Supporting Materials and Methods Expression

Supporting Materials and Methods Expression constructs Plasmids for transient overexpression under the CMV promoter were produced by standard molecular biological methods. A monomeric variant of GFP (mGFP, carrying the single amino acid substitution A206K to prevent dimerization (1)) was fused N-terminally to the sequence of full length rat SNAP25B (AB003992) or SNAP25B constructs in a vector based on pEGFP-C1 (Clontech, Mountain View, CA) (GenBank accession No. U55763). All fusion proteins contained a linker of 5 amino acids (RSRAL) between mGFP and the N-terminus of SNAP25B. The constructs used for transient overexpression coded for the following fusion proteins: SNAP25 [mGFP + SNAP25B-(1-206)]; SNAP25-NL0 [mGFP + SNAP25B-(1-142 + 203-206)]; SNAP25-0LC [mGFP + SNAP25B-(113 + 80-206)]; SNAP25-0L0 [mGFP + SNAP25B-(1-13 + 80-142 + 203-206)]; SNAP25-NLN [mGFP + SNAP25B(1-142 + 14-79 + 203-206)]; SNAP25-CLC [mGFP + SNAP25B-(1-13 + 143-202 + 80-206)]; SNAP25-0LN [mGFP + SNAP25B-(1-13 + 80-142 + 14-79)]; SNAP25-M32P, V36P [mGFP + SNAP25B-(1-206 carrying the mutations M32P and V36P]; SNAP25-I60P, M64P [mGFP + SNAP25B-(1-206 carrying the mutations I60P and M64P]; SNAP25-! -7 to -5 [mGFP + SNAP25B-(1-25 + 37-206)]; SNAP25-! -7 to -2 [mGFP + SNAP25B-(1-25 + 47206)]; SNAP25-! -7 to +3 [mGFP + SNAP25B-(1-25 + 65-206); SNAP25-G43D [mGFP + SNAP25B-(1-206 carrying the mutations G43D]. The C-terminally CFP tagged syntaxin 1A (Sx1A-CFP [Sx1A-(1-288) + CFP]) and syntaxin 4 (Sx4-CFP [Sx4-(1298) + CFP]) constructs are based on the expression vector pECFP-N1 (Clontech, Mountain View, CA). A 12 amino acids linker (LVPRARDPPVAT) connects the corresponding syntaxin to ECFP. Coding sequences of all constructs have been verified by sequencing using for SNAP25B, syntaxin 1A and syntaxin 4 the according rat sequences as references. FRAP experiments PC12 cells expressing corresponding constructs were mounted at RT in a chamber containing Ringer solution (130 mM NaCl, 4 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 48 mM D(+)glucose, 10 mM HEPES pH 7.4). For imaging, we used an inverted confocal laser scanning microscope (TCS-SP5; Leica Microsystems, Mannheim) equipped with a 63x 1.4NA (numerical aperture) plan apochromat oil objective. Pixel size was adjusted to 68.6 nm x 68.6 nm and scanning was performed at 700 to1000 Hz. For GFP we used for excitation the 488 nm line (laser power 2%, for bleaching 75% was used) of an Argon ion laser and collected emission (1 Airy disc pinhole size) from 500-600 nm. When also CFP-constructs were expressed, CFP-fluorescence of the corresponding basal plasma membrane was monitored using for excitation the 458 nm line of the Argon ion laser (laser power of 25%) and emission was collected between 461-480 nm. For FRAP experiments we applied the FRAP-wizard routine from the Leica Application Suite-Advanced Flourescence (LAS-AF) (Version 1.6.3 Build 1163; Leica Microsystems, Mannheim). In brief, recordings started by 10-15 frames taken at maximal speed. Then a 3 "m x 3 "m region of interest (ROI) was bleached twice using the ‘zoom in’ option and 20-30 frames were recorded at maximum speed (approximately up to 2 frames per second at 1000 Hz) followed by 40 frames at a rate of 1 frame per 3 seconds. Finally, 10 frames were recorded at maximum speed (not shown). For data analysis the programs SigmaPlot 9.01 (Systat software, Inc.) and Origin 7.5 SR1 (OriginLab Corporation, Northampton, MA) were used. First, signals in the ROI were background corrected and the pre-bleach value was determined by averaging the 10-15 frames recorded before bleaching. Then values of the recovery trace were normalized to the averaged pre-bleach value. Experiments exhibiting a strong z drift, as detected by a 20% decrease or 5% increase in fluorescence intensity in an unbleached control region, were excluded from the analysis. In Figure 1C for example, for one experiment and condition 4-7 individual traces were averaged and fitted as described (2) applying an hyperbola function (t1/2 = half-time of recovery, a = offset and b = amplitude of recovery):

