1 Running head: Membrane anchor of StREM1.3 Remorin · 25 Bordeaux, 146 rue Léo Saignat, 33076 Bordeaux Cedex, France (SR, AP, JLC and 26 SM) 27 28 The Sainsbury Laboratory, John

1

Running head: Membrane anchor of StREM1.3 Remorin 1

2

Correspondence: 3

Dr Sébastien MONGRAND, phD 4

Address: Laboratoire de Biogenèse Membranaire (LBM) 5

CNRS UMR 5200 / Université Bordeaux Segalen 6

146, rue Léo Saignat - Case 92 7

33076 Bordeaux Cédex 8

Email: [email protected] 9

Telephone: + 33 (0) 5 57 57 14 35 10

11

Journal research area: 12

Biochemical Processes and Macromolecular Structures 13

Cell Biology 14

Plant Physiology Preview. Published on August 1, 2012, as DOI:10.1104/pp.112.200519

Copyright 2012 by the American Society of Plant Biologists

www.plantphysiol.orgon November 30, 2018 - Published by Downloaded from Copyright © 2012 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

2

Plasma membrane localization of StREM1.3 Remorin is mediated by 15

conformational changes in a novel C-terminal anchor and required for the 16

restriction of PVX movement 17

18

Artemis Perraki, Jean-Luc Cacas, Jean-Marc Crowet, Laurence Lins, Michel 19

Castroviejo, Sylvie German-Retana, Sébastien Mongrand1* and Sylvain Raffaele1 20

21

22

Laboratoire de Biogenèse Membranaire, (LBM) Unité Mixte de Recherche 5200 23

(UMR 5200) Centre National de la Recherche Scientifique (CNRS)-Université de 24

Bordeaux, 146 rue Léo Saignat, 33076 Bordeaux Cedex, France (SR, AP, JLC and 25

SM) 26

27

The Sainsbury Laboratory, John Innes Centre, Norwich NR4 7UH, UK (SR) 28

29

Centre de Biophysique Moléculaire Numérique, ULg, Gembloux Agro Bio Tech, 30

Passage des Déportés, 2, B-5030 Gembloux, Belgium (JMC and LL) 31

32

Laboratoire de Microbiologie Fondamentale et Pathogénicité Unité Mixte de 33

Recherche 5234 (UMR 5234) Centre National de la Recherche Scientifique (CNRS) 34

146 rue Léo Saignat 33076 Bordeaux Cedex, France (MC) 35

36 Equipe de Virologie, INRA and Université de Bordeaux, UMR 1332 Biologie du Fruit 37

et Pathologie, BP81, 33883 Villenave d'Ornon Cedex, France. (SGR) 38



3

Footnotes : 39

This work was supported by a Marie Curie IEF to S.R. (contract 255104). We 40

acknowledge the French ‘Agence Nationale pour la Recherche’ (ANR) for financial 41

support (contracts NT09_517917 PANACEA to S.M. and JLC). L.L. is senior 42

research assistant at the Belgian National Funds for Research (FNRS). LL and JMC 43

thank IAP project (iPROS) for financial support. 44

45

1 - SR and SM must be considered as co-last authors 46

47

SR Present address: Laboratoire des Interactions Plantes Micro-organismes, Chemin 48

de Borde Rouge BP 52627, 31326 Castanet Tolosan, France. 49

50

JLC Present address: Université de Bourgogne, UMR Plante-Microbe-Environnement 51

1088, Institut National de la Recherche Agronomique (INRA), 5184 CNRS, 21065 52

Dijon Cedex, France. 53



4

ABSTRACT 54

The formation of plasma membrane (PM) micro-domains plays a crucial role in 55

the regulation of membrane signalling and trafficking. Remorins are a plant-specific 56

family of proteins organized in six phylogenetic groups, and Remorins of the group 1 57

are among the few plant proteins known to specifically associate with membrane 58

rafts. As such, they are valuable to understand the molecular bases for PM lateral 59

organization in plants. However, little is known about the structural determinants 60

underlying group 1 Remorins specific association with membrane rafts. We used a 61

structure-function approach to identify a short C-terminal anchor (RemCA) 62

indispensable and sufficient for tight direct binding of Solanum tuberosum REMORIN 63

1.3 (StREM1.3) to the PM. RemCA switches from unordered to an alpha-helical 64

structure in a non-polar environment. Protein structure modelling indicates that 65

RemCA folds into a tight hairpin of amphipathic helices. Consistently, mutations 66

reducing RemCA amphipathy abolished StREM1.3 PM localization. Furthermore, 67

RemCA directly binds to biological membranes in vitro, shows higher affinity for 68

Detergent-Insoluble Membranes (DIM) lipids, and targets YFP to DIMs in vivo. 69

Mutations in RemCA resulting in cytoplasmic StREM1.3 localization abolish 70

StREM1.3 function in restricting potato virus X movement. The mechanisms 71

described here provide new insights on the control and function of lateral segregation 72

of plant PM. 73



5

INTRODUCTION 74

Protein-lipid interactions are increasingly recognised as key regulatory 75

processes for signal perception and cellular signalling cascades (Cho and Stahelin, 76

2005). During signal transduction and trafficking, a number of soluble proteins 77

dynamically associate with plasma membranes (PMs) to deliver their cargo and to 78

recruit pathway components to the sites of action (Seong et al., 2011). For such 79

proteins, membrane association can be critical for function (Porter and Koelle, 2010). 80

81

PM targeting of peripheral proteins is achieved through (i) binding to integral 82

membrane proteins, (ii) post-translational modifications, or (iii) directly by intrinsic 83

membrane anchor domains. Post-translational modifications function as auxiliary 84

modifications for transient or weak association of soluble proteins to the intracellular 85

face of the PM. In plants, these include N-myristoylation, S-palmitoylation, prenylation 86

by farnesyl or geranylgeranyl moieties, or attachment of glycosylphosphatidylinositol 87

(GPI) anchors (Thompson and Okuyama, 2000). GPI-anchors, for example, tightly 88

associate proteins to the extracellular face of PMs by interaction of the inositol head 89

group of the membrane lipid phosphatidylinositol with a glucosamine residue linked to 90

the C-terminal amino acid of the protein (Paulick and Bertozzi, 2008). As an 91

alternative mechanism, globular structures either recognize phospholipids in a 92

stereospecific manner or associate with membranes by their biophysical properties 93

(reviewed in Lemmon, 2008). Other proteins expose unstructured clusters of basic 94

and hydrophobic residues to mediate PM binding (McLaughlin et al., 2002; 95

McLaughlin and Murray, 2005). 96

97

Selective recognition of membrane compartments or domains by protein 98

anchors can be critical in triggering the appropriate downstream trafficking and 99

signalling events (for review: Gruenberg, 2003; De Matteis and Godi, 2004). 100

Membrane domain selectivity can be specified by the anchoring post-translational 101

modification or by a protein anchor domain. For instance proteins carrying GPI-102

anchors are over-represented in membrane rafts, indicating that addition of this lipid 103

anchor directs proteins to these micro-domains (Cordy et al., 2003; Kierszniowska et 104

al., 2009). Membrane rafts are enriched in highly-saturated long chain sphingolipids, 105

sterols and saturated phospholipids, creating tightly-packed domains, designated as 106

‘Liquid-ordered’. These lipids display a stronger affinity to saturated acyl chains as 107



6

found in GPI anchored and acylated proteins (Brown, 2006). The composition of 108

membrane rafts also prevents solubilisation by detergent at low temperature with 109

non-ionic detergent, and allows the partial purification of raft in so called Detergent-110

Insoluble Membrane (DIM) fractions, supposedly biochemical counterparts of 111

membrane rafts. Many signalling proteins are found in membrane rafts supporting the 112

hypothesis that they serve as key platforms for cellular signal transduction and cell-113

to-cell communication (Lingwood and Simons, 2010; Simon-Plas et al., 2011). For 114

example, in human cells, key soluble signalling components such as the Ser/Thr 115

kinase Akt (protein kinase B) are recruited to membrane rafts where they activate 116

signal transduction cascades (Lasserre et al., 2008). Nevertheless, few protein motifs 117

were described to contribute to raft targeting (Rossin et al., 2010) while the six amino 118

acid long raft target signal from human tyrosine phosphatase Src homology 2-119

containing phosphatase 1 (SHP-1) is the only known motif sufficient to anchor soluble 120

proteins specifically to domains of the intracellular face of the PM (Sankarshanan et 121

al., 2007). However, the anchoring mechanism itself remains to be unravelled. In 122

plants even though DIMs also exist (Mongrand et al., 2010) and functional PM 123

domains have been reported (Bhat et al., 2005), the molecular basis for specific 124

targeting and binding of proteins to membrane rafts has never been described. 125

126

Remorins form a diverse family of plant-specific proteins organized in six 127

distinct phylogenetic groups (Raffaele et al., 2007). Remorins from the group 1 have 128

been reported to localize to the PM despite their overall hydrophilic nature (Reymond 129

et al., 1996; Raffaele et al., 2007). Moreover, group 1 Remorins almost exclusively 130

associate to DIMs and localize to membrane micro-domains in a sterol-dependent 131

manner (Lefebvre et al., 2007; Kierszniowska et al., 2009; Raffaele et al., 2009). The 132

function of Remorins is mostly unknown, but we showed in a previous study that 133

StREM1.3 (Solanum tuberosum Remorin from group 1, homolog 3, initially described 134

in (Reymond et al., 1996) regulates cell-to-cell propagation of the potato virus X 135

(PVX) likely by directly interacting with the viral movement protein TGBp1 (Raffaele 136

et al., 2009). StREM1.3 localizes to the inner leaflet of PMs and along 137

plasmodesmata, bridges connecting neighbour cells essential for cell-to-cell 138

communication in plants (Maule, 2008). Other members of the Remorin family group 139

1 are likely involved in innate immune responses (Liu et al., 2009; Widjaja et al., 140

2009; Keinath et al., 2010). Remorins from group 2 are involved in the control of 141



7

infection by symbiotic bacteria at nodular infection threads and the peribacteroid 142

membrane (Lefebvre et al., 2010) These data suggest general roles for Remorins in 143

regulating signalling in plant-microbe interactions (Jarsch and Ott, 2011). 144

145

Elucidating the mechanisms driving StREM1.3 association with PM micro-146

domains therefore provides a unique opportunity for understanding the regulation 147

and function of membrane lateral segregation in plants. StREM1.3 does not contain 148

predictable transmembrane or membrane associated domains. The bases for its 149

association to PMs and selective targeting to DIMs are unknown. Here we identified a 150

novel membrane anchor domain required for StREM1.3 tight and direct association 151

with detergent insoluble fraction of the PM. We combined biophysics, in silico 152

analysis and directed mutagenesis to unravel the molecular bases of StREM1.3 153

membrane binding and its biological significance in the control of PVX propagation. 154

155

RESULTS 156

157

StREM1.3 is a strongly associated peripheral membrane protein 158

To identify structural domains potentially involved in PM localization of 159

StREM1.3, we performed in silico domain analysis and secondary structure 160

prediction. Proteins of the Remorin family typically contain a variable N-terminal 161

region (Marín and Ott, 2012) and a conserved C-terminal domain, termed Remorin_C 162

(PF03763), of approximately 120 amino acid (aa) residues that encompasses a 163

predicted coiled-coil domain. In StREM1.3, the Remorin_C domain extends from aa 164

85 to 195 with a coiled-coil domain predicted between aa 116 to 152 (Figure 1A). 165

StREM1.3 C-terminal region is predicted to be mostly alpha-helical composed of a 84 166

aa long coiled alpha helix ranging from position 85 to 169. This helix is predicted to 167

be interrupted at Glycine residue 171. A short (~12 aa) alpha helix is predicted 168

between residues 172 and 187. Finally the most C-terminal residues are predicted to 169

form a random coil except for residues 192 to 195 where a short alpha helix is 170

predicted. The N-terminal region of StREM1.3 contains the proline-rich Remorin_N 171

domain (PF03766) extending between aa 27 to 84 and is mostly unordered (Marín 172

and Ott, 2012). No recognizable transmembrane domains or any other membrane-173

targeting signals could be detected, indicating that StREM1.3 is a peripheral 174

membrane protein with an atypical membrane-binding domain. 175



8

To test the nature of Remorin association with biological membranes we 176

purified native microsomes and PMs from tobacco leaves and tested its affinity by 177

removal of extrinsic membrane proteins. In accordance with published data PM 178

stripping resulted in ~50% loss of total protein, representing peripheral proteins with 179

weak affinity to the PM (Santoni et al., 2003). Typical of peripheral membrane 180

proteins is Actin, which was almost completely removed from the PM after treatment 181

