Arabidopsis thaliana Col-0 ecotype (i.e., WT Col-0) was used in all experiments. Mutants and transgenic lines used in this study include rpm1/rps2/rin4 (r124), rpm1/rps2/rin4/exo70B1-1 (r124b1), rpm1-3/rps2-101c/exo70B1-1 (r12b1), gRIN4p::T7-RIN4/r124, gRIN4p::T7-RIN4^T166A /r124, gRIN4p::T7-RIN4^T166D/r124, r124 co-expressing 35S::RFP-PIP2A and YFP fused RIN4 variants (RIN4, RIN4^T166A, and RIN4^T166D), r124 co-expressing 35S::RFP-ST and YFP fused RIN4 variants (RIN4, RIN4^T166A, and RIN4^T166D), 35S::YFP-RIN4/r124b1, 35S::YFP-RIN4^T166A /r124b1, 35S::YFP-RIN4^T166D/r124b1, exo70B1-1 (CS2103358, ABRC), exo70B1-2 (CS414954, ABRC), exo70B2 (CS2103357, ABRC), exo70B1-1/exo70B2 (CS2103356, ABRC), 35S::YFP-RIN4/Col-0, 35S::YFP-RIN4/r124, 35S::YFP-RIN4^T166A/r124, 35S::YFP-RIN4^T166D/r124, 35S::YFP-RIN4/exo70B1-1, and 35S::HA-RIN4^{T21E/S160E/T166D}/r124. To generate transgenic lines expressing YFP- or HA-tagged RIN4 derivatives, the coding sequence of RIN4 was inserted into pENTR/D-TOPO followed by site-directed mutagenesis using the Phusion Site-Directed Mutagenesis Kit (Thermo Scientific). The mutated and non-mutated coding sequences of RIN4 were subsequently recombined into the pEarleyGate104 or pGWB515 binary vector with the Gateway LR Clonase Kit (Thermo Scientific). The final constructs were introduced into WT Col-0, or mutants plants (e.g., r124, exo70B1-1), using the floral dip method (Zhang et al., 2006). The seeds of T7-tagged lines were provided by Dr. Jeff Dangl (UNC-Chapel Hill, USA). Mutants were identified by PCR and DNA sequencing. Primers used for genotyping are listed in Supplementary Table S1 . Arabidopsis and Nicotiana benthamiana plants were grown in a growth room at 20°C and ~60% relative humidity in a 12-h-light/12-h-dark day/night cycle.

For transient protein expression analyses, leaf mesophyll protoplasts were isolated from 4-week-old Arabidopsis plants and transfected using the polyethylene glycol method (Wu et al., 2009). Microscopic images were captured under a fluorescence microscope (Nikon A1R-TIRF-STORM) that equipped with a 40x oil objective after 16 h. For CHX (cycloheximide) treatments, methods previously described were followed (Maldonado-Bonilla et al., 2014). For the Latrunculin-B (LatB) treatment assay, protoplasts were treated with 0.1 μM LatB for 30 minutes at room temperature (ca. 25°C). For the stable transgenic lines expressing fluorescence protein, the cotyledons were visualized under a Nikon A1R-TIRF-STORM fluorescence microscope. The fluorescence of YFP, GFP, RFP, and chlorophyll were observed using the following excitation/emission wavelengths: 513/520-550 nm, 488/500-525 nm, 561/580-620 nm, and 488/680-700 nm, respectively. The fluorescence intensity was measured by using the “Measure” tool, Fiji software (https://fiji.sc/), and data were analyzed using GraphPad Prism 6 software.

A solution of Pst DC3000 hrcC^- bacteria were syringe-infiltrated into leaves of 4-week-old Arabidopsis plants (final OD_600nm = 0.002). Bacterial growth was determined at 0- and 3-days post-inoculation. The flg22-activated bacterial growth suppression assay was performed as described by Knepper et al (Knepper et al., 2011). Four-to-five-week-old Arabidopsis plants were hand-infiltrated with 1 µM of flg22 or distilled water on the lower leaf surface, Pst DC3000 bacteria were hand-infiltrated at 1×10⁶ cfu/mL (OD_600nm = 0.002) 24 h post flg22-treatment. Bacterial growth was determined 2-days post-inoculation. Statistical difference in bacterial growth was analyzed by pairwise comparisons for all means using one-way ANOVA with GraphPad Prism 6 software.