. In Figure 1C half times of recovery were obtained from three independent experiments and in each experiment values from the different conditions were normalized to the control value (SNAP25). Fitting of hyperpola functions to traces from individual cells was also performed in order to analyze the dependency of half time of recovery and maximum recovery from GFP-SNAP25 expression levels (Figure S1A) and to compare the maximal recovery of SNAP25 and SN25-0L0. The average maximal recovery for SNAP25 and SN25-0L0 (including cells analyzed for Figure 1C) was 75.8 ± 2.3 % (n = 14, mean ± SEM) and 74.6 ± 2.1 % (n = 13, mean ± SEM), respectively. Regarding maximal recovery, referring to control regions of interest placed outside the bleached area, we estimate that in this experiment for SNAP25 about 8% (or on average about 0.1% per frame) of the lack of recovery is due to bleaching, resulting in a pool of immobilized SNAP25 of approximately 16%. For experiments shown in Figure 1F and G, Figure 2A and B, Figure 3 and Supporting Figure S3A and B, half times of recovery from individual cells were determined as above and plotted versus the corresponding cellular expression level of syntaxin 1A-CFP or syntaxin 4-CFP (only shown in Figure 1F). Then a slope value was calculated for each cell by subtracting first from its half time of recovery, the mean half time of recovery value of all measured cells (including 3 - 6 cells) from the same experiment that overexpressed the corresponding SNAP25 construct but had no detectable signal for syntaxin-CFP (the mean value indicates the background mobility without additional syntaxin). Then, the background corrected half time of recovery was divided by the CFP-expression level. For each condition, individual slope values (from 3 - 10 cells) were averaged resulting in the mean value from one independent experiment. For calculation of the average recovery time of SN25-0L0, we included the three experiments shown in Figure 1C and the three experiments shown in Figure 1G (in this case we used the mean recovery times from those cells that had no signal in the CFP-channel), resulting in an average recovery time of 5.3 s (n=6). Biochemistry on membrane sheets For preparation of membrane sheets, BHK cells expressing SNAP25 based GFP-fusion protein constructs were grown on poly-L-lysine coated coverslips and disrupted as previously described (3) using a 100 ms ultrasoundpulse in ice-cold KGlu-sonication-buffer (120 mM potassium glutamate, 20 mM potassium acetate, 10 mM EGTA and 20 mM Hepes pH 7.2). Membrane sheets were then reacted with 4 "M recombinant syntaxin1A (1-262 + 263Cys) in KGlu containing 3% BSA and 2 mM DTT at 37°C in a humid chamber for about 10 minutes. Then membrane sheets were washed in KGlu at RT for 30-40 seconds and fixed in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) (150 mM NaCl, 10 mM Na2HPO4, 10 mM NaH2PO4, pH 7.2) for about an hour. Then PFA was quenched using 50 mM NH4Cl in PBS for about 20-30 minutes and bound syntaxin1A was visualized by immunostaining using HPC-1 (4) as primary antibody (dilutes 1:100) and goat-anti-mouse Cy3 (Dianova, Hamburg, Germany) as secondary antibody (diluted 1:200) essentially as previously described (5). The glass-coverslips were mounted in a microscopy chamber filled with PBS containing 1-(4- trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene (TMA-DPH) (Molecular Probes, Eugene, Oregon, United States) for the visualization of phospholipid-membranes. For conventional, diffraction limited fluorescence microscopy an inverted Zeiss Axiovert 100 TV fluorescence microscope with a 100x 1.4 numerical aperture plan apochromat oil objective (Zeiss, Göttingen, Germany) was used. Illumination was provided by a XBO 75 xenon lamp. For imaging, we applied a back-illuminated charge-coupled device camera (Princeton Instruments, Princeton, NJ) with a magnifying lens (2.5x Optovar, Zeiss) to avoid spatial undersampling by large pixels. The focal position was controlled using a low voltage piezo translator device and a linear variable transformer displacement sensor/controller (Physik Instrumente, Waldbronn, Germany). Appropriate filter sets were used for TMA-DPH [excitation bandpass (BP) 360/50, beamsplitter (BS) 400–420, and emission longpass (LP) 420], GFP [excitation