(Figure 1B). However, endogenous StREM1.3 remained almost entirely associated 182

with the PM, similar to true integral proteins such as the aquaporin PIP2 that was 183

used as an integral membrane protein control. Similar results were obtained when 184

purified PMs were washed under high ionic strength conditions, increasing EDTA 185

concentrations, and increasing pH for 30 minutes (Figure S1A). These results 186

suggest that the association of StREM1.3 with the membrane is not limited to 187

electrostatic interactions. 188

189

Remorin C-terminal Anchor (RemCA) is required and sufficient for plasma 190

membrane targeting in vivo. 191

A series of full length and truncated protein variants was generated and fused 192

to G/YFP to analyse the structural requirements for StREM1.3 PM localization. The 193

full-length protein localized to the PM (Figure 1C). This localization is consistent with 194

immunolocalization studies reported earlier (Raffaele et al., 2009; Lefebvre et al., 195

2010) suggesting that the fluorescent tag did not affect localization of the fusion 196

proteins. Deletion of the C-terminal residues (StREM1.31-170) resulted in an entire 197

loss of PM association indicating the presence of a PM anchor motif within these 198

residues (Figure 1C). We will refer to this 28 amino acid region as the “Remorin C-199

terminal Anchor” (RemCA). This hypothesis was confirmed by reciprocal experiments 200

where expression of soluble YFP tagged with RemCA led to clear YFP association 201

with the PM (Figure 1C). We designed a variety of different constructs to express the 202

fluorescently tagged RemCA but we were not able to prevent a partial cleavage of 203

the tag, resulting in a residual nucleo-cytoplasmic fluorescence in vivo. To confirm 204

that RemCA is sufficient to anchor soluble proteins to the PM, we performed 205

biochemical fractionation experiments on cells expressing StREM1.3 G/YFP fusion 206

constructs (Figure 1D). These experiments showed that the GFP:StREM1.3 is fully 207

microsomal whereas YFP:StREM1.31-170 is entirely soluble. The YFP:RemCA fusion 208

was recovered exclusively in the microsomal fraction, with only cleaved YFP in the 209



9

soluble fraction. These results demonstrate that RemCA is required for StREM1.3 210

PM localization and sufficient to fully associate soluble proteins with the PM. 211

212

Non-polar environments trigger conformational changes in RemCA. 213

To analyse the structural properties of RemCA we obtained chemically 214

synthesized tag-free ultra pure peptides that were assessed using circular dichroism 215

(CD) spectroscopy. CD spectra were acquired in 100% 2,2,2-Trifluoroethanol (TFE), 216

a solvent mimicking biological membranes hydrophobicity (causing partial 217

desolvation of the peptide, a situation that occurs in membranes and during protein-218

protein and protein-lipid interactions), or in semi-hydrophobic solvent (50% TFE, 50% 219

Tris-HCl). CD spectra showed distinct minima at 208 and 222 nm, with a marked 220

maximum at 195 nm, consistent with RemCA forming an alpha helical structure 221

(Figure 2A). By contrast the spectrum of RemCA in aqueous solution (100% Tris-222

HCl) showed a reduced minimum at 208 nm and reduced concavity around 195 nm, 223

indicating that alpha helical contributions to the overall RemCA structure are reduced 224

under these conditions. We estimated RemCA protein secondary structure 225

composition from CD spectra (Figure 2B) and found that it undergoes transition from 226

15% alpha-helical content in aqueous solution to 50% in hydrophobic non-polar 227

solvents. Such estimations remain challenging and should thus be taken with 228

caution. Nevertheless, these data imply that the RemCA peptide is largely unordered 229

in aqueous solution and folds into a mostly alpha helical structure in hydrophobic 230

environments. 231

To gain further insights into the structural mechanisms underlying RemCA 232

binding to PMs, we performed homology 3D modelling of this peptide using the I-233

TASSER server. The predicted structure for RemCA corresponds to two alpha 234

helices arranged as a tight hairpin (Figure 2C, Supplemental File 1). This structure 235

is reminiscent of a recently published structure for Influenza hemagglutinin fusion 236

peptide (HAfp) (Lorieau et al., 2010) (Figure S2). The predicted structure has a high 237

helical content, therefore probably corresponding to the conformation adopted by 238

RemCA in hydrophobic environments. This hypothesis is consistent with the structure 239

of the HAfp which forms a tight hairpin of alpha helices in the presence of lipids 240

(Lorieau et al., 2010). The distribution of the relative hydrophobicity of amino-acids 241

reveal an accumulation of hydrophobic aa at the centre of the hairpin, with the side 242

chains of hydrophilic charged amino acids pointing outward (Figure 2C). The turn 243



10

connecting the two alpha helices is composed of the palindromic neutral sequence 244

Ala-Thr-Gly-Thr-Ala. 245

246

Amphipathy of RemCA alpha helices is required for Remorin anchoring to the 247

PM in vivo. 248

The predicted hydrophobic core at the centre of the RemCA hairpin and 249

solvent-exposed charged surfaces (Figure 2C) implies an asymmetry in the 250

distribution of hydrophobic and charged residues around the alpha-helices, a 251

property known as amphipathy. Amphipathic helices are crucial for transient 252

anchoring of some peripheral membrane proteins, notably viral fusion proteins (e.g 253

Lins and Brasseur, 2008; Drin and Antonny, 2009). We used the heliquest server to 254

calculate the difference, named hydrophobic moment (µH), in hydrophobicity 255

between the two faces of each alpha helix forming RemCA. The E172-A186 helix had a 256

µH of 0.153, similar to values obtained for viral fusion peptides (e.g 0.171 for the 257

Simian influenza virus -SIV- fusion peptide), whereas helical fragment P191-F198 258

had a µH of 0.661, with each helix organized with clearly distinct charged and 259

hydrophobic faces (Figure 3A). 260

To test the importance of alpha-helices amphipathy in the function of RemCA, 261

we designed a StREM1.3* variant harbouring a mutated RemCA* with reduced 262

amphipathy (Figure 3B, Supplemental File 2). The transient expression of a 263

YFP:StREM1.3* in tobacco leaves showed that the mutations completely abolished 264

membrane localization, leading to a fully cytoplasmic protein (Figure 3C). Loss of PM 265

association was confirmed biochemically by protein fractionation where the 266

YFP:StREM1.3* protein was exclusively detected in the soluble fraction, while 267

endogenous Remorin remained in the PM fraction (Figure 3D). 268

269

RemCA directly binds to lipid bilayers in vitro. 270

To test whether RemCA anchors StREM1.3 to the PM independently of post-271

translational modifications or association with other proteins, recombinant 272

6His:StREM1.3 (+), 6His:StREM1.31-170 (-, deletion of RemCA), and 6His:StREM1.3* 273

(*, mutated RemCA, see above) were purified from E. coli, and tested for binding to 274

liposomes mimicking plant PM lipid composition in vitro. Liposomes were separated 275

at the top of a sucrose gradient by centrifugation. The presence of StREM1.3 in the 276

liposome-bound or unbound fraction was assessed by Western Blot and lipids were 277



11

analysed by TLC (Figures 4A and S3) (Figure. StREM1.3 bound to liposomes in the 278

absence of any other proteins (Figure 4A). Although similar amounts of liposomes 279

were isolated for each sample (Figure S3B), binding was completely abolished in 280

StREM1.31-170 and StREM1.3* variants (Figure 4A). Furthermore, and in agreement 281

with our data obtained using purified PMs, treatment with denaturing agents did not 282

significantly reduce the amount of 6His:StREM1.3 bound to liposomes (Figure S1B). 283

These results demonstrate that RemCA confers direct membrane binding to 284

StREM1.3. Loss of membrane binding in StREM1.3* indicates that amphipathy plays 285

a crucial role in this process. 286

Insertion of a peptide into membranes results in a partial destabilization of 287

artificial bilayers and can thus be taken as a measure to discriminate between 288

physical insertion and outer association (Longo et al., 1997). To test whether RemCA 289

induces membrane destabilization, we performed lipid-mixing assays in which 290

application of purified peptides to a mixture of liposomes, labelled or not by the 291

lipophilic fluorescent dye octadecyl rhodamine chloride (R18), leads to a de-292

quenching of fluorescence. We confirmed the functionality of this assay by the 293

application of the well-known Simian immunodeficiency virus (SIV) fusion peptide 294

(Lorin et al., 2008) to liposomes (see materials and methods). By contrast liposome 295

treatment with TFE, used as a solvent for the peptides, did not result in lipid mixing. 296

Application of synthetic ultra pure RemCA peptides resulted in a ~40% de-quenching 297

of fluorescence relative to the values obtained with the SIV fusion peptide (Figure 298

4B) indicating destabilization of the membrane and insertion of the peptide in the 299

bilayer. 300

To verify that the increase in fluorescence was due to liposome destabilization rather 301

than liposome aggregation, we performed liposome leakage assays, in which 302

liposomes entrapping the fluorescent dye 1-hydroxypyrene-3,6,8-trisulfonate (HPTS) 303

and the corresponding cationic quencher α,α'-dipyridinium p-xylene dibromide (DPX) 304

were used. Application of fusion peptides leads to a de-quenching of fluorescence 305

due to physical separation of both components (Ellens et al., 1985). We used 306

liposomes composed of either negatively charged lipids or neutral lipids. As 307

expected, the use of the detergent TX100 entirely disrupted the liposomes, leading to 308

maximal fluorescence (Figure 4C). Application of the SIV fusion peptide induced 309

~90% leakage of negatively charged liposomes, and only ~9% leakage of neutral 310

liposomes, well above leakage induced by solvents used for the peptides (leakage 311



12

<2% by either DMSO or TFE for negatively charged and neutral liposomes 312

respectively). Application of synthetic RemCA peptide yielded in ~5% of leakage of 313

negatively charged liposomes, but did not cause any significant leakage of neutral 314

liposomes (Figure 4C). These experiments demonstrate that RemCA physically 315

inserts into membranes independently of other proteins, and predominantly binds 316

negatively charged lipids, suggesting a differential affinity for classes of lipids. 317

318

RemCA confers lipid-specific binding to StREM1.3 319

To test whether RemCA binds selectively to certain lipids, we probed 320

immobilized lipids with purified recombinant 6His:StREM1.3 and 6His:StREM1-170 in 321

lipid overlay assays. StREM1.3 showed differential affinity to 5 of the 7 lipids tested 322

(Figure 5A). Affinity of StREM1.3 was strong to PI(3,4)P2 phosphoinositol (3,4)-323

bisphosphate and sphingosine-1-phosphate (S1P), moderate to phosphatidic acid 324

(PA) and PI(3,5)P2, and weak to phosphatidylserine (PS). Binding to these 325

immobilized lipids was dependent on RemCA since it was almost completely 326

abolished in the StREM1.31-170 protein, with only a weak residual binding to PI(3,4)P2 327

remaining. Similar results were obtained with the purified recombinant 328

6His:StREM1.3* mutant (Figure S4A). These results indicate that RemCA has 329

affinity for minor anionic phospholipids and negatively charged lipids, and that 330

RemCA confers selectivity to the range of lipids bound in vitro. Interestingly, these 331

lipids are enriched in membrane rafts from plants (Vermeer et al., 2009; Furt et al., 332

2010) or animal cells (Hope and Pike, 1996; Pike et al., 2002), suggesting that 333

RemCA lipid-specificity contributes to driving the segregation of StREM1.3 into 334

membrane micro-domains. 335

To support this hypothesis, we first tested the binding of 6His:StREM1.3 to 336

liposomes with lipid composition mimicking either the PM or DIMs in vitro (see 337

materials and methods). Although the total amount of liposomes pulled down was 338

similar for all samples (Figure S4B), the relative quantification of 6His:StREM1.3 339

bound demonstrated a higher affinity for DIM-like liposomes compared to PM-like 340

liposomes (Figure 5B). 341

Finally, to test the specificity of RemCA-mediated binding in vivo, we extracted 342

DIMs from PM purified from tobacco plants expressing RemCA tagged with the 343

soluble YFP. As expected, ~80% of total proteins were excluded from the DIM 344

fraction (Figure 5C, S4C) (see also Raffaele et al., 2009) but the endogenous 345



