An easy, fast and “low-tech”-equipment-requiring alternative method to optimize immunolabelling conditions for pre-embedding immunogold electron microscopy and to correlate light and electron microscopical immunogold labelling results

Journal of Immunological Methods 444 (2017) 7–16

Contents lists available at ScienceDirect

Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim

Research paper

An easy, fast and “low-tech”-equipment-requiring alternative method to optimize immunolabelling conditions for pre-embedding immunogold electron microscopy and to correlate light and electron microscopical immunogold labelling results Shweta Suiwal, Gabriele Kiefer, Frank Schmitz ⁎, Karin Schwarz Saarland University, Department of Neuroanatomy, Institute of Anatomy and Cell Biology, Kirrbergerstrasse, 66421 Homburg/Saar, Germany

a r t i c l e

i n f o

Article history: Received 18 November 2016 Received in revised form 2 February 2017 Accepted 7 February 2017 Available online 13 February 2017 Keywords: Pre-embedding Immunogold Ribbon synapse Retina RIBEYE Correlative light and electron microscopy (CLEM)

a b s t r a c t Correlating light microscopic immunolabelling results with electron microscopic data is of great interest in many fields of biomedical research but typically requires very specialized, expensive equipment and complex procedures which are not available in most labs. In this technical study, we describe an easy and “low-tech”equipment-requiring pre-embedding immunolabelling approach that allows correlation of light microscopical immunolabelling results with electron microscopic (EM) data as demonstrated by the example of immunolabelled synaptic ribbons from retinal rod photoreceptor synapses. This pre-embedding approach does not require specialized embedding devices but only commonly available equipment. The cryostat sectionbased procedure allows optimization of the pre-embedding immunolabelling conditions at the less laborious and time-consuming light microscopic (LM) level before the ultrastructural analyses of the immunolabelled structures can be performed on the same sample after ultrathin sectioning without further modification. The same photoreceptor synapse that has been first studied at the light microscopic level can be subsequently analyzed with this approach at the electron microscopic level at individual ultrathin sections or serial ultrathin sections from individual, identical synapses. Higher resolution EM analyses of the immunolabelled synapses can be performed with only minor modifications of the combined LM/EM procedure. The detergent-free procedure is applicable even for weakly fixed cryostat sections which is a relevant aspect for many antibodies that do not work with more strongly fixed biological samples. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Correlating light microscopic immunolabelling results with electron microscopic data by CLEM (correlative light and electron microscopy) procedures is of great interest in many fields of biomedical research (for review, see Perkovic et al., 2014; Fonta and Humbel, 2015; De Boer et al., 2015; Begemann and Galic, 2016; Karreman et al., 2016). Different approaches have been developed that allow correlation of light and electron microscopic data at different levels of resolution (for review, see Watanabe and Jorgensen, 2012; Fonta and Humbel, 2015; De Boer et al., 2015; Begemann and Galic, 2016). DAB-dependent methods employ photo-oxidation (Grabenbauer et al., 2005; Meisslitzer-Ruppitsch et al., 2009, 2013; Grabenbauer, 2012; Horstmann et al., 2013) or enzymatic oxidation (Schikorski, 2010) of a light microscopic signal to generate an electron-dense reaction product that can be detected by electron microscopy. Other CLEM procedures involve ultracryotome sections that can be analyzed both at the light and ⁎ Corresponding author. E-mail address: [email protected] (F. Schmitz).

http://dx.doi.org/10.1016/j.jim.2017.02.003 0022-1759/© 2017 Elsevier B.V. All rights reserved.

electron microscopic level (Tokuyasu, 1973, 1980; Van Rijnsoever et al., 2008). With ultracryotomy, ultrathin frozen sections are obtained from frozen tissue samples by sectioning at ≈−100 °C with an ultracryotome. The frozen ultrathin sections are then processed for immunolabelling (Tokuyasu and Singer, 1976; Slot and Geuze, 2007). Ultracryotomy is technically challenging (Tokuyasu, 1980; Christensen and Komorowski, 1985) and requires specialized equipment. Recently, also sophisticated fixing and embedding procedures have been developed that preserve fluorescence in resin-embedded samples (Watanabe et al., 2011, 2014; for review, see Watanabe and Jorgensen, 2012). For the CLEM analyses, these resin-persisting fluorescent signals were first analyzed by super-resolution microscopy to obtain highly resolved light microscopic signals in the sample sections. Later on, these highly resolved light microscopic signals are overlayered onto the electron microscopic images. This approach provides a powerful, very high resolution approach to CLEM (nano-fluorescence electron microscopy; nano-fEM). But these procedures also require specialized and expensive equipment not readily available in most labs, e.g. high-pressure freeze substitution device/freeze substitution devices (Watanabe et al., 2011, 2014; Watanabe and Jorgensen, 2012); ultracryotomy (Al-Amoudi et