Total RNA was extracted using the RNeasy Plant Mini Kit (Qiagen) and converted to first-strand cDNA using the SuperScript III reverse transcriptase Kit (Invitrogen). The IPP2 gene (At3g02780) was used as the normalization standard (Marcolino-Gomes et al., 2015) and the comparative Ct method (2^{-ΔΔ C t}). The primers used for RT-PCR are listed in Supplementary Table S1 . The expression levels of RIN4, Exo70B1, and Exo70B2 were analyzed with GraphPad Prism 6 software.

Plant tissue samples were prepared as previously described. Total PM protein was extracted following the kit protocol (Minute Plasma Membrane Protein Isolation Kit for Plants, Invent Biotechnologies). Golgi were isolated with Minute™ Plant Golgi Apparatus Enrichment Kit (Invent Biotechnologies). Protein samples were resolved by electrophoresis on 4-12% Bis-Tris Protein gels. After transferring to nitrocellulose membranes, the membranes were blocked in 5% nonfat dry milk diluted in phosphate-buffered saline (PBS) for 1-2 h at room temperature. The membrane was then transferred into a 5% dry milk solution containing primary antibody and incubated for 1 h at room temperature. After washing with PBS buffer for 10 minutes, 3 times, the membrane was incubated in incubation solution with secondary antibody (1:10,000) for 1 h and then washed with PBS containing 1% Tween20 for 10 minutes, 3 times. The SuperSignal West Pico PLUS Chemiluminescent Substrate Kit (Pierce) was used to detect the protein signal. Actin was detected using an anti-Actin antibody (Sigma-Aldrich; cat. #A0480, 1:2000). RIN4 was detected using the anti-RIN4 antibody (1:2000). YFP was detected using the anti-GFP antibody (Sigma, SAB4301138; 1:2000). Golgi were detected using the anti-Sec21p antibody (Agrisera, AS08 327; 1:2000).

The full-length RIN4 coding sequence, as well as those of the phospho-derivatives of RIN4 (e.g., RIN4^T166D, and RIN4^{T21ES160ET166D}), were cloned into the bait vector pDEST32. EXO70B1 was cloned into the prey vector pDEST22. The interaction between RIN4 and pDEST22 empty vector, and the interaction between pDEST32 empty vector with Exo70B1, served as negative controls. The interaction between RIN4 and AHA1 was used as positive control. The bait and prey vectors were co-transformed into the yeast strain AH109 in different combinations, and the transformants were selected on synthetic dextrose (SD) medium without Leu and Trp (SD-Leu-Trp). Positive clones were plated on SD-Leu-Trp-Ura-His medium.

Agrobacterium tumefaciens strain GV3101 expressing the desired constructs (pSPYNE::RIN4, pSPYCE::Exo70B1, and pSPYCE::Exo70A1) were diluted in infiltration buffer (10 mM MgSO₄·7H₂O with 100 µM acetosyringone) at an optical density OD_600nm = 0.05. The Agrobacterium mixture was infiltrated into the abaxial leaf surface of 2-week-old N. benthamiana leaves, as previously described (Sassmann et al., 2018). Fluorescence was observed at 72 h post-infiltration as previously described.