13

tobacco Remorin proteins were almost exclusively found in DIM fractions (Figure 346

5C). Strikingly, the majority of YFP:RemCA proteins were also detected in DIMs 347

(Figure 5C). These results demonstrate that RemCA is sufficient to target soluble 348

proteins specifically to DIMs of plant PM. 349

350

StREM1.3 forms homo-trimers in vitro independently of RemCA 351

Previous work showed by direct electron microscopy observations and 352

chemical crosslinking with glutaraldehyde that group1 Remorins forms homo-353

polymeric filaments (Bariola et al., 2004). To test whether RemCA plays a role in the 354

assembly or stability of these higher-order polymers, we used exclusion column 355

chromatography to estimate the native molecular weight (MW) of 6His:StREM1.3 and 356

6His:StREM1.3* recombinant proteins. Chromatography revealed a peak of elution at 357

~1.25 mL (Figure S5A). Western blot analysis after denaturing SDS-PAGE of the 358

corresponding fractions confirmed that 6His:StREM1.3 eluted at this volume (Figure 359

S5B). Calibration of the gel filtration column with standards of known MW allowed to 360

estimate 6His:StREM1.3 native MW to ~120 kDa, suggesting that 6His:StREM1.3 361

forms meta-stable trimers in vitro (Figure 6A). Neither monomers nor dimers were 362

detected by gel filtration. Chromatography performed with 6His:StREM1.3* protein 363

showed similar elution pattern, indicating that a non-functional RemCA does not alter 364

the assembly of StREM1.3 into trimers. Interestingly, as discussed previously (Figure 365

S2), RemCA shows structural similarities with influenza virus hemagglutinin (HA) 366

fusion peptide. Influenza HA adopts a central trimeric alpha-helical coiled-coil 367

structure, and the formation of trimers of fusion peptide hairpins is critical for 368

membrane fusion (Cross et al., 2001). To get insights on the possible consequences 369

of StREM1.3 trimer formation on its membrane-binding properties, we used 370

homology modelling and geometry based docking to produce a 3D model of 371

StREM1.3 C-terminal domain (E87-F198) trimer (Figure 6B). This model suggests 372

that RemCAs are organized circularly, forming a hydrophobic core at the centre of 373

RemCA trimer of hairpins and exposing patches of positively charged residues all 374

around the C-terminus of the trimer. This conformation would create a strong 375

hydrophobic moment directed inward RemCA trimer, and likely increases the 376

strength and lipid-specificity of the membrane binding. 377

378



14

Mutations in RemCA* abolish StREM1.3 restriction of PVX cell to cell 379

movement 380

To test for the biological significance of StREM1.3 PM localization, we 381

exploited the previously reported role of StREM1.3 in restricting the cell-to-cell 382

propagation of the potato virus X (PVX) in tomato leaves (Raffaele et al., 2009). 383

Here, we transiently co-expressed in N. benthamiana leaves 35S:StREM1.3 or 384

35S:StREM1.3* with a highly diluted suspension of Agrobacterium tumefaciens 385

harbouring an infectious PVX:GFP clone, so that only isolated individual cells were 386

infected with the PVX:GFP. The accumulation of PVX:GFP in distinct infection foci 387

was visible 5 days post infection (dpi). StREM1.3 or StREM1.3* expression was 388

verified by Western blot (Figure S6). The effect of the expression of the StREM1.3 389

proteins was assessed by visualizing the virus infection foci and counting the number 390

of fluorescent cells of at least 30 infection foci by comparison with plants infiltrated 391

with A. tumefaciens carrying the PVX:GFP clone only. At 5 dpi, plants infiltrated with 392

A. tumefaciens carrying PVX:GFP alone showed infection foci covering 17 cells in 393

average (Figure 7A, B). As expected, the co-expression of StREM1.3 lead to a 394

significant decrease in PVX:GFP spreading, with foci covering 12 cells in average 395

(Figure 7A, B). By contrast, co-expression of StREM1.3* did not significantly alter 396

the size of the PVX:GFP infection foci (Figure 7A, B). StREM1.3 was shown to 397

directly interact with the TGBp1 viral movement protein during PVX infection 398

(Raffaele et al., 2009). To determine if the loss of activity in StREM1.3* was due a 399

loss of interaction with TGBp1, we tested the interaction of the StREM1.3* and 400

StREM1.31-170 mutants with TGBp1 in a Split Ubiquitin assay. In both case, a clear 401

interaction was observed, similar to the wild type StREM1.3 (Figure 7C). These 402

results clearly demonstrate that RemCA-mediated PM anchoring of StREM1.3 is 403

required for the function of this protein in the control of virus propagation and indicate 404

that this activity takes place at the PM. 405

406

DISCUSSION 407

In this paper we showed that StREM1.3 directly binds to the PM via a short C-408

terminal anchor domain (RemCA). This domain is unordered in aqueous solution and 409

spontaneously folds, probably into a tight hairpin of amphipathic alpha helices, in 410

non-polar environments. The amphipathy of RemCA helices is required for StREM1.3 411

PM localization. RemCA selectively binds to lipids enriched in DIMs and targets the 412



15

soluble YFP to DIMs in vivo. StREM1.3 forms trimers in vitro. This does not require a 413

functional RemCA and may increase the stability of PM anchoring in vivo. PM 414

anchoring is crucial for StREM1.3 function in restricting PVX movement since 415

mutations in RemCA abolishing PM binding also abolish StREM1.3 function towards 416

PVX. Molecular mechanisms controlled by raft-mediated signalling and invasion of 417

host cells via membrane micro-domains may be altered by directly targeting proteins 418

to these sites. Using RemCA as an anchor may thus provide the opportunity to 419

modulate raft-mediated signalling processes or alter host resistance towards 420

pathogenic microbes. 421

422

Molecular bases for the assembly of Remorin micro-domains at the PM: a 423

model. 424

Our results indicate that the folding of RemCA into amphipathic alpha helices, 425

probably organized as a tight hairpin, is required for StREM1.3 PM anchoring and 426

function. Our CD data showed that RemCA switches from unordered to alpha helical 427

folding in non-polar environments. RemCA folding could be triggered by proximity 428

with the hydrophobic vicinity of the PM (Figure 8A). Similar observations were made 429

for Fis1 Tail-anchor protein (Suzuki et al., 2003), the ArfGAP1 lipid sensor motif 430

(Bigay et al., 2005), and for the amphipathic membrane-binding helix of the epsin N-431

terminal homology (ENTH) family of clathrin adaptors (Ford et al., 2002). 432

Alternatively, the formation of StREM1.3 trimers would result in the formation of a 433

hydrophobic core, thereby creating an environment favourable to RemCA folding 434

(Figure 8B, C). Association of group 1 Remorins into higher-order complexes has 435

been reported in vitro (Bariola et al., 2004) and in vivo (Raffaele et al., 2009; 436

Lefebvre et al., 2010), but whether this is triggered by external signals remains 437

unknown. We propose that RemCA anchoring occurs as described for other lipid-438

binding proteins (Stahelin et al., 2003; Peter et al., 2004) in a two-step manner. First, 439

positively charged residues E172-A186 increase the affinity of RemCA for negatively 440

charged lipids of the inner leaflet of membrane rafts (Figure 8A, D). This destabilizes 441

the internal monolayer of the PM and allows RemCA to insert deeper in the 442

membrane (Figure 8B, E). The complete folding of RemCA results in the formation of 443

a tight hairpin of alpha helices defining a hydrophobic pocket at its core. As described 444

for Influenza HAfp (Han et al., 2001), formation of this hairpin could result in a deeper 445

insertion of RemCA explaining the intrinsic-like behaviour of Remorins (see Figure 446



16

1B). Considering that Influenza HAfp is non fusogenic if not properly kinked (Lai and 447

Tamm, 2007), the tight closure of RemCA hairpin of helices might be required for 448

stable anchoring in the PM. Finally, the possibility that the formation of StREM1.3 449

multimers occurs at the PM, eventually under control of external signals (Figure 8B, 450

E), would allow the aggregation of membrane rafts into higher order clusters, as 451

suggested during the response to pathogen (Raffaele et al., 2009). 452

453

Biophysical properties defining RemCA lipid-specificity 454

In this work, we have shown that RemCA alone is able to anchor soluble 455

hydrophilic proteins to the PM, such as G/YFP (Figure 1). We cannot exclude that 456

additional structural features of StREM1.3 protein modulate this activity, such as 457

post-translational modifications (e.g. phosphorylation), and the formation of 458

StREM1.3 multimers that would exacerbate local gradients of hydrophobicity and 459

charge. It is remarkable that RemCA preferentially binds to negatively-charged lipids 460

abundant in membrane rafts and target the soluble YFP to DIMs. A number of 461

globular domains is known to bind to the surface of the PM by specifically recognizing 462

certain lipids, particularly discriminating between different phospholipid types 463

(Lemmon, 2008; Kutateladze, 2010). A combination of steric recognition of specific 464

lipid polar head groups, electrostatic attraction to negatively charged lipids and 465

membrane destabilization-mediated penetration govern the function of these 466

domains. Some of those such as the Annexin, ENTH, ANTH (‘AP180 N-terminal 467

Homology’), BAR and PX (‘Phox homology’) domains adopt an alpha-helical fold. 468

Overall the hydrophobicity, the hydrophobic moment, the net charge and helix-469

breaking amino acids contribute to the membrane binding properties of amphipathic 470

helices (Drin and Antonny, 2009). In small GTPases, polybasic clusters provide PM 471

specificity by binding to negatively charged PI(4,5)P2 and PI(3,4,5)P3 of the PM (Heo 472

et al., 2006). The charge of the polar face of ArfGAP1 amphipathic helices has been 473

associated with the ability to perceive membrane curvature, with the introduction of 474

lysine residues at the polar-non polar interface reducing curvature sensitivity and 475

allowing binding to large liposomes (Drin et al., 2007). It is therefore tempting to 476

speculate that RemCA membrane binding could sense membrane curvature as a 477

mean to specifically target certain domains in the membrane. In addition, large 478

hydrophobic residues on both sides of the kink are required for setting the angle of 479

the hairpin structure and for function in Influenza HAfp (Lai and Tamm, 2007). The 480



17

L174, L180, I194 and F198 could play this role in RemCA. The overall conformation 481

of RemCA as well as individual residue side chains are likely to contribute to its 482

membrane binding ability, and further investigations will be required to dissect how 483

this activity is mediated. 484

485

Towards understanding the molecular bases of specific association with 486

membrane rafts 487

RemCA drives the soluble YFP to DIM PM fractions, and lipids bound by 488

RemCA are enriched in DIMs. Indeed, polyphosphoinositides and sphingolipids are 489

enriched in membrane rafts from plants (Vermeer et al., 2009; Furt et al., 2010) and 490

animal cells (Hope and Pike, 1996), the latter being also enriched in PS (Pike et al., 491

2002). The ability to specifically target DIMs might therefore be related to the 492

differential affinity of RemCA to certain negative lipids, as known for SNARE proteins 493

(Mima and Wickner, 2009). However, It should be noted that the presence of a 494

protein in DIMs is not sufficient to associate it with membrane rafts (Munro, 2003; 495

Simons and Gerl, 2010; Tanner et al., 2011). In the case of StREM1.3, we previously 496

reported that enrichment in DIMs correlate with clustering into microdomains of 70 497

nm in the PM detected by immunogold-labeling (Raffaele et al., 2009). Importantly, 498

both DIM enrichment and clustering at the PM were phytosterol-dependent 499

(Kierszniowska et al., 2009; Raffaele et al., 2009; Tanner et al., 2011). Since lipid-500

lipid interactions are sufficient to drive the segregation of liquid-ordered and liquid-501

disordered phases in model membranes (Dietrich et al., 2001; Baumgart et al., 2003), 502

it can be hypothesized that RemCA selective lipid binding properties could be 503

sufficient to restrict the localization of StREM1.3 to membrane rafts. 504

Nevertheless, it is reasonable to assume that in addition to RemCA other 505

domains or motifs may determine the organization of StREM1.3 into PM micro-506

domains. Indeed, the association of Remorins with DIMs is regulated quantitatively 507

by signals such as cold (Minami et al., 2009) and pathogen-associated molecular 508

patterns (PAMPs) (Keinath et al., 2010). The molecular bases of this regulation are 509

not fully understood but may likely involve protein-protein interactions or post 510

translational modifications. It is also not known whether the high-ordered organization 511

of Remorins into multimers is regulated by external signals. In any case, protein-512

protein interactions mediated by the coiled-coil or the N-terminal domain of Remorins 513



