Contrasting Effects of Cytoskeleton Disruption on Plasma Membrane Receptor Dynamics: Insights from Single-Molecule Analyses

cytoskeleton, actin filaments, microtubules , p lant membrane protein dynamics, 11 super-resolution microscopy, motion classification, sptPALM, NASTIC, DC-MSS 12 13

14 Traditional models such as the fluid mosaic model or the lipid raft hypothesis have shaped our 15 understanding of plasma membrane ( PM) organization. However, recent discoveries have extended 16 these paradigms by pointing to the existence of micro - and nanodomains. Here, we investigated the 17 role of the cytoskeleton in general and whether the picket fence model, established in animal cells, is 18 transferable to the plant cell system . By using single-particle tracking photoactivated localization 19 microscopy ( sptPALM) in combination with genetically encoded enzymatic tools for the targeted 20 disruption of the cytoskeleton, we studied the dynamics and nanoscale organization of a selection of 21 PM receptor-like kinases (RLKs) and receptor -like proteins (RLPs) . Our findings show that the 22 disintegration of actin filaments leads to decreased diffusion, more restrictive motion patterns , and 23 enlarged clusters, whereas the disintegration of microtubules results in increased diffusion, more 24 unconstrained diffusive behavior , and decreased cluster sizes of the tested RLKs and RLPs . These 25

underscore the potential unique regulatory functions of cytoskel eton components in plants 26 and suggest an altered mechanism compared to the picket fence model of the animal cell system. Our 27 qualitative data can serve as the foundation for further investigation s aimed at developing a 28 comprehensive and refined model of protein dynamics and organization in plant cells. 29 30 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint

31 The plasma membrane (PM), together with the cell wall, functions as the first selective barrier 32 between the cell and the environment ( Gronnier et al., 2018; Jaillais and Ott, 2020). Singer and 33 Nicolson (1972) emphasized the fundamental significance of biological membranes and their 34 organization in general and proposed the widely known fluid mosaic model. Their assumption was 35 that proteins can laterally diffuse within the membrane without major restrictions. However, this 36 would result in uniformly distributed membrane-embedded proteins such as receptors, independent 37 of the entire PM proteome, which is unequivocal ly not the case. One significant expansion of the 38 model was the in troduction of the “lipid raft” hypothesis, which suggests that lipid rafts, containing 39 high levels of cholesterol and sphingolipids, serve as platforms with a high molecular order of proteins 40 and lipids. These platforms facilitate , for instance, selective interactions between signaling proteins 41 and effector molecules (Simons and Ikonen, 1997). The hypothesis found support in results from the 42 mammalian field, which indicated a binary characteristic of membranes that are partitioned in to 43 detergent-resistant and detergent-sensitive fractions (Brown and Rose, 1992; Yu et al., 1973). Similar 44 research was later conducted in plants, delivering comparable results (Borner et al., 2005; Laloi et al., 45 2007; Lefebvre et al., 2007; Mongrand et al., 2004; Morel et al., 2006). These studies showed that the 46 detergent-resistant membrane protein profile is distinct from that of the whole PM. However, the 47 isolation of detergent-resistant fractions and the detergents used may cause changes in the PM itself. 48 Moreover, results derived from the application of newer methods, such as fluorescent microscopy 49 techniques, raised the question of whether detergent-resistant membrane areas indeed define 50 functional membrane rafts (Kusumi et al., 2005 ; Raffaele et al., 2009 ; Tanner et al., 2011 ). Although 51 the existence of “lipid rafts” in plants, now referred to as “micro- or nanodomains”, is undoubtedly 52 a ccepted, their dynamics, organization and regulation still require further research. 53 Besides the plant-specific family of remorin proteins serving as scaffolding factors ( Jarsch and Ott, 54 2011; Raffaele et al., 2009), the asymmetric localization and order of lipids within the PM (Gronnier et 55 al., 2018), as well as the cell wall-PM continuum (Martiniere et al., 2012), the cytoskeleton is believed 56 to play a vital regulatory role in the organization of plant PMs (Jaillais and Ott, 2020). This hypothesis 57 is based on the so -called picket fence model from the mammalian field (Kusumi et al., 2012 ). In this 58 model, the cortical cytoskeleton, composed of actin filaments in mammalian cells, defines membrane 59 domains by acting as a fence that restricts the lateral diffusion of lipids and proteins within these 60 domains. The model postulates additional pickets, which are represented by transmembrane proteins 61 that are anchored either by the cytoskeleton in the cytosol or the extracellular matrix. However, it is 62 important to note that the model, as being based on animal cells, does not include specific properties 63 of plants, such as the existence of cortical microtubules that c ould act as an additional fence (Jaillais 64 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint and Ott, 2020 ). McKenna et al. (2019) demonstrated an impact of actin and microtubules on the 65 diffusion of some but not all proteins in the PM. Moreover, additional studies revealed a change in or 66 loss of nanodomain organization of proteins after cytoskeleton disruption, again for some but not for 67 all tested proteins (Bücherl et al., 2017; Danek et al., 2020; Jarsch et al., 2014; Konrad et al., 2014; Lv 68 et al., 2017; Raffaele et al., 2008; Szymanski et al., 2015 ). Chemicals such as latrunculin or oryzalin 69 were used in these studies to disintegrate either the actin or the microtubule cytoskeleton. However, 70 these chemicals are difficult to fine-tune in terms of their tissue-specific activity and concentration. In 71 contrast, we utilized recently developed genetically encoded, enzymatic tools that mimic the function 72 of those chemicals. Specifically , we employed the Salmonella enterica effector SpvB for actin 73 cytoskeleton disintegration (Harterink et al., 2017 ; Vilches Barro et al., 2019) and a 74 phosphatase-inactive variant of the atypical tubulin kinase PROPYZAMIDE-HYPERSENSITIVE 75 1 (PHS1ΔP) for microtubule cytoskeleton disintegration (Fujita et al., 2013; Vilches Barro et al., 2019). 76 To address the question of how the cytoskeleton may interfere with the dynamics and organization of 77 plant PM proteins, we focused on a group of proteins associated with signaling mechanisms , namely 78 the receptor -like kinases (RLKs) BRASSINOSTEROID INSENSITIVE 1 (BRI1), PHYTOSULFOKIN 79 RECEPTOR 1 (PSKR1), FLAGELLIN-SENSITIVE 2 (FLS2) and BRI1-ASSOCIATED RECEPTOR KINASE (BAK1), 80 along with the RECEPTOR-LIKE PROTEIN 44 (RLP44). RLP44 is proposed to be a cell wall integrity sensor 81 that controls cell wall homeostasis through an interplay with BRI1 and its co-receptor BAK1. The three 82 proteins form a ternary receptor complex in the PM of plant cells (Glöckner et al., 2022). Additionally, 83 RLP44 has been associated with phytosulfokine signaling, as it forms a complex with the corresponding 84 P SKR1 receptor and its co-receptor BAK1 (Garnelo Gomez et al., 2021; Holzwart et al., 2018). FLS2 was 85 chosen as a protein that is not connected to RLP44. Although FLS2 interacts with BAK1, it has been 86 demonstrated that th is interaction takes place at a minimum distance of 11.1 nm from the 87 BRI1-BAK1-RLP44 complexes (Glöckner et al., 2022). 88 We used SpvB and PHS1 ΔP to analyze the cytoskeleton influence on the dynamic properties of the 89 four above-mentioned PM receptors using single- particle tracking with photoactivated localization 90 microscopy (sptPALM) in a transient Nicotiana benthamiana system. T o gain insight into the 91 underlying mechanisms, we focused our investigation on three key parameters: (i) The diffusion 92 coefficients, (ii) the organization of the respective receptors into nanoscale -like protein clusters, and 93 (iii) the classification of molecular movement into transient movement types. It is worth mentioning 94 that the three parameters are not inherently linked, for example, a reduced diffusion coefficient does 95 not necessarily result in larger nanoscale protein clusters and/or immobile movement patterns. 96 Consequently, their separate evaluation can give insights into different regulatory mechanisms. 97 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint We were able to show a clear link between manipulated microtubule formation, reflected by an 98 increase in the diffusion coefficient for most of the tested proteins. Conversely, disintegration of actin 99 filaments predominantly led to reduc ed diffusion coefficients. W e furthermore demonstrated that 100 destroyed microtubules predominantly led to decreased cluster sizes of the analyzed proteins clusters, 101 while actin destruction resulted in increased cluster sizes. 102 Additionally, the classification of the protein tracks into the motion types (i) free diffusion, (ii) confined 103 diffusion, (iii) immobility and (iv) directed diffusion revealed another influence of the cytoskeleton. 104 Our data show that proteins spend more time in free diffusion states in the absence of microtubules. 105 Conversely, disintegration of actin filaments resulted in an overall more confined behavior. 106 The opposing effects on diffusion, cluster sizes and motion patterns demonstrate the potentially 107 unique regulatory functions of cytoskeletal components in plants and suggest that the picket fence 108 model is not directly transferable to plant systems. However, our research may provide a basis for 109 further investigation to translate these findings into an extended or revised functional model. 110