Plants were grown for 4 weeks, and fully expanded leaves were infiltrated with flg22 (1 µM). Leaf samples were then collected at 0, 5-, 10-, 15-, and 30-min time points for WT Col-0, exo70B1-1, exo70B1-2, exo70B2, and exo70B1-1/exo70B2. Each sample weighed 0.1 g and then was ground in liquid nitrogen. After grinding, 100 µl protein extraction buffer (50 mM Tris–HCl (pH 7.5), 5 mM EDTA, 5 mM EGTA, 150 mM NaCl, 1 mM DTT, 10 mM Na₃VO₄, 10 mM NaF, 50 mM β-glycerolphosphate, 1 mM phenylmethylsulfonyl fluoride (PMSF), plant protease inhibitor cocktail (Sigma-Aldrich), 10% glycerol, 0.1% NP-40) for each tube was added and well mixed immediately. After centrifugation, the protein supernatant was mixed with Laemmli sample buffer and boiled for 5 min.

MAPK activity was detected by western blotting using p44/42 MAPK (Erk1/2) Antibody (Cell Signaling Technology, #9102S; dilution, 1:1000) as primary antibody and goat anti-rabbit-HRP (Amersham Biosciences) as secondary antibody. Actin was detected using an anti-actin antibody (Sigma-Aldrich; cat. #A0480) as the primary antibody and goat anti-mouse IgG (H+L) as the secondary antibody (Invitrogen; cat. #31430).

The leaf discs from the fully expanded leaves of 4-week-old WT Col-0, fls2, exo70B1-1, exo70B1-2, and exo70B1-1/exo70B2 were placed into a white 96-well plate, which was with 150 μl of water per well. After overnight incubation, 100 μl of reaction mixture (17 μg/mL of Luminol (Sigma), 10 μg/mL of Horseradish Peroxidase (HRP; Sigma), 100 nM flg22), was substituted for the water to monitor ROS burst (Chung et al., 2014). Luminescence was measured immediately with 1 s integration and 2 min interval over 60 min by using a BioTek Synergy H1 plate reader. Every leaf was sampled for only one disc of 20 mm diameter, each plant was punched for two leaf discs, and eight plants of each line were used for sampling. GraphPad Prism 6 was used for data analysis.

To investigate the subcellular localization of RIN4 and the RIN4 phospho-variants in planta, the fully expanded leaves of 4-week-old plant were dissected into 1 mm² leaf discs in fixation solution (4% (v/v) paraformaldehyde and 0.5% (v/v) glutaraldehyde in PBS buffer), followed by the application of a vacuum ensure complete infiltration of the fixative. The samples were embedded in LR White (Zechmann et al., 2005). Ultra-thin sections (80 nm) were collected on formvar-coated nickel grids. The grids were pre-incubated with the blocking solution (0.01M PBS, 5% BSA, and 5% normal goat serum) for 10-30 minutes at room temperature, then washed three times with incubation buffer (0.01M PBS and 0.1% BSA) for 5 minutes each. The grids were then incubated with the primary antibody (anti-RIN4 antibody, diluted 1:500 in incubation buffer) for 1 hour at room temperature. Afterward, the grids were washed three times with incubation buffer (5 minutes per wash), followed by incubation with the secondary antibody (Goat anti-mouse IgG coupled with 10 nm colloidal gold; Sigma, G7652), which was diluted 1:20 in incubation buffer, for 1 h. The grids were washed in incubation buffer three times (5 minutes each), followed by three washes in distilled water (5 minutes each). Ultra-thin sections were stained with uranyl acetate followed by lead citrate. The samples were embedded and cut into sections in the Center for Advanced Microscopy (CAM) at Michigan State University. Immunogold labeling and TEM analyses were also conducted in the CAM. The measurements of the size of RIN4 aggregation particles were performed using Fiji (https://fiji.sc/). The data were analyzed using GraphPad Prism 6.