18

(Lefebvre et al., 2010; Tóth et al., 2012) likely modulate the localization of Remorins 514

in vivo. 515

516

517

Implications for the mechanisms underlying PVX cell-to-cell movement 518

In a previous study we reported that StREM1.3 restricts potato virus X (PVX) 519

cell-to-cell movement independently of viral replication and that StREM1.3 physically 520

associates in vivo and in vitro with the triple gene bloc (TGB) protein 1 of the virus 521

(Raffaele et al., 2009). Here we show that restriction of PVX movement is abolished 522

in StREM1.3* cytoplasmic mutant. We also show that deletion or mutation of RemCA 523

do not alter StREM1.3 interaction with PVX TGBp1 in a split ubiquitin assay, which is 524

consistent with evidences that RemCA would insert into the PM. This indicates that 525

RemCA is not directly involved in the interaction with TGBp1 but rather that 526

StREM1.3 –TGBp1 interaction takes place at the PM. The sub-cellular localization of 527

TGBp1 is complex and depends on the other TGB proteins, viral RNA, and on the 528

host plant species (Verchot-Lubicz et al., 2010). When expressed alone in N. 529

benthamiana, TGBp1 is mostly nuclear and cytoplasmic. However in the presence of 530

viral RNA, TGBp1 could traffic along the ER with TGBp2 and TGBp3 granules, and 531

reach plasmodesmata. Given that StREM1.3 is anchored on the cytosolic face on the 532

PM, it could associate with cytoplasmic TGBp1 at the periphery of the cell, explaining 533

why we could detect association when TGBp1 was expressed alone. Future work will 534

investigate the specificity of StREM1.3 association with TGBp1 and the impact of 535

viral RNA on this interaction. The recent report of the role of Arabidopsis 536

synaptotagmin 1 (SYT1) in the control of Begomovirus and Tobamovirus cell-to-cell 537

movement (Lewis and Lazarowitz, 2010) shed new light on the involvement host 538

membrane proteins in virus movement. SYT1 resides to the cortical endoplasmic 539

reticulum, endosomes, in DIMs at the PM and in PDs (Schapire et al., 2008; Minami 540

et al., 2009; Lewis and Lazarowitz, 2010; Yamazaki et al., 2010), and regulates 541

endocytosis and PM repair processes (Schapire et al., 2008; Yamazaki et al., 2008; 542

Lewis and Lazarowitz, 2010). These studies support a link between DIMs and PDs, 543

and provide evidence for a role of DIM proteins in cell-to-cell communication via PDs. 544

545

Considering the diversity of proteins in the Remorin family (Raffaele et al., 546

2007), there is probably a corresponding diversity of sub-cellular localizations and 547



19

functions to be discovered. Analysing the variability the RemCA region in relation with 548

Remorins sub-cellular location, the variability of the membrane lipid composition 549

according to the nature of membrane compartments and plant species, and specific 550

functions devoted to each protein in the Remorin family will provide valuable insights 551

in the mechanisms regulating lateral heterogeneity of plant PM. 552

553

MATERIALS AND METHODS 554

Molecular cloning and sequence analyses 555

StREM1.3 was cloned from Potato Solanum tuberosum cv. Desiree cDNAs by 556

standard Gateway techniques using the pDONR221 entry vector and primers 557

containing the recombination site sequences and the gene specific sequences 5’-558

ATGGCAGAATTGGAAGCT-3’ (forward) and 5’-TCAAAATATTCCAAGGAT-3’ 559

(reverse for N-terminal fusion), 5’-AAATATTCCAAGGAT-3’ (reverse for C-terminal 560

fusion), or 5’- TCAACGTTTAGCTTCAATCAT-3’ (reverse for YFP :StREM1-170). A 561

forward primer containing the gene specific sequence 5’-GGAGAAGATCTTCTC-3’ 562

was used to clone the RemCA constructs. StREM1.3* was generated by PCR 563

amplification using a reverse primer containing the sequence 5’- 564

TCAAAATATTACAGAGATTTTCTTTGGAGAAGTTACAGTGGAACGGACTTTTGC-565

3’. Mutated residues were selected at the C-terminus of RemCA to facilitate PCR-566

based mutations. The position of Glycines and the aromatic residue were shown to 567

be important in HA fusion peptide function (Cross et al., 2001), therefore G188V and 568

G196V were introduced to preserve small size while introducing hydrophobic amino-569

acids and Y184V to remove the aromatic residue. A186S and A190S were 570

introduced to preserve the tiny size of amino acids while removing hydrophobic 571

residues and L195S allowed reducing further the local hydrophobicity of RemCA 572

peptide. These mutations allowed reducing the hydrophobic moment from 0.182 to 573

0.112 for the whole RemCA peptide. These entry clones were recombined into the 574

pDEST17 (6His fusions), pH7YWG2, pH7WGY2 or pk7WGF2 (YFP/GFP fusions) 575

vectors. Secondary structure predictions were ran on SSPro 576

(http://download.igb.uci.edu/sspro4.html) and JPred 3 577

(http://www.compbio.dundee.ac.uk/www-jpred/) servers. Coiled-coil domains were 578

predicted using Marcoil1.0 server (http://www.isrec.isb-579

sib.ch/webmarcoil/webmarcoilC1.html).. Amphipathic helix analyses were performed 580

with heliquest server (http://heliquest.ipmc.cnrs.fr/) and 3D modelling via the I-581



20

TASSER server (http://zhanglab.ccmb.med.umich.edu/I-TASSER/) (Roy et al., 2010). 582

StREM1.3 C-terminal domain models were assembled into a trimer using the 583

SymmDock server (Schneidman-Duhovny et al., 2005). Models were rendered using 584

UCSF Chimera (Pettersen et al., 2004). 585

586

Plant transformation and fluorescence microscopy 587

Four week old tobacco plants (Nicotiana tabacum cv. Xanthi) were used for 588

Agrobacterium tumefaciens (strain GV3101)-mediated transient expression (Batoko 589

et al., 2000). Confocal imaging was performed two days after agroinfiltration using 590

Leica TCS SP2/5 confocal microscopes with 20x to 63x oil/water immersion 591

objectives. Laser and image settings were as described by (Raffaele et al., 2009). 592

593

Preparation of pure PM and Detergent Insoluble Membrane (DIM) 594

Microsomal fractions and PMs from tobacco leaves were purified described in 595

(Mongrand et al., 2004). For DIM preparation, TX100 (10 % w/v) was added to a ratio 596

detergent-to-PM proteins of 10 (at 1% w/v final concentration), and the PM 597

preparations were solubilized at 4°C for 30min. Treated PMs were brought to a final 598

concentration of 48 % sucrose (w/w), overlaid successively with layers of 40, 35 and 599

30 % sucrose in TBS buffer (w/w), and then centrifuged for 16 h at 200,000 g in 600

TST41 rotor (SORVALL). DIMs could be recovered above the 30-35 % layers as an 601

opaque band and this fraction was washed in 4 ml of TBS buffer to remove residual 602

sucrose. The protein concentration was determined with a BCA protein assay to 603

avoid TX100 interference, using bovine serum albumin as a protein standard. 604

605

Protein analysis, Western blots and protein purification 606

6-Histidine tagged constructs were expressed in E. coli BL21 DE3 cells. Cells were 607

lysed by ultasonication, and the crude extract was centrifuged at 10,000g for 15 min. 608

6-Histidine tagged proteins were mostly recovered in the soluble fraction. Soluble 609

proteins were then and purified using fast flow chelating sepharose resin (Amersham) 610

according to manufacturer’s instructions. Ultrapure RemCA peptides were obtained 611

by de novo peptide synthesis (GL Biochem Shanghai Ltd; www.glschina.com). 612

Western blot analyses were performed using either anti-REM (Raffaele et al., 2009), 613

anti-PMA (Lefebvre et al., 2004), anti-PIP (Santoni et al., 2003), anti-actin or 614

commercial anti-GFP (Millipore/Roche) antibodies. 615



21

616

Extrinsic PM proteins stripping 617

Stripping on PM preparations and microsomal fractions was performed as described 618

in (Santoni et al., 2003) with minor modifications. Membranes (200 μg) were 619

incubated in 40 ml/1 ml of 5 mM EDTA, 5 mM EGTA, 4 M urea and 5 mM Tris-HCl, 620

pH 9.5, for 5 min on ice before being centrifuged for 20 min at 100,000 g. The 621

subsequent pellet was suspended in 20 mM NaOH, and centrifuged at 100,000 g for 622

20 min. The membranes were then washed in 2 mM EDTA, 2 mM EGTA, 100 mM 623

NaCl and 5 mM Tris-HCl, pH 8, centrifuged at 100 000 g for 20 min, and finally 624

resuspended in loading buffer for SDS-PAGE and Western blot where samples were 625

loaded at equal volumes. The mean protein-purification yield was ~50%. A negative 626

control treated with TBS was added (mock), yielding 95% recovery. Total protein 627

amounts were quantified by Bradford or by image quantification of the Coomassie-628

stained gels. 629

630

Liposome and lipid binding assays 631

Liposomes mimicking the lipid content of tobacco PM were freshly prepared in TBS 632

buffer according to (Doms et al., 1985). For PM-like liposome, the lipid content was 633

phosphatidylethanolamine (PE)/ phosphatidylcholine (PC)/ phosphatidic acid (PA) / 634

glucosylcerebrosides (gluCER) / tobacco mix of free sterols / acyl steryl glucosides 635

(ASG)/ steryl glucosides (SG) at molar ratio 1/0.6/0.2/0.6/0.8/0.5/0.5 and for DIM-like 636

liposomes same lipids at molar ratio 1/0.6/0.2/1.2/2.5/3/1 (Mongrand et al., 2004; Furt 637

et al., 2010). Signals were quantified by the ImageJ software. Standard plant lipids 638

were purchased from SIGMA and MATREYA. 5 µg of purified 6His:StREM1.3, 639

6His:StREM1.31-170 or 6His:StREM1.3* was incubated for 30 min at 30°C with 5 mM 640

liposomes. The samples were further processed in 62% sucrose, and placed at the 641

bottom of a discontinuous sucrose gradient (35/30/15/5% w/w). After centrifugation, 642

liposomes floating at the 15-5% interface (top of the gradient) and fractions at the 643

bottom of the gradient were collected, proteins were precipitated with 15% 644

trichloracetic acid (TCA) and analysed by western blot. To control for liposome 645

recovery, lipids were extracted and analysed by Thin Layer Chromatography as 646

described previously (Mongrand et al., 2004). 647

648

Circular dichroism (CD) measurements 649



22

CD spectra were recorded on a Jasco J-815 CD spectrometer with 10-mm path 650

length quartz cuvettes. Ten scans were taken and automatically averaged in the 651

wavelength range from 190 to 250 nm. Peptide secondary structures were 652

determined using CDpro software package, which involved CDSSTR, SELCON3, 653

and CONTINLL methods (Adam et al., 2004; Zakharov et al., 2004). Percentages 654

were calculated by averaging the percentages provided by the three methods. The 655

peptides stock solutions used for the measurements were diluted in 1 mM Tris buffer 656

at pH 7.4, or in TFE. 657

658

Liposome leakage assays 659

Membrane perturbation and vesicle release can be measured by the assay of (Ellens 660

et al., 1985) based on the quenching of HPTS by DPX. HPTS and DPX were co-661

encapsulated in the aqueous phase of the same liposomes. When leakage of vesicle 662

content occurs, the quenching by DPX stops and subsequently the fluorescence of 663

HPTS increases. Small unilamellar vesicles (SUVs) were used in our experiments. 664

These vesicles were prepared from a solution of multilamellar vesicles (MLV) 665

obtained after hydration for 1 h at 37°C of dry lipid films. These films were mixtures 666

by weight of 26.6 % PC, 26.6 % SM, 26.6 % PE, and 20.2 % CHOL for neutral 667

liposomes and 30 % PC, 30 % PE, 2.5 % PI, 10 % PS, 5 % SM, and 22.5% CHOL for 668

negatively charged liposomes. Large unilamellar vesicles (LUVs) were prepared by 669

the extrusion technique of Mayer et al. The MLV suspension was submitted to five 670

successive cycles of freezing and thawing and thereafter extruded 10 times through 671

stacked polycarbonate filters (pore size, 0.08 μm), under a nitrogen pressure of 20 672

bars using an extruder (Lipex Biomembranes, Vancouver Canada). The 673

concentration of the liposome suspensions was determined by phosphorus analysis. 674