111 Genetically encoded, enzymatic tools for cytoskeleton disruption are functional 112 Initially, SpvB and PHS1 ΔP (Vilches Barro et al., 20 19) were modified for our needs to be applied in 113 sptPALM experiments. To avoid potential compatibility problems with the used sptPALM 114 fluorophores, we decided to generate versions of the genetically encoded enzymatic tools without 115 fluorescent tags but instead with a hemagglutinin (HA) tag. In addition, the expression of SpvB and 116 PHS1ΔP was designed to be under the control of constitutive promoter s (Figure 1C and F). The 117 modified protein tools were tested for their ability to destroy the integrity of the cytoskeleton by 118 co-expressing them with corresponding marker proteins in N. benthamiana. To label actin filaments, 119 we used actin-binding domain 2 (ABD2) of Arabidopsis fimbrin 1 fused to GFP at the C - and 120 N- terminus (Wang et al., 2008). To test the state of the microtubules, we applied the MICROTUBULE-121 ASSOCIATED PROTEIN 65 -8 (MAP65-8) fused to RFP, which is known to bind cortical 122 microtubules (Smertenko et al., 2008 ). Epidermal cells of N. benthamiana leaves were investigated 123 under the confocal microscope three days post infiltration (dpi) with the constructs -containing 124 Agrobacteria. The clear actin filament and microtubule cytoskeleton disassembly was observed in the 125 pr esence of the corresponding genetically encoded, enzymatic tool (Figure 1B and E), proving their 126 applicability as disintegration tools in epidermal N. benthamiana leaf cells. 127 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Disintegration of actin filaments predominantly leads to reduced protein mobility in 128 the PM 129 To determine the influence of the actin cytoskeleton on the dynamics and the nanoscale organization 130 of RLP44, BRI1, PSKR1, FLS2 and BAK1, we first expressed corresponding mEos3.2-tagged versions 131 under the control of their native promoter in the absence or presence of HA-SpvB in N. benthamiana 132 epidermal leaf cells. Subsequently, we used sptPALM and used the recently introduced OneFlowTraX 133 software package for the analysis of protein dynamics and complex organization (Rohr et al., 2024 ). 134 Independent of the co-expression with HA-SpvB, some general findings are worth mention ing: While 135 the mEos3.2 fusions of RLP44, BRI1, PSKR1 and FLS2 showed one population of mobility each , 136 BAK1-mEos3.2 presents a more confined and a more mobile variety ( Figure 2A). The diffusion 137 coefficients of RLP44-mEos3.2, BRI1-mEos3.2, PSKR1-mEos3.2 and FLS2-mEos3.2, as well as the one 138 of the slower population of BAK1-mEos3.2, were comparable with other confined receptor proteins 139 such as the PM intrinsic protein (PIP) 2;1 (D = 0,0047 µm²/s) reported before (Bayle et al., 2021; Hosy 140 et al., 2015) . In contrast, the more mobile variety of BAK1 -mEos3.2 showed a diffusion coefficient 141 about ten times higher than the other proteins (Figure 2B). 142 In the presence of HA-SpvB, we observed a significant reduction of the diffusion coefficient for 143 RLP44-mEos3.2, PSKR1 -mEos3.2, BRI1 -mEos3.2 and the mobile population of BAK1 -Eos3.2. In 144 contrast, actin disruption did not affect FLS2-mEos3.2 dynamics. Moreover, the less mobile population 145 of BAK1-mEos3.2 tended to have an even more reduced mobility. Additionally, the bimodal mobility 146 distribution of BAK1 -mEos3.2 shifted overall when the actin cytoskeleton was des troyed with 147 HA-SpvB, increasing the fraction of the more restricted variant of BAK1-mEos3.2 compared to control 148 cells. 149 Disintegration of the microtubule cytoskeleton predominantly leads to enhanced 150 protein mobility in the PM 151 While microtubules arise from centrosomes in animal cells , plants possess cortical 152 microtubules (Farquharson, 2009). Their presence needs to be considered when a putative effect of 153 the “membrane skeleton” on protein dynamics is investigated. Therefore, the identical set of RLK 154 fusion proteins used for the actin approach (see above) were co -expressed with PHS1 ΔP-HA, which 155 specifically causes depolymerization of cortical microtubules (Fujita et al., 2013) . While the 156 mEos3.2-fusions of RLP44, BRI1, PSKR1 and FLS2 displayed one population for the diffusion coefficient, 157 BAK1-mEos3.2 again showed two populations of different mobility, independent of the absence or 158 presence of PHS1 ΔP-HA (Figure 3A). In contrast to the data obtained after the destruction of actin 159 filaments, the manipulation of cortical microtubules caused an increased diffusion of RLP44-mEos3.2 160 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint and PSKR1 -mEos3.2. However , t he mobility of BRI1-mEos3.2 and FLS2-mEos3.2 remained nearly 161 unaffected in the presence of PHS1Δ P-HA (although a trend of increased diffusion was present 162 for BRI1-mEos3.2). For BAK1-mEos3.2 a significant increase in the mobility was solely observed for the 163 less mobile population (Figure 3B). Additionally, in the absence of polymerized microtubules , there 164 was a slight shift in the BAK1 distribution between the two populations: More BAK1 -mEos3.2 was 165 present in the faster fraction and less in the slower one compared to the measurements where 166 PHS1ΔP-HA was not present. 167 The nanoscale organization of most PM proteins is changed by cytoskeleton 168 disintegration 169 By manipulating the cytoskeleton structures, it is of great interest to know whether the nanoscale 170 organization of protein clusters, commonly referred to as nanodomains, is affected as well (Jaillais and 171 Ott, 2020; McKenna et al., 2019). To determine the cluster properties on the basis of spt data, several 172 algorithms are available, such as Voronoi tessellation (Andronov et al., 2016; Levet et al., 2015 ), 173 density-based spatial clustering of applications with noise (DBSCAN) (Ester et al., 1996) and the 174 recently introduced nanoscale spatiotemporal indexing clustering (NASTIC) (Wallis et al., 2023 ). For 175 our studies, we decided to apply the NASTIC algorithm since its approach is the least influenced by 176 user-defined parameters and exclusively deals with track data as a whole (Rohr et al., 2024; Wallis et 177 al., 2023). However, it is, like all the other algorithms, partly influenced by the set parameters and the 178 raw data quality . Thus, we want to emphasize that only the relative changes in the nanoscale 179 organization of clusters should be considered rather than absolute values. 180 Applying the NASTIC algorithm (see Material and Methods ) to our tracking data in the absence or 181 presence of actin-disintegrating SpvB, a predominantly decreased cluster diameter was observed for 182 all tested PM proteins, with the exception of RLP44 -mEos3.2 and FLS2 -mEos3.2 (Figure 4A ). In 183 contrast, improper polymerization of cortical microtubules predominantly led to enlarged cluster 184 diameters for all tested protein fusions, except for FLS2-mEos3.2 , where the cluster siz e remained 185 unaffected (Figure 4B). 186 In summary, the destruction of the actin filaments predominantly led to reduced mobility and smaller 187 clusters, while the disintegration of the microtubule cytoskeleton resulted in higher mobility and 188 enlarged clusters of most of the tested PM proteins. 189 190 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Classification of motion behavior reveals changes upon the disintegration of 191 cytoskeleton components 192 In typical sptPALM experiments, the so -called short -range diffusion coefficient is extracted from 193 molecular trajectories to obtain a value that is as independent as possible of directional motion, 194 obstacles and boundaries (Saxton, 1997). Consequently, transient interactions of a diffusing protein 195 with structural elements such as the cytoskeleton are not necessarily detectable by the diffusion 196 coefficient alone. However , the long- term temporal evolution of these trajectories can be used to 197 classify the type of movement (e.g., confined, directed, or free movement). For this purpose, several 198 freely available software solutions and algorithms can be used (Das et al., 2009; Helmuth et al., 2007; 199 Persson et al., 2013; Wagner et al., 2017). 200 Vega et al. (2018) introduced an efficient transient mobility analysis framework called 201 "divide and conquer moment scaling spectrum" (DC-MSS), which was used to analyze the 202 spatiotemporal organization of cell surface receptors and proved to be a suitable tool to analyze our 203 data as well. 204 We focused on four major motion types of proteins, which are commonly used for studying biological 205 systems. Proteins can (i) diffuse freely, for example in large unilamellar vesicles and in membrane 206 blebs (Jaqaman and Grinstein, 2012 ; Kusumi et al., 20 05), (ii) or become confined within structural 207 corrals (Fujiwara et al., 2016). Furthermore, they can get (iii) anchored or immobilized when they bind 208 to static components (Komura et al., 2016 ) or (iv) exhibit directed motion, for example, by transport 209 processes via cytoskeleton components (Serge et al., 2003 ). Importantly, a protein may switch 210 between different motion types during its lifespan. This was also suggested previously for plant PM 211 proteins, such as PIP2,1 (Li et al., 2011). 212 In our analysis pipeline, the motion type classification was based on the same track data that were 213 used before to evaluate the diffusion coefficients and cluster sizes. Individual tracks were classified 214 with DC-MSS, and motion types were assigned either to entire tracks or track segments if the protein 215 changed its motion type during the recording time. This allowed for an overall analysis of the relative 216 time proteins spent in the following states: immobility, confined diffusion, free diffusion, and directed 217 diffusion (Figure 5). A negligible number of tracks or track segments could not be classified (< 0.5 %) 218 and were therefore excluded from the calculations. 219 Under control conditions, RLP44-mEos3.2 mostly exhibited confined (49 %) and free behavior (37 %). 220 In the presence of SpvB, i.e., in the absence of actin filaments, there was a decrease in the free 221 diffusive behavior (- 11 %) while immobile and confined movement increased (+ 5 % and 222 + 6 %, respectively). A comparable effect was observable for PSKR1-mEos3.2 : C onfined behavior 223 represented the majority (50 %) under control conditions, and the disintegration of actin filaments by 224 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint SpvB led to a shift in the motion patterns: The free diffusive motions decreased ( - 18 %) while 225 immobility and confined movement increased (+ 12 % and + 6 % respectively). These observations 226 hold true for BRI1 -mEos3.2 as well, although with less pronounced shifts as compared to 227 RLP44 mEos3.2 and PSKR1 -mEos3.2. In contrast, the consequences of actin disintegration on 228 FLS2-mEos3.2 were contrary to those of the mEos3.2 -fusions of BRI1, PSKR1 and RLP44 . Here, a 229 decrease in the immobile proportion (- 11 %) and an increase in the free diffusive behavior (+ 9 %) 230 were detectable. Finally, BAK1 -mEos3.2 exhibited a minor effect in response to actin filament 231 disintegration, with shifts smaller than 2 %, suggesting a less eminent role of actin filaments on the 232 motion behavior of BAK1 (Figure 5A and B). 233 As shown for the analysis of the diffusion coefficient and the cluster sizes, the disintegration of actin 234 filaments and microtubules had contrary effects. This was also the case when evaluating the motion 235 patterns. 236 In the presence of intact microtubules , RLP44-mEos3.2 mostly exhibited free diffusive or confined 237 behavior (approx. 44 % each). However, with disintegrated microtubules, RLP44-mEos3.2 spent more 238 time in a free diffusive state (+ 12 %) while confined and immobile behavior was less present (- 8 % 239 and - 4 %, respectively). A comparable trend was observable for PSKR1-mEos3.2, where, under control 240 conditions, confined motions represented the majority with 50 % followed by free diffusion (27 %). 241 Again, with the co -expression of PHS1ΔP, there was in an increase in free diffusive behavior (+ 8 %), 242 while immobile (- 5 %) and confined states (- 3 %) decreased. The relative changes for FLS2 -mEos3.2 243 showed a decrease of immobil ity (- 5 %) in the absence of microtubules , a slight increase in the free 244 diffusive proportion (+ 3 %) and a negligible effect on the confined behavior (+ 2 %). Compared to the 245 strength of the shifts that RL P44-, PSKR1 - and FLS2 -mEos3.2 showed, BRI1’s changes were less 246 pronounced, indicating a minor role of microtubules on the motion behavior of BRI1. This holds true 247 as well for BAK1 -mEos3.2, which exhibit s shifts in the motion patterns below 1 % for all mobility 248 classes (Figure 5C and D). All studied protein did not display substantial directed diffusive behavior, 249 which suggests that the cytoskeleton components may not play a major role in transport 250 processes (Figure 5). 251 In summary, we evaluated three parameters that gave insights into the organizational function of the 252 cytoskeleton on PM proteins. Interestingly, diffusion coefficients, cluster sizes and motion classes 253 were influenced in an opposite manner by the disintegration of actin filaments and microtubules. 254 While the disintegration of actin filaments resulted predominantly in decreased diffusion but bigger 255 clusters, the absence of microtubules increased the diffusion and led to smaller clusters. Concerning 256 the motion classes, actin disruption predominantly caused a shift to more immobility and confined 257 movement, while free diffusive behavior was reduced. Contrar ily, the disintegration of microtubules 258 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint resulted in increased free diffusive behavior, while immobility and the time proteins spent in confined 259 states decreased. For BRI1 and BAK1, a less pronounced effect of the cytoskeleton on the motion 260 behavior can be assumed, as the shifts between control conditions and the absence of actin filaments 261 or microtubules were relatively low compared to the other PM proteins. 262

