the subcellular localization of RFP–EXO70B1. RFP–EXO70B1 without RIN4 co-localized in the cytoplasm with both free YFP and the N-terminal RIN4 fragment (N-RCS2). Scale bars=50 µm.

Fig. 4.

DTT causes relocalization of both GFP–RIN4 and RFP–EXO70B1 to the cytoplasm. Shown are the confocal microscopy images of fluorescently labeled constructs transiently expressed in N. benthamiana leaves. Upper panels depict control cells; lower panels show cells treated with 50 mM DTT. Scale bars=50 µm.

Fig. 5.

P. syringae AvrRpt2 causes relocalization of both GFP–RIN4 and RFP–EXO70B1 to the cytoplasm. Shown are Z projections of confocal images of cells transiently expressing GFP–RIN4 (left panels) together with RFP–EXO70B1 (right panels) that were co-infiltrated either with the mutant strain of Pseudomonas (Pto DC3000 HrpH–; upper two panels) or with a strain harboring the AvrRpt2 protease (Pto DC3000 AvrRpt2; lower two panels). See the Materials and Methods for further details. Scale bars=50 µm.

Fig. 6.

RIN4 does not recruit EXO70B2 to the plasma membrane. GFP–RIN4 (upper left panel) was overexpressed together with RFP–EXO70B2 (upper right panel) in N. benthamiana leaves. Asterisks mark a cell expressing RFP–EXO70B2 only. Panels are Z projections of 17 confocal sections. Scale bar=50 µm. Quantification of the membrane to cytoplasm fluorescence ratio from constructs transiently expressed in N. benthamiana leaves is shown in the bottom part of the figure. Fluorescence was measured as described in the Materials and Methods in the cytoplasmic strands and the membrane portion of the cells. Shown are the average and the SD (error bars) from 11 cells for each combination. Cytoplasmic strands were seen only occasionally in (a) with most RFP–EXO70B1 signal being on the plasma membrane when co-expressed with GFP–RIN4. The average ratio for this combination is therefore underestimated. Three asterisks denote a significant difference as determined by a t-test (P-value <0.001) and N.S. indicates a difference that is not statistically significant (t-test P-value >0.05).

Fig. 7.

Neither GFP–RIN4 nor GFP–RIN4 co-expressed with HA-EXO70B1 recruit SEC6–RFP to the PM. Shown is the quantification of membrane to cytoplasm fluorescence ratio of constructs transiently expressed in N. benthamiana leaves. Fluorescence was measured as described in the Materials and Methods in the cytoplasmic strands and the membrane portion of the cell. Depicted are the means and the SD (error bars) from eight (SEC6–RFP alone and with GFP–RIN4) and nine (SEC6–RFP with GFP–RIN4 and HA-EXO70B1) cells. Differences between the means are not statistically significant (N.S.; ANOVA P-value=0.599).

Fig. 8.

RIN4 does not recruit SEC6 even in the presence of EXO70B1. EXO70B1–CFP was co-expressed with GFP–RIN4 and SEC6–RFP transiently in N. benthamiana leaves. Shown are the confocal images and composite images. Cytoplasmic strands are clearly visible in the SEC6–RFP channel. Scale bar=50 µm.

Fig. 9.

RIN4 recruits EXO70B1 to the PM in Arabidopsis stomatal guard cells. Confocal images of epidermal pavement and guard cells of the first true leaves of Arabidopsis seedlings show red fluorescence of RFP–EXO70B1 and green fluorescence of GFP–RIN4. Note the cytoplasmic and nuclear signal when RFP–EXO70B1 is expressed alone. Images in the upper two rows are maximal projections of 24 (RFP–EXO70B1) and nine (RFP–EXO70B1+GFP–RIN4) optical sections, respectively. The scale bar is 50 µm for the images in the upper two rows and 10 µm for the third row.