The peptides, dissolved in TFE or HFP/TFE, were added to a mixture of liposomes 675

with encapsulated HPTS and DPX. HPTS fluorescence was measured on a Perkin-676

Elmer LS- 50B fluorimeter (excitation, 360 nm; emission, 520 nm). Liposomes were 677

prepared as described above but were rehydrated with 1 ml of 12.5 mM HPTS (45 678

mM NaCl), 45 mM DPX (20 mM NaCl), and 10 mM Tris-HCl at pH 7.4 and passed 679

through a Sephadex G-75 column to removed un-encapsulated material. 680

681

Lipid mixing assays 682



23

The induction of vesicular lipid mixing by the different peptides was tested with 683

liposomes (SUVs) made of PC/PE/SM/chol at 26.6/26.6/26.6/20.2 molar ratio and 684

PC/PE/PI/PS/SM/CHOL at a molar ratio 30/30/2.1/10/5/22.5. Labelled liposomes 685

were obtained by incorporating R18 in the dry lipid film at a concentration of 6.3 % of 686

the total lipid weight. Labelled and unlabelled liposomes were mixed at a weight ratio 687

of 1:4 and a final concentration of 50 mM in 10mM Tris-HCl, 150 mM NaCl, 0.01% 688

EDTA, 1 mM NaN3 (pH 8). Mixing of liposome membranes was followed by 689

measuring the fluorescence increase of R18, a lipid soluble probe, occurring after the 690

fusion of labelled and unlabelled liposomes, as described in (Lins et al., 2002). 691

Incubation of labelled and unlabelled vesicles in buffer alone did not modify the 692

fluorescence intensity. Fluorescence was recorded at room temperature (excitation: 693

560 nm, emission: 590 nm) on an LS-50B PerkinElmer fluorimeter. 694

695

Size-exclusion chromatography 696

Size-exclusion chromatography was performed on an AKTApurifier apparatus (GE 697

Healthcare). 6His:StREM1.3 and 6his:StREM1.3* purified from E. Coli cultures 698

(1.2g/l) were centrifuged for 5 min at 20,000 g before being processed. The 699

molecular size of the proteins was analysed by chromatography on a Superdex 75 700

10/30 column (GE Healthcare) equilibrated with 20 mM of HEPEs, pH 7,4, 300mM 701

NaCl. Proteins (200 µl) were eluted with a flow rate of 0.2 ml/min and recorded by 702

continuously monitoring the absorbance at 280nm. The column was calibrated with 703

the standard proteins described in Fig.6. 704

705

Viral spreading and split ubiquitin assays 706

To assess spreading of PVX:GFP in N. benthamiana leaves Agrobacterium 707

tumefaciens strain GV3101 carrying the respective StREM1.3 constructs were 708

infiltrated at a final OD600= 0.2 together with the same strain carrying the plasmid 709

pGr208, which expresses the PVX:GFPcDNA, as well as the helper plasmid pSoup 710

(Peart et al., 2002) at final OD600 = 0.001. Spreading of PVX:GFP was visualized by 711

confocal laser-scanning microscopy at 5 days post infection and the number of 712

fluorescent cells of at least 30 PVX:GFP infection foci were counted. Protein extracts 713

from the infiltrated N. benthamiana leaves were subjected to cell fractionation and 714

Western Blot. The experiment was repeated 3 times with same results. Split ubiquitin 715

assays were performed as described in (Raffaele et al., 2009) with yeasts grown on 716



24

SD medium supplement with uracile (50 mg.L-1), leucine (380 mg.L-1), lysine (76 717

mg.L-1) and 5’fluoro-orotic acid (1 g.L-1). 718

719

ACKNOWLEDGEMENTS 720

We would like to thank Mohammed-Amine Belka, Christelle Flore and Jeanny 721

Laroche-Traineau their valuable contributions. We are very grateful to Claudia Popp 722

and Thomas Ott (LMU Munich) for inspiring discussions, shared data, contagious 723

enthusiasm, and a lot of help. We thank all other members of the labs of Sébastien 724

Mongrand (LBM Bordeaux) and Thomas Ott (LMU Munich) for fruitful discussions. 725

We acknowledge the platform Métabolome- Fluxome-Lipidome of Bordeaux 726

(http://www.biomemb.cnrs.fr/INDEX.html 727

https://www.bordeaux.inra.fr/umr619/RMN_index.htm) for contribution to lipid 728

analysis and the unit of plant imaging of Bordeaux Imaging Center (BIC) for confocal 729

imaging. This work has been funded by the French National Agency for Research, 730

ANR program (NT09_517917, ANR PANACEA to SM, JLC). 731

AUTHOR CONTRIBUTION 732

A.P., L.L., S.M. and S.R. designed research. A.P., JL.C., JM.C., S.M. and S.R. 733

performed research. JM.C., L.L. and S.G-R. contributed new reagents and analytic 734

tools. A.P., JM.C, S.M. and S.R. analysed data. S.M. and S.R. wrote the paper. All 735

authors read, edited and approved the manuscript. 736

737

CONFLICT OF INTEREST 738

The authors declare no conflict of interest. 739



25

REFERENCES 740

Adam B, Lins L, Stroobant V, Thomas A, Brasseur R (2004) Distribution of hydrophobic residues is 741 crucial for the fusogenic properties of the Ebola virus GP2 fusion peptide. J Virol 78: 2131-742 2136 743

Bariola P, Retelska D, Stasiak A, Kammerer R, Fleming A, Hijri M, Frank S, Farmer E (2004) 744 Remorins form a novel family of coiled coil-forming oligomeric and filamentous proteins 745 associated with apical, vascular and embryonic tissues in plants. Plant molecular biology 55: 746 579-594 747

Batoko H, Zheng HQ, Hawes C, Moore I (2000) A rab1 GTPase is required for transport between the 748 endoplasmic reticulum and golgi apparatus and for normal golgi movement in plants. Plant 749 Cell 12: 2201-2218 750

Baumgart T, Hess ST, Webb WW (2003) Imaging coexisting fluid domains in biomembrane models 751 coupling curvature and line tension. Nature 425: 821-824 752

Bhat RA, Miklis M, Schmelzer E, Schulze-Lefert P, Panstruga R (2005) Recruitment and 753 interaction dynamics of plant penetration resistance components in a plasma membrane 754 microdomain. Proceedings of the National Academy of Sciences of the United States of 755 America 102: 3135 756

Bigay J, Casella JF, Drin G, Mesmin B, Antonny B (2005) ArfGAP1 responds to membrane 757 curvature through the folding of a lipid packing sensor motif. Embo J 24: 2244-2253 758

Brown DA (2006) Lipid rafts, detergent-resistant membranes, and raft targeting signals. Physiology 759 21: 430 760

Cho W, Stahelin RV (2005) Membrane-protein interactions in cell signaling and membrane trafficking. 761 Annu. Rev. Biophys. Biomol. Struct. 34: 119-151 762

Cordy JM, Hussain I, Dingwall C, Hooper NM, Turner AJ (2003) Exclusively targeting beta-763 secretase to lipid rafts by GPI-anchor addition up-regulates-site processing of the amyloid 764 precursor protein. Proceedings of the National Academy of Sciences 100: 11735 765

Cross KJ, Wharton SA, Skehel JJ, Wiley DC, Steinhauer DA (2001) Studies on influenza 766 haemagglutinin fusion peptide mutants generated by reverse genetics. The EMBO journal 20: 767 4432-4442 768

De Matteis MA, Godi A (2004) PI-loting membrane traffic. Nature cell biology 6: 487-492 769

Dietrich C, Bagatolli L, Volovyk Z, Thompson N, Levi M, Jacobson K, Gratton E (2001) Lipid rafts 770 reconstituted in model membranes. Biophysical journal 80: 1417-1428 771

Doms RW, Helenius A, White J (1985) Membrane fusion activity of the influenza virus hemagglutinin. 772 The low pH-induced conformational change. J Biol Chem 260: 2973-2981 773

Drin G, Antonny B (2009) Amphipathic helices and membrane curvature. FEBS Lett 584: 1840-1847 774

Drin G, Casella JF, Gautier R, Boehmer T, Schwartz TU, Antonny B (2007) A general amphipathic 775 -helical motif for sensing membrane curvature. Nature 200: 7 776

Ellens H, Bentz J, Szoka FC (1985) H+- and Ca2+-induced fusion and destabilization of liposomes. 777 Biochemistry 24: 3099-3106 778

Ford MGJ, Mills IG, Peter BJ, Vallis Y, Praefcke GJK, Evans PR, McMahon HT (2002) Curvature 779 of clathrin-coated pits driven by epsin. Nature 419: 361-366 780

Furt F, Konig S, Bessoule JJ, Sargueil F, Zallot R, Stanislas T, Noirot E, Lherminier J, Simon-781 Plas F, Heilmann I, Mongrand S (2010) Polyphosphoinositides are enriched in plant 782 membrane rafts and form microdomains in the plasma membrane. Plant Physiol 152: 2173-783 2187 784

Gruenberg J (2003) Lipids in endocytic membrane transport and sorting. Current opinion in cell 785 biology 15: 382-388 786



26

Han X, Bushweller JH, Cafiso DS, Tamm LK (2001) Membrane structure and fusion-triggering 787 conformational change of the fusion domain from influenza hemagglutinin. Nature Structural & 788 Molecular Biology 8: 715-720 789

Heo WD, Inoue T, Park WS, Kim ML, Park BO, Wandless TJ, Meyer T (2006) PI (3, 4, 5) P3 and PI 790 (4, 5) P2 lipids target proteins with polybasic clusters to the plasma membrane. Science 314: 791 1458 792

Hope HR, Pike LJ (1996) Phosphoinositides and phosphoinositide-utilizing enzymes in detergent-793 insoluble lipid domains. Mol Biol Cell 7: 843-851 794

Jarsch IK, Ott T (2011) Perspectives on Remorin Proteins, Membrane Rafts, and Their Role During 795 Plant-Microbe Interactions. Molecular Plant-Microbe Interactions 24: 7-12 796

Keinath NF, Kierszniowska S, Lorek J, Bourdais G, Kessler SA, Shimosato-Asano H, 797 Grossniklaus U, Schulze WX, Robatzek S, Panstruga R (2010) PAMP (pathogen-798 associated molecular pattern)-induced changes in plasma membrane compartmentalization 799 reveal novel components of plant immunity. Journal of Biological Chemistry 285: 39140 800

Kierszniowska S, Seiwert B, Schulze WX (2009) Definition of Arabidopsis sterol-rich membrane 801 microdomains by differential treatment with methyl-beta-cyclodextrin and quantitative 802 proteomics. Mol Cell Proteomics 8: 612-623 803

Kutateladze TG (2010) Translation of the phosphoinositide code by PI effectors. Nature chemical 804 biology 6: 507-513 805

Lai AL, Tamm LK (2007) Locking the kink in the influenza hemagglutinin fusion domain structure. 806 Journal of Biological Chemistry 282: 23946 807

Lasserre R, Guo XJ, Conchonaud F, Hamon Y, Hawchar O, Bernard AM, Soudja SMH, Lenne 808 PF, Rigneault H, Olive D (2008) Raft nanodomains contribute to Akt/PKB plasma membrane 809 recruitment and activation. Nature chemical biology 4: 538-547 810

Lefebvre B, Batoko H, Duby G, Boutry M (2004) Targeting of a Nicotiana plumbaginifolia H+ -811 ATPase to the plasma membrane is not by default and requires cytosolic structural 812 determinants. Plant Cell 16: 1772-1789 813

Lefebvre B, Furt F, Hartmann MA, Michaelson LV, Carde JP, Sargueil-Boiron F, Rossignol M, 814 Napier JA, Cullimore J, Bessoule JJ, Mongrand S (2007) Characterization of lipid rafts 815 from Medicago truncatula root plasma membranes: a proteomic study reveals the presence of 816 a raft-associated redox system. Plant Physiol 144: 402-418 817