263 The PM plays a vital role for cell properties and a variety of biological processes. Its organization and 264 associated functions have been the subject of several theories and considerations (Simons and Ikonen, 265 1997; Singer and Nicolson, 1972). Among them, the picket fence hypothesis is one of the most recent 266 models that was initially studied in animal cells. Mainly due to the dynamic nature of the 267 submembrane actin meshwork, no tractable experimental model for the mechanistic investigation of 268 the fence or picket model is available. However, recently it was shown that actin rings in neurons 269 compartmentalize the PM, acting as fences and confining membrane proteins (Rentsch et al., 2024). 270 The direct transfer of the picket fence model from animal to plant cell biology is challenging since 271 plant-specific properties need to be considered, such as the presence of cortical 272 microtubules (Farquharson, 2009). Nevertheless, the model has also been tested and discussed as 273 important for the dynamics and organization of plant PM protein dynamics and organization in several 274 research articles and reviews (Jaillais and Ott, 2020; Martiniere et al., 2012; McKenna et al., 2019). 275 By applying genetically encoded enzymatic tools, we investigated the influence of the actin and 276 microtubule cytoskeleton disintegration on the dynamics (i.e., via the diffusion coefficient and motion 277 classification) and nanoscale organization of selected plant PM integral protein fusions, namely 278 RLP44-mEos3.2, BRI1 -mEos3.2, PSKR1 -mEos3.2, FLS2 -mEos3.2 and BAK1 -mEos3.2, in the tobacco 279 epidermal leaf cell system (Fujita et al., 2013; Harterink et al., 2017; Vilches Barro et al., 2019). 280 After the disintegration of the actin cytoskeleton by SpvB, w e observed a decreased diffusion 281 coefficient for all proteins , except for FLS2-mEos3.2 . Furthermore, we exclusively detected two 282 protein populations with different mobilities for BAK1-mEos3.2. The presence of such subpopulations 283 usually indicates diverse molecular states of the protein . For instance, the subpopulation with lower 284 m obility could be bound to cellular structures or other molecules, while the second one moves more 285 freely. This may subsequently provide information about the biological roles of the protein (Hansen 286 et al., 2018). Here, we speculate that the slow subpopulation of BAK1 -mEos3.2 may be involved in 287 signaling processes in restricted nanodomains, while the faster subpopulation is freely available for 288 possible other interaction partners. Interestingly , the two subpopulations responded differently to 289 actin filament disintegration via SpvB. Whereas no significant change was observed for the 290 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint slower-moving BAK1-mEos3.2 population, the faster one showed a decreased diffusion comparable 291 to the one of the other analyzed fusion proteins. 292 Furthermore, we observed that in the presence of intact actin filaments, the overall mobility 293 distribution of BAK1 mEos3.2 features a higher fraction of the more mobile variant. The destruction 294 of actin increases the relative amount of slower BAK1 -mEos3.2, resulting in comparably equal 295 amounts for both populations. Evidently, the functional meaning of these changes requires future 296 experiments. 297 We additionally observed an increase in the cluster sizes for all tested proteins , except for 298 RLP44-mEos3.2, after SpvB-mediated actin filament disintegration. This observation is in line with the 299 picket fence model. In addition, the results can be integrated into a broader context: Recently, it was 300 shown that the chemical destruction of actin leads to an increase in salicylic acid (SA) levels and the 301 activation of SA -responsive genes in A. thaliana (Kalachova et al., 2019 ; Leontovycova et al., 2019 ; 302 Matouskova et al., 2014 ). Moreover, the external application of SA constrained the diffusion of the 303 PM auxin efflux carrier PIN-FORMED 2 (PIN2), followed by its condensation into PIN2 hyperclusters; a 304 process mediated by remorins (Ke et al., 2021 ). We observed the same phenomenon for our fusion 305 proteins: r educed diffusion with enlarged clusters. Whether these changes in the dynamics and 306 nanoscale organization are related to SA requires further study. 307 However, so far , none of the available studies have shown a decrease in PM protein diffusion as a 308 direct consequence of actin cytoskeleton disintegration (Hosy et al., 2015; McKenna et al., 2019). 309 For BRI1 and RLP44, no experiments have been published yet that tested the influence of the actin 310 cytoskeleton on their dynamics. Interestingly, according to Lanza et al. (2012) , brassinosteroid 311 application modifies the actin cytoskeleton in an auxin-like manner by unbundling actin filaments. This 312 implicates a contribution of BR signaling to the reorganization of the actin cytoskeleton. Moreover, a 313 link between BR signaling and the actin cytoskeleton is further substantiated by the observation that 314 the root -waving phenotype of the Arabidopsis act2-5 mutant copies that of wildtype Arabidopsis 315 s eedlings treated with brassinolide (Lanza et al., 2012). 316 In contrast, the application of flg22 , being the ligand of FLS2, increased the formation of actin 317 filaments within three hours. This process is proposed to be the result of a well-orchestrated signaling 318 cascade that (i) triggers the local high-order assembly of remorins that (ii) recruit formins, components 319 comparable to mammalian integrins, which finally (iii) results in increased actin polymerization (Ma et 320 al., 2021; Ma et al., 2022; Ma et al., 2023) 321 A long-standing paradigm that was still recently under debate was the direct involvement of actin 322 during clathrin-mediated endocytosis. For example, it has been reported that FLS2 internalization 323 depends on actin directly (Beck et al., 2012). However, Narasimhan et al. (2020) showed that actin is 324 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint neither involved in membrane bending and scission nor in the initiation of endocytic processes. 325 Rather, it is only required to transport endocytic vesicles after vesicle scission . Although this 326 perspective seems to be acknowledged in the scientific community now (Kraus et al., 2024) , current 327 publications still refer to the old paradigm (Lu et al., 2023), stifling the debate. Whether the changed 328 mobility and nanoscale organization of our tested proteins are a result of disturbed transport after 329 vesicle scission due to actin filament disintegration requires further investigation. 330 In contrast to our results, McKenna et al. (2019) reported that the mobility of FLS2 -GFP in epidermal 331 hypocotyl cells of A. thaliana is enhanced after latrunculin treatment. However, we used SpvB, a 332 bacterial effector from Salmonella enterica, as an actin disintegration tool. Although both latrunculin 333 and SpvB are thought to specifically address actin, their modes of action differ: While SpvB acts almost 334 exclusively on G actin by ADP-ribosylation (Hochmann et al., 2006), more recent publications indicate 335 that latrunculin functions via direct binding to G actin (Spector et al., 1999 ) and can also affect 336 F actin (Fujiwara et al., 2018 ). In addition, tool -specific, pleiotropic effects on cell physiology cannot 337 be excluded. Both different modes of action and tool-specific site effects may have an impact on FLS2 338 dynamics. Another aspect that should be considered is that our experimental setup investigates 339 proteins from A. thaliana in the heterologous cell environment of N. benthamiana. Therefore, the 340 transfer of our results to cells of transgenic A. thaliana plants is not possible on a one-to -one basis, 341 because the physiological, cellular and biochemical contexts of the N. benthamiana and A. thaliana 342 cells are certainly not identical. 343 Besides studying the dynamics of the tested proteins via the diffusion coefficient, we analyzed their 344 motion classes, too. We observed that all proteins show predominantly confined behavior under 345 control conditions as well as in situations where the actin filaments were disintegrated. However, 346 there was a substantial shift in the proportions of the motion classes for some but not all fusion 347 proteins. In the absence of actin filaments, RLP44-mEos3.2 showed increased immobility and confined 348 behavior, while free motion was reduced. The same holds true for PSKR1-mEos3.2 and BRI1-mEos3.2, 349 although the effect on BRI1 - mEos3.2 was less pronounced. In contrast , FLS2-mEos3.2 showed less 350 immobile behavior but increased free motion in the absence of actin filaments. BAK1 -mEos3.2 was 351 nearly unaffected in its diffusive behavior when actin was disrupted. 352 In consideration of the combined results of the actin disintegration within the framework of the picket 353 fence model, a direct transfer from the animal to the plant cell system does not appear to be possible 354 without the inclusion of further aspects. 355 Due to the lack of actin filaments, an increase in the diffusion coefficient accompanied by enlarged 356 nano-sized cluster structures and less confined and/or immobile behavior is expected (Fujiwara et al., 357 2002; Fujiwara et al., 2016; Jaqaman and Grinstein, 2012 ). However, our data prompt for a more 358 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint sophisticated mechanism, as exclusively FLS2’ s increase in cluster sizes and motion behavior in the 359 absence of actin fits into the picket fence model. As actin is involved in a variety of processes, such as 360 cell growth, cell division, cytokinesis, and various intracellular trafficking events (Szymanski and 361 Staiger, 2018 ), pleiotropic effects cannot be excluded that are not linked to the role of actin in 362 membrane organization. Additionally, in the absence of actin, a compensatory cytoskeletal interaction 363 could take place through increased microtubule associations that would trap the proteins in confined 364 corrals. Thus, a combined disruption of actin and microtubules will be of special interest in the future. 365 So far, in the presented study, we focused on individual manipulations. 366 Through the disintegration of microtubules via PHS1ΔP (Fujita et al., 2013; Vilches Barro et al., 2019), 367 we observed a contrary effect compared to that of the actin cytoskeleton disturbance: With the 368 exception of BRI1 -mEos3.2 and FLS2 -mEos3.2, a significant increase in the diffusion coefficient was 369 observed for the tested fusion proteins. Furthermore, reduced cluster sizes were observed for all 370 tested receptors, again apart from FLS2-mEos3.2. Interestingly , for BAK1-mEos3.2 , the significantly 371 changed diffusion coefficient was this time observable for the slower subpopulation, while the faster 372 one remained unaffected. This suggests a stronger connection between microtubules and the slower 373 subpopulation of BAK1 than for the faster subpopulation. 