BP 480/40, BS LP 505, and emission BP 527/30] and Cy3 [excitation BP 525/30, BS LP 550, and emission BP 575/30]. Image acquisition was performed with Metamorph 5.01 (Universal Imaging, West Chester, PA). Membrane sheets were quickly screened for GFP expression and when found, images were acquired in the TMADPH channel documenting the integrity of the membrane sheets. Then the image was acquired in the GFP channel to quantify the expression levels of GFP-labelled SNAP25B or SNAP25B constructs. Then the third image was taken in the Cy3 channel for quantifying bound syntaxin1A. Exposure time for all three channels was one second. The images were analysed using Metamorph imaging version 4.0. An on average approximately 10 µm2 large region of interest (ROI) was randomly placed in the TMA-DPH channel and average fluorescence intensity and background values were quantified in the GFP and Cy3 channels. Then green (GFP-construct expression level) was plotted versus red (bound syntaxin) fluorescence (using Sigmaplot 9.0). For each experiment and condition 15-25 membrane sheets were analysed. STED microscopy GFP tagged SNAP25 or SN25-0L0 was expressed in PC12 cells and membrane sheets were prepared and immunostained essentially as described (5). For experiments shown in Supporting Figure S5, prior to fixation membrane sheets were incubated in a humid chamber for 15 min at 37°C in KGlu containing 3% BSA (in the absence or presence of 10 µM synaptobrevin 2 (1-96)) and then briefly washed at RT for 30 - 40 s in KGlu. As primary antibodies we used rabbit polyclonal antibody R31 (5) for syntaxin 1, a rabbit polyclonal antibody for syntaxin 4 (Cat#110 042, Synaptic Systems, Göttingen) and a monoclonal anti-GFP antibody (clone 3E6, Molecular Probes, Invitrogen) for GFP-tagged SNAP25 and SN25-0L0. In the case of GFP-tagged SNAP25, using an anti-GFP antibody prevents the preferential visualization of the uncomplexed form of SNAP25 (for incompatibility of anti-SNAP25 antibody binding and SNAP25 complex formation see e.g. (6)). As secondary antibodies goat anti-rabbit-Cy3 (Dianova, Hamburg, Germany) and Atto647N-labelled sheep anti-mouse (provided by Dept. of Nanobiophotonics, Max-Planck Institute for Biophysical Chemistry, Göttingen, Germany) were used. After immunostaining samples were incubated for about 30 min at RT with 40 nm green fluorescent beads (F8795, Molecular Probes, Invitrogen) that were adsorbed to the glass cover-slips and used as a reference for focusing to the plane of the plasma membrane sheets. The samples were embedded in Mowiol (6 g Glycerol (AR No. 4094, Merck, Darmstadt, Germany), 2.4 g Mowiol 4-88 -Hoechst, Frankfurt, Germany, 6 ml water, 12 ml 200 mM Tris, pH 7.2 buffer) (provided by Dept. of Nanobiophotonics, Max-Planck Institute for Biophysical Chemistry, Göttingen, Germany) and imaged using a TCS STED (Stimulated Emission Depletion) superresolution fluorescence microscope from Leica Microsystems GmbH (Mannheim, Germany) equipped with a 1.4 NA 100x objective (Leica Microsystems GmbH, Mannheim, Germany). Plasma membrane sheets with expressed GFPconstructs were selected and focussed using as reference the glass adsorbed green fluorescent beads. Multichannel confocal images were then obtained simultaneously at 100 Hz scan speed for Cy3 (syntaxin staining), GFP (GFP-SNAP25 or GFP-SN25-0L0) and Atto647N (staining for GFP), followed by imaging Atto647N applying STED resolution (at 10 Hz scan speed). Pixel size was 10.1 nm (Supporting Figures S5 and S7) or 21.6 nm (Figure 5 and Supporting Figure S4). STED excitation was performed with a 635 nm diode laser, and depletion was achieved via a Spectra-Physics MaiTai tunable laser at 750 nm. Emission was collected at 645-720 nm for the Atto647N and at 551-602 nm (Figure 5 and Supporting Figure S4) or 562-632 nm (Supporting Figures S5 and S7) for Cy3 and at 498-540 nm (Figure 5 and Supporting Figure S4) or 511-540 nm (Supporting Figures S5 and S7) for GFP. Confocal imaging was performed using PMTs (photomultiplier tubes) and an Avalanche PhotoDiode was used for STED image aquisition. The system resolution limit was approximately 70–90 nm,