Previous studies suggest that the suppression of PTI by secreted pathogen effector AvrRpt2 require the release of RIN4 from the PM (Takemoto and Jones, 2005; Afzal et al., 2011). To further evaluate this possible mechanism, we first examined the subcellular localization of a set of RIN4 phosphomimetic mutants – T21E, S160E, and T166D – the phosphorylation of which have been reported to be induced by the bacterial pathogen effector AvrB (Liu et al., 2011; Xu et al., 2017). As shown in Figure 1A , expression of RIN4^T166D resulted in the formation of punctate structures, whereas other single site phosphomimetic mutants of RIN4 displayed WT PM localization. This aberrant subcellular localization of RIN4^T166D was further confirmed in Arabidopsis transgenic lines producing the YFP-RIN4^T166D chimeric protein ( Figure 1B ). In contrast, the phospho-dead RIN4 mutants, in which T21, S160, or T166 residues were substituted with alanine, showed proper PM localization when expressed in WT protoplasts ( Supplementary Figure S1A ). Since AvrB triggers RIN4 hyperphosphorylation at Thr21, Ser160, and Thr166 (Xu et al., 2017), we next examined the subcellular localization of double- and triple-site phosphomimetic mutants of RIN4. It was found that all the RIN4 multiple phospho-variants had disrupted PM localization ( Supplementary Figure S1B ).

Next, we investigated whether the RIN4^T166D variant restricts the activation of plant innate immunity. Transgenic plants expressing T7-tagged RIN4 variants in the r124 (rpm1/rps2/rin4) genetic background were generated (Belkhadir et al., 2004), and the flg22-priming assay was performed to evaluate the function of T166 in the PTI response. We found that plants expressing the RIN4^T166D isoform supported higher bacterial growth as compared to those expressing WT RIN4 or the RIN4^T166A variant ( Supplementary Figure S2A ), which is supported by published data (Chung et al., 2014). We further evaluated the activation of plant basal defense responses in these transgenic lines using the mutant bacteria Pst DC3000 hrcC^- . As shown in Supplementary Figure S2B , plants expressing the RIN4^T166D variant allowed more bacterial growth than those expressing WT RIN4 or the phospho-dead variant RIN4^T166A. Meanwhile, the RIN4 isoforms were produced at similar levels in these transgenic lines ( Supplementary Figure S2C ). Taken together, our results suggest that the phospho-status of T166 plays a significant role in modulating host PTI response and shows gain-of-function phenotype.

To determine the nature of the YFP-RIN4^T166D puncta, we next performed co-localization analysis by detecting the pavement cells from stable transgenic lines co-expressing YFP fused RIN4 variants and RFP fused markers in r124 background. We found that the YFP-labelled RIN4 and RIN4^T166A show perfect colocalization with the PM marker PIP2A, while RIN4^T166D displays both PM and cytoplasmic localization ( Supplementary Figure S3 ). Then the Golgi localization of the RIN4^T166D protein was confirmed by full co-localization with the RFP labelled sialyltransferase (RFP-ST) at the Golgi apparatus ( Figure 2A ). To determine whether the cellular distribution of YFP-RIN4^T166D was affected by protein degradation, total protein extracts were isolated from transgenic lines and evaluated using anti-GFP antibody produced in rabbit (Sigma, SAB4301138) by western blot. The result showed specific bands for YFP-RIN4 isoforms, confirming protein stability ( Figure 2B ). Additionally, we extracted Golgi from stable transgenic lines expressing RIN4 native promoter driven T7-fused RIN4 variants (e.g., RIN4, RIN4^T166D) and evaluated protein levels using anti-RIN4 antibody. As shown in Figure 2C , the protein level of RIN4^T166D was notably higher in Golgi compared to the WT RIN4 protein, which is in line with the observations of Figure 2A . These results indicated that the phosphorylation of RIN4 on T166 residue might block the transport of RIN4 to the PM, which induced protein aggregation in Golgi.