Lefebvre B, Timmers T, Mbengue M, Moreau S, Herve C, Toth K, Bittencourt-Silvestre J, Klaus 818 D, Deslandes L, Godiard L, Murray JD, Udvardi MK, Raffaele S, Mongrand S, Cullimore 819 J, Gamas P, Niebel A, Ott T (2010) A remorin protein interacts with symbiotic receptors and 820 regulates bacterial infection. Proc Natl Acad Sci U S A 107: 2343-2348 821

Lemmon MA (2008) Membrane recognition by phospholipid-binding domains. Nature Reviews 822 Molecular Cell Biology 9: 99-111 823

Lewis JD, Lazarowitz SG (2010) Arabidopsis synaptotagmin SYTA regulates endocytosis and virus 824 movement protein cell-to-cell transport. Proceedings of the National Academy of Sciences 825 107: 2491-2496 826

Lingwood D, Simons K (2010) Lipid rafts as a membrane-organizing principle. Science 327: 46-50 827

Lins L, Brasseur R (2008) Tilted peptides: a structural motif involved in protein membrane insertion? 828 J Pept Sci 14: 416-422 829

Lins L, Flore C, Chapelle L, Talmud PJ, Thomas A, Brasseur R (2002) Lipid-interacting properties 830 of the N-terminal domain of human apolipoprotein C-III. Protein Eng 15: 513-520 831

Liu J, Elmore JM, Fuglsang AT, Palmgren MG, Staskawicz BJ, Coaker G (2009) RIN4 functions 832 with plasma membrane H+-ATPases to regulate stomatal apertures during pathogen attack. 833 PLoS biology 7: e1000139 834

Longo ML, Waring AJ, Hammer DA (1997) Interaction of the influenza hemagglutinin fusion peptide 835 with lipid bilayers: area expansion and permeation. Biophys J 73: 1430-1439 836



27

Lorieau JL, Louis JM, Bax A (2010) The complete influenza hemagglutinin fusion domain adopts a 837 tight helical hairpin arrangement at the lipid:water interface. Proc Natl Acad Sci U S A 107: 838 11341-11346 839

Lorin A, Lins L, Stroobant V, Brasseur R, Charloteaux B (2008) The minimal fusion peptide of 840 simian immunodeficiency virus corresponds to the 11 first residues of gp32. Journal of Peptide 841 Science 14: 423-428 842

Marín M, Ott T (2012) Phosphorylation of intrinsically disordered regions in remorin proteins. Front 843 Plant Sci. 3:86. 844

Maule AJ (2008) Plasmodesmata: structure, function and biogenesis. Current opinion in plant biology 845 11: 680-686 846

McLaughlin S, Murray D (2005) Plasma membrane phosphoinositide organization by protein 847 electrostatics. Nature 438: 605-611 848

McLaughlin S, Wang J, Gambhir A, Murray D (2002) PIP2 and proteins: interactions, organization, 849 and information flow. Annual review of biophysics and biomolecular structure 31: 151-175 850

Mima J, Wickner W (2009) Complex lipid requirements for SNARE- and SNARE chaperone-851 dependent membrane fusion. J Biol Chem 284: 27114-27122 852

Minami A, Fujiwara M, Furuto A, Fukao Y, Yamashita T, Kamo M, Kawamura Y, Uemura M (2009) 853 Alterations in detergent-resistant plasma membrane microdomains in Arabidopsis thaliana 854 during cold acclimation. Plant Cell Physiol 50: 341-359 855

Mongrand S, Morel J, Laroche J, Claverol S, Carde JP, Hartmann MA, Bonneu M, Simon-Plas F, 856 Lessire R, Bessoule JJ (2004) Lipid rafts in higher plant cells: purification and 857 characterization of Triton X-100-insoluble microdomains from tobacco plasma membrane. J 858 Biol Chem 279: 36277-36286 859

Mongrand S, Stanislas T, Bayer EMF, Lherminier J, Simon-Plas F (2010) Membrane rafts in plant 860 cells. Trends in plant science 15: 656-663 861

Munro S (2003) Lipid rafts: elusive or illusive? Cell 115: 377-388 862

Paulick MG, Bertozzi CR (2008) The Glycosylphosphatidylinositol Anchor: A Complex Membrane-863 Anchoring Structure for Proteins†. Biochemistry 47: 6991-7000 864

Peart JR, Lu R, Sadanandom A, Malcuit I, Moffett P, Brice DC, Schauser L, Jaggard DAW, Xiao 865 S, Coleman MJ (2002) Ubiquitin ligase-associated protein SGT1 is required for host and 866 nonhost disease resistance in plants. Proceedings of the National Academy of Sciences 99: 867 10865 868

Peter BJ, Kent HM, Mills IG, Vallis Y, Butler PJ, Evans PR, McMahon HT (2004) BAR domains as 869 sensors of membrane curvature: the amphiphysin BAR structure. Science 303: 495-499 870

Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE (2004) 871 UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 872 25: 1605-1612 873

Pike LJ, Han X, Chung KN, Gross RW (2002) Lipid rafts are enriched in arachidonic acid and 874 plasmenylethanolamine and their composition is independent of caveolin-1 expression: a 875 quantitative electrospray ionization/mass spectrometric analysis. Biochemistry 41: 2075-2088 876

Porter MY, Koelle MR (2010) RSBP-1 Is a Membrane-targeting Subunit Required by the G {alpha} q-877 specific But Not the G {alpha} o-specific R7 Regulator of G protein Signaling in Caenorhabditis 878 elegans. Molecular biology of the cell 21: 232 879

Raffaele S, Bayer E, Lafarge D, Cluzet S, German Retana S, Boubekeur T, Leborgne-Castel N, 880 Carde JP, Lherminier J, Noirot E, Satiat-Jeunemaitre B, Laroche-Traineau J, Moreau P, 881 Ott T, Maule AJ, Reymond P, Simon-Plas F, Farmer EE, Bessoule JJ, Mongrand S (2009) 882 Remorin, a solanaceae protein resident in membrane rafts and plasmodesmata, impairs 883 potato virus X movement. Plant Cell 21: 1541-1555 884

Raffaele S, Mongrand S, Gamas P, Niebel A, Ott T (2007) Genome-wide annotation of remorins, a 885 plant-specific protein family: evolutionary and functional perspectives. Plant Physiol 145: 593-886 600 887



28

Reymond P, Kunz B, Paul-Pletzer K, Grimm R, Eckerskorn C, Farmer EE (1996) Cloning of a 888 cDNA encoding a plasma membrane-associated, uronide binding phosphoprotein with 889 physical properties similar to viral movement proteins. Plant Cell 8: 2265-2276 890

Rossin A, Kral R, Lounnas N, Chakrabandhu K, Mailfert S, Marguet D, Hueber AO (2010) 891 Identification of a lysine-rich region of Fas as a raft nanodomain targeting signal necessary for 892 Fas-mediated cell death. Experimental cell research 316: 1513-1522 893

Roy A, Kucukural A, Zhang Y (2010) I-TASSER: a unified platform for automated protein structure 894 and function prediction. Nat Protoc 5: 725-738 895

Sankarshanan M, Ma Z, Iype T, Lorenz U (2007) Identification of a novel lipid raft-targeting motif in 896 Src homology 2-containing phosphatase 1. The Journal of Immunology 179: 483 897

Santoni V, Vinh J, Pflieger D, Sommerer N, Maurel C (2003) A proteomic study reveals novel 898 insights into the diversity of aquaporin forms expressed in the plasma membrane of plant 899 roots. Biochem J 373: 289-296 900

Schapire AL, Voigt B, Jasik J, Rosado A, Lopez-Cobollo R, Menzel D, Salinas J, Mancuso S, 901 Valpuesta V, Baluska F (2008) Arabidopsis synaptotagmin 1 is required for the maintenance 902 of plasma membrane integrity and cell viability. The Plant Cell 20: 3374-3388 903

Schneidman-Duhovny D, Inbar Y, Nussinov R, Wolfson HJ (2005) PatchDock and SymmDock: 904 servers for rigid and symmetric docking. Nucleic acids research 33: W363-W367 905

Seong J, Ouyang M, Kim T, Sun J, Wen PC, Lu S, Zhuo Y, Llewellyn NM, Schlaepfer DD, Guan 906 JL (2011) Detection of focal adhesion kinase activation at membrane microdomains by 907 fluorescence resonance energy transfer. Nature Communications 2: 406 908

Simon-Plas F, Perraki A, Bayer E, Gerbeau-Pissot P, Mongrand S (2011) An update on Plant 909 Membrane rafts. Current opinion in plant biology 14: 642-649 910

Simons K, Gerl MJ (2010) Revitalizing membrane rafts: new tools and insights. Nature Reviews 911 Molecular Cell Biology 11: 688-699 912

Stahelin RV, Long F, Peter BJ, Murray D, De Camilli P, McMahon HT, Cho W (2003) Contrasting 913 membrane interaction mechanisms of AP180 N-terminal homology (ANTH) and epsin N-914 terminal homology (ENTH) domains. Journal of Biological Chemistry 278: 28993 915

Suzuki M, Jeong SY, Karbowski M, Youle RJ, Tjandra N (2003) The solution structure of human 916 mitochondria fission protein Fis1 reveals a novel TPR-like helix bundle. J Mol Biol 334: 445-917 458 918

Tanner W, Malinsky J, Opekarová M (2011) In plant and animal cells, detergent-resistant 919 membranes do not define functional membrane rafts. The Plant Cell 23: 1191-1193 920

Thompson GA, Jr., Okuyama H (2000) Lipid-linked proteins of plants. Prog Lipid Res 39: 19-39 921

Tóth K, Stratil TF, Madsen EB, Ye J, Popp C, Antolín-Llovera M, Grossmann C, Jensen ON, 922 Schüßler A, Parniske M (2012) Functional Domain Analysis of the Remorin Protein 923 LjSYMREM1 in Lotus japonicus. PLoS One 7: e30817 924

Verchot-Lubicz J, Torrance L, Solovyev AG, Morozov SY, Jackson AO, Gilmer D (2010) Varied 925 movement strategies employed by triple gene block-encoding viruses. Molecular Plant-926 Microbe Interactions 23: 1231-1247 927

Vermeer JE, Thole JM, Goedhart J, Nielsen E, Munnik T, Gadella TW, Jr. (2009) Imaging 928 phosphatidylinositol 4-phosphate dynamics in living plant cells. Plant J 57: 356-372 929

Widjaja I, Naumann K, Roth U, Wolf N, Mackey D, Dangl JL, Scheel D, Lee J (2009) Combining 930 subproteome enrichment and Rubisco depletion enables identification of low abundance 931 proteins differentially regulated during plant defense. Proteomics 9: 138-147 932

Yamazaki T, Kawamura Y, Minami A, Uemura M (2008) Calcium-dependent freezing tolerance in 933 Arabidopsis involves membrane resealing via synaptotagmin SYT1. The Plant Cell 20: 3389-934 3404 935



29

Yamazaki T, Takata N, Uemura M, Kawamura Y (2010) Arabidopsis synaptotagmin SYT1, a type I 936 signal-anchor protein, requires tandem C2 domains for delivery to the plasma membrane. 937 Journal of Biological Chemistry 285: 23165 938

Zakharov SD, Kotova EA, Antonenko YN, Cramer WA (2004) On the role of lipid in colicin pore 939 formation. Biochim Biophys Acta 1666: 239-249 940

941



30

FIGURE LEGENDS 942

943 Figure 1. A C-terminal anchor peptide (RemCA) from StREM1.3 Remorin is 944

required and sufficient for strong plasma membrane (PM) anchoring in vivo. (A) 945

Predicted secondary structure and protein domains in StREM1.3. Numbers indicate 946

amino acid position. (B) Stripping of peripheral PM proteins is not sufficient to wash 947

out StREM1.3. Purified PMs from tobacco leaves were washed according to Santoni 948

et al. 2003 (Stripped) or treated with TBS (Mock). In average, 50% (n=3) of the PM-949

associated proteins, such as Actin, were stripped whereas the quantity of Remorin 950

protein associated to the PM was unaffected, as for the Plasma membrane Intrinsic 951