374 Additionally, we analyzed the motion patterns of the fusion proteins. In the absence of microtubules, 375 RLP44-mEos3.2, PSKR1-mEos3.2 and FLS2-mEos3.2 showed more free diffusive behavior, while BRI1-376 mEos3.2 and BAK1-mEos3.2 were barely affected. 377 In particular, the increased diffusion coefficients after microtubule destruction as well as the shift to 378 more mobile motion patterns in combination with decrease d immobility and confined motion are in 379 good agreement with the picket fence model. 380 However, considering that the absence of physical barriers in the form of microtubules enables a less 381 restricted motion of the fusion proteins, the decrease in cluster sizes after microtubule disintegration 382 is difficult to interpret . Again, this change in nanoscale organization might be an indirect pleiotropic 383 effect, as microtubules participate in a variety of processes in plant cells, such as the guidance of the 384 cellulose synthase complexes to the PM (Paredez et al., 2006) or the maintenance of pavement cell 385 morphogenesis (B elteton et al., 2018). Thus, alterations in the microtubule cytoskeleton organization 386 may change the properties at the cell wall-PM interface that interfere with the dimension of the PM 387 protein clusters. 388 Additionally, McKenna et al. (2019) showed that FLS2-GFP exhibits an enhanced diffusion coefficient 389 after the disintegration of microtubule s by oryzalin. In general, these observations are confirmed by 390 our data, although in our case, the increase was less pronounced. 391 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint In summary, we studied three parameters that changed upon the disintegration of actin filaments and 392 microtubules. Interestingly, the manipulated cytoskeleton components influenced the studied 393 proteins in a contrary manner and resulted in distinct trends for diffusion coefficients, cluster sizes 394 and motion patter ns. While some aspects of either the actin or microtubule disruption match the 395 picket fence model, others do not. Consequently, the model cannot be directly transferred from the 396 animal field to the plant cell system. We conclude that this might be caused by the regulatory and 397 functional responsibility of two cortical components, namely actin and microtubules, instead of 398 cortical actin exclusively in animal cells. To further unravel a potential compensatory effect, the 399 depletion of both structures will be a main task in the future. Additionally, studies in A. thaliana with 400 inducible manipulations of the cytoskeleton will enable more background -free observation in the 401 native organism. 402 We also started integrating our experimental data into a computational model generated by Smoldyn, 403 a particle-based spatial simulation software (Andrews, 2009). Although modeling approaches have the 404 potential to advance plant science to a great extent, the biggest challenge is still its underutilization in 405 plant biology and thus the lack of comparative approaches (Dale et al., 2021). Recently, we predicted 406 a new component of the fast brassinosteroid signaling pathway by computational modeling and 407 emphasized its strengths in combination with wetlab experiments (Grosseholz et al., 2022). Based on 408 our experience, we believe that the modeling of spatiotemporal dynamics may also reveal so -far 409 hidden aspects in its regulation by the cytoskeleton. 410 411

412 Plasmid construction 413 The genetically encoded enzymatic tools for the cytoskeleton manipulation, namely SpvB and PHS1ΔP, 414 were provided by Prof. Alexis Maizel (COS Heidelberg) (Vilches Barro et al., 2019). For the generation 415 of the expression constructs , the desired plasmid DNA was first amplified by PCR and then either a 416 BP reaction into pDONR207 (for SpvB) or a blunt -end cloning reaction into pJET1.2 (Thermo Fisher 417 Scientific) (for PHS1ΔP) was performed. Subsequently, an LR reaction was performed according to the 418 manufacturer’s manual for SpvB into pEG201 (Earley et al., 2006 ) and cut-ligations with the needed 419 Level I constructs into BB10 (Binder et al., 2014 ) were executed for PHS1 ΔP. While SpvB cloning 420 resulted in an N-terminally HA-tagged version under the control of the 35S promoter, PHS1 ΔP is 421 controlled by a 2x 35Sω promoter and C -terminally HA-tagged (Figure 1C and F). The BRI1-mEos3.2 422 a nd RLP44 -mEos3.2 constructs are described in Ro hr et al. (2 024) and the construction of 423 PSKR1-mEos3.2 was performed according to their protocol. BAK1-mEos3.2 and FLS2 -mEos3.2 were 424 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint provided by Dr. Birgit Kemmerling (ZMBP, Tübingen). The cytoskeleton marker s GFP-ABD2-GFP and 425 MAP65-8-RFP were kindly provided by Dr. Pantelis Livanos (FAU Erlangen). 426 Plant material and growth conditions 427 All experiments conducted in this study were performed in transiently transformed N. benthamiana 428 plants, cultivated under controlled greenhouse conditions. The desired proteins were transiently 429 expressed using the AGL1 Agrobacterium tumefaciens strain (Lifeasible), as previously 430 described (Hecker et al., 2015; Ladwig et al., 2015), without the washing step with sterile water. The 431 plants were infiltrated with the respective constructs at an OD600 of 0.1 in a ratio of 1:1 or 1:1:1 with 432 the silencing inhibitor p19. After watering, the plants were kept in ambient conditions and were 433 imaged three days after infiltration. 434 Confocal imaging 435 To confirm the functionality of the genetically encoded enzymatic tools for the cytoskeleton 436 manipulation, the constructs were co -expressed with corresponding markers, namely 437 GFP-ABD2-GFP (for actin) and MAP65-8-RFP (for microtubules). Subsequently, their localization inside 438 epidermal leaf cells of N. benthamiana was investigated using confocal laser scanning microscopy on 439 a SP8 laser scanning microscope (Leica Microsystems GmbH) with HyD detectors and a HC PL APOCS2 440 63 x/1.20 WATER objective three days post infiltration. For detection of the GFP signal, a 488 nm argon 441 laser was used. The detection range was set to 500 nm – 550 nm. The images in Figure 1 are maximum 442 projections that covered a z range of ~ 15 µm obtained with a step size of 1 µm. The resulting images 443 were processed with the help of the Leica Application Suite X (Version 3.3.0.16x). The detection of the 444 RFP signal was performed as described above, using a detection range from 600 nm – 650 nm and a 445 561 nm diode pumped solid state laser. 446 Sample preparation and movie acquisition for sptPALM measurements 447 All sptPALM measurements with transiently transformed N. benthamiana were performed three days 448 post infiltration. For the acquisition, a small leaf area was cut out, excluding veins , and was placed 449 between two coverslips (Epredia 24x50 mm #1 or equivalent) with a drop of water. The 450 “coverslip sandwich” was then placed on the specimen stage, lightly weighted down by a brass ring to 451 help flatten uneven cell layers. The composition of the custom-built microscope platform is described 452 in detail in Rohr et al. ( 2024). For our purposes , the following filters for mEos3.2 were inserted into 453 the emission beam path: 568 LP Edge Basic Longpass Filter, 584/40 ET Bandpass. The excitation power 454 a rriving at the sample was quantified (PM100D with S120C, Thorlabs) in epifluorescence mode after 455 the objective to maintain consistency across experimental sets. Photoconversion of mEos3.2 was 456 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint executed by applying moderate low intensities using 405 nm excitation. The signal of the red version 457 of mEos3.2 was then obtained by excitation at 561 nm with 1800 µW. The magnification of the optical 458 system was adjusted so that the length of one camera pixel corresponds to 100 nm in the sample 459 plane. To identify feasible regions, larger areas of 51.2 x 51.2 µm were utilized by adjusting the focal 460 plane and the VAEM angle with a frame rate of 10 Hz. In contrast, the recording was conducted in 461 smaller regions of 12.8 x 12.8 µm and frame rates of 20 Hz by streaming between 2,500 and 5,000 462 frames per movie. A series of dark images were recorded at the same frame rate as the corresponding 463 movies to correct for noise in data processing. 464 Raw data processing and analysis of sptPALM movies 465 The experimental data sets were imported into OneFlowTraX (Rohr et al., 2024) to assess their quality. 466 Samples exhibiting obvious outliers during the quality assessment were excluded from further 467 analysis. The remaining files were analyzed using the “ Batch Analysis” function of OneFlowTraX , 468 according to the settings introduced by Ro hr et al. (2 024) for localization, tracking, and mobility 469 analysis. The diffusion coefficients were calculated for all samples except BAK1 by fitting the data 470 distribution with a Gaussian function and subsequently extracting the average diffusion coefficient 471 from its peak center. For BAK1, two populations were clearly visible, necessitating the use of a 472 two-component Gaussian mixture model to estimate their respective diffusion coefficients and 473 relative fractions. The nanoscale organization of protein clusters was investigated using the NASTIC 474 algorithm from Wal lis et al. (2 023) (also available in OneFlowTraX ) with the following parameters : 475 radius factor: 1.2 and at least three tracks per cluster. For the comparisons between the different 476 cytoskeleton disintegration scenarios , the cluster diameter (nm) was used. This is calculated by 477 treating the localiza tions in a cluster as a point cloud that is fitted by a two -dimensional Gaussian 478 function. Its full width at half maximum (FWHM) values for x and y are then averaged to provide one 479 value that is defined as the cluster’s diameter. The subsequent analysis steps were processed with 480 custom-built R applications. Clusters with diameters greater than 2,500 nm were excluded from 481 further analysis. Given that cluster data are log -normally distributed, specific statistical tests were 482 employed to identify significant differences (Zhou et al., 1997). The respective figure (Figure 4) report 483 the corrected, transformed mean as recommended by the aforementioned study. 484 485 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Motion classification with DC-MSS 486 The motion classification of individual (sub-)trajectories was conducted using the “divide-and-conquer 487 moment scaling spectrum” (DC-MSS) algorithm (Vega et al., 20 18). In short, trajectories that consist 488 of at least 20 localizations are initially divided into segments of potentially disparate motion classes 489 based on the extent of molecular movement. Subsequently, a movement scaling spectrum analysis is 490 employed for the classification of these segments, utiliz ing threshold values to minimi ze the 491 probability of misclassification among adjacent motion types. Intermediate refining steps are 492 incorporated to enhance the confidence of both the trajectory segmentation and their classification. 493 Acknowledgments and Funding 494 Our research was supported by the German Research Foundation (DFG) via the CRC 1101 495 (projects D02 and Z02) to S.z.O.