measured by analysis of 20 nm crimson-fluorescent beads (F-8782, Molecular Probes, Invitrogen). Images were taken from three independent sample preparations. Image analysis was performed using self-written routines in Matlab (The Mathworks, Inc). Briefly, the Cy3 and STED resolved channel were aligned using the alignment of the Atto647N STED image to the Atto647N confocal image as reference. For alignment of these images (Atto647N STED and Atto647N confocal) we used a routine in Matlab and alignment was manually verified and corrected if necessary. Regions of interest (for example, for analyses as shown in Fig. S4 ranging from 2 to 41 µm2 (on average around 10 µm2)) on the membrane sheets were manually selected and the Pearson's correlation coefficient (see also (7)) was determined between the Cy3 (syntaxin 1) and STED resolved (staining for GFP) images, that were filtered by using an unsharp filter, performed using the imopen function (the structural element, strel, being a disk of 10-pixel radius for STED images and 15pixel radius for confocal images) in Matlab (The Mathworks, Natick, MA) in order to remove any uneven background. For the STED images analysed for Supporting Figure S4 we also extracted data from the autocorrelation analysis that was applied using a routine based on the corr2 function of Matlab. Means of the values from 3 independent experiments (at least 6 membrane sheets were analyzed for each experiment) were plotted with ± SEM using Sigmaplot 9.0. For illustration of images in Figure 5 and Supporting Figures S4, S5 and S7, images were manually aligned using Metamorph (Universal Imaging, West Chester, PA) and scaled arbitrarily to illustrate spotty image features. Co-immunoprecipitation of SNAP25 using syntaxin 1 and syntaxin 4 antibodies PC12 cells, grown in 10-cm dishes were incubated for 20 min at RT in high calcium Ringer solution (130 mM NaCl, 4 mM KCl, 5 mM CaCl2, 1 mM MgCl2, 48 mM D(+)glucose, 10 mM HEPES pH7.3), washed with ice-cold PBS II (150 mM NaCl, 20 mM Na2HPO4, pH 7.4) and solubilized in 1 ml extraction buffer (PBS II containing 1% Triton X-100, 5 mM EGTA and 5 mM EDTA) containing 1 mM PMSF and inhibitor cocktail (Roche, Basel, Switzerland). After protein extraction at 4oC for 90 min, insoluble material was removed by centrifugation at 200.000 x g for 30 min. For syntaxin 1, immunoprecipitation reactions were performed with the cleared supernatants by adding either 2 "l of polyclonal anti-syntaxin sera (R31, (5)) or 7 "l of anti-syntaxin monoclonal ascites fluids - 78.3 (8) or HPC-1 (4) - per 0.35 mg of total protein (500 "l reaction volume). Also one condition without antibodies was prepared (for the loading of starting material). The reaction tubes were incubated overnight at 4°C, under continuous head-overhead rotation. Protein G Sepharose beads (Amersham Biosciences) were used to isolate the antibody-protein complexes. Before use, Sepharose beads were washed three times (pelleting with 3 min-centrifugation steps at 700 # g) and were resuspended in ice-cold extraction buffer, to a final suspension of 50% beads in extraction buffer. 50 "l of the bead suspension was added to each immunoprecipitation reaction and samples were further incubated for 2h at 4°C. The beads were pelleted at 700 # g for 3 min and the supernatants were collected. Beads were washed three times with ice-cold extraction buffer and proteins were eluted with addition of 50 "l 2x sample buffer (100 mM Tris, 8% SDS, 24% glycerol, 0,02% Serva Blue G, 4% $-mercaptoethanol, pH 6.8). Appropriate volume of 5x sample buffer was added to the supernatants. Samples analyzed by SDS-PAGE and Western blotting. The supernatant of the sample containing no antibody is considered as starting material. 10% of IP and 1.66% of starting material was loaded. SNAP25 was detected by using the monoclonal antibody 71.1 (Synaptic Systems, Göttingen, Germany) for the precipitants of the polyclonal antibody and the polyclonal antibody Casanova (9), when the monoclonal anti-syntaxin antibodies were used for immunoprecipitation. For detection of syntaxin we used a polyclonal anti-syntaxin antibody (R31) when IPs were performed with monoclonal antibodies and a monoclonal antibody (78.3 when for immunoprecipitation polyclonal antibodies were used. As secondary antibodies we used HRP-conjugated goat anti-mouse and rabbit anti-mouse antibodies (BioRad Laboratories). The bands were detected by enhanced chemiluminescence (ECL, Perkin Elmer, Boston)