Protein 2 (PIP2) aquaporin intrinsic PM protein. (C) Subcellular localization of YFP-952

tagged StREM1.3, truncated StREM1.3 comprising amino-acids 1 to 170, and 953

truncated StREM1.3 comprising amino acids 171 to 198 (RemCA peptide only) 954

transiently expressed in tobacco epidermal cells. Scale bars indicate 20 µm. (D) Cell 955

fractionation of the constructs shown in C. Crude protein extracts (CE) from tobacco 956

leaf cells expressing each construct were fractionated into soluble (sol.) and 957

microsomal (µ) compartments by centrifugation (15µg of proteins per lane). In spite of 958

various attempts to solve this problem, we were not able to prevent partial cleavage 959

of the YFP:RemCA construct, resulting in a free YFP signal as shown. 960

961

Figure 2. Non-polar environments trigger conformational changes in the 962

RemCA peptide. (A) Circular dichroism spectra of synthetic ultrapure RemCA 963

peptide in various environments. The spectral signatures of alpha helices (at 195, 964

208 and 222nm) are indicated by arrows. Spectra were recorded in aqueous (100% 965

Tris-HCl), semi-hydrophobic (50% TFE) and hydrophobic (100% TFE) environments. 966

(B) Estimated distribution of secondary folds in RemCA peptide in the different 967

environments assayed. (C) Sequence and predicted structure of the RemCA peptide. 968

Homology modeling predicts a tight hairpin of alpha helices for RemCA, consistent 969

with the circular dischroism spectra collected in hydrophobic environments. The 970

hydrophobicity of residues (Kyte-Doolitle scale) is indicated by a color-code. TFE, 971

2,2,2 trifluoroethanol. 972

973

Figure 3. Amphipathy of RemCA alpha-helices plays a critical role in anchoring 974

StREM1.3 Remorin to the plasma membrane. (A) A helical wheel projection of 975



31

RemCA as a hairpin of α-helices showing amphipathy of each helix. Overall helix 976

hydrophobicity (H) and hydrophobic moment (µH) are indicated in the middle of the 977

helices, with an arrow showing the direction and intensity of the hydrophobic 978

moments. (B) Ribbon and surface models of the RemCA and mutated RemCA* 979

peptides with hydrophobicity (Kyte-Doolitle scale) shown by a color-code. The 980

surface model of RemCA shows the predicted hydrophobic core in red, and charged 981

surfaces in blue. Substitutions introduced in the RemCA* mutant aimed at reducing 982

the amphipathy of the α-helices by altering the distribution of hydrophobic residues 983

(Y184V, A190S, and L195S) and positively charged residues (A186S, G188V and 984

G196V). (C) Maximum projection of confocal scans through N. benthamiana 985

epidermal cells expressing YFP:StREM1.3* mutant, with plastids shown in red. Scale 986

bar indicates 20 µm. (D) Cell fractionation analysis of cells expressing GFP tagged 987

StREM1.3 with native (+) and mutated (*) RemCA sequence, confirming that 988

mutations introduced in StREM1.3* abolish PM localization. 989

990

Figure 4. RemCA confers intrinsic lipid binding to StREM1.3 Remorin. (A) 991

Affinity purified 6His:StREM1.3 (WT) binds to protein-free liposomes in vitro in a 992

RemCA-dependent manner. Liposome binding is abolished in 6His:StREM1.3 993

recombinant protein lacking RemCA (∆CA) or harbouring a mutated RemCA (*). The 994

soluble (Sol.) and liposome-bound (Bound) fractions recovered after centrifugation 995

are shown. Lipid mixing assays (B) and liposome leakage assays (C) show that 996

RemCA ultrapure peptides insert into liposome membranes containing negatively 997

charged lipids, as indicated by an increase in fluorescence de-quenching. By 998

contrast, liposome leakage assays (C) showed that RemCA is not able to insert into 999

neutral liposome membranes. SIV, Simian influenza virus fusion peptide; TFE, 1000

trifluoroethanol; TX100, Triton X-100. Solvent used in (C) was dimethyl sulfoxide for 1001

negatively charged liposomes and TFE for neutral liposomes. Error bars show 1002

standard deviation of 6-8 independent measurements. 1003

1004

Figure 5. RemCA targets soluble proteins to the PM in a lipid-specific manner. 1005

(A) Lipid overlay assays performed with 6His tagged StREM1.3 and StREM1.31-170 1006

show a differential affinity of StREM1.3 for various classes of lipids, dependent on 1007

RemCA. The right panel shows Western Blot of the protein input material illustrating 1008

equal loading for StREM1.3 and StREM1.31-170. PA, phosphatidic acid; PIP3, 1009



32

phosphoinositol 3,4,5, triphosphate; PIP2, Phosphoinositol bisphosphate, the position 1010

of phosphate groups is indicated below; S1P, sphingosine 1-phosphate. (B) 1011

6His:StREM1.3 shows higher affinity for liposomes with a composition mimicking 1012

detergent insoluble membranes (DIM) compared to plasma membrane (PM - 1013

described in materials and methods). Histograms show quantifications and standard 1014

error based on Western blot signals for three independent experiments. Differential 1015

affinity is significant with p<0.05 in a Student t-test. (C) RemCA-tagged GFP 1016

transiently expressed in tobacco leaves segregates into the DIM fractions of the PM, 1017

similar to the endogenous Remorin. Total proteins were quantified by bicinchoninic 1018

acid assay (BCA) relative to total proteins in purified PM (error bars show standard 1019

deviation from three independent replicates). µ, microsomal fraction; DSM, detergent 1020

soluble membranes. 1021

1022

Figure 6. StREM1.3 Remorin forms homo-trimers independently of RemCA in 1023

vitro. (A) Exclusion column chromatography of 6His:StREM1.3, 6His:StREM1.3* and 1024

elution standards allows to estimate the native MW of 6His:StREM1.3 to ~120 kDA, 1025

suggesting that it assembles into trimers, and shows that mutations introduced in 1026

6His:StREM1.3* do not alter the ability to form trimers. (B) A trimer model of the C-1027

terminal domain of StREM1.3. Views from the C-terminal end of the trimer with 1028

residues colored according to hydrophobicity (Kyte-Doolitle scale) reveal the 1029

formation of a hydrophobic core (in red) at the centre of the trimer, with positive 1030

surface patches surrounding the RemCA region, in blue. Standards used in A: 1, blue 1031

dextran (2MDa); 2, Thyroglobuline (670kDa); 3, Catalase (232kDa); 4, Ovalbumine 1032

(43kDa); 5, Chymotrypsinogene (25kDa); 6, soybean Trypsin inhibitor (21.5kDa); 7, 1033

Ribonuclease (13.7kDa); 8, Cytochrome C (12.3kDa). 1034

1035

Figure 7. RemCA-mediated plasma membrane binding is required for StREM1.3 1036

restriction of PVX cell to cell propagation. (A) Expression of PVX:GFP alone and 1037

in combination with StREM1.3 or the mutated StREM1.3* variant unable to bind to 1038

the plasma membrane. Images of infection foci were taken at 5 days post infection 1039

(dpi) of N. benthamiana leaves. Scale bars indicate 150µm. (B) Quantification of PVX 1040

cell to cell spreading showing the average number of cells in the infection foci at 5 1041

dpi. For each treatment, 90 pictures were analysed, with error bars showing standard 1042

deviation. Significance was assessed by Dunn’s multiple comparison test (***, 1043



33

p<0.001; n=3). (C) Split ubiquitin assays showing that StREM1.31-170 and StREM1.3* 1044

interact with PVX TGBp1 movement protein similar to the wild type StREM1.3. 1045

1046

Figure 8. Model for the molecular mechanisms leading to the assembly of 1047

StREM1.3 micro-domains at the plasma membrane (PM). Conformational 1048

changes in RemCA, leading to folding into amphipathic alpha helices, are required to 1049

anchor StREM1.3 to the PM. These conformational changes are triggered by non-1050

polar environments, such as the vicinity of the PM (A). Helix E172-A186 folds 1051

exposing clusters of positively charged residues (+). These residues establish 1052

electrostatic interactions with the negatively charged lipids of the inner leaflet of 1053

membrane rafts, destabilizing the PM and allowing insertion of RemCA in the bilayer 1054

(B). Hydrophobicity triggers RemCA hairpin formation, creating an hydrophobic 1055

pocket (small red star) that deepens and strengthen insertion in the membrane. 1056

Alternatively, an external signal may trigger the formation of StREM1.3 trimers (C). 1057

The hydrophobic core of the trimer (large yellow star) favours folding of helix E172-1058

A186 of RemCA (D), leading to PM anchoring as previously described. The formation 1059

of trimers, either prior to or after PM anchoring, is likely to increase the strength of 1060

the lipid binding, may promote membrane domain aggregation, lead to membrane 1061

deformation or mediate the assembly of protein complexes at the PM via Remorin 1062

protein-protein interaction domains (E). 1063

1064

SUPPLEMENTAL DATA 1065

1066

Figure S1. Membrane washing experiments show that StREM1.3 strongly binds 1067

to purified plasma membrane and liposomes. (A) Anti-Remorin Western blot on 1068

tobacco leaf purified plasma membranes incubated for 30 min in water, in Tris-HCl 1069

buffer saline (TBS) containing NaCl, urea, TBS of increasing pH and TBS containing 1070

Ethylene diamine tetra acetic acid (EDTA). (B) Anti-Remorin Western blot on 1071

liposomes supplemented with 1 µg of purified 6His:StREM1.3 (input). These samples 1072

were incubated in TBS in H2O or in TBS containing 1 M NaCl or 6 M Urea for 30 min 1073

at 4°C, and centrifuged. 1074

1075

Figure S2. Sequence and structure comparison between StREM1.3 RemCA and 1076

Influenza hemagglutinin fusion peptide (HAfp). StREM1.3 homology model (in 1077



34

blue) was manually superimposed with HAfp structure (in yellow, Lorieau et al., 1078

2010). Kyte-Doolittle scale is used to illustrate the hydrophobicity of residues along 1079

the alignment and in one of the structure views. 1080

1081

Figure S3. Controls for the production of purified recombinant proteins and 1082

liposome binding assays. (A) 6His-tagged StREM1.3 constructs were expressed in 1083

BL21 DE3 E. Coli cells, with induction controlled by addition of isopropyl β-D-1-1084

thiogalactopyranoside (IPTG - top left panel). Red arrowheads indicate the size 1085

expected for StREM1.3 constructs. Cells were lysed by ultrasonication (CE) and 1086

centrifuged. StREM1.3 variants were mostly recovered in the supernatant (Sup.) but 1087

not the pellet (Pel.). Soluble proteins were then processed using fast flow chelating 1088

sepharose resin with immidazole concentrations as indicated on the figure (mM). (B) 1089

To control the recovery of the liposomes, the lipids of half of the liposomes 1090

supplemented with StREM1.3, StREM1.31-170 or StREM1.3* were extracted and 1091

analysed by Thin Layer Chromatography as described in Mongrand et al., 2004. 1092

Lipids were identified by comparison with commercial standards. ASG, acyl steryl 1093

glucoside; DIM, detergent insoluble membranes; gluCER, glucosyl ceramide; PA, 1094

phosphatidic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; SG, 1095

Steryl glucoside. 1096

1097

Figure S4. Controls for lipid-specific binding of StREM1.3. (A) Lipid overlay 1098

assay comparing phospholipid binding properties of recombinant 6His:StREM1.3 and 1099

6His:StREM1.3*. Total protein used in the assay is shown in the top panel. (B) To 1100

control for the recivery of the liposomes in the liposome binding assays, the lipids of 1101

half the liposomes mimicking PM or DIM lipid composition were extracted and 1102

analyzed by Thin Layer Chromatography as described in Mongrand et al. 2004. 1103

Lipids were identified by comparison with commercial standards. ASG, acyl steryl 1104

glucoside; DIM, detergent insoluble membranes; gluCER, glucosyl ceramide; PA, 1105

phosphatidic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; SG, 1106

Steryl glucoside. (C) Segregation of the PMA proton pump and Remorin along a 1107

sucrose gradient showing distribution of these proteins between Detergent Insoluble 1108

Membranes (DIM, fraction 1) and Detergent Soluble Membranes (DSM, fractions 2 to 1109