According to previous studies, there are two potential pathways through which RIN4 might localize to the PM. One is exocytosis, which transports protein to the PM after synthesis (Bassham and Raikhel, 1996; Ponnambalam and Baldwin, 2003; Chia and Gleeson, 2014). The other is protein endocytic recycling, where plasma membrane-localized protein enters cytoplasmic through endocytosis and then re-localizes to the PM (Grant and Donaldson, 2009). As actin has been linked in multiple ways with exocytosis and endocytosis, we hypothesized that cytoskeletal actin plays a role in mediating RIN4 PM localization (Chakrabarti et al., 2021). To test this hypothesis, we treated mesophyll protoplasts that were transiently expressing YFP-RIN4 with the actin antagonist LatB and then evaluated the subcellular localization of RIN4. As shown in Figure 3A , we observed the rapid formation of puncta in proximity to the PM upon treatment, suggesting that disruption of the host actin cytoskeleton negatively impacts membrane trafficking of RIN4 and leads to aggregation of the YFP-RIN4 protein ( Figure 3A ). When protoplasts are pretreated with the protein biosynthesis inhibitor cycloheximide (CHX), we hypothesized that the PM localization of RIN4 would follow the endocytic recycling pathway, as noted above, and the punctate structure(s) would be detected following LatB application. Interestingly, the impact of LatB on RIN4 location was eliminated when protoplasts were pretreated with CHX ( Figure 3B ). These observations demonstrate that PM localization of RIN4 does not rely on endocytic recycling, but rather depends on exocytosis. We then employed transgenic lines stably expressing YFP-RIN4 and observed that the mis-localization of RIN4 protein induced by LatB was also abolished by CHX pretreatment ( Figure 3C ). Taken together, our data suggests that membrane tethering of RIN4 relies on actin cytoskeleton-involved exocytosis.

To further investigate the mechanisms(s) underpinning this activity, we next asked how the function of the phosphomimic derivative RIN4^T166D intersects with the localization-dependent function of RIN4. As a first step, we asked if the exocytosis pathway contributes to PM localization of RIN4; this line of investigation is based on previous results demonstrating that RIN4 co-localizes with Exo70B1 in N. benthamiana (Sabol et al., 2017). After confirming the interaction using a split-YFP approach in N. benthamiana ( Supplementary Figure S4A ), we used a Y2H approach to demonstrate that both phosphomimetic mutants RIN4^T166D and RIN4^{T21E/S160E/T166D} abolished the interaction between RIN4 and Exo70B1 ( Supplementary Figure S4B ). Taken together, these data support the hypothesis that altered subcellular localization of RIN4^T166D ( Figure 1 ) and the RIN4^{T21E/S160E/T166D} triple mutant ( Supplementary Figure S1B ) is likely due to a loss in the interaction between RIN4 and Exo70B1.

We then asked whether Exo70B1 was required for PM-localization of RIN4. To do this, we detected the natural distribution of RIN4 protein in exo70 mutants and RIN4 variants expression lines using immunogold labelling assay and collected the images with transmission electron microscope ( Figure 4 ). The leaf-sections from 4-week-old plants were specific labeled with RIN4 antibody, and the results showed that the RIN4 displayed a similar distribution in WT Col-0, exo70B2, and 35S::YFP-RIN4 in r124 where gold particles gave specific and clear labeling along the PM and non-specific labeling in the cytoplasm. However, in the transgenic line expressing the T166 phosphomimetic variant of RIN4, YFP-RIN4^T166D, RIN4 was detected in aggregates ( Figure 4A ).

To quantify the protein aggregation level of each line, we first define a valuation standard as the shape of RIN4 (or RIN4^T166D) aggregation is irregular ( Supplementary Figure S5 ). Firstly, we drew lines between outermost nanoparticles of an aggregation and then got a closed graph. A line (named lineA) that links two furthest points was drawn and measured. After that, a rectangular coordinate system, whose y-axis parallels lineA and x-axis passes through the midpoint of lineA, was made. Two points where graph crossed the x-axis were named 1 and 1’, and the distance between these two points was measured (lineB). The average of lineA and lineB (Size=(“length lineA” + “length lineB”)/2) was calculated and the outputs were divided into 3 groups based on size: 50-100 nm, 100-200 nm, and >200 nm. Figure 4B shows the average number of each type per image and r124 expressing gRIN4p::T7-RIN4^T166D contains more RIN4^T166D aggregations, whose sizes were between 50-100 nm (2.77 ± 0.03) and 100-200 nm (0.92 ± 0.03), than other lines ( Figure 4B ). This result is consistent with our hypothesis that the phosphomimetic mutant RIN4^T166D induces RIN4 aggregation and blocks its tethering to the PM. Interestingly, mutants lacking Exo70B1 showed similar labeling patterns, with abundant RIN4 aggregations detected in the cytoplasm. It should also be noted that the RIN4 aggregates in exo70B1 knockout mutants bound RIN4 antibody less heavily than those of the RIN4^T166D transgenic line, which might because the RIN4^T166D was overexpressed in r124 background so higher protein level leads to heavy aggregation ( Figure 4C ).

