EPON

Muscle tissue was prepared as described in Schwarz et al. (2000) (link). For the preparation of rodent brain tissue the animals were perfused transcardially first with 30 ml of phosphate-buffered saline and then with 40 ml of fixative solution (4% paraformaldehyde in 0.1M PBS [pH 7.4]). The brain tissue was then removed and kept in fixative over night at 4 °C. After being washed twice in PBS, tissue slices (0.2 to 1.5 mm thick) were cut on a vibratome (752 M Vibroslice, Campden Instruments, Leichester, United Kingdom) and kept for 24 h in PBS at 4 °C. Pieces about 1.5 mm in size were then excised and washed three times for 30 min each in cacodylate buffer at pH 7.4.The tissue was postfixed for 2 h in 2% osmium tetroxide/1.5% potassium ferric cyanide in aqueous solution at room temperature. Then the tissue was subjected to a contrast enhancement step by soaking it over night in a solution of 4% uranyl acetate in a 25% methanol/75% water mixture (Stempak and Ward 1964 (link)) at room temperature. After that the tissue was dehydrated in a methanol sequence (25%, 70%, 90%, and 100% for 30 min each) followed by infiltration of the epoxy (Spurr, Epon 812, or Araldite, all from Serva, Heidelberg, Germany) monomer (epoxy/methanol 1:1, for 3 h rotation at room temperature; epoxy/methanol 3:1, overnight at 4°C; pure epoxy, 3 h rotating at room temperature). Polymerization was 48 h at 60 °C for Epon and at 70 °C for Spurr and Araldite. The block face was trimmed to a width of several hundred microns and a length of about 500 μm using either a conventional microtome or a sharp knife. SEM images of the untrimmed block face can be used to select the desired field of view before the final trimming step producing the desired small cutting pyramid.

Experimental datasets used in this study:

Vesicles are giant unilamellar vesicles made of DOPC, supplemented with 0.1% DOPE-Atto647N (ref AD-647N, Atto-tec, Germany) and 0.03% DSPE-PEG(2000) Biotin (ref 880129, Avanti Polar Lipids, USA) electroformed during 1 h at 1V RMS [44 (link)] in a sucrose buffer at 250 milliosmoles. Vesicules were adhered on avidin coated glass coverslips, deflated with an hyperosomotic shock due to buffer evaporation and imaged with a Yokogawa spinning-disc CSU-X1 mounted on a Nikon Ti-Eclipse microscope stand using a 100x objective with NA 1.3 (z spacing 340 nm, xy pixel size 122 nm).

MRI dataset was acquired from a normal healthy person, using a FLAIR sequence.

FIB-SEM 80% confluent HeLa cells were rinsed once with PBS, fixed for 3h on ice using 2.5% glutaraldehyde/2% paraformaldehyde in buffer A (0.15M cacodylate, 2mM CaCl2). Then cells were extensively washed on ice in buffer A, pelleted and incubated 1h on ice in 2% osmium tetroxide and 1.5% potassium Ferro cyanide in buffer A and finally rinsed 5 times in distilled water at room temperature. Cells were then incubated 20min at room temperature in 0.1M thiocarbohydrazide, which had been passed through a 0.22 μm filter, and extensively washed with water. Samples were incubated overnight at 4° C protected from light in 1% uranyl-acetate, washed in water, further incubated in 20mM lead aspartame for 30min at 60°C and finally washed in water. Samples were dehydrated in a graded series ethanol, embedded in hard Epon and incubated for 60h at 45°C then for 60 h at 60°C. A small bloc was cut and mounted on a pin, coated with gold and inserted into the chamber the HELIOS 660 Nanolab DualBeam SEM/FIB microscope (FEI Company, Eindhoven, Netherlands). ROI were prepared using focused ion beam (FIB) and ROI set to be approximatively 20 microns wide. For imaging, electrons were detected using Elstar In-Column secondary electrons Detector (ICD). During acquisition process, the thickness of the FIB slice between each image acquisition was 5 nm.

The drosophila egg chamber is dissected from a drosophila ovary. Cell nuclei were stained with DAPI and cell membranes labeled with the fusion proteins Nrg::GFP and Bsg::GFP [45 (link)]. The egg chamber was embedded in Vectashield and spacers were used to prevent tissue deformation. Images were acquired using an inverted Olympus point scanning confocal microscope IX81 with a 60x objective NA 1.42(z spacing 750 nm, xy pixel size 265 nm).