4). 1110

1111



35

Figure S5. StREM1.3 Remorin forms trimers in vitro. (A) Exclusion column 1112

chromatogram obtained for 6His:StREM1.3. (B) Fractions A8, A9, B4, B5, C3 and C4 1113

were analysed by SDS-PAGE (denaturing conditions) followed by anti-Remorin 1114

western blot showing that StREM1.3 elutes in fractions B4 and B5. 1115

1116

Figure S6. Western Blot controls for the expression of StREM1.3 and 1117

StREM1.3* in PVX:GFP propagation experiments. Protein extracts (10 µg) from 1118

infected tobacco leaves with PVX:GFP alone or co-expressing StREM1.3 or 1119

StREM1.3* were used for Western Blot analysis. 1120

1121

Supplemental File 1. Protein Databank file containing atom coordinates for the 1122

predicted StREM1.3 structure. 1123

1124

Supplemental File 2. Protein Databank file containing atom coordinates for the 1125

predicted StREM1.3* structure. 1126



Figure 1. A C-terminal anchor peptide (RemCA) from StREM1.3 Remorin is required and sufficient for strong plasma membrane (PM) anchoring in vivo. (A) Predicted secondary structure and protein domains in StREM1.3. Numbers indicate amino acid position. (B) Stripping of peripheral PM proteins is not sufficient to wash out StREM1.3. Purified PMs from tobacco leaves were washed according to Santoni et al. 2003 (Stripped) or treated with TBS (Mock). In average, 50% (n=3) of the PM-associated proteins, such as Actin, were stripped whereas the quantity of Remorin protein associated to the PM was unaffected, as for the Plasma membrane Intrinsic Protein 2 (PIP2) aquaporin intrinsic PM protein. (C) Subcellular localization of YFP-tagged StREM1.3, truncated StREM1.3 comprising amino-acids 1 to 170, and truncated StREM1.3 comprising amino acids 171 to 198 (RemCA peptide only) transiently expressed in tobacco epidermal cells. Scale bars indicate 20 μm. (D) Cell fractionation of the constructs shown in C. Crude protein extracts (CE) from tobacco leaf cells expressing each construct were fractionated into soluble (sol.) and microsomal (μ) compartments by centrifugation (15μg of proteins per lane). In spite of various attempts to solve this problem, we were not able to prevent partial cleavage of the YFP:RemCA construct, resulting in a free YFP signal as shown.

YFP:StREM1.3 YFP:StREM1.31-170 YFP:StREM1.3171-198

(RemCA)

A B

C

D

100

75

50

25

α-Rem36 KDa

Total PMproteins

StrippedMock

α-Actin42 KDa

α-PIP230 KDa

- GFP:StREM1.3 (63 kDa)

CE sol. μ

α-GFP

- YFP:StREM1.31-170 (60 kDa)

- YFP:RemCA (30 kDa)- YFP (27 kDa)

1

198

85 169

172 187

192 195

alpha helixno secondary structure

116 152

predictedcoiled-coil

Remorin_N(PF03766)

Remorin_C(PF03763)

27 84

RemCA

Pfamdomains:

N C



Figure 2. Non-polar environments trigger conformational changes in the RemCA peptide. (A) Circular dichroism spectra of synthetic ultrapure RemCA peptide in various environments. The spectral signatures of alpha helices (at 195, 208 and 222nm) are indicated by arrows. Spectra were recorded in aqueous (100% Tris-HCl), semi-hydrophobic (50% TFE) and hydrophobic (100% TFE) environments. (B) Estimated distribution of secondary folds in RemCA peptide in the di�erent environments assayed. (C) Sequence and predicted structure of the RemCA peptide. Homology model-ing predicts a tight hairpin of alpha helices for RemCA, consistent with the circular dischroism spectra collected in hydrophobic environments. The hydrophobicity of residues (Kyte-Doolitle scale) is indicated by a color-code. TFE, 2,2,2 tri�uoroethanol.

GEDLLKAEELAAKYRATGTAPKKILGIF

Wavelength (nm)190 250200 220 240

100

-50

0

50

Mol

ar E

llipt

icity

B

A

171

G171

210 230

C Hydrophobicity:

-4.5 +4.5

181 191

0

20

40

60

80

100

Estimated proportion of RemCA peptide (%)

100%TrisHCL

50% TFE 100% TFE

Solvent for CD spectra collection

Unordered

TurnBeta sheet

Alpha helix

E172

D173 L174

L175

K176

A177

E178

E179

L180 A181

A182

K183

Y184

R185A186

T187

G188

T189

A190

P191

K192

K193

I194

L195

G196

I197

F198

90°

Tris-HCl

100% TFE50% TFE



S

YFP:StREM1.3* D

GFP:StREM1.3

endogenousREM WT * WT * WT *

CE μ Soluble

A

C

hydrophobic core

charged surface

N

C

90°

-36

-70

BGE

LD

L

L

K

A

E

E

A

A

K

Y

R

A

T

GT

APP

KK

I

L

G

I

N

C

Hydrophobicity: -4.5 +4.5

chargedface

chargedface

hydrophobicface

hydrophobicface

H = 0.104μH = 0.153

H = 0.682μH = 0.661

F

Rem

CARe

mCA

*

L195SG196V

A190S

G188V

Y184VA186SN

C

Figure 3. Amphipathy of RemCA alpha-helices plays a critical role in anchoring StREM1.3 Remorin to the plasma membrane. (A) A helical wheel projection of RemCA as a hairpin of α-helices showing amphipathy of each helix. Overall helix hydrophobicity (H) and hydrophobic moment (μH) are indicated in the middle of the helices, with an arrow showing the direction and intensity of the hydrophobic moments. (B) Ribbon and surface models of the RemCA and mutated RemCA* peptides with hydrophobicity (Kyte-Doolitle scale) shown by a color-code. The surface model of RemCA shows the predicted hydrophobic core in red, and charged surfaces in blue. Substitutions introduced in the RemCA* mutant aimed at reducing the amphipathy of the α-helices by altering the distribution of hydrophobic residues (Y184V, A190S, and L195S) and positively charged residues (A186S, G188V and G196V). (C) Maximum projec-tion of confocal scans through N. benthamiana epidermal cells expressing YFP:StREM1.3* mutant, with plastids shown in red. Scale bar indicates 20 μm. (D) Cell fractionation analysis of cells expressing GFP tagged StREM1.3 with native (+) and mutated (*) RemCA sequence, confirming that mutations introduced in StREM1.3* abolish PM localization.



Figure 4. RemCA confers intrinsic lipid binding to StREM1.3 Remorin. (A) A�nity puri�ed 6His:StREM1.3 (WT) binds to protein-free liposomes in vitro in a RemCA-dependent manner. Liposome binding is abolished in 6His:StREM1.3 recombinant protein lacking RemCA (∆CA) or harbouring a mutated RemCA (*). The soluble (Sol.) and liposome-bound (Bound) fractions recovered after centrifugation are shown. Lipid mixing assays (B) and liposome leakage assays (C) show that RemCA ultrapure peptides insert into liposome membranes containing negatively charged lipids, as indicated by an increase in �uorescence de-quenching. By contrast, liposome leakage assays (C) showed that RemCA is not able to insert into neutral liposome membranes. SIV, Simian in�uenza virus fusion peptide; TFE, tri�uoroethanol; TX100, Triton X-100. Solvent used in (C) was dimethyl sulfoxide for negatively charged liposomes and TFE for neutral liposomes. Error bars show standard deviation of 6-8 independent measurements.

A

-40

α-Re

m

B

TX100

100

10

1

100

75

0 SIV

Rem

CA

TFE

50

25

Relat

ive �u

ores

cenc

ede

-que

nchin

g

C

SIV Solvent RemCA

Relat

ive �u

ores

cenc

ede

-que

nchin

g

negatively chargedliposomes

neutral liposomes

WT*∆CAInput

BoundSol. BoundSol. BoundSol.

-37

∆CA * WT



Figure 5. RemCA targets soluble proteins to the PM in a lipid-speci�c manner. (A) Lipid overlay assays performed with 6His tagged StREM1.3 and StREM1.31-170 show a di�erential a�nity of StREM1.3 for various classes of lipids, dependent on RemCA. The right panel shows Western Blot of the protein input material illustrating equal loading for StREM1.3 and StREM1.31-170. PA, phosphatidic acid; PIP3, phosphoinositol 3,4,5, triphosphate; PIP2, Phosphoinositol bisphosphate, the position of phosphate groups is indicated below; S1P, sphingosine 1-phosphate. (B) 6His:StREM1.3 shows higher a�nity for liposomes with a composition mimicking detergent insoluble membranes (DIM) compared to plasma membrane (PM - described in materials and methods). Histograms show quanti�cations and standard error based on Western blot signals for three independent experiments. Di�erential a�nity is signi�cant with p<0.05 in a Student t-test. (C) RemCA-tagged GFP transiently expressed in tobacco leaves segregates into the DIM fractions of the PM, similar to the endogenous Remorin. Total proteins were quanti�ed by bicinchoninic acid assay (BCA) relative to total proteins in puri�ed PM (error bars show standard deviation from three independent replicates). µ, microsomal fraction; DSM, detergent soluble membranes.

BLiposomes + 6his:StREM1.3Input

PM-likeliposomes

DIM-likeliposomes

µ PM DSM DIM

α-GFP

α-REM

30

20

10

0

% o

f inp

ut b

ound

PM-like DIM-like

endogenousREM

RemCA:GFP

C

-40

-36

-30

Blank PS PA PIP3

PIP2

S1P3,54,5 3,4

StREM1.3

StREM1.31-170

A

Input

StREM1.3

StREM1.31-170

-40-37

**

0

50

100% of totalproteins in

PM



Figure 6. StREM1.3 Remorin forms homo-trimers independently of RemCA in vitro. (A) Exclusion column chromatography of 6His:StREM1.3, 6His:StREM1.3* and elution standards allows to estimate the native MW of 6His:StREM1.3 to ~120 kDA, suggesting that it assembles into trimers, and shows that mutations introduced in 6His:StREM1.3* do not alter the ability to form trimers. (B) A trimer model of the C-terminal domain of StREM1.3. Views from the C-terminal end of the trimer with residues colored according to hydrophobicity (Kyte-Doolitle scale) reveal the formation of a hydrophobic core (in red) at the centre of the trimer, with positive surface patches surrounding the RemCA region, in blue. Standards used in A: 1, blue dextran (2MDa); 2, Thyroglobuline (670kDa); 3, Catalase (232kDa); 4, Ovalbumine (43kDa); 5, Chymotrypsinogene (25kDa); 6, soybean Trypsin inhibitor (21.5kDa); 7, Ribonuclease (13.7kDa); 8, Cytochrome C (12.3kDa).

N

N N

C

C

C

B E87

E87

E87F198

RemCA

90°F198

F198

F198

Monomer #3

Monomer #1Monomer #2

180°

Hydrophobicity:

-4.5

+4.5

hydrophobic corepositive surfacepatches

A

F198F198

0.8 0.9 1.010

100

1,000

10,000

Elution volume (mL)M

olec

ular

Wei

ght (

log

kDa)

1.1 1.2 1.3 1.4 1.5 1.6 1.7

1

23

45

6

7 8

6His:StREM1.36His:StREM1.3*



Number of PVX:GFP infected cells

0

5

10

15

20

25

***

Control p35S-StREM1.3

p35S-StREM1.3*

B

Figure 7. RemCA-mediated plasma membrane binding is required for StREM1.3 restriction of PVX cell to cell propagation. (A) Expression of PVX:GFP alone and in combination with StREM1.3 or the mutated StREM1.3* variant unable to bind to the plasma membrane. Images of infection foci were taken at 5 days post infection (dpi) of N. bentha-miana leaves. Scale bars indicate 150µm. (B) Quanti�cation of PVX cell to cell spreading showing the average number of cells in the infection foci at 5 dpi. For each treatment, 90 pictures were analysed, with error bars showing standard deviation. Signi�cance was assessed by Dunn’s multiple comparison test (***, p<0.001; n=3). (C) Split ubiquitin assays showing that StREM1.31-170 and StREM1.3* interact with PVX TGBp1 movement protein similar to the wild type StREM1.3.

StREM1.3

StREM1.3*

O.D.

1 10-1 10-2

Ø TGBp1

ØStREM1.31-170

StREM1.31-170 TGBp1

Bait Prey

TGBp1

TGBp1

A PVX:GFP

PVX:GFP /StREM1.3

PVX:GFP /StREM1.3*

C





Documents

1 Running head: Membrane anchor of StREM1.3 Remorin · 25 Bordeaux, 146 rue Léo Saignat, 33076 Bordeaux Cedex, France (SR, AP, JLC and 26 SM) 27 28 The Sainsbury Laboratory, John