The PTI response was repressed in r124 plants that produce the RIN4^T166D mutant protein ( Supplementary Figure S2 ). To continue this line of investigation, we pretreated exo70B mutants with flg22 followed by inoculation with the virulent pathogen Pst DC3000. Simultaneously, we infiltrated exo70B leaves with Pst DC3000 hrcC^- to evaluate induction of the PTI response. Both exo70B1-1 and exo70B1-1/exo70B2 plants were more susceptible to the bacteria, while exo70B2 showed comparable susceptibility to WT plants ( Figures 5A, B ). Moreover, an additional mutant allele exo70B1-2 phenocopied exo70B1-1 in regard to their PTI responses ( Supplementary Figures S6A, B ). To address whether the increased susceptibility of exo70B1 mutants was caused by enhanced expression of RIN4, we compared the transcription level of RIN4 in exo70B1, exo70B2, and WT plants. It was found that the transcript abundances of RIN4 in the exo70 mutants were comparable to that in WT Col-0 ( Supplementary Figure S6C ). We next sought to determine whether the disease resistance was repressed further by expressing the triple phospho-variant in r124, since the disruption of RIN4 PM localization was enhanced when expressing the RIN4^{T21E/S160E/T166D} mutant ( Supplementary Figure S1B ). As shown in Figure 5C , susceptibility of the two transgenic lines expressing RIN4^{T21E/S160E/T166D} was comparable to that of exo70B1 mutants. The data showed that flg22-induced resistance is reduced in RIN4^T166D and RIN4^{T21E/S160E/T166D} expression lines and exo70B1 genetic mutants.

Given that rin4 plants are reduced in the flg22-induced ROS burst (Lee et al., 2015), we analyzed the flg22-triggered ROS burst in WT Col-0, exo70B1-1, exo70B1-2, exo70B1-1/exo70B2, and fls2. The fls2 mutant was used as negative control, and WT Col-0 was the positive control. Upon flg22 treatment, similar observation of reduced ROS level was made in exo70B1-1, exo70B1-2, and exo70B1-1/exo70B2 mutants ( Figure 6A ), which is consistent with previous report (Stegmann et al., 2012). Since PTI response also includes the activation of mitogen-activated protein (MAP) kinases, we tested MAPK activity levels in exo70B1-1 and WT Col-0 at 0-, 10-,20-, 30-, 60-min upon flg22 infiltration. MAPK activation was also reduced in exo70B1-1 plants when compared with WT Col-0 ( Figure 6B ). Taken together, we posit that PTI responses are dampened.

To confirm the enhanced susceptibility of exo70B1 mutants to Pseudomonas syringae pv. tomato is related to RIN4 phosphorylation at T166 site, the stable transgenic lines overexpressing RIN4 variants in rpm1-3/rps2-101c/rin4/exo70B1-1 (r124b1) mutant background were prepared and checked with the PTI response and basal defense. By choosing r124b1 quadruple mutant, not exo70B1 mutant, as transgenic background, we eliminate the side effect of ETI induced by RIN4 variants. Compared to r124, rpm1-3/rps2-101c/exo70B1-1 (r12b1) was more susceptible to bacteria like rpm1-3/rps2-101c/rin4/exo70B1-1 (r124b1), which indicates RIN4 and Exo70B1 are in the same PTI regulation pathway and Exo70B1 has a bigger effect on PTI than RIN4. Interestingly, overexpressing of RIN4 or RIN4^T166A in the r124b1 background reduced pathogen growth compared to r12b1, while RIN4^T166D-expression did not ( Figure 7 ). This not only proves RIN4 plays a role downstream of Exo70B1 and proper PM localization of RIN4 requires Exo70B1, but also indicates that the gain-of-function of phospho-RIN4^T166D requires Exo70B1 even though their interaction is disrupted by this RIN4 mutation. Additionally, we noticed that the phenotype of r124b1 was complemented in 35S::YFP-RIN4/r124b1 ( Figure 7 ). This may be due to the constitutive expression of RIN4 leading to a higher protein level than in WT Col-0, with some of proteins localized to the PM. Taken together, these analyses support the model that the PM-localization of RIN4, which is mediated by Exo70B1, plays a role in PTI response, and the phosphorylation of RIN4 at T166 site represses PTI by inhibiting RIN4-Exo70B1 interaction.