8

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

al., 2004; Van Rijnsoever et al., 2008; Cortese et al., 2012) and challenging, complex analyses procedures or use of genetically enhanced probes (for review, see Watanabe and Jorgensen, 2012; De Boer et al., 2015; Kuipers et al., 2015; Souslova et al., 2016). In the present study, we describe an alternative, technically very easy approach for pre-embedding immuno electron microscopy (immuno EM) that also allows correlation of light microscopic (LM) immunosignals with electron microscopic (EM), ultrastructural data in an uncomplicated manner using standard equipment, probes and procedures. It does not require specialized equipment and provides a simple, straightforward approach to obtain CLEM data also for non CLEM-specialized labs with a good level of resolution and ultrastructural preservation. The usefulness and performance of the cryostat section-based procedure was exemplified by the analysis of the (bovine) retina. The retina is particularly suitable for morphological analyses because of its clear organization into well characterized and histologically clearly distinguishable layers. To establish the procedure, we analyzed photoreceptor synapses of the retina with a well characterized antibody against RIBEYE, a protein component of synaptic ribbons (Schmitz et al., 2000; for review, see Lagnado and Schmitz, 2015). Synaptic ribbons are relatively large, electron-dense structures associated with the active zone of ribbon synapses (for review, see Schmitz, 2009; Matthews and Fuchs, 2010). Ribbon synapses are tonically active synapses in the retina, inner ear and pineal gland that are able to maintain exocytosis for prolonged periods of time (Heidelberger et al., 2005; Matthews and Fuchs, 2010). RIBEYE is the main component of synaptic ribbons (Schmitz et al., 2000; Zenisek et al., 2004; Magupalli et al., 2008; Maxeiner et al., 2016). RIBEYE consists of a N-terminal A-domain and a carboxyterminal B-domain (Schmitz et al., 2000). The A-domain predominantly has a structural role in scaffolding the synaptic ribbon while the B-domain is most likely exposed on the surface of the synaptic ribbon (Magupalli et al., 2008). At that site, it can interact with proteins of the presynaptic terminal (Alpadi et al., 2008; Dembla et al., 2014; Wahl et al., 2016). The synaptic ribbon is considered to organize vesicle trafficking at the continuously active ribbon synapses and to provide the active zone with many release-ready vesicles (Matthews and Fuchs, 2010; Lagnado and Schmitz, 2015). It also contributes to the organization of active zone components (Maxeiner et al., 2016). In the retina, photoreceptors and bipolar cells form ribbon synapses to communicate with their postsynaptic targets. Photoreceptor synapses are located in the outer plexiform layer (OPL) and have particularly large synaptic ribbons. Photoreceptor synaptic ribbons are plate-like, horseshoe-shaped structures in side views (with a contour length of about 1.5–2 μm) and EM reconstructions (for review, see Schmitz, 2009; Matthews and Fuchs, 2010). In EM cross-sections, they typically appear as bar-shaped structures. The OPL is dominated by ribbon synapses whereas the inner synaptic layer of the retina, the inner plexiform layer (IPL), contains a mixed population of ribbon synapses and “conventional”, non-ribbon-containing synapses. Synaptic ribbons in the IPL are smaller than in the OPL (Heidelberger et al., 2005; Schmitz, 2009; Sterling, 2013). 2. Materials and methods 2.1. Materials 2.1.1. Bovine retina Immunolabelling experiments were performed with bovine retinas freshly isolated from bovine eyes (see below). Bovine eyes were obtained from a local slaughterhouse. 2.1.2. Primary antibodies 2.1.2.1. Anti-RIBEYE. Polyclonal rabbit antiserum (U2656) raised against RIBEYE(B)-domain. This antibody has been extensively characterized previously (Schmitz et al., 2000; Magupalli et al., 2008; Alpadi et al.,

2008; Maxeiner et al., 2016). Omitting primary antibody (i.e. using only secondary antibody) or irrelevant primary antibodies served as negative control incubations. 2.1.3. Secondary antibodies 2.1.3.1. Secondary antibodies conjugated to nanogold particles for light microscopy and electron microscopy. Goat anti-rabbit IgG conjugated to ultrasmall gold particles (1.4 nm gold particle size) (NanoProbes order #2003; via Biotrend, Cologne, Germany); used for pre-embedding immunolabelling at a 1:40 dilution in PBS to which 0.5% BSA has been added. Silver enhancement kit:HQ SILVER™ Enhancement kit (NanoProbes order #2012; via Biotrend, Cologne, Germany). 2.1.4. Epon embedding medium Epon embedding medium was prepared by mixing 6.5 g of epoxy embedding medium (Epon 812, Fluka-Sigma-Aldrich; Munich, Germany) with 2.75 g 2-Dodecenyl Succinic Anhydride (DDSA; Electron Microscopy Sciences; Science Services; Munich, Germany) and 4 g Methyl-5-Norbornene-2,3-Dicarboxylic Anhydride (MNA; Electron Microscopy Sciences). 2,4,6-Tri(dimethylaminomethyl)phenol (DMP-30; Electron Microscopy Sciences) served as accelerator. After thoroughly mixing of Epon 812, DDSA and MNA, 0.1 g of DMP-30 was added to the mixture. Epon resin was degased using an exsiccator and directly used for embedding. 2.2. Methods 2.2.1. Gelatin coating of glass slides Standard glass slides were coated with a relatively thick layer of gelatin. This was done 1.) for allowing a better attachment of the cryostat sections and semithin sections to the glass surface and 2.) to promote the detachment of the cryostat/semithin sections-containing Epon capsules from the surface of the glass slides at the very end of the pre-embedding procedure (see below). For gelatin coating, a 1% gelatin solution (w/v) in deionized H2O was freshly prepared just before use with gentle heating (60 °C). To 100 ml of this solution, 0.05 g of potassium chromium-III-sulfate was added. Dust-free glass slides were inserted into cuvette-holders. For coating, the cuvette-holders with the loaded glass slides were submerged into cuvettes containing the freshly prepared gelatin solution and were incubated at RT for 5– 15 min with gentle agitation. Excess gelatin solution was drained from the slides. The slides were incubated at 60 °C for ≈1 h to completely dry the gelatin-coated glass slides. This coating procedure was repeated five times. After the last drying step slides were placed in dust-free slide boxes and kept at RT until needed. This method allowed a very flat mounting of the immunolabelled cryostat sections/semithin sections. A flat embedding of the immunolabelled semithin sections is particularly important for making serial ultrathin sections from these samples (Fig. 3). 2.2.2. Ultra-small immuno gold staining of cryostat sections for pre-embedding light microscopy and electron microscopy Freshly isolated bovine retinas were flash-frozen in liquid nitrogencooled isopentane, as previously described (Schmitz et al., 1996). In a liquid nitrogen-filled styrofoam box, an aluminum tray filled with isopentane was fixed in the way that the aluminum tray was surrounded by the liquid nitrogen. For freezing, small pieces of retina were rapidly inserted into the solidifying isopentane using metal tweezers. Frozen retinas were stored at − 80 °C until the samples were mounted in the cryostat for sectioning. The samples were transported from the −80 °C freezer to the cryostat in a liquid nitrogen-containing dewar. Frozen retinas were mounted in freezing medium (Thermo Scientific Richard-Allan Scientific Freezing medium “NEG-50”) at −20 °C and 20 μm-thick cryosections were prepared with a LEICA CM 1950 cryostat (LEICA Biosystems, Wetzlar, Germany). For sectioning, an anti-role plate (Leica, #14047742497) was used to obtain flat, well-