-K and K.H. and by individual DFG grants to K.H. (HA 2146/22, 496 HA 2146/23 We also thank the DFG for grants for scientific equipment (FUGG: INST 37/991-1, 497 INST 37/992-1, INST 37/819-1, INST 37/965-1). 498 Author contributions 499 Conceptualization, L.R. (Leander Rohr) , K.H. and S.z.O.-K.; Data curation, L.R. and 500 Ll.R. (Luiselotte Rausch); Formal Analysis, L.R. and Ll.R.; Funding acquisition, K.H.; Investigation, L.R. 501 and Ll.R.; Methodology, S.z.O.-K.; Project administration, L.R., K.H. and S.z.O.-K.; Resources, Ll.R. and 502 S.z.O.-K.; Software, S.z.O.-K.; Supervision, K.H.; Visualization, L.R.; Writing – original draft, L.R., K.H 503 and S.z.O.-K.; Writing – review & editing, L.R., Ll.R., K.H and S.z.O.-K. 504 505 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Figure legends 506 Figure 1 | Overview of genetically encoded, enzymatic tools for cytoskeleton disintegration. 507 (A) Exemplary confocal microscopy image of epidermal leaf cells of Nicotiana benthamiana (N. benthamiana) expressing the 508 actin marker GFP-ABD2-GFP with the GFP channel on the left and the corresponding transmission light channel on the right. 509 Intact actin filaments are clearly visible. Scale bar = 10 µm (B) Exemplary image of the co -expression of the disruption tool 510 HA-SpvB and the actin marker GFP -ABD2-GFP in the GFP channel (left) and the corresponding transmission light channel 511 (right). The co-expression with the disruption tool leads to removal of F-actin cables in all cells as shown before (Vilches Barro 512 et al., 2019) . Scale bar = 10 µm. (C) Schematic plasmid structure of the genetically encoded SpvB tool: By Gateway® 513 technology SpvB was inserted into the pEG201 backbone (Earley et al., 2 006) which contains a 35S promoter and an 514 N-terminal HA -tag. (D) Exemplary confocal microscopy image of epidermal leaf cells of N. benthamiana expressing the 515 microtubules marker MAP65 -8-RFP with the RFP channel on the left and the corresponding transmission light channel on 516 the right. Intact microtubules are observable. Scale bar = 10 µm. (E) Exemplary image of the co-expression of the disruption 517 tool PHS1ΔP-HA and the microtubules marker MAP65 -8-RFP in the RFP channel (left) and the corresponding transmission 518 light channel (right). The co -expression with the disruption tool leads to the destabilization of cortical microtubules. 519 Scale bar = 10 µm. (F) Schematic plasmid structure of the genetically encoded PHS1 ΔP tool: The plasmid was generated by 520 GoldenGate cloning (Binder et al., 2014) using Level I modules which were subsequently assembled in the Level II backbone 521 of BB10. PHS1ΔP is under the control of a 2x 35Sω promoter module and fused C-terminally to an HA-tag. For the generation 522 of higher order assemblies, BB10 contains Bpi I recognition sites. 523 524 Figure 2 | Disintegration of actin filaments primarily results in reduced protein dynamics in the plasma membrane. 525 (A) Distribution of diffusion coefficients (D) represented as log(D) and plotted against their occurrence [%] over all quantified 526 cells for RLP44-, BRI1-, PSKR1-, FLS2- and BAK1-mEos3.2. For each protein fusion, two distributions are shown: ( i) In black, 527 values obtained from epidermal N. benthamiana leaf cells expressing the respective fusion alone ( - SpvB) and ( ii) in blue 528 values from the co-expression of the respective protein fusions with the genetically encoded, enzymatic tool SpvB (+ SpvB). 529 For RLP44-, BRI1- and PSKR1-mEos3.2 a slight shift to lower log(D) values is observable, when SpvB is co-expressed. The effect 530 is barely visible for FLS2-mEos3.2 and BAK1-mEos3.2. All measurements were performed three days post infiltration. Please 531 note that all protein fusions show a bell -shape distribution (i.e., one mobility population), except for BAK1 -mEos3.2 that 532 presents a slower and a faster variety (two Gaussian fit). When co-expressed with SpvB, the slow fraction of BAK1-mEos3.2 533 is increased. (B) Representation of the peak D values of RLP44 -, BRI1-, PSKR1-, FLS2- and BAK1-mEos3.2. with same color 534 code as in (A). The peak values of individual cells (illustrated as single dots or triangles; n ≥ 17) were obtained by normal fits 535 of distributions comparable to (A) expect of BAK1 -mEos3.2 where a two -component Gaussian mixture model was applied 536 (see Material and Methods). The separation of the BAK1 fractions was done according to this model with the peaks of the 537 first maxima representing the slow fraction and the peaks of the second maxima the fast fraction, respectively. In the absence 538 of intact actin filaments (blue; +SpvB) the diffusion coefficient is significantly decreased for the RLP44 -, BRI1 - and 539 PSKR1-mEos3.2 fusions, while the reduction for FLS2 -mEos3.2 is not significant. For BAK1 -mEos3.2, only a decrease in the 540 fast fraction is observable. For statistical evaluation, the data were checked for normal distribution and unequal variances 541 and then analyzed according to the results of the test by applying either a Mann -Whitney U test or a One -way ANOVA. 542 Whiskers show the data range excluding outliers, while the boxes represent the 25 -75 percentile. p ≤ 0.001 (***); 543 p ≤ 0.01 (**); p ≤ 0.05 (*); p > 0.05 (n.s.). All statistical analyses were performed with custom-made R scripts. 544 545 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Figure 3 | Disintegration of microtubule filaments primarily results in increased protein dynamics in the plasma 546 membrane. 547 (A) Distribution of diffusion coefficients (D) represented as log(D) and plotted against their occurrence [%] over all quantified 548 cells for RLP44-, BRI1-, PSKR1-, FLS2- and BAK1-mEos3.2. For each protein fusion, two distributions are shown: ( i) In black, 549 values obtained from epidermal N. benthamiana leaf cells expressing the respective fusion alone ( - PHS1ΔP) and (ii) in blue 550 values from the co -expression of the respective protein fusions with the genetically encoded, enzymatic tool PHS1 ΔP 551 (+ PHS1ΔP). RLP44, PSKR1, and BAK1 -mEos3.2 show a slight shift to higher log(D) values when co -expressed with PHS1ΔP. 552 The effect is barely visible for the other protein fusions. Measurement conditions are as described in Figure 2. Again, all 553 protein fusions show a bell-shape distribution (i.e., one mobility population), except for BAK1-mEos3.2 that presents a slower 554 and a faster variety (two Gaussian fit). When co -expressed with PHS1 ΔP, the fast fraction of BAK1 -mEos3.2 is increased. 555 (B) Representation of the peak D values of RLP44 -, BRI1-, PSKR1-, FLS2- and BAK1-mEos3.2. with same color code as in (A) 556 and obtained by n ≥ 25 cells. The evaluation was performed as described for Figure 2. In the absence of intact microtubules 557 (blue; + PHS1ΔP) the diffusion coefficient is significantly increased for the RLP44 - and PSKR1-mEos3.2 fusions as well as for 558 the slow fraction of BAK1-mEos3.2, without a significant effect on the fast variety. Although BRI1- and FLS2-mEos3.2, as well 559 as the fast variety of BAK1-mEos3.2, show no significant effect, a trend of increasing diffusion coefficients after microtubule 560 disintegration is observable . Statistical analyses were conducted as in Figure 2, with the same box representation. 561 p ≤ 0.001 (***); p ≤ 0.01 (**); p ≤ 0.05 (*); p > 0.05 (n.s.). 562 563 Figure 4 | Actin and microtubules disintegration lead to opposing effects on nanocluster sizes. 564 (A) Representation of corrected, transformed means of nanocluster diameter sizes (nm) of the tested protein fusions based 565 on the track trajectories, also used for Figure 2 and Figure 3. The sizes were either determined for the protein fusions 566 expressed alone (- SpvB, black) or together with the genetically encoded disruption tool for actin, SpvB (+ SpvB, blue). The 567 analyses were performed by applying the NASTIC algorithm (Wallis et al., 2023) available in OneFlowTraX (Rohr et al., 2024) 568 with a radius factor of 1.3 and at least three tracks per cluster. The corrected, transformed means are based on the 569 log-normal distribution of all determined clusters among all evaluated cells (n as in Figure 2 and Figure 3), not considering 570 clusters larger than 2,500 nm in diameter ( ≥ 1630). Except for RLP44 -mEos3.2, all other tested fusions proteins show 571 significantly increased cluster sizes upon the disintegration of actin filaments. Statistical analyses were performed according 572 to Zhou et al. (1 997). (B) Representation of corrected, transformed means of nanocluster diameter sizes (nm) as in (A) but 573 here either with intact microtubules (i.e., expressed alone; - PHS1ΔP, black) or upon microtubules disintegration (+ PHS1ΔP, 574 blue). Parameters and filtering as in (A) with ≥ 4356 evaluated cluster. Except for FLS2- mEos3.2, all other tested fusions 575 proteins show significantly decreased cluster sizes upon the disintegration of microtubules filaments. Statistical analyses 576 were performed as in (A). p ≤ 0.001 (***); p ≤ 0.01 (**); p ≤ 0.05 (*); p > 0.05 (n.s.). 577 578 Figure 5 | Classification of segment motion patterns depending on the cytoskeleton integrity. 579 (A) Proportion of motion behavior (in percent) depending on the status of the actin cytoskeleton (either intact ( - SpvB) or 580 disintegrated (+ SpvB)), namely, directed (magenta), free (grey), confined (yellow) and immobile (blue). All studied proteins 581 reveal primarily confined behavior, independent of the actin integrity. Immobility and free diffusive behavior are present in 582 varying amounts amongst the proteins. Directed movement is barely present. The classification was performed according to 583 Vega et al. (2018) . (B) Relative shifts in the motion patterns in the absence of actin filaments with the same color code as 584 in (A). RLP44-, BRI1 - and PSKR1-mEos3.2 show a clear decrease in free diffusive behavior, while immobility and confined 585 behavior (not for BRI1-mEos3.2) increases. However, FLS2-mEos3.2 shows a contrary effect by a decrease in immobility and 586 an increase in free diffusive behavior. BAK1 -mEos3.2 is nearly unaffected by the manipulation of the actin cytoskeleton. 587 (C) Proportion of motion behavior (in percent) depending on the status of the microtubule cytoskeleton (either intact 588 (- PHS1ΔP) or disintegrated (+ PHS1ΔP) with same behavior classes and color code as in (A) and (B). Again, all studied proteins 589 reveal primarily confined behavior, independent of the microtubule disintegration except for RLP44 -mEos3.2 (+ PHS1ΔP) 590 that shows mostly free diffusive behavior. Immobility and free diffusive behavior are present again with varying amounts 591 amongst the proteins, with distinct differences between RLP44 -mEos3.2 (+ PHS1ΔP) and the other fusion proteins under 592 manipulated conditions. Again, directed movement is barely present. (D) Relative shifts in the motion patterns in the absence 593 of the microtubule cytoskeleton with the same color code as before. RLP44 - and PSKR1-mEos3.2 show comparable effects 594 with increased free diffusive behavior as well as with decreased confined movement and immobility. BRI1-mEos3.2 exhibits 595 only minor changes while FLS2-mEos3.2 shows a slight increase in free and confined motion, while immobility is decreased, 596 too. BAK1 is barely affected. 597 598 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint

599 600 Andrews, S.S. (2009). Accurate particle-based simulation of adsorption, desorption and partial 601 transmission. Phys Biol 6:046015. 602 DOI: https://doi.org/10.1088/1478-3975/6/4/046015 603 604 Andronov, L., Orlov, I., Lutz, Y., Vonesch, J.L., and Klaholz, B.P. (2016). ClusterViSu, a method for 605 clustering of protein complexes by Voronoi tessellation in super-resolution microscopy. Sci Rep 606 6:24084. 607 DOI: https://doi.org/10.1038/srep24084 608 609 Bayle, V., Fiche, J.B., Burny, C., Platre, M.P., Nollmann, M., Martiniere, A., and Jaillais, Y. (2021). 610 Single-particle tracking photoactivated localization microscopy of membrane proteins in living plant 611 tissues. Nat Protoc 16:1600-1628. 612 DOI: https://doi.org/10.1038/s41596-020-00471-4 613 614 Beck, M., Zhou, J., Faulkner, C., MacLean, D., and Robatzek, S. (2012). Spatio-temporal cellular 615 dynamics of the Arabidopsis flagellin receptor reveal activation status-dependent endosomal 616 sorting. Plant Cell 24:4205-4219. 617 DOI: https://doi.org/10.1105/tpc.112.100263 618 619 Belteton, S.A., Sawchuk, M.G., Donohoe, B.S., Scarpella, E., and Szymanski, D.B. (2018). 620 Reassessing the Roles of PIN Proteins and Anticlinal Microtubules during Pavement Cell 621 Morphogenesis. Plant Physiol 176:432-449. 622 DOI: https://doi.org/10.1104/pp.17.01554 623 624 Binder, A., Lambert, J., Morbitzer, R., Popp, C., Ott, T., Lahaye, T., and Parniske, M. (2014). A 625 modular plasmid assembly kit for multigene expression, gene silencing and silencing rescue in plants. 626 PLoS One 9:e88218. 627 DOI: https://doi.org/10.1371/journal.pone.0088218 628 629 Borner, G.H., Sherrier, D.J., Weimar, T., Michaelson, L.V., Hawkins, N.D., Macaskill, A., Napier, J.A., 630 Beale, M.H., Lilley, K.S., and Dupree, P. (2005). Analysis of detergent-resistant membranes in 631 Arabidopsis. Evidence for plasma membrane lipid rafts. Plant Physiol 137:104-116. 632 DOI: https://doi.org/10.1104/pp.104.053041 633 634 Brown, D.A., and Rose, J.K. (1992). Sorting of GPI-anchored proteins to glycolipid-enriched 635 membrane subdomains during transport to the apical cell surface. Cell 68:533-544. 636 DOI: https://doi.org/10.1016/0092-8674(92)90189-j 637 638 Bücherl, C.A., Jarsch, I.K., Schudoma, C., Segonzac, C., Mbengue, M., Robatzek, S., MacLean, D., 639 Ott, T., and Zipfel, C. (2017). Plant immune and growth receptors share common signalling 640 components but localise to distinct plasma membrane nanodomains. Elife 6:e25114. 641 DOI: https://doi.org/10.7554/eLife.25114 642 643 Dale, R., Oswald, S., Jalihal, A., LaPorte, M.F., Fletcher, D.M., Hubbard, A., Shiu, S.H., Nelson, 644 A.D.L., and Bucksch, A. (2021). Overcoming the Challenges to Enhancing Experimental Plant Biology 645 With Computational Modeling. Front Plant Sci 12:687652. 646 DOI: https://doi.org/10.3389/fpls.2021.687652 647 648 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Danek, M., Angelini, J., Malinska, K., Andrejch, J., Amlerova, Z., Kocourkova, D., Brouzdova, J., 649 Valentova, O., Martinec, J., and Petrasek, J. (2020). Cell wall contributes to the stability of plasma 650 membrane nanodomain organization of Arabidopsis thaliana FLOTILLIN2 and HYPERSENSITIVE 651 INDUCED REACTION1 proteins. Plant J 101:619-636. 652 DOI: https://doi.org/10.1111/tpj.14566 653 654 Das, R., Cairo, C.W., and Coombs, D. (2009). A hidden Markov model for single particle tracks 655 quantifies dynamic interactions between LFA-1 and the actin cytoskeleton. PLoS Comput Biol 656 5:e1000556. 657 DOI: https://doi.org/10.1371/journal.pcbi.1000556 658 659 Earley, K.W., Haag, J.R., Pontes, O., Opper, K., Juehne, T., Song, K., and Pikaard, C.S. (2006). 660 Gateway-compatible vectors for plant functional genomics and proteomics. Plant J 45:616-629. 661 DOI: https://doi.org/10.1111/j.1365-313X.2005.02617.x 662 663 Ester, M., Kriegel, H.-P., Sander, J., and Xu, X. (1996). A density-based algorithm for discovering 664 clusters in large spatial databases with noise. Proceedings of the Second International Conference on 665 Knowledge Discovery and Data Mining. AAAI Press. 666 667 Farquharson, K.L. (2009). On the origin of cortical microtubules. Plant Cell 21:2193. 668 DOI: https://doi.org/10.1105/tpc.109.210812 669 670 Fujita, S., Pytela, J., Hotta, T., Kato, T., Hamada, T., Akamatsu, R., Ishida, Y., Kutsuna, N., 671 Hasezawa, S., Nomura, Y., et al. (2013). An atypical tubulin kinase mediates stress-induced 672 microtubule depolymerization in Arabidopsis. Curr Biol 23:1969-1978. 673 DOI: https://doi.org/10.1016/j.cub.2013.08.006 674 675 Fujiwara, I., Zweifel, M.E., Courtemanche, N., and Pollard, T.D. (2018). Latrunculin A Accelerates 676 Actin Filament Depolymerization in Addition to Sequestering Actin Monomers. Curr Biol 28:3183-677 3192. 678 DOI: https://doi.org/10.1016/j.cub.2018.07.082 679 680 Fujiwara, T., Ritchie, K., Murakoshi, H., Jacobson, K., and Kusumi, A. (2002). Phospholipids undergo 681 hop diffusion in compartmentalized cell membrane. J Cell Biol 157:1071-1081. 682 DOI: https://doi.org/10.1083/jcb.200202050 683 684 Fujiwara, T.K., Iwasawa, K., Kalay, Z., Tsunoyama, T.A., Watanabe, Y., Umemura, Y.M., Murakoshi, 685 H., Suzuki, K.G., Nemoto, Y.L., Morone, N., et al. (2016). Confined diffusion of transmembrane 686 proteins and lipids induced by the same actin meshwork lining the plasma membrane. Mol Biol Cell 687 27:1101-1119. 688 DOI: https://doi.org/10.1091/mbc.E15-04-0186 689 690 Garnelo Gomez, B., Holzwart, E., Shi, C., Lozano-Duran, R., and Wolf, S. (2021). Phosphorylation-691 dependent routing of RLP44 towards brassinosteroid or phytosulfokine signalling. J Cell Sci 692 134:jcs259134. 693 DOI: https://doi.org/10.1242/jcs.259134 694 695 Glöckner, N., Zur Oven-Krockhaus, S., Rohr, L., Wackenhut, F., Burmeister, M., Wanke, F., 696 Holzwart, E., Meixner, A.J., Wolf, S., and Harter, K. (2022). Three-Fluorophore FRET Enables the 697 Analysis of Ternary Protein Association in Living Plant Cells. Plants (Basel) 11:2630. 698 DOI: https://doi.org/10.3390/plants11192630 699 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint 700 Gronnier, J., Gerbeau-Pissot, P., Germain, V., Mongrand, S., and Simon-Plas, F. (2018). Divide and 701 Rule: Plant Plasma Membrane Organization. Trends Plant Sci 23:899-917. 702 DOI: https://doi.org/10.1016/j.tplants.2018.07.007 703 704 Grosseholz, R., Wanke, F., Rohr, L., Glöckner, N., Rausch, L., Scholl, S., Scacchi, E., Spazierer, A.J., 705 Shabala, L., Shabala, S., et al. (2022). Computational modeling and quantitative physiology reveal 706 central parameters for brassinosteroid-regulated early cell physiological processes linked to 707 elongation growth of the Arabidopsis root. Elife 11:e73031. 708 DOI: https://doi.org/10.7554/eLife.73031 709 710 Hansen, A.S., Woringer, M., Grimm, J.B., Lavis, L.D., Tjian, R., and Darzacq, X. (2018). Robust 711 model-based analysis of single-particle tracking experiments with Spot-On. Elife 7:e33125. 712 DOI: https://doi.org/10.7554/eLife.33125 713 714 Harterink, M., da Silva, M.E., Will, L., Turan, J., Ibrahim, A., Lang, A.E., van Battum, E.Y., 715 Pasterkamp, R.J., Kapitein, L.C., Kudryashov, D., et al. (2017). DeActs: genetically encoded tools for 716 perturbing the actin cytoskeleton in single cells. Nat Methods 14:479-482. 717 DOI: https://doi.org/10.1038/nmeth.4257 718 719 Hecker, A., Brand, L.H., Peter, S., Simoncello, N., Kilian, J., Harter, K., Gaudin, V., and Wanke, D. 720 (2015). The Arabidopsis GAGA-Binding Factor BASIC PENTACYSTEINE6 Recruits the POLYCOMB-721 REPRESSIVE COMPLEX1 Component LIKE HETEROCHROMATIN PROTEIN1 to GAGA DNA Motifs. Plant 722 Physiol 168:1013-1024. 723 DOI: https://doi.org/10.1104/pp.15.00409 724 725 Helmuth, J.A., Burckhardt, C.J., Koumoutsakos, P., Greber, U.F., and Sbalzarini, I.F. (2007). A novel 726 supervised trajectory segmentation algorithm identifies distinct types of human adenovirus motion 727 in host cells. J Struct Biol 159:347-358. 728 DOI: https://doi.org/10.1016/j.jsb.2007.04.003 729 730 Hochmann, H., Pust, S., von Figura, G., Aktories, K., and Barth, H. (2006). Salmonella enterica SpvB 731 ADP-ribosylates actin at position arginine-177-characterization of the catalytic domain within the 732 SpvB protein and a comparison to binary clostridial actin-ADP-ribosylating toxins. Biochemistry 733 45:1271-1277. 734 DOI: https://doi.org/10.1021/bi051810w 735 736 Holzwart, E., Huerta, A.I., Glöckner, N., Garnelo Gomez, B., Wanke, F., Augustin, S., Askani, J.C., 737 Schurholz, A.K., Harter, K., and Wolf, S. (2018). BRI1 controls vascular cell fate in the Arabidopsis 738 root through RLP44 and phytosulfokine signaling. Proc Natl Acad Sci U S A 115:11838-11843. 739 DOI: https://doi.org/10.1073/pnas.1814434115 740 741 Hosy, E., Martiniere, A., Choquet, D., Maurel, C., and Luu, D.T. (2015). Super-resolved and dynamic 742 imaging of membrane proteins in plant cells reveal contrasting kinetic profiles and multiple 743 confinement mechanisms. Mol Plant 8:339-342. 744 DOI: https://doi.org/10.1016/j.molp.2014.10.006 745 746 Jaillais, Y., and Ott, T. (2020). The Nanoscale Organization of the Plasma Membrane and Its 747 Importance in Signaling: A Proteolipid Perspective. Plant Physiol 182:1682-1696. 748 DOI: https://doi.org/10.1104/pp.19.01349 749 750 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Jaqaman, K., and Grinstein, S. (2012). Regulation from within: the cytoskeleton in transmembrane 751 signaling. Trends Cell Biol 22:515-526. 752 DOI: https://doi.org/10.1016/j.tcb.2012.07.006 753 754 Jarsch, I.K., and Ott, T. (2011). Perspectives on remorin proteins, membrane rafts, and their role 755 during plant-microbe interactions. Mol Plant Microbe Interact 24:7-12. 756 DOI: https://doi.org/10.1094/mpmi-07-10-0166 757 758 Jarsch, I.K., Konrad, S.S., Stratil, T.F., Urbanus, S.L., Szymanski, W., Braun, P., Braun, K.H., and Ott, 759 T. (2014). Plasma Membranes Are Subcompartmentalized into a Plethora of Coexisting and Diverse 760 Microdomains in Arabidopsis and Nicotiana benthamiana. Plant Cell 26:1698-1711. 761 DOI: https://doi.org/10.1105/tpc.114.124446 762 763 Kalachova, T., Leontovyčová, H., Iakovenko, O., Pospíchalová, R., Maršík, P., Klouček, P., Janda, M., 764 Valentová, O., Kocourková, D., Martinec, J., et al. (2019). Interplay between phosphoinositides and 765 actin cytoskeleton in the regulation of immunity related responses in Arabidopsis thaliana seedlings. 766 Environmental and Experimental Botany 167:103867. 767 DOI: https://doi.org/10.1016/j.envexpbot.2019.103867 768 769 Ke, M., Ma, Z., Wang, D., Sun, Y., Wen, C., Huang, D., Chen, Z., Yang, L., Tan, S., Li, R., et al. (2021). 770 Salicylic acid regulates PIN2 auxin transporter hyperclustering and root gravitropic growth via 771 Remorin-dependent lipid nanodomain organisation in Arabidopsis thaliana. New Phytol 229:963-772 978. 773 DOI: https://doi.org/10.1111/nph.16915 774 775 Komura, N., Suzuki, K.G., Ando, H., Konishi, M., Koikeda, M., Imamura, A., Chadda, R., Fujiwara, 776 T.K., Tsuboi, H., Sheng, R., et al. (2016). Raft-based interactions of gangliosides with a GPI-anchored 777 receptor. Nat Chem Biol 12:402-410. 778 DOI: https://doi.org/10.1038/nchembio.2059 779 780 Konrad, S.S., Popp, C., Stratil, T.F., Jarsch, I.K., Thallmair, V., Folgmann, J., Marin, M., and Ott, T. 781 (2014). S-acylation anchors remorin proteins to the plasma membrane but does not primarily 782 determine their localization in membrane microdomains. New Phytol 203:758-769. 783 DOI: https://doi.org/10.1111/nph.12867 784 785 Kraus, M., Pleskot, R., and Van Damme, D. (2024). Structural and Evolutionary Aspects of Plant 786 Endocytosis. Annu Rev Plant Biol:521-550. 787 DOI: https://doi.org/10.1146/annurev-arplant-070122-023455 788 789 Kusumi, A., Fujiwara, T.K., Chadda, R., Xie, M., Tsunoyama, T.A., Kalay, Z., Kasai, R.S., and Suzuki, 790 K.G. (2012). Dynamic organizing principles of the plasma membrane that regulate signal 791 transduction: commemorating the fortieth anniversary of Singer and Nicolson's fluid-mosaic model. 