on a FujiFilm LAS-1000 imaging station (Fuji Photo Film, Tokyo, Japan). The intensity of each band was measured using a routine custom-written in Matlab (The Mathworks Inc., Natick, MA). For syntaxin 4, the immunoprecipitation was performed as for syntaxin 1 using 3 "l of anti-syntaxin 4 polyclonal antibody (Cat#110 042, Synaptic Systems, Göttingen) per 0.35 mg of total protein (500 "l reaction volume). Co-immunoprecipitation of GFP-constructs using syntaxin 1-antibody PC12 cells were grown in 10 cm culture dishes and were transfected with the indicated GFP-SNAP25 constructs using Lipofectamine 2000 (Invitrogen). Forty-eight hours after transfection, cells were collected in extraction buffer. The immunoprecipitation was performed using 10 "l of anti-syntaxin monoclonal HPC-1 (ascites fluids) per approximately 1-2 mg of total protein (1.2 ml reaction volume). For Supporting Figure S2B, PC12 cells transfected with the different GFP-SNAP25 constructs were collected 48h post-transfection in 2x sample buffer. Equal volumes were loaded and analyzed by SDS-PAGE and Western blot. GFP was detected by using a polyclonal anti-GFP antibody (Cat# 132 002, Synaptic Systems, Göttingen).

References 1. Zacharias DA, Violin JD, Newton AC, Tsien RY. Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science 2002;296(5569):913-916. 2. Ficz G, Heintzmann R, Arndt-Jovin DJ. Polycomb group protein complexes exchange rapidly in living Drosophila. Development 2005;132(17):3963-3976. 3. Avery J, Ellis DJ, Lang T, Holroyd P, Riedel D, Henderson RM, Edwardson JM, Jahn R. A cell-free system for regulated exocytosis in PC12 cells. J Cell Biol 2000;148(2):317-324. 4. Barnstable CJ, Hofstein R, Akagawa K. A marker of early amacrine cell development in rat retina. Brain Res 1985;352(2):286-290. 5. Lang T, Bruns D, Wenzel D, Riedel D, Holroyd P, Thiele C, Jahn R. SNAREs are concentrated in cholesteroldependent clusters that define docking and fusion sites for exocytosis. Embo J 2001;20(9):2202-2213. 6. Xu T, Rammner B, Margittai M, Artalejo AR, Neher E, Jahn R. Inhibition of SNARE complex assembly differentially affects kinetic components of exocytosis. Cell 1999;99(7):713-722. 7. Sieber JJ, Willig KI, Heintzmann R, Hell SW, Lang T. The SNARE motif is essential for the formation of syntaxin clusters in the plasma membrane. Biophys J 2006;90(8):2843-2851. 8. Blasi J, Chapman ER, Yamasaki S, Binz T, Niemann H, Jahn R. Botulinum neurotoxin C1 blocks neurotransmitter release by means of cleaving HPC-1/syntaxin. Embo J 1993;12(12):4821-4828. 9. Aguado F, Majo G, Ruiz-Montasell B, Canals JM, Casanova A, Marsal J, Blasi J. Expression of synaptosomalassociated protein SNAP-25 in endocrine anterior pituitary cells. Eur J Cell Biol 1996;69(4):351-359.