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

preserved cryosections. Cryosections were collected with the gelatinecoated slides that were kept at room temperature prior to the collection of the cryosections. This was done to promote adhesion of the cryosections to the gelatine-coated glass slides. Sections were heat-fixed at 50 °C for 10 min. Next, sections were treated with freshly prepared 2% paraformaldehyde (PFA) in PBS for 2 min at RT and washed in PBS for 4×15 min at RT. In the next step, sections were pre-treated with 0.5% BSA in PBS for 1 h at RT to block non-specific protein-binding sites. After blocking, rabbit anti-RIBEYE polyclonal antibody (U2656 immune serum; Schmitz et al., 2000) was added at a 1:50 dilution in 0.5% BSA in PBS at 4 °C for 24 h. Unbound antibody was removed by washing with PBS for 3 × 10 min at RT. Incubations in which the primary antibodies were omitted or in which irrelevant rabbit primary antibodies were used served as control incubations. After several washes with PBS/ BSA, the samples were incubated with the goat anti-rabbit secondary antibody conjugated to 1.4 nm sized gold particles (Nanoprobes) at a 1:40 dilution in 0.5% BSA in PBS for 24 h at 4 °C. Then, sections were washed with PBS 3 × 10 min at RT. Sections were postfixed with 2% glutaraldehyde in PBS at RT for 10 min. After several washes with PBS, sections were rinsed with deionized water for 5 × 5 min at RT. For enhancement of ultrasmall 1.4 nm gold particles according to Danscher (1981), a commercial silver enhancement reagent (HQ SILVER™ Nanoprobes) was used (Danscher, 1981; Hayat, 1995; Hacker et al., 1996; Lackie, 1996; Roth, 1996; Bendayan, 2000; Oliver, 2010) exactly as described by the manufacturer: Equal amounts of the three kit components were mixed. Initiator solution (solution A) was dispensed into a clean eppendorf tube. An equal volume of moderator solution (solution B) was added next and thoroughly mixed before an equal volume of activator solution (solution C) was added and mixed thoroughly again to prepare the complete reagent. Specimens were enhanced for 20 min if the purpose was to analyze the samples for both light and electron microscopy (combined LM/EM procedure) or for only 15 min (refinement procedure; for EM only) to get a more discrete silver-enhanced gold labelling at the EM level. Enhancement was performed in the dark at RT (26 °C) followed by 5 × 5 min washes with deionized water at RT. At this point, cryostat sections were analyzed and documented for light microscopy (LM). For this purpose, the immunolabelled sections were mounted with a drop of PBS, covered with a glass coverslip and documented for light microscopy. Images were taken with a Zeiss Axiophot microscope (40X Plan Neofluo-R oil objective, numerical aperture, 1.30). After LM documentation, the glass slides were transferred into a Petri dish filled with PBS. The glass slides containing the attached immunolabelled cryostat section (covered by the coverslip) were completely submerged in PBS. In the PBSfilled Petri dish, the coverslips were gently removed by lateral movements in order to not damage the immunolabelled cryostat sections. After removal of the coverslips, the sections were further processed for electron microscopy as described. Sections were postfixed with 1% osmium tetroxide (OsO4) in deionized water for 10 min at RT. Sections were washed three times with deionized water and next incubated with 2% uranyl acetate in deionized water for 5 min at RT in the dark. After several washes with deionized water, the slides were then dehydrated with increasing alcohol concentrations 50%, 70%, 80%, 90%, 95% for 5 min and 100% for 10 min at RT. Finally, slides were equilibrated with propylene oxide for 5 min. A drop of Epon was put onto the section as soon as they were taken out of the propylene oxide to avoid drying of sections. A gelatin capsule completely filled with liquid Epon mixture was rapidly inverted onto the immunolabelled cryostat section and remained on the section in that inverted position during polymerization of the resin. Epon was polymerized in a 60 °C incubator for 16 h. 2.2.3. Sectioning First, the Epon-embedded capsules containing the immunolabelled cryostat/semithin sections were removed from the gelatinized glass slides by plunging the glass slides with the attached Epon capsule into liquid nitrogen for a few seconds (≈15 s). After removal from the liquid

9

nitrogen, the Epon capsule could be mechanically detached from the gelatin-coated glass slide. The detached surface of the Epon capsule was carefully inspected for traces of glass that could damage the diamond knife during sectioning. Capsules were then trimmed to obtain a trapezoid surface shape using a Reichert trimming machine (C, Reichert TM60, Austria). For light microscopy, semithin sections (0.5 μm in thickness) were cut using Reichert ultramicrotome (Leica AG; Reichert division 702501, Austria). Sections were cut from this trimmed trapezoid shaped capsule using a semithin diamond knife (Histo diamond knife; Diatome) with a distilled water filled boat. Sections were transferred onto a drop of water on clean glass slides. Glass slides were then placed on a 80 °C hot plate for 5 min which allows sections to adhere on slides. After drying, sections were stained with Richardson Blue staining solution (1:1 (v/v) mixture of solution A (1% Methylene Blue dye in 1% borax (di-sodium tetraborate decahydrate) in distilled H2O) and solution B (1% Azur II dye in distilled H2O) at 80 °C (2 min). The glass slide was then rinsed with distilled water and mounted. For standard transmission electron microscopy of the immunogold labelled samples, ultrathin sections (≈ 70 nm in thickness) were cut from the same trapezoid-shaped, sample-containing capsule with a Reichert ultramicrotome (Ultracut S; Leica Microsystems). Sections were cut using a diamond knife (45° Ultra diamond knife; Diatome AG; Biel, Switzerland; 6° sectioning angle) with a distilled water filled boat. Sections were picked up on Formvar-coated slot grids. Grids were then allowed to dry properly. Sections were stained with lead citrate (Reynold's lead citrate; 80 mM lead nitrate, 136 mM sodium citrate, adjusted with 1 N NaOH to pH 12.0) for 2 min at RT according to Reynolds (1963). For CLEM purposes, i.e. to analyze the very same synapse both at the light as well as at the electron microscopic level, 0.5 μm-thick semithin sections were collected on gelatinized glass slides (see above), stained with Richardson-Blue as described above and documented by standard widefield light microscopy. At that step, landmarks were identified and documented (e.g. a characteristic, easily re-cognizable arrays of nuclei in the ONL (see Fig. 2) or other conspicuous tissue features or alternative marks like sectioning scratches). Also the silver-enhanced immunogold signals were easily visible by conventional light microscopy at that step (see Figs. 1, 2). The described landmark structures together with the silver-enhanced immunogold signals were used later on in the electron microscopic analyses to locate the very same synapse identified in light microscopy also at the ultrastructural level (see below). To further ease the finding of the landmark structures for later-on analyses, we generated small drawings (“maps”) in which the position of the landmarks was noted at low magnification also in relation to some easily visible structures within the sections (e.g. sectioning scratches, large vessels, tissue kinks). After light-microscopic documentation of the Richardson-Bluestained semithin section, the very same semithin section was prepared for electron microscopy. For this purpose, a gelatin capsule completely filled with liquid Epon mixture was rapidly inverted onto the immunolabelled semithin section and remained on the section in the inverted position during polymerization of the resin. Epon was polymerized in a 60 °C incubator for 16 h. After re-polymerization, the reembedded semithin sections were detached from the glass slides by plunging into liquid nitrogen, as decribed in detail above in Section 2.2.3. Afterwards, the re-embedded semithin sections were ultrathinsectioned with a diamond knife and analyzed by transmission electron microscopy. 2.2.4. Analysis of immunolabelled semithin sections for light microscopy (LM) For light microscopy, 0.5 μm thick (“semithin”) sections were cut from the polymerized Epon capsules that contained the flat-embedded immunolabelled samples. These samples were silver enhanced for 20 min and semithin sections were stained with Richardson-Blue staining solution. Images were either taken with Leica DM2500 microscope (40X N-Plan objective; numerical aperture, 0.65; and Leica Application