792 Annu Rev Cell Dev Biol 28:215-250. 793 DOI: https://doi.org/10.1146/annurev-cellbio-100809-151736 794 795 Kusumi, A., Nakada, C., Ritchie, K., Murase, K., Suzuki, K., Murakoshi, H., Kasai, R.S., Kondo, J., and 796 Fujiwara, T. (2005). Paradigm shift of the plasma membrane concept from the two-dimensional 797 continuum fluid to the partitioned fluid: high-speed single-molecule tracking of membrane 798 molecules. Annu Rev Biophys Biomol Struct 34:351-378. 799 DOI: https://doi.org/10.1146/annurev.biophys.34.040204.144637 800 801 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Ladwig, F., Dahlke, R.I., Stuhrwohldt, N., Hartmann, J., Harter, K., and Sauter, M. (2015). 802 Phytosulfokine Regulates Growth in Arabidopsis through a Response Module at the Plasma 803 Membrane That Includes CYCLIC NUCLEOTIDE-GATED CHANNEL17, H+-ATPase, and BAK1. Plant Cell 804 27:1718-1729. 805 DOI: https://doi.org/10.1105/tpc.15.00306 806 807 Laloi, M., Perret, A.M., Chatre, L., Melser, S., Cantrel, C., Vaultier, M.N., Zachowski, A., Bathany, K., 808 Schmitter, J.M., Vallet, M., et al. (2007). Insights into the role of specific lipids in the formation and 809 delivery of lipid microdomains to the plasma membrane of plant cells. Plant Physiol 143:461-472. 810 DOI: https://doi.org/10.1104/pp.106.091496 811 812 Lanza, M., Garcia-Ponce, B., Castrillo, G., Catarecha, P., Sauer, M., Rodriguez-Serrano, M., Paez-813 Garcia, A., Sanchez-Bermejo, E., T, C.M., Leo del Puerto, Y., et al. (2012). Role of actin cytoskeleton 814 in brassinosteroid signaling and in its integration with the auxin response in plants. Dev Cell 22:1275-815 1285. 816 DOI: https://doi.org/10.1016/j.devcel.2012.04.008 817 818 Lefebvre, B., Furt, F., Hartmann, M.A., Michaelson, L.V., Carde, J.P., Sargueil-Boiron, F., Rossignol, 819 M., Napier, J.A., Cullimore, J., Bessoule, J.J., et al. (2007). Characterization of lipid rafts from 820 Medicago truncatula root plasma membranes: a proteomic study reveals the presence of a raft-821 associated redox system. Plant Physiol 144:402-418. 822 DOI: https://doi.org/10.1104/pp.106.094102 823 824 Leontovycova, H., Kalachova, T., Trda, L., Pospichalova, R., Lamparova, L., Dobrev, P.I., Malinska, 825 K., Burketova, L., Valentova, O., and Janda, M. (2019). Actin depolymerization is able to increase 826 plant resistance against pathogens via activation of salicylic acid signalling pathway. Sci Rep 9:10397. 827 DOI: https://doi.org/10.1038/s41598-019-46465-5 828 829 Levet, F., Hosy, E., Kechkar, A., Butler, C., Beghin, A., Choquet, D., and Sibarita, J.B. (2015). SR-830 Tesseler: a method to segment and quantify localization-based super-resolution microscopy data. 831 Nat Methods 12:1065-1071. 832 DOI: https://doi.org/10.1038/nmeth.3579 833 834 Li, X., Wang, X., Yang, Y., Li, R., He, Q., Fang, X., Luu, D.T., Maurel, C., and Lin, J. (2011). Single-835 molecule analysis of PIP2;1 dynamics and partitioning reveals multiple modes of Arabidopsis plasma 836 membrane aquaporin regulation. Plant Cell 23:3780-3797. 837 DOI: https://doi.org/10.1105/tpc.111.091454 838 839 Lu, Y., Zhang, Y., Lian, N., and Li, X. (2023). Membrane Dynamics Regulated by Cytoskeleton in Plant 840 Immunity. Int J Mol Sci 24:6059. 841 DOI: https://doi.org/10.3390/ijms24076059 842 843 Lv, X., Jing, Y., Xiao, J., Zhang, Y., Zhu, Y., Julian, R., and Lin, J. (2017). Membrane microdomains and 844 the cytoskeleton constrain AtHIR1 dynamics and facilitate the formation of an AtHIR1-associated 845 immune complex. Plant J 90:3-16. 846 DOI: https://doi.org/10.1111/tpj.13480 847 848 Ma, Z., Zhu, K., Gao, Y.G., Tan, S.M., and Miao, Y. (2023). Molecular condensation and 849 mechanoregulation of plant class I formin, an integrin-like actin nucleator. FEBS J 290:3336-3354. 850 DOI: https://doi.org/10.1111/febs.16571 851 852 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Ma, Z., Sun, Y., Zhu, X., Yang, L., Chen, X., and Miao, Y. (2022). Membrane nanodomains modulate 853 formin condensation for actin remodeling in Arabidopsis innate immune responses. Plant Cell 854 34:374-394. 855 DOI: https://doi.org/10.1093/plcell/koab261 856 857 Ma, Z., Liu, X., Nath, S., Sun, H., Tran, T.M., Yang, L., Mayor, S., and Miao, Y. (2021). Formin 858 nanoclustering-mediated actin assembly during plant flagellin and DSF signaling. Cell Rep 34:108884. 859 DOI: https://doi.org/10.1016/j.celrep.2021.108884 860 861 Martiniere, A., Lavagi, I., Nageswaran, G., Rolfe, D.J., Maneta-Peyret, L., Luu, D.T., Botchway, S.W., 862 Webb, S.E., Mongrand, S., Maurel, C., et al. (2012). Cell wall constrains lateral diffusion of plant 863 plasma-membrane proteins. Proc Natl Acad Sci U S A 109:12805-12810. 864 DOI: https://doi.org/10.1073/pnas.1202040109 865 866 Matouskova, J., Janda, M., Fiser, R., Sasek, V., Kocourkova, D., Burketova, L., Duskova, J., 867 Martinec, J., and Valentova, O. (2014). Changes in actin dynamics are involved in salicylic acid 868 signaling pathway. Plant Sci 223:36-44. 869 DOI: https://doi.org/10.1016/j.plantsci.2014.03.002 870 871 McKenna, J.F., Rolfe, D.J., Webb, S.E.D., Tolmie, A.F., Botchway, S.W., Martin-Fernandez, M.L., 872 Hawes, C., and Runions, J. (2019). The cell wall regulates dynamics and size of plasma-membrane 873 nanodomains in Arabidopsis. Proc Natl Acad Sci U S A 116:12857-12862. 874 DOI: https://doi.org/10.1073/pnas.1819077116 875 876 Mongrand, S., Morel, J., Laroche, J., Claverol, S., Carde, J.P., Hartmann, M.A., Bonneu, M., Simon-877 Plas, F., Lessire, R., and Bessoule, J.J. (2004). Lipid rafts in higher plant cells: purification and 878 characterization of Triton X-100-insoluble microdomains from tobacco plasma membrane. J Biol 879 Chem 279:36277-32686. 880 DOI: https://doi.org/10.1074/jbc.M403440200 881 882 Morel, J., Claverol, S., Mongrand, S., Furt, F., Fromentin, J., Bessoule, J.J., Blein, J.P., and Simon-883 Plas, F. (2006). Proteomics of plant detergent-resistant membranes. Mol Cell Proteomics 5:1396-884 1411. 885 DOI: https://doi.org/10.1074/mcp.M600044-MCP200 886 887 Narasimhan, M., Johnson, A., Prizak, R., Kaufmann, W.A., Tan, S., Casillas-Perez, B., and Friml, J. 888 (2020). Evolutionarily unique mechanistic framework of clathrin-mediated endocytosis in plants. 889 Elife 9:eLife.52067. 890 DOI: https://doi.org/10.7554/eLife.52067 891 892 Paredez, A.R., Somerville, C.R., and Ehrhardt, D.W. (2006). Visualization of cellulose synthase 893 demonstrates functional association with microtubules. Science 312:1491-1495. 894 DOI: https://doi.org/10.1126/science.1126551 895 896 Persson, F., Linden, M., Unoson, C., and Elf, J. (2013). Extracting intracellular diffusive states and 897 transition rates from single-molecule tracking data. Nat Methods 10:265-269. 898 DOI: https://doi.org/10.1038/nmeth.2367 899 900 Raffaele, S., Vailleau, F., Leger, A., Joubes, J., Miersch, O., Huard, C., Blee, E., Mongrand, S., 901 Domergue, F., and Roby, D. (2008). A MYB transcription factor regulates very-long-chain fatty acid 902 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint biosynthesis for activation of the hypersensitive cell death response in Arabidopsis. Plant Cell 903 20:752-767. 904 DOI: https://doi.org/10.1105/tpc.107.054858 905 906 Raffaele, S., Bayer, E., Lafarge, D., Cluzet, S., German Retana, S., Boubekeur, T., Leborgne-Castel, 907 N., Carde, J.P., Lherminier, J., Noirot, E., et al. (2009). Remorin, a solanaceae protein resident in 908 membrane rafts and plasmodesmata, impairs potato virus X movement. Plant Cell 21:1541-1555. 909 DOI: https://doi.org/10.1105/tpc.108.064279 910 911 Rentsch, J., Bandstra, S., Sezen, B., Sigrist, P., Bottanelli, F., Schmerl, B., Shoichet, S., Noe, F., 912 Sadeghi, M., and Ewers, H. (2024). Sub-membrane actin rings compartmentalize the plasma 913 membrane. J Cell Biol 223:e202310138. 914 DOI: https://doi.org/10.1083/jcb.202310138 915 916 Rohr, L., Ehinger, A., Rausch, L., Glöckner Burmeister, N., Meixner, A.J., Gronnier, J., Harter, K., 917 Kemmerling, B., and Zur Oven-Krockhaus, S. (2024). OneFlowTraX: a user-friendly software for 918 super-resolution analysis of single-molecule dynamics and nanoscale organization. Front Plant Sci 919 15:1358935. 920 DOI: https://doi.org/10.3389/fpls.2024.1358935 921 922 Saxton, M.J. (1997). Single-particle tracking: the distribution of diffusion coefficients. Biophys J 923 72:1744-1753. 924 DOI: https://doi.org/10.1016/s0006-3495(97)78820-9 925 926 Serge, A., Fourgeaud, L., Hemar, A., and Choquet, D. (2003). Active surface transport of 927 metabotropic glutamate receptors through binding to microtubules and actin flow. J Cell Sci 928 116:5015-5022. 929 DOI: https://doi.org/10.1242/jcs.00822 930 931 Simons, K., and Ikonen, E. (1997). Functional rafts in cell membranes. Nature 387:569-572. 932 DOI: https://doi.org/10.1038/42408 933 934 Singer, S.J., and Nicolson, G.L. (1972). The fluid mosaic model of the structure of cell membranes. 935 Science 175:720-731. 936 DOI: https://doi.org/10.1126/science.175.4023.720 937 938 Smertenko, A.P., Kaloriti, D., Chang, H.Y., Fiserova, J., Opatrny, Z., and Hussey, P.J. (2008). The C-939 terminal variable region specifies the dynamic properties of Arabidopsis microtubule-associated 940 protein MAP65 isotypes. Plant Cell 20:3346-3358. 941 DOI: https://doi.org/10.1105/tpc.108.063362 942 943 Spector, I., Braet, F., Shochet, N.R., and Bubb, M.R. (1999). New anti-actin drugs in the study of the 944 organization and function of the actin cytoskeleton. Microscopy Research and Technique 47:18-37. 945 DOI: https://doi.org/10.1002/(sici)1097-0029(19991001)47:1%3C18::aid-jemt3%3E3.0.co;2-e 946 947 Szymanski, D., and Staiger, C.J. (2018). The Actin Cytoskeleton: Functional Arrays for Cytoplasmic 948 Organization and Cell Shape Control. Plant Physiol 176:106-118. 949 DOI: https://doi.org/10.1104/pp.17.01519 950 951 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Szymanski, W.G., Zauber, H., Erban, A., Gorka, M., Wu, X.N., and Schulze, W.X. (2015). Cytoskeletal 952 Components Define Protein Location to Membrane Microdomains. Mol Cell Proteomics 14:2493-953 2509. 954 DOI: https://doi.org/10.1074/mcp.M114.046904 955 956 Tanner, W., Malinsky, J., and Opekarova, M. (2011). In plant and animal cells, detergent-resistant 957 membranes do not define functional membrane rafts. Plant Cell 23:1191-1193. 958 DOI: https://doi.org/10.1105/tpc.111.086249 959 960 Vega, A.R., Freeman, S.A., Grinstein, S., and Jaqaman, K. (2018). Multistep Track Segmentation and 961 Motion Classification for Transient Mobility Analysis. Biophys J 114:1018-1025. 962 DOI: https://doi.org/10.1016/j.bpj.2018.01.012 963 964 Vilches Barro, A., Stockle, D., Thellmann, M., Ruiz-Duarte, P., Bald, L., Louveaux, M., von Born, P., 965 Denninger, P., Goh, T., Fukaki, H., et al. (2019). Cytoskeleton Dynamics Are Necessary for Early 966 Events of Lateral Root Initiation in Arabidopsis. Curr Biol 29:2443-2454. 967 DOI: https://doi.org/10.1016/j.cub.2019.06.039 968 969 Wagner, T., Kroll, A., Haramagatti, C.R., Lipinski, H.G., and Wiemann, M. (2017). Classification and 970 Segmentation of Nanoparticle Diffusion Trajectories in Cellular Micro Environments. PLoS One 971 12:e0170165. 972 DOI: https://doi.org/10.1371/journal.pone.0170165 973 974 Wallis, T.P., Jiang, A., Young, K., Hou, H., Kudo, K., McCann, A.J., Durisic, N., Joensuu, M., Oelz, D., 975 Nguyen, H., et al. (2023). Super-resolved trajectory-derived nanoclustering analysis using 976 spatiotemporal indexing. Nat Commun 14:3353. 977 DOI: https://doi.org/10.1038/s41467-023-38866-y 978 979 Wang, Y.S., Yoo, C.M., and Blancaflor, E.B. (2008). Improved imaging of actin filaments in transgenic 980 Arabidopsis plants expressing a green fluorescent protein fusion to the C- and N-termini of the 981 fimbrin actin-binding domain 2. New Phytol 177:525-536. 