The function of RIN4 in plant immunity is regulated by phosphorylation, which is induced by both virulence factors of pathogens and host enzymes. Phosphorylation at specific residues of RIN4 has different effects on the activation or suppression of PTI signaling. It’s well known that flg22-induced phosphorylation of RIN4 at S141 by host BIK1 and PBS1-like 1 kinases enhances PTI signaling (Lu et al., 2010; Zhang et al., 2010; Chung et al., 2014; Li et al., 2014). However, the effect of S141 phosphorylation is suppressed by phosphorylation of T166, which leads to PTI repression (Chung et al., 2014). Phosphorylation at RIN4^T166, induced by AvrRpm1 and AvrB, requires CDG1 (CONSTITUTIVE DIFFERENTIAL GROWTH1) and RIPK (RPM1-induced protein kinase) (Chung et al., 2011; Liu et al., 2011; Yang et al., 2021). Besides this, RIN4 is hyperphosphorylated by RIPK at T21, S160, and T166 and then activates RPM1-mediated ETI (Liu et al., 2011; Lee et al., 2015; Xu et al., 2017). To date, studies on phosphorylation of RIN4 have been focusing on the molecular mechanisms underlying pathological phenotype, while little attention has been paid to the cellular biological regulation. It is noteworthy that a previous study claimed that the nonmembrane-tethered RIN4 derivatives suppress PTI more potently, but no further mechanism was explored (Afzal et al., 2011). In our work, we demonstrated that the subcellular localization of RIN4 is altered by phosphorylation at T166, and/or at two residues of T21, S160, and T166 residues ( Figure 1 ; Supplementary Figure S1 ). Transgenic lines expressing RIN4^T166D displayed gain-of-function PTI phenotypes ( Supplementary Figure S2 ), suggesting that proper PM localization of RIN4 is required for plant immunity. Then, the question raised by our work is whether pathogen or PAMP treatment alters the PM localization of RIN4. To investigate this, we infiltrated leaves of stable transgenic plants expressing YFP-RIN4 with H₂O, AvrB, and AvrPphB and detected the subcellular localization of RIN4 at 1 h and 3 h post-treatment. We found that both AvrB and AvrPphB treatments led to altered localization of RIN4, and no significant difference was detected between the two treatments.

The exocyst complex, a conserved complex found in yeast, mammals, and plants, plays a role in protein delivery to the PM (Liu and Guo, 2012). The exocyst complex consists of eight subunits: Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, Exo70, and Exo84 (Elias et al., 2003; Hsu et al., 2004; Hala et al., 2008; Cvrckova et al., 2012; Martin-Urdiroz et al., 2016; Mei and Guo, 2018; Mei et al., 2018; Zhu et al., 2018). In Arabidopsis, each exocyst subunit has gene duplications (Cvrckova et al., 2012), and in the case of Exo70, it has evolved into 23 paralogs (Zarsky et al., 2020). These isoforms possess specific expression patterns and biological functions that further define and differentiate the distinct functions of exocyst (Saeed et al., 2019; Pecenkova et al., 2020). Given that RIN4 interacts with Exo70B1 and RIN4 pT166 eliminates the RIN4/Exo70B1 association (Sabol et al., 2017; Redditt et al., 2019; Yuen et al., 2023), we first confirmed this conclusion by performing BiFC and yeast-two-hybrid ( Supplementary Figure S4 ). Furthermore, the PM localization of RIN4 was also altered in exo70B1 mutants and in RIN4^T166D stable transgenic line, coinciding with suppressed PTI responses, including ROS burst and MAPK activation ( Figures 4 – 6 ; Supplementary Figure S6 ). These results support the hypothesis that the interaction between Exo70B1 and RIN4 is required for PM localization of RIN4 as well as the activation of the RIN4-dependent PTI response.