10

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

Suite (LAS) V4.2 software; Leica Microsystems Wetzlar, Germany) or with a Zeiss Axiophot microscope (40X Plan Neofluo-R oil objective, numerical aperture, 1.30; 100X Plan Apochromat oil objective, numerical aperture 1.4; Carl Zeiss Jena; Germany) equipped with a colour view II Olympus XC30 camera and cellSens Standard software; vs.1.8.1 (Olympus Life Science Solutions; Hamburg, Germany).

2.2.5. Analysis of immunolabelled sections for electron microscopy (EM) For electron microscopy, ultrathin sections were cut from polymerized Epon capsules that contained the immunolabelled silver-enhanced cryostat sections or re-embedded semithin sections. The ultrathin sections were collected on Formvar-coated slot grids (no mesh grids) to exclude that the synapses of interest could be obscured by metal grids.

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

11

Fig. 2. (A–C) Correlative light and electron microscopy. Semithin sections obtained from cryostat sections immunolabelled with antibodies against RIBEYE were counter-stained with Richardson-Blue staining solution for light microscopy first. In the Richardson-Blue-stained sections, the immunolabelled synaptic ribbons are clearly visible at the light microscopic level as dark black structures. The synaptic ribbons often displayed the typical horseshoe-shaped immunolabelling pattern (“sr” in (A), see also Fig. 1). The Richardson-Blue staining revealed the tissue “background” and allowed the identification of characteristic landmarks, e.g. arrangement of nuclei in the outer nuclear layer (ONL): “n1” … “n4” in the light microscopic sections (in A). The very same region and the very same synaptic ribbon (e.g. synaptic ribbon 1, “sr1” in A–C) can be easily identified in the electron microscopic analyses first at low magnification (B) and then at higher magnification (C). The detailed EM analysis (B,C) revealed that the apparently single immunolabelled complex from light microscopy (red arrow in A) actually consists of a group of small, RIBEYE-immunolabelled synaptic ribbons (encircled in B,C). The asterisks mark a possible freezing artifact (clear, empty space around the nuclei in the ONL). Abbreviations: “n1”, “n2” … “n4” are “landmark” photoreceptor nuclei in the outer nuclear layer (ONL) used to localize the region of interest in the electron microscope (EM). These landmark nuclei of light microscopy (A) are readily visible also at the electron microscopic level (B,C) and allow identification of the very same immunolabelled structures at the light- and electron microscopic level; sr, synaptic ribbon; sv, synaptic vesicles; n, photoreceptor nucleus; OPL, outer plexiform layer. Scale bars: 1 μm.

Typically, four to five 70 nm-thick ultrathin sections were obtained from a re-embedded semithin section of 0.5 μm thickness (often up to six ultrathin sections per semithin section, Fig. 3E). The above mentioned landmarks were used to identify the same region that was previously analyzed by light microscopy (see Fig. 2). Images were taken using a transmission electron microscope (TEM) (Tecnai 12 Biotwin; FEI, Eindhoven, The Netherlands) equipped with a Megaview III digital camera (Gatan) and controlled by iTEM acquisition software (Olympus, Hamburg, Germany). The TEM microscope was operated at 100 kV. Images were assembled and labelled with Adobe Illustrator.

2.2.6. Conventional transmission electron microscopy (TEM) Conventional TEM was performed exactly as previously described (Maxeiner et al., 2016) using freshly isolated mouse retinal tissue isolated within 5 min post mortem and immediately fixed with 4% freshly dissolved paraformaldehyde in PBS and 2.5% glutaraldehyde in PBS (overnight, 4 °C). All embedding steps were performed exactly as described (Maxeiner et al., 2016).

3. Results Cryostat sections of the bovine retina - immunolabelled with antiRIBEYE primary antibody (or control antibody) - were incubated with goat anti-rabbit secondary antibody conjugated to ultrasmall (1.4 nm) gold particles. Subsequently, the ultrasmall gold particles bound to the primary antibodies were enhanced for a time period of 20 min with a commercially available silver enhancement kit (as described in detail in the Methods section). After initial analyses (Fig. 1A,B), the immunolabelled cryostat sections were embedded in Epon, and 0.5 μm thick (“semithin”) sections were obtained from the immunolabelled samples (Fig. 1C–F). To relate the silver-enhanced immunogold product to the retinal layers, sections were stained with Richardson-Blue to visualize the retinal layers. In these semithin sections of the anti-RIBEYE-immunolabelled retinas, a strong discrete black labelling product was readily visible and strongly enriched in the synaptic layers, the inner plexiform layer (IPL) and particularly in the outer plexiform layer (OPL) (Fig. 1C,E). This immunolabelling pattern could already be observed at the level of the (non Epon-embedded) 20 μm-thick cryostat section (Fig. 1A)