For live imaging, confocal images colocalization analysis was performed using ImageJ/Fiji Software, JAcoP plugin, Manders coefficients. The mitochondrial morphology was analyzed with the macro “Mito-morphology” from ImageJ/Fiji Software.
PLA images were analyzed manually with “Threshold” and “Analyze particles” commands from ImageJ/Fiji Software (http://imagej.nih.gov/ij/).
FIB-SEM images were first preprocessed by denoising using the BM4D algorithm (Maggioni et al., 2013 (link)). Then, images were first segmented manually with MIB (microscopy image browser), and then the manual segmentation was used to optimize a semi-automatic segmentation analysis. For the automated analysis, an iterative procedure based on U-Net (Falk et al., 2019 (link)) and manual correction was employed. First, annotations of the datasets were obtained manually for the mitochondria and semi-automatically for the ER (particularly thresholding, morphological cleaning, and manual correction). These annotations were used to train an initial 2D U-Net (3D volumes are processed slice by slice). Next, the dataset was segmented with this model, followed by small particle removal and manual correction. The U-Net model was then trained on the three annotated datasets to improve generalization performance and finally applied on the remaining datasets. These datasets were then post-processed by removing small particles and manual correction to obtain the final segmentation of the mitochondria and ER. Segmented images were then imported in Imaris 9.9 and analyzed with Surface-Surface contact plugin. In Imaris, Surface-area and volume were calculated using the Statistic features.
For the quantification of ER–mitochondria contact sites in HRP-stained Epon sections, the total circumference of each mitochondria and the length of the multiple HRP-positive ER segments closely associated (<30 nm) with them were measured by manual drawing using the iTEM software (Olympus), as in Galmes et al. (2016 (link)), on acquired micrographs of HeLa cells for each cell profile. Cells were randomly selected for analysis without prior knowledge of transfected plasmid or siRNA. All data are presented as mean (%) ± SEM.
For the quantification of E-Syt1 immunogold labeling on ultrathin cryosections, 300 gold particles were counted in randomly selected cell profiles and assigned to either reticular ER (ER), ER in contact with mitochondria (EMCS) and ER in contact with PM (ER-PM) using the iTEM software. All data are presented as mean (%) ± SEM of three technical replicates.
All data are represented as mean ± SEM. Data distribution was assumed to be normal but this was not formally tested. Statistical significance between two groups was determined by standard unpaired t-test with F-testing or one sample t test. Statistical significance between multiple groups was determined by one-way ANOVA to ensure comparable variance, then individual comparisons performed by Tukey’s post-hoc test. Analysis was done in Prism v9.0f, GraphPad.
*, P < 0.05; **, P < 0.01; ***, P <0.001; and ****, P <0.0001, where a P value <0.05 is considered significant.

HeLa cells expressing HRP-KDEL-myc were fixed on coverslips with 1.3% glutaraldehyde in 0.1 M cacodylate buffer, washed in 0.1 M ammonium phosphate [pH 7.4] buffer for 1 h, and HRP was visualized with 0.5 mg/ml DAB and 0.005% H2O2 in 0.1 M ammonium phosphate [pH 7.4] buffer. Development of HRP (DAB dark reaction product) took between 5 to 20 min and was stopped by extensive washes with cold water. Cells were post-fixed in 2% OsO4+1% K3Fe(CN)6 in 0.1 M cacodylate buffer at 4°C for 1 h, washed in cold water and then contrasted in 0.5% uranyl acetate for 2 h at 4°C, dehydrated in an ethanol series and embedded in epon as for conventional EM. Ultrathin sections were counterstained with 2% uranyl acetate and observed under a FEI Tecnai 12 microscope, 4,800X magnification, equipped with a CCD (SiS 1kx1k keenView) camera and iTEM acquisition software.

Primary B cells (unstimulated or 24 hr IgM stimulated) were seeded onto poly-L-lysine coated cell culture dishes and fixed with 1% glutaraldehyde in 0.1 M sodium phosphate buffer, pH 7.3. After fixation, the cells were washed with 100 mM Tris (pH 7.2) and 160 mM sucrose for 30 mins. The cells were washed twice with phosphate buffer (150 mM NaCl, 5 mM KCl, 10 mM Na3PO4, pH 7.3) for 30 min, followed by treatment with 1% OsO4 in 140 mM Na3PO4 (pH 7.3) for 1 hr. The cells were washed twice with water and stained with saturated uranyl acetate for 1 hr, dehydrated in ethanol, and embedded in Epon (Electron Microscopy Sciences, Hatfield, PA). Roughly 60 nm sections were cut and stained with uranyl acetate and lead nitrate. The stained grids were analyzed using a Philips CM-12 electron microscope (FEI; Eindhoven, The Netherlands) and photographed with a Gatan Erlangshen ES1000W digital camera (Model 785, 4 k 3 2.7 k; Gatan, Pleasanton, CA).