Chitin perception in plasmodesmata identifies subcellular, context-specific immune signalling in plants

The plasma membrane (PM) that lines plasmodesmata has a distinct protein and lipid composition, underpinning specific regulation of these connections between cells. The plasmodesmal PM can integrate extracellular signals differently from the cellular PM, but it is not known how this specificity is established or how a single stimulus can trigger independent signalling cascades in neighbouring membrane domains. Here we have used the fungal elicitor chitin to investigate signal integration and responses at the plasmodesmal PM. We found that the plasmodesmal PM employs a receptor complex composed of the LysM receptors LYM2 and LYK4 which respectively change their location and interactions in response to chitin. Downstream, signalling is transmitted via a specific phosphorylation signature of an NADPH oxidase and localised callose synthesis that causes plasmodesmata closure. This demonstrates the plasmodesmal PM deploys both plasmodesmata-specific components and differential activation of PM-common components to independently integrate an immune signal.


96
Chitin-triggered plasmodesmata closure is dependent on LYK4 and LYK5 97 We previously identified that LYM2 is a GPI-anchored, LysM receptor protein that is 98 resident in the plasmodesmal PM . As LYM2 has no intracellular 99 domains we reasoned that it must interact with other proteins to initiate downstream signals  is deposited at plasmodesmata in response to flg22 (Xu et al., 2017). Therefore, we examined 134 callose deposition at plasmodesmata in response to chitin in Arabidopsis to determine if this 135 is common to pathogen-triggered plasmodesmata closure. We quantified aniline blue-stained   immunoprecipitated Citrine-LYM2 from membrane fractions with anti-GFP beads (Fig. 2B).

173
LYK4-HA was co-immunoprecipitated with LYM2 from both water and chitin treated tissue 174 suggesting that LYM2 associates with LYK4 in a chitin independent manner. The negative 175 control BRI1-RFP did not co-immunoprecipitate with Citrine-LYM2 (Fig. S3A).

176
Given that LYM2 and LYK4 are present in plasmodesmata we thought that these two 177 proteins could directly execute plasmodesmal PM chitin signalling. However, LYK5 is also 178 required for chitin-triggered plasmodesmata closure and thus, we tested the dependence of 179 the interaction between LYM2 and LYK4 on LYK5. For this we transformed Arabidopsis IPs. We observed that LYK4-RFP immunoprecipitated with Citrine-LYM2 in a chitin-182 independent manner from both Col-0 and lyk5-2 protoplasts. While this clearly demonstrates 183 that the interaction between LYM2 and LYK4 is LYK5 independent, these assays also 184 identified that LYK4-RFP appears approximately 10 kDa smaller on SDS-PAGE gels when 185 extracted from lyk5-2 protoplasts, suggesting that LYK4 is modified in a LYK5 dependent 186 manner. Further, following chitin treatment the dominant LYK4-RFP band detected in lyk5-2 187 protoplasts appears approximately 30kDa smaller than the dominant band in the WT 188 protoplasts suggesting that the 'non-modified' variant of LYK4-RFP in the lyk5-2 mutant is 189 subject to degradative processing in response to chitin. Given the genetic dependence of  and LYK5-RFP showed that LYK5-RFP co-immunoprecipitates with LYK4-GFP (Fig. 3A).

203
This suggests that LYK4 and LYK5 associate in the PM.     this assay, lym2-1 mutants were able to close their plasmdesmata in response to H2O2 258 demonstrating that any role for ROS signalling in chitin-triggered plasmodesmata closure 259 occurs independently, or downstream of LYM2 activity. We also further established that 260 transient expression of LYM2, LYK4 and LYK5 enhanced ROS production in N. 261 benthamiana leaf discs following chitin treatment, suggesting that each of these LysM 262 proteins can mediate ROS signalling.

263
The rapid production of ROS in response to chitin is associated with the NADPH Among the CPKs, CPK4, 5, 6 and 11 have been described to be involved in the 294 regulation of RBOHD-dependent ROS burst, and CPK6 has been shown to phosphorylate 295 both Ser133 and S347 (Kadota et al., 2014). Therefore, we tested the involvement of CPK6 296 in chitin-triggered plasmodesmata closure. Bombardment assays showed that two 297 independent cpk6 mutants, cpk6-1 and cpk6-2, were unable to close their plasmodesmata in 298 response to chitin (Fig. 5A) confirming CPK6 functions in plasmodesmal signalling.

299
LYM2 is a GPI anchored protein and therefore unlikely to directly interact with 300 intracellular CPKs. Having shown previously that LYM2 associates with LYK4 (Fig 2), we 301 investigated the interaction between CPK6 and LYK4. We generated a GFP-tagged   We found that in addition to LYM2, the RK LYK4 is also critical for plasmodesmal 346 PM chitin signalling. LYK4 associates with LYM2 and is detected in plasmodesmata but 347 does not accumulate at plasmodesmata in response to chitin. LYK4 associates with the RK 348 LYK5 in the PM but in response to chitin treatment we observed that the mobile fraction of 349 LYK4 increased and the association between LYK4 and LYK5 in the PM decreased. When X-X-S/T-X-B) while Ser163 is categorised as a motif 2 site (φ-X-B-X-X-S-X-X-X-φ) linking 407 chitin-triggered plasmodesmata closure to motif 1 sites. The significance of these motifs and 408 their relevance to plasmodesmal signalling is unknown; they may correlate to specific CPKs, 409 or to an independent mechanism for tuning RBOHD activity.    Table S2. Icorr is the corrected intensity, Inorm is the normalised intensity, Idecay is the modelled intensity 629 from the acquisition decay curve. transiently expressing the desired construct, were ground in liquid nitrogen to a fine powder.