Fig. 1. (A,B) Wide field light microscopic images from the (non-embedded) 20 μm-thick cryostat sections immunolabelled with RIBEYE antibodies (in A) and without primary antibody (in B, control incubation). Arrows in (A) point to immunolabelled synaptic ribbons in the OPL where photoreceptor ribbon synapses are located. Arrowheads point to immunolabelled synaptic ribbons in the IPL. (C‐F) Semithin sections (0.5 μm-thick) obtained from immunolabelled cryostat sections as shown in (A,B) (light microscopy of semithin sections after embedding in Epon). (C,E) show semithin sections immunolabelled with RIBEYE antibodies; (D,F) are control incubations (primary antibody omitted). Arrows in (C) point to single immunolabelled synaptic ribbons in the OPL. Arrowheads point to immunolabelled synaptic ribbons in the IPL. The dashed circle in (E) shows a single immunolabelled synaptic ribbon in a rod synapse (r). (G–M) show micrographs from the same samples as shown in (C‐F) but at the electron microscopic (EM) level after ultrathin sectioning. Already at low EM magnification (G–I), synaptic ribbons are easily visible as dark black immunolabelled structures in the outer portion of the OPL (arrows in G–I). The arrows in (G–I) denote single immunolabelled synaptic ribbons which are typical for rod synapses (see also the encircled structure in E,G–I). The white arrows in (J–L) indicate immunolabelled synaptic ribbons. The dark arrows denote the active zone to which the synaptic ribbon is anchored. In (J), the synaptic ribbon is shown in a side view, revealing its horseshoe-shaped appearance; in (K,L) the synaptic ribbon is cross-sectioned. Please note that all images in Fig. 1 were obtained from samples which have been processed for a 20 min long silver intensification step. (M) shows a representative negative control incubation. No immunosignal is detected in this case. Abbreviations: IS, inner segments; OS, outer segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; n, nucleus of a photoreceptor cell in the ONL; pre, presynaptic; po, postsynaptic; sr, synaptic ribbon; sv, synaptic vesicles; r, single synaptic ribbon in a rod synapse. Scale bars: 10 μm (A–F); 5 μm (G–I); 500 nm (J); 100 nm (K); 200 nm (L); 5 μm (M).

12

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

Fig. 3. (A-E) Correlative light and electron microscopy in serial ultrathin sections. The samples were processed as in Fig. 2 with the exception that serial ultrathin sections were sectioned from the re-embedded semithin section. Similar as in Fig. 2 landmark structures could be easily identified in the Richardson Blue-stained semithin sections, e.g. “n1” … “n4” in addition to the silver-enhanced immunogold labelled synaptic ribbons (A). Using these landmark structures that could also be readily identified at the electron microscopic level at low magnification (B), the very same immunolabelled synaptic ribbons (e.g. “sr1” and “sr2”) could be easily identified in electron microscopic serial sections (C–D). The serial sections (from which two sections are shown) were collected on Formvar-coated slot grids (E). The asterisks mark a possible freezing artifact (clear, empty space around the nuclei in the ONL). Abbreviations: “n1”, “n2” … “n5” are “landmark” photoreceptor nuclei in the outer nuclear layer (ONL) used to localize the region of interest in the electron microscope (EM). These landmark nuclei of light microscopy (A) are readily visible also at the electron microscopic level (B,C) and allow identification of the very same immunolabelled structures at the light- and electron microscopic level; sr, synaptic ribbon; OPL, outer plexiform layer. Scale bars: 2 μm.

though with a less good resolution as compared to semithin sections (Fig. 1C,E) due to the differences in section thickness. No silver-enhanced immunogold product was visible in the synaptic layers in the control experiments at the light microscopic level (Fig. 1B,D,F; and

data not shown). In the OPL, where the photoreceptor ribbon synapses are located, the silver-enhanced immunogold product was particularly strong, corresponding to the bigger size of the ribbons in the OPL vs. IPL, and displayed a horseshoe-shaped immunolabelling pattern in the

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

OPL (Fig. 1E). This horseshoe-shaped pattern is typical for single, immunolabelled photoreceptor synaptic ribbons (Schmitz, 2009; Wahl et al., 2013, 2016; Dembla et al., 2014). Single synaptic ribbons could be readily discriminated at high resolution in the semithin sections. These single synaptic ribbons are typical for rod synapses that only possess a single active zone (Fig. 1E). Groups of synaptic ribbons are typically found in cone synapses (Schmitz, 2009; Regus-Leidig and Brandstätter, 2012). Next, for routine TEM purposes, we processed the very same sample for electron microscopy by sectioning 70 nm thick (“ultrathin”) sections (Fig. 1G–M). Similar to the light microscopical results, we observed a very strong and dense, silver-enhanced immunogold signal in the photoreceptor synapses in the OPL. The staining product was strongly enriched at the synaptic ribbon (Fig. 1G–L). For CLEM purposes, i.e. for correlating light and electron microscopic immunolabelling anaylses at the very same synapse, we re-embedded the immunolabelled and Richardson Blue-stained semithin section documented for light microscopy (Fig. 2) for electron microscopy as described in “Materials and methods”. Since the semithin sections can be counter-stained with Richardson-Blue without compromising the subsequent electron microscopic analyses, it is easy to identify “landmarks”, e.g. characteristic arrangements of nuclei in the ONL, that allow to find that region in the light microscope and to identify the very same immunolabelled structure also in the electron microscope with ease (Fig. 2). Furthermore, the immunolabelled silver-enhanced immunogold signals are readily visible as dark immunostained structures both at the light and electron microscopic level without any further processing. Using the described procedure explained above, the very same synapse could be easily identified in the electron microscope with the help of the Richardson Blue-stained light microscopic landmark structures and the silver-enhanced immunogold signals (Fig. 2). The electron microscopic analyses (Fig. 2B,C) can provide ultrastructural informations/details that light microscopic analyses alone (Fig. 2A) cannot deliver and can assign to the immunosignals a specific subcellular ultrastructural context. This different degree of spatial resolution is exemplified in Fig. 2 in which the same synapse was analyzed both at the light- and electron microscopic level. While the light microscopic analysis of the immunolabelled synaptic region highlighted in Fig. 2A indicated a single immunolabelled structure detected by the RIBEYE antibody (red arrow in Fig. 2A), the ultrastructural analysis demonstrated that the apparantly single immunolabelled structure actually consisted of a bundle of small immunolabelled synaptic ribbons (encircled structures in Fig. 2B,C). These synaptic ribbons were surrounded by a halo of synaptic vesicles (Fig. 2C). From the re-embedded semithin sections (0.5 μm thickness), we typically obtained 4–5 ultrathin sections of ≈70 nm thickness (often up to six ultrathin sections; Fig. 3E). This is possible due to the flat embedding of the re-embedded semithin section on the even glass surface. Therefore, the described CLEM method can be further expanded towards a more detailed three-dimensional analysis of the immunosignals in the z-direction by the analysis of the serial ultrathin sections that can be sectioned from the re-embedded semithin section (Fig. 3). Up to 6 ultrathin sections were collected as a single ribbon on a Formvar-coated slot grid in identical orientation (Fig. 3E). With this procedure and the landmark strategy described above, it is easy to identify the same immunolabelled structure, i.e. synaptic ribbon, in subsequent serial ultrathin sections (as exemplified in Fig. 3C,D). This three-dimensional informations obtained from the analyses of serial sections can provide additional informations about the three-dimensional distribution of the immunosignal which is relevant particularly in the case that an antigen should have a heterogeneous distribution in the three-dimensional space (see also Discussion). The 20 min long enhancement procedure generated a very dense silver-enhanced immunogold product that is easily visible at the light microscopic level (Fig. 1A,C,E) and also at the electron microscopic level (Fig. 1G–L). To get a more discrete immunolabelling signal on the synaptic ribbon and thus also a better resolution, we used the same