In support of the above, we observed higher bacterial growth in exo70b1 compared to the RIN4^T166D transgenic line. Interestingly, in planta bacterial growth loads were comparable to those observed in the RIN4^{T21E/S160E/T166D} line ( Figure 5 ). This indicates that the repression of PTI by phosphorylated RIN4 might depend on the interaction affinity between RIN4 and Exo70B1. It should be noted that our observations differ from previous studies (Zhao et al., 2015; Liu et al., 2017), and it is worth pointing out that the repressed PTI responses observed in exo70B1-1 and exo70B1-2 are contrary to the previous conclusions which reported enhanced disease resistance in exo70B1-3.

Based on our observation that all three exo70B1 knockout mutants exhibited HR-like cell death and that both exo70B1-1 and exo70B1-2 showed reduced resistance against virulent Pst and Pst hrcC- (Kulich et al., 2013; Stegmann et al., 2013), we propose that additional genomic mutations may be present and warrant further investigation. Previous research reported that exo70B1-3 had lower FLS2 accumulation at the PM compared to WT Col-0 (Wang et al., 2020). As FLS2 is a receptor kinase for flg22, a decrease in its PM levels correlates with reduced flg22 responses (Wang et al., 2014; Orosa et al., 2018). However, to determine if the reduced FLS2 levels in exo70B1-3 are due to the absence of Exo70B1 or other genomic mutations, it is essential to investigate FLS2 levels at the PM in exo70B1-1 and exo70B1-2 as well. Moreover, assessing pathogen growth in RIN4^T166A/r124b1 compared to r12 could provide insights. If the repressed PTI response in exo70B1-1 and exo70B1-2 results from reduced FLS2 rather than solely disrupted RIN4-Exo70B1 interaction, RIN4^T166A/r124b1 should phenocopy r12. Notably, phosphorylation at Thr-166 on RIN4 enhances its interaction with EXO70E2, another exocyst subunit involved in flg22-induced callose deposition (Redditt et al., 2019). This interaction inhibits callose deposition in the RIN4^T166D variant, suggesting that EXO70E2 may contribute to PTI suppression in stable transgenic lines expressing RIN4^T166D.

As shown in Figure 7 , we observed that r12b1 supported similar pathogen growth level as r124b1, and that transgenic expression of RIN4^T166D in the r124b1 background did not complement the enhanced bacterial growth of r124b1. However, expression of WT RIN4 and RIN4^T166A did complement ( Figure 7 ). This data provides further evidence that the proper PM localization of RIN4 is required for the execution of Exo70B1-related PTI responses, downstream of Exo70B1. Thus, we propose that RIN4 plays a critical role in the spatiotemporal control of complex immune signaling dynamics required for membrane trafficking and plant secretory pathways. In addition, the higher susceptibility of r12b1 to Pst hrcC^- and DC3000 than r124, indicating that Exo70B1 affects more factors in PTI process besides RIN4, requires further exploration.

In summary, the data presented here provide new insights into the mechanisms connecting effector-induced RIN4^T166 phosphorylation with PTI suppression. A central aspect of this mechanism involves the precise modulation of the interaction between RIN4 and Exo70B1. This study enhances our understanding of the spatiotemporal coordination between RIN4 and exocytosis in plant-pathogen interactions, shedding light on both conserved and unique functions of RIN4 in plant immunity.