Figure S1. Correlation of expression levels and maximal recovery and half time (A) The GFP-SNAP25 expression level in individual cells was plotted against maximal recovery (left) or half time of recovery (right). Included are cells from Figures 1C, 1F/G, 2A and B, 3 and S2A (only cells without syntaxin 1CFP co-expression are included). The analysis shows that both maximal recovery and half time of recovery are independent from GFP-SNAP25 overexpression level. (B) The syntaxin 1-CFP expression level in individual cells was plotted against maximal recovery (blue dots) of GFP-SNAP25 or GFP-SNAP25 constructs (from left to right: GFP-SNAP25, SNAP25-NL0 and SNAP25-G43D). Included are cells from Figures 1F/G, 2A and B, 3 and S2A. Green dots indicate average maximal recovery values obtained from cells showing no syntaxin-CFP expression (see also green reference line). At increased expression levels, syntaxin 1-CFP tends to decrease maximal recovery of SNAP25 but not of SN25-NL0 and SNAP25-G43D. This could indicate that for this effect the C-terminal SNARE motif of SNAP25 is needed in addition to the unmutated N-terminal SNARE-motif, and that at artificially very high syntaxin:SNAP25 ratios SNAP25 is driven into complexes different from a simple two-helix-QaQb-complex.

Figure S2. Co-immunoprecipitation of GFP-labeled SNAP25/SNAP25-constructs by anti-syntaxin 1 antibody. (A) PC12 cells expressing either GFP-SNAP25, GFP-SN25-NL0, GFP-SN25-0LC or GFP-SN25-0L0 were lysed (using 1% Triton X-100) and it was analysed if GFP-labelled constructs co-immunoprecipitate with syntaxin 1 using a monoclonal anti-syntaxin 1 antibody (HPC-1). For immunoblotting of GFP (using a polyclonal anti-GFP antibody) and syntaxin 1 (using monoclonal antibody 78.3), 24% of the immunoprecipitate and 0.8 % of the supernatant were loaded. Boxes mark areas where co-immunoprecipitated bands should appear. For each construct, the percentage of GFP in the precipitate was calculated and multiplied by the efficiency (1 = 100%) of the IP as judged from the percentage of precipitated syntaxin 1 in each case. From duplicate blots we determined that 2.09% of GFP-SNAP25, 2.29% of GFP-SN25-NL0, 0.28% of GFP-SN25-0LC and -0.25% of GFP-SN25-0L0 co-immunoprecipitated with syntaxin 1. (B) Control experiment blotting directly lysed cells showing that degradation bands evident in panel A (especially in the supernatants) are not present in freshly lysed cells, showing that degradation occurs during IP processing.

Figure S3. Syntaxin 4 interacts with SNAP25 as efficient as syntaxin 1 and SNAP25-syntaxin complexes require no neuronal co-factors. Experiments as in Figure 1G showing the dependence of recovery times of SNAP25 or SN25-0LC from syntaxin 1 or syntaxin 4 in neuroendocrine PC12 cells (A) or in fibroblast BHK cells (B). (A) As in vitro, also in vivo syntaxin 4 forms non-cognate SNARE-complexes, because syntaxin 4 slows down SNAP25 mobility as syntaxin 1. (B) Although BHK cells lack neuronal co-factors, complexes between syntaxin 1 and SNAP25 form efficiently, indicating that at this stage no neuronal co-factors are required for the syntaxin 1-SNAP25 interaction. Also in BHK cells no difference between syntaxin 1 and syntaxin 4 is observed. Values are means ± SEM (n = 3 experiments).

Figure S4. No difference between SNAP25 and SN25-0L0 can be detected regarding overlap with syntaxin clusters or overall signal distribution (A) STED-microscopic imaging of SN25-0L0. Membrane sheet from a cell expressing GFP-labelled SN25-0L0 immunostained for syntaxin 1 and GFP (used for visualization of SN25-0L0, see Supplementary Materials and Methods). Upper panel, syntaxin 1- and SN25-0L0-staining at confocal resolution (left and middle) and SN25-0L0staining at STED-resolution (right). Lower panel, magnified views from syntaxin and STED-resolved SN25-0L0 (left and middle) and overlay (right). (B) and (C), analysis of images as shown in Figure 5 and (A). (B) The degree of similarity between syntaxin and STED-resolved images was quantified by calculating the Pearson-correlation coefficient. For each condition, mirrored images were also analysed as a control. (C) SNAP25 or SN25-0L0 signal distribution characterized by autocorrelation analysis. The mean object sizes are reflected in the characteristics of the decay curve. No difference in signal distribution between SNAP25 and SN25-0L0 could be detected at this level. Values are means ± SEM (n = 3 experiments, for each experiment 8-24 membrane sheets were analyzed for SNAP25 and 11-18 membrane sheets for SN25-0L0).