13

procedure as described above, but with a 15 min enhancement step, i.e. 25% shorter enhancement time with the silver enhancement kit. This shorter enhancement procedure generated a less strong signal on the target structure, the synaptic ribbon, and allowed a higher resolution analysis of the immunolabelled structures (Fig. 4A–C). The RIBEYE immunosignal was strongly enriched at the synaptic ribbon and on its surface (Fig. 4A–C). The ultrastructural preservation of the photoreceptor synapses is satisfactory as judged by comparison with similar views of photoreceptor synapses obtained from conventional transmission electron microscopy from strongly fixed synapses (Fig. 4E–G). 4. Discussion In this manuscript, we present a straightforward, convenient and versatile immunolabelling method that is suitable to visualize immunosignals both at the light- and electron microscopic level and to correlate these signals at the light and electron microscopic level even within the same sample. In a first step, weakly fixed cryostat sections are processed for light microscopy using nanogold-conjugated secondary antibodies and silver-enhancement. This step can be rapidly performed and allows an early, first judgement on the suitability of the antibody for the pre-embedding procedure and/or the possible need for optimization steps of the immunolabelling protocol at the light microscopic level. In general, pre-embedding immuno electron microscopy can often be quite laborious and time-consuming because many steps have to be optimized to make an antibody work for pre-embedding immuno EM (for review, see Begemann and Galic, 2016). Therefore, the described combined LM/EM can be very time-saving and shorten the efforts for obtaining successful ultrastructural immunogold labelling results. Once the light microscopic immunosignals are judged as satisfactory, the very same sample can be directly processed for electron microscopy, either for conventional EM or correlative EM (see below). The electron microscopic analyses can put the immunosignals into the subcellular ultrastructural context and provide ultrastructural details that cannot be provided by light microscopy alone. An important advantage of the presented method is that it uses the same staining method. i.e. silver-enhanced immunogold signals for both light- and electron microscopy which are directly visible for both methods. This advantage is shared with the bi-functional immunolabelling reagent FluoroNanogold that can be visualized both at the light- and electron microscopic level (Takizawa and Robinson, 2000; Takizawa et al., 2015). Our approach complements other approaches that use vibratome sections of stronger fixed tissue and detergent-containing solutions for the ultrastructural localization of antigens in pre-embedding electron microscopy (Schikorski, 2010). Also in our approach, the fixation conditions can be easily adjusted and optimized at the light microscopic level first and then, the optimized procedure can be directly transferred to the electron microscopic level in the next step. The presented method is technically easy to apply. Comparable approaches like immunolabelling of ultrathin cryosections (Tokuyasu, 1980; Robinson et al., 2000; Takizawa and Robinson, 2003; Van Rijnsoever et al., 2008) are technically much more demanding and require sophisticated, expensive equipment which might not be available in many laboratories (see also below). Another advantage of the presented method is that it employs cryostat sections which allow easy access of the antibody to its target antigen. Access of the antibody to its subcellular target is an important problem in many pre-embedding protocols (for review, see Danscher, 1981; Kuipers et al., 2015). The method employs ultrasmall gold particles which infiltrate into the tissue much more readily than larger size gold particles (Van de Plas and Leunissen, 1993; Oliver, 2010; Melo et al., 2014). In comparison to immunoperoxidase procedures, the employed method does not generate a diffusable reaction product and thus can generate a higher resolution of the immunosignal at the electron microscopic level. With the described simple procedure it is possible with relatively minor efforts to correlate light microscopical immunolabelling data

14

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

Fig. 4. (A–D) Transmission electron micrographs from ultrathin sections of photoreceptor synapses that were immunolabelled with antibodies against RIBEYE (A–C). In (A,B), the immunolabelled synaptic ribbon complex is cross-sectioned. The arrows in (A,B) point to the active zone to which the synaptic ribbon is anchored. (C) shows a tangentially cut immunolabelled synaptic ribbon. Therefore, the active zone is not visible in (C). All micrographs in Fig. 4 were obtained from samples that were silver-enhanced for 15 min (instead of 20 min as in Figs. 1–3) to obtain a less dense silver-enhanced immunogold signal at the target structures, the synaptic ribbons. (D) shows a negative control incubation in which the primary antibody was omitted. No immunosignal is detected in (D). (E–F) Samples embedded for conventional transmission electron microscopy (TEM) for comparison. Samples in (E–G) were strongly fixed with 4% paraformaldehyde, 2.5% glutaraldehyde and OsO4 as described in “Materials and methods” in the chapter 2.2.4. Abbreviations: sr, synaptic ribbon; sv, synaptic vesicles; pm, presynaptic plasma membrane. Scale bars: 100 nm (A–D); 200 nm (E–G).