982 DOI: https://doi.org/10.1111/j.1469-8137.2007.02261.x 983 984 Yu, J., Fischman, D.A., and Steck, T.L. (1973). Selective solubilization of proteins and phospholipids 985 from red blood cell membranes by nonionic detergents. J Supramol Struct 1:233-248. 986 DOI: https://doi.org/10.1002/jss.400010308 987 988 Zhou, X.-H., Gao, S., and Hui, S.L. (1997). Methods for Comparing the Means of Two Independent 989 Log-Normal Samples. Biometrics 53:1129-1135. 990 DOI: https://doi.org/10.2307/2533570 991 992 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint Figures 993 994 995 Figure 1 | Overview of genetically encoded, enzymatic tools for cytoskeleton disintegration. 996 (A) Exemplary confocal microscopy image of epidermal leaf cells of Nicotiana benthamiana (N. benthamiana) expressing the 997 actin marker GFP-ABD2-GFP with the GFP channel on the left and the corresponding transmission light channel on the right. 998 Intact actin filaments are clearly visible. Scale bar = 10 µm (B) Exemplary image of the c o-expression of the disruption tool 999 HA-SpvB and the actin marker GFP -ABD2-GFP in the GFP channel (left) and the corresponding transmission light channel 1000 (right). The co-expression with the disruption tool leads to removal of F-actin cables in all cells as shown before (Vilches Barro 1001 et al., 2019) . Scale bar = 10 µm. (C) Schematic plasmid structure of the genetically encoded SpvB tool: By Gateway® 1002 technology SpvB was inserted into the pEG201 backbone (Earley et al., 2 006) which contains a 35S promoter and an 1003 N-terminal HA -tag. (D) Exemplary confocal microscopy image of epidermal leaf cells of N. benthamiana expressing the 1004 microtubules marker MAP65 -8-RFP with the RFP channel on the left and the corresponding transmission light channel on 1005 the right. Intact microtubules are observable. Scale bar = 10 µm. (E) Exemplary image of the co-expression of the disruption 1006 tool PHS1ΔP-HA and the microtubules marker MAP65 -8-RFP in the RFP channel (left) and the corresponding transmission 1007 light channel (right) . The co -expression with the disruption tool leads to the destabilization of cortical microtubules. 1008 Scale bar = 10 µm. (F) Schematic plasmid structure of the genetically encoded PHS1 ΔP tool: The plasmid was generated by 1009 GoldenGate cloning (Binder et al., 2014) using Level I modules which were subsequently assembled in the Level II backbone 1010 of BB10. PHS1ΔP is under the control of a 2x 35Sω promoter module and fused C-terminally to an HA-tag. For the generation 1011 of higher order assemblies, BB10 contains Bpi I recognition sites. 1012 1013 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint 1014 Figure 2 | Disintegration of actin filaments primarily results in reduced protein dynamics in the plasma membrane. 1015 (A) Distribution of diffusion coefficients (D) represented as log(D) and plotted against their occurrence [%] over all quantified 1016 cells for RLP44-, BRI1-, PSKR1-, FLS2- and BAK1-mEos3.2. For each protein fusion, two distributions are shown: ( i) In black, 1017 values obtained from epidermal N. benthamiana leaf cells expressing the respective fusion alone ( - SpvB) and ( ii) in blue 1018 values from the co-expression of the respective protein fusions with the genetically encoded, enzymatic tool SpvB (+ SpvB). 1019 For RLP44-, BRI1- and PSKR1-mEos3.2 a slight shift to lower log(D) values is observable, when SpvB is co-expressed. The effect 1020 is barely visible for FLS2-mEos3.2 and BAK1-mEos3.2. All measurements were performed three days post infiltration. Please 1021 note that all protein fusions show a bell -shape distribution (i.e., one mobility population), except for BAK1-mEos3.2 that 1022 presents a slower and a faster variety (two Gaussian fit). When co-expressed with SpvB, the slow fraction of BAK1-mEos3.2 1023 is increased. (B) Representation of the peak D values of RLP44-, BRI1 -, PSKR1-, FLS2- and BAK1-mEos3.2. with same color 1024 code as in (A). The peak values of individual cells (illustrated as single dots or triangles; n ≥ 17) were obtained by normal fits 1025 of distributions comparable to (A) expect of BAK1 -mEos3.2 where a two -component Gaussian mixture model was applied 1026 (see Material and Methods). The separation of the BAK1 fractions was done according to this model with the peaks of the 1027 first maxima representing the slow fraction and the peaks of the second maxima the fast fraction, respectively. In the absence 1028 of intact actin filaments (blue; +SpvB) the diffusion coefficient is significantly decreased for the RLP44 -, BRI1 - and 1029 PSKR1-mEos3.2 fusions, while the reduction for FLS2 -mEos3.2 is not significant. For BAK1 -mEos3.2, only a decrease in the 1030 fast fraction is observable. For statistical evaluation, the data were checked for normal distribution and unequal variances 1031 and then analyzed according to the results of the test by applying either a Mann -Whitney U test or a One-way ANOVA. 1032 Whiskers show the data range excluding outliers, while the boxes represent the 25 -75 percentile . p ≤ 0.001 (***); 1033 p ≤ 0.01 (**); p ≤ 0.05 (*); p > 0.05 (n.s.). All statistical analyses were performed with custom-made R scripts. 1034 1035 1036 1037 1038 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint 1039 Figure 3 | Disintegration of microtubule filaments primarily results in increased protein dynamics in the plasma 1040 membrane. 1041 (A) Distribution of diffusion coefficients (D) represented as log(D) and plotted against their occurrence [%] over all quantified 1042 cells for RLP44-, BRI1-, PSKR1-, FLS2- and BAK1-mEos3.2. For each protein fusion, two distributions are shown: ( i) In black, 1043 values obtained from epidermal N. benthamiana leaf cells expressing the respective fusion alone ( - PHS1ΔP) and (ii) in blue 1044 values from the co -expression of the respective protein fusions with the genetically encoded, enzymatic tool PHS1 ΔP 1045 (+ PHS1ΔP). RLP44, PSKR1, and BAK1 -mEos3.2 show a slight shift to higher log(D) values when co -expressed with PHS1ΔP. 1046 The effect is barely visible for the other protein fusions . Measurement conditions are as described in Figure 2. Again, all 1047 protein fusions show a bell-shape distribution (i.e., one mobility population), except for BAK1-mEos3.2 that presents a slower 1048 and a faster variety (two Gaussian fit). When co -expressed with PHS1 ΔP, the fast fraction of BAK1 -mEos3.2 is increased. 1049 (B) Representation of the peak D values of RLP44 -, BRI1-, PSKR1-, FLS2- and BAK1-mEos3.2. with same color code as in (A) 1050 and obtained by n ≥ 25 cells. The evaluation was performed as described for Figure 2. In the absence of intact microtubules 1051 (blue; + PHS1ΔP) the diffusion coefficient is significantly increased for the RLP44- and PSKR1-mEos3.2 fusions as well as for 1052 the slow fraction of BAK1-mEos3.2, without a significant effect on the fast variety. Although BRI1- and FLS2-mEos3.2, as well 1053 as the fast variety of BAK1-mEos3.2, show no significant effect, a trend of increasing diffusion coefficients after microtubule 1054 disintegration is observable . Statistical analyses were conducted as in Figure 2, with the same box representation . 1055 p ≤ 0.001 (***); p ≤ 0.01 (**); p ≤ 0.05 (*); p > 0.05 (n.s.). 1056 1057 1058 1059 1060 1061 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint 1062 Figure 4 | Actin and microtubules disintegration lead to opposing effects on nanocluster sizes. 1063 (A) Representation of corrected, transformed means of nanocluster diameter sizes (nm) of the tested protein fusions based 1064 on the track trajectories, also used for Figure 2 and Figure 3. The sizes were either determined for the protein fusions 1065 expressed alone (- SpvB, black) or together with the genetically encoded disruption tool for actin, SpvB (+ SpvB, blue). The 1066 analyses were performed by applying the NASTIC algorithm (Wallis et al., 2023) available in OneFlowTraX (Rohr et al., 2024) 1067 with a radius factor of 1.3 and at least three tracks per cluster . The corrected, transformed means are based on the 1068 log-normal distribution of all determined clusters among all evaluated cells (n as in Figure 2 and Figure 3), not considering 1069 clusters larger than 2,500 nm in diameter ( ≥ 1630). Except for RLP44-mEos3.2 , all other tested fusions proteins show 1070 significantly increased cluster sizes upon the disintegration of actin filaments. Statistical analyses were performed according 1071 to Zhou et al. (19 97). (B) Representation of corrected, transformed means of nanocluster diameter sizes (nm) as in (A) but 1072 here either with intact microtubules (i.e., expressed alone; - PHS1ΔP, black) or upon microtubules disintegration (+ PHS1ΔP, 1073 blue). Parameters and filtering as in (A) with ≥ 4356 evaluated cluster. Except for FLS2-mEos3.2, all other tested fusions 1074 proteins show significantly decreased cluster sizes upon the disintegration of microtubules filaments. Statistical analyses 1075 were performed as in (A). p ≤ 0.001 (***); p ≤ 0.01 (**); p ≤ 0.05 (*); p > 0.05 (n.s.). 1076 1077 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint 1078 Figure 5 | Classification of segment motion patterns depending on the cytoskeleton integrity. 1079 (A) Proportion of motion behavior (in percent) depending on the status of the actin cytoskeleton (either intact ( - SpvB) or 1080 disintegrated (+ SpvB)), namely, directed (magenta), free (grey), confined (yellow) and immobile (blue). All studied proteins 1081 reveal primarily confined behavior, independent of the actin integrity. Immobility and free diffusive behavior are present in 1082 varying amounts amongst the proteins. Directed movement is barely present. The classification was performed according to 1083 Vega et al. (20 18). (B) Relative shifts in the motion patterns in the absence of actin filaments with the same color code as 1084 in (A). RLP44-, BRI1 - and PSKR1-mEos3.2 show a clear decrease in free diffusive behavior, while immobil ity and confined 1085 behavior (not for BRI1-mEos3.2) increases. However, FLS2-mEos3.2 shows a contrary effect by a decrease in immobility and 1086 an increase in free diffusi ve behavior. BAK1-mEos3.2 is nearly unaffected by the manipulation of the actin cytoskeleton. 1087 (C) Proportion of motion behavior (in percent) depending on the status of the microtubule cytoskeleton (either intact 1088 (- PHS1ΔP) or disintegrated (+ PHS1ΔP) with same behavior classes and color code as in (A) and (B). Again, all studied proteins 1089 reveal primarily confined behavior, independent of the microtubule disintegration except for RLP44-mEos3.2 (+ PHS1ΔP) 1090 that shows mostly free diffusive behavior. Immobility and free diffusive behavior are present again with varying amounts 1091 amongst the proteins, with distinct differences between RLP44-mEos3.2 (+ PHS1ΔP) and the other fusion proteins under 1092 manipulated conditions. Again, directed movement is barely present. (D) Relative shifts in the motion patterns in the absence 1093 of the microtubule cytoskeleton with the same color code as before. RLP44 - and PSKR1-mEos3.2 show comparable effects 1094 with increased free diffusive behavior as well as with decreased confined movement and immobility. BRI1-mEos3.2 exhibits 1095 only minor changes while FLS2-mEos3.2 shows a slight increase in free and confined motion, while immobility is decreased, 1096 too. BAK1 is barely affected. 1097 1098 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 9, 2024. ; https://doi.org/10.1101/2024.09.09.612020doi: bioRxiv preprint