Figure S5. Upon SNARE-complex formation overlap increases between syntaxin 1 and SNAP25 but not between syntaxin 1 and SN25-0L0. (A and B) Membrane sheets from cells expressing either GFP-labelled SNAP25 (upper panels in A and B) or GFP-labelled SN25-0L0 (lower panels in A and B) were incubated for 15 min at 37°C in (A) absence or (B) presence of recombinant 10 µM synaptobrevin 2 to drive SNAREs into (A) SNAP25-syntaxin-complexes and cisSNARE-complexes or (B) cis-SNARE-complexes (see also (25). They were then immunostained for syntaxin 1 and GFP (immunostaining of GFP was used for visualization of the GFP-constructs at STED-resolution). (A and B) As indicated, overviews and magnified views show immunostainings for syntaxin 1- and SNAP25 (upper panels) or for syntaxin 1 and SN25-0L0 (lower panels). (C) Pearson-correlation coefficients between syntaxin 1 and STED-resolved images. For each condition, mirrored images were analysed as a control. Values are means ± SEM (n = 3 experiments; for (A) for each experiment 912 membrane sheets were analyzed for SNAP25 and 7-9 membrane sheets for SN25-0L0, for (B) 7-11 membrane sheets were analyzed for SNAP25 and 7-9 membrane sheets for SN25-0L0,).

Figure S6. Co-immunoprecipitation (IP) of SNAP25 with (A) syntaxin 1 or (B) syntaxin 4 antibodies from PC12 cell extracts. (A) PC12 cells were solubilized in Triton X-100 and for co-IP of SNAP25 three different anti-syntaxin 1 antibodies were used (rabbit polyclonal R31 and mouse monoclonals 78.3 and HPC-1). For immunoblot analysis, 10% from the immunoprecipitate and 1.66% from the starting material were loaded. For immunoblotting, SNAP25 and syntaxin were detected using either polyclonal antibodies (Casanova for SNAP25 and R31 for syntaxin when for IP a monoclonal antibody was used) or monoclonal antibodies (71.1 for SNAP25 and 78.3 for syntaxin when a polyclonal antibody was used for immunoprecipitation). The percentage of SNAP25 in the precipitate was calculated and multiplied by the efficiency (1 = 100%) of the IP as judged from the percentage of precipitated syntaxin. For each syntaxin antibody, three independent experiments were performed. On average, 19% (15.4%, 15.2% and 26.5% for R31, 78.3 and HPC-1, respectively) of SNAP25 co-immunoprecipitated with syntaxin. (B) Co-IP from Triton X-100 solubilized PC12 cells using an anti-syntaxin 4 antibody. For immunoblotting SNAP25 (using monoclonal 71.1 as antibody), 10 % of the immunoprecipitate and 0.8 % of the supernatant were loaded; for the syntaxin 4 (using polyclonal anti-synaxin 4 antibody) blot, 10% of the immunoprecipitate and 2.4% of the supernantant were loaded. The percentage of SNAP25 in the precipitate was calculated and multiplied by the efficiency (1 = 100%) of the IP as judged from the percentage of precipitated syntaxin 4. On average (n=3 experiments), 2.04% of SNAP-25 coimmunoprecipitated with syntaxin4.

Figure S7. Syntaxin 4 overlap with SNAP25 or SN25-0L0. (A + B) Membrane sheets from cells expressing either GFP-labelled SNAP25 (A) or GFP-labelled SN25-0L0 (B) immunostained for syntaxin 4 and GFP (immunostaining of GFP was used for visualization of the GFP-constructs

at STED-resolution). Upper panels in A and B, syntaxin 4-(left), (A) SNAP25- or (B) SN25-0L0-stainings at confocal resolution (middle) and (A) SNAP25- or (B) SN25-0L0-stainings at STED-resolution (right). Lower panels in A and B, magnified views from syntaxin 4 (left), STED-resolved (A) SNAP25 or (B) SN25-0L0 (middle) and overlay (right). (C) Pearson-correlation coefficients between syntaxin 4 and STED-resolved images. For each condition, mirrored images were analysed as a control. Values are means ± SEM (n = 3 experiments, for each experiment 6-11 membrane sheets were analyzed for SNAP25 and 6-16 membrane sheets for SN25-0L0).