with electron microscopic data with high precision. The very same immunosignal can first be analyzed at the light microscopic level using simple widefield light microscopy with a routine light microscope (no super-resolution microscopy needed) and then analyzed at the ultrastructural level in the same sample. Since the immunolabelled semithin section can be counter-stained with light microscopic dyes, i.e. Richardson-Blue, it is easy to obtain light microscopic “landmarks” (e.g. conspicuous configurations of somata/nuclei in the ONL) that can later be used to find the very same immunolabelled structure identified in the previous light microscopic analyses also in the electron microscopic analyses. CLEM procedures often require quite specialized and expensive equipment not readily available in most labs (e.g. high-pressure freezing/freeze substitution/progressive lowering of temperature devices (Watanabe et al., 2011, 2014, Watanabe and Jorgensen, 2012); ultracryotomy (Takizawa and Robinson, 2003; Slot and Geuze, 2007; Van Rijnsoever et al., 2008) and complex, challenging analyses procedures, e.g. overlayering and aligning super-resolution fluorescent signals onto the ultrastructural context obtained by electron microscopy or using genetically enhanced probes (Takizawa and Robinson, 2000; Watanabe et al., 2011, Watanabe and Jorgensen, 2012; for review, see De Boer et al., 2015; Kuipers et al., 2015; Takizawa et al., 2015). In terms of the required technical equipment and simplicity of use, the current procedure represents a complementary approach to preexisting techniques and can be performed by a non-CLEM-specialized lab. The approach does not require specialized equipment but only generally available equipment like a cryostat. It also does not require additional steps (e.g. photoconversion) or additional probes to transform the light microscopic signal into a signal that is detectable by electron microscopy. The method works with plunge-frozen unfixed retinal tissue that is mildly fixed after cryostat sectioning with low concentrations

of PFA for a short time (2% PFA for only 2 min, RT). Still, ultrastructural preservation and membrane contrast appear satisfactory if compared to samples fixed with high concentrations of PFA/glutaraldehyde and processed with routine methods for conventional transmission electron microscopy (Fig. 4E–G). The good preservation of membranes is due to the absence of any detergents in the applied solutions similar also to methods that performed pre-embedding techniques with immunogold (Schmitz et al., 1996) or immunoperoxidase procedures (Van den Pol, 1985). The method is versatile and easily adjustable. If the silver-enhanced immunogold signal is too strong for high-resolution electron microscopy in the described combined LM/EM procedure, the silver enhancement step can be shortened to produce a less enhanced product. This minor modification allows to obtain a higher spatial resolution at the EM level (pls. compare Fig. 4A–C with Fig. 1G–L). The described approach also offers the possibility to obtain three-dimensional (3D) ultrastructural informations by the analysis of the immunosignals in serial ultrathin sections. Serial ultrathin sections for electron microscopy can be easily generated from the re-embedded semithin sections because of the flat embedding. The procedure is easier than other approaches that generate 3D informations, such as EM tomography, focused ion beam scanning electron microscopy (FIBSEM) and serial block face scanning electron microscopy (Kopek et al., 2012; Cortese et al., 2012; Koning et al., 2014; Maco et al., 2014; Bushong et al., 2015). These 3D methods can provide important informations about the spatial profile in the distribution of the antigen. The main purpose of the study was to develop a simple CLEM-compatible immunolabelling approach that can be performed also by a nonCLEM-specialized lab. Clearly, this simple approach has limitations. Freezing artefacts can occur within the sample that can result from the relative low freezing speed of manual freezing. In order to minimize

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

freezing artefacts, the volume of the tissue block should be kept as small as possible. Keeping the sample volume as small as possible is also an important issue if other tissues than the retina are used, e.g. different parts of the brain or non-neuronal tissue. Stronger fixation of the tissue might be needed to analyze soluble cytosolic proteins. For this purpose, it might be necessary to perform perfusion fixation and subsequent cryoprotection of the tissue e.g. by sucrose or other cryoprotectants, before cryosectioning. 5. Conclusions Our approach provides an alternative, straightforward “low-tech” equipment-requiring method for the correlation of light microscopic with electron microscopic data. This CLEM approach can also be performed by a non-CLEM-specialized lab with standard lab equipment. The approach is easy, fast and versatile. The same sample can be analyzed with the same procedure both at the light- as well as at the electron microscopic level without complex analyses procedures and without modifications. Identical immunolabelled structures can be easily studied at the light and electron microscopic level with the described technique. Conflict of interest None of the authors has a conflict of interests related to this work. Acknowledgements We thank Sylvia Brundaler for excellent technical assistance. Work of the authors was supported by the DFG (Graduate school GRK1326, SFB894, FOR2289 [Schm797/7-1] and Schm797/8-1). References Al-Amoudi, A., Chang, J.J., Leforestier, A., McDowall, A., Salamin, L.M., Norfén, L.P.O., Richter, K., Blanc, N.S., Studer, D., Dubochet, J., 2004. Cryo-electron microscopy of vitreous sections. EMBO J. 23, 3583–3588. Alpadi, K., Magupalli, V.G., Käppel, S., Köblitz, L., Schwarz, K., Seigel, G.M., Sung, C.H., Schmitz, F., 2008. RIBEYE recruits Munc119, a mammalian ortholog of the Caenorhabditis elegans protein unc119, to synaptic ribbons of photoreceptor synapses. J. Biol. Chem. 283, 26461–26467. Begemann, I., Galic, M., 2016. Correlative light electron microscopy: connecting synaptic structure and function. Front. Synaptic Neurosci. 8, 28. Bendayan, M., 2000. A review of the potential and versatility of colloidal gold cytochemical labeling for molecular morphology. Biotech. Histochem. 75, 203–242. Bushong, E.A., Johnson Jr., D.D., Kim, K.Y., Terada, M., Hatori, M., Peltier, S.T., Panda, S., Merkle, A., Ellisman, M.H., 2015. X-ray microscopy as an approach to increasing accuracy and efficiency of serial block-face imaging for correlated light and electron microscopy of biological specimens. Microsc. Microanal. 21, 231–238. Christensen, A.K., Komorowski, T.E., 1985. The preparation of ultrathin frozen sections for immunocytochemistry at the electron microscope level. J. Electron. Microsc. Technol. 2, 497–507. Cortese, K., Vicidomini, G., Gagliani, M.C., Boccacci, P., Diaspro, A., Tacchetti, C., 2012. 3D HDO-CLEM: cellular compartment analysis by correlative light-electron microscopy on cryosections. Methods Cell Biol. 111, 95–115. Danscher, G., 1981. Localization of gold in biological tissue. A photochemical method for light and electron microscopy. Histochem. 71, 81–88. De Boer, P., Hoogenboom, J.P., Giepmans, B.N.G., 2015. Correlated light and electron microscopy: ultrastructure lights up! Nat. Methods 12, 503–513. Dembla, M., Wahl, S., Katiyar, R., Schmitz, F., 2014. ArfGAP3 is a component of the photoreceptor synaptic ribbon complex and forms a NAD(H)-regulated, redox-sensitive complex with RIBEYE that is important for endocytosis. J. Neurosci. 34, 5245–5260. Fonta, C.L., Humbel, B.M., 2015. Correlative microscopy. Arch. Biochem. Biophys. 581, 98–110. Grabenbauer, M., 2012. Correlative light and electron microscopy of GFP. Methods Cell Biol. 111, 117–138. Grabenbauer, M., Geerts, W.J., Fernandez-Rodriguez, J., Hoenger, A., Koster, A.J., Nilsson, T., 2005. Correlative microscopy and electron tomography of GFP through photooxidation. Nat. Methods 2, 857–862. Hacker, G.W., Muss, W.H., Hauser-Kronberger, C., Danscher, G., Rufner, R., Gu, J., Su, H., Andreasen, A., Stoltenberg, M., Dietze, O., 1996. Electron microscopical autometallography: immunogold-silver staining (IGSS) and heavy metal histochemistry. Methods 10, 257–269. Hayat, M.A., 1995. Immunogold-silver staining. Principles, Methods and Applications. CRC Press Boca, Raton (ISBN 0-8493-2449-1).

15

Heidelberger, R., Thoreson, W.B., Witkovsky, P., 2005. Synaptic transmission of retinal ribbon synapses. Prog. Retin. Eye Res. 24, 682–720. Horstmann, H., Vasileva, M., Kuner, T., 2013. Photooxidation-guided ultrastructural identification and analysis of cells in neuronal tissue labeled with green fluorescent protein. PLoS One 8 (5), e64764. Karreman, M.A., Hyenne, V., Schwab, Y., Goetz, J.G., 2016. Intravital correlative microscopy: imaging life at the nanoscale. Trends Cell Biol. 26, 848–863. Koning, R.I., Celler, K., Willemse, J., Bos, E., van Wezel, G.P., Koster, A.J., 2014. Correlative cryo-fluorescence light microscopy and cryo-electron tomography of streptomyces. Methods Cell Biol. 124, 217–239. Kopek, B.G., Shtengel, G., Xu, S., Clayton, D.A., Hess, H.F., 2012. Correlative 3D superresolution fluorescence and electron microscopy reveal the relationship of mitochondrial nucleoids to membranes. Proc. Natl. Acad. Sci. USA 109, 6136–6141. Kuipers, J., van Ham, T.J., Kalicharan, R.D., Veenstra-Algra, A., Sjollema, K.A., Fijk, F., Schnell, U., Giepmans, B.N.G., 2015. FLIPPER, a combinatory probe for correlative live imaging and electron microscopy, allows identification and quantitative analysis of various cells and organelles. Cell Tissue Res. 360, 61–70. Lackie, P.M., 1996. Immunogold silver staining for light microscopy. Histochem. Cell Biol. 106, 9–17. Lagnado, L., Schmitz, F., 2015. Ribbon synapses and visual processing in the retina. Annu. Rev. Vis. Sci. 1, 235–262. Maco, B., Holtmaat, A., Jorstad, A., Fua, P., Knott, G.W., 2014. Correlative in vivo 2-photon imaging and focused ion beam scanning electron microscopy: 3D analysis of neuronal ultrastructure. Methods Cell Biol. 124, 339–361. Magupalli, V.G., Schwarz, K., Alpadi, K., Natarajan, S., Seigel, G.M., Schmitz, F., 2008. Multiple RIBEYE - RIBEYE interactions create a dynamic scaffold for the formation of synaptic ribbons. J. Neurosci. 28, 7954–7967. Matthews, G., Fuchs, P., 2010. The diverse roles of ribbon synapses in sensory neurotransmission. Nat. Rev. Neurosci. 11, 812–822. Maxeiner, S., Luo, F., Tan, A., Schmitz, F., Südhof, T.C., 2016. How to make a synaptic ribbon: RIBEYE deletion abolishes ribbons in retinal synapses and disrupts neurotransmitter release. EMBO J. 35, 1098–1114. Meisslitzer-Ruppitsch, C., Röhrl, C., Neumüller, J., Pavelka, M., Ellinger, A., 2009. Photooxidation technology for correlated light and electron microscopy. J. Microsc. 235, 322–335. Meisslitzer-Ruppitsch, C., Röhrl, C., Ranftler, C., Stangl, H., Neumüller, J., Pavelka, M., Ellinger, A., 2013. Photooxidation technology for correlative light and electron microscopy. Methods Mol. Biol. 931, 423–436. Melo, R.C., Morgan, E., Monahan-Earley, R., Dvorak, A.M., Weller, P., 2014. Pre-embedding immunogold labeling to optimize protein localization at subcellular compartments and membrane microdomains of leucocytes. Nat. Protoc. 9, 2328–2394. Oliver, C., 2010. Pre-embedding labeling methods. Methods Mol. Biol. 115, 341–345. Perkovic, M., Kunz, M., Endesfelder, U., Bunse, S., Wigge, C., Yu, Z., Hodirnau, V.V., Scheffer, M.P., Seybert, A., Malkusch, S., Schuman, E.M., Heileman, M., Frangakis, A.S., 2014. Correlative light and electron microscopy with chemical tags. J. Struct. Biol. 186, 205–213. Regus-Leidig, H., Brandstätter, J.H., 2012. Structure and function of a complex sensory synapse. Acta Physiol (Oxford) 204, 479–486. Reynolds, E.S., 1963. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol. 17, 208–212. Robinson, J.M., Takizawa, T., Vandré, D.D., 2000. Application of gold cluster compounds in immunocytochemistry and correlative microscopy: Comparison with colloidal gold. J. Microsc. 199, 163–179. Roth, J., 1996. The silver anniversary of gold: 25 years of colloidal gold marker system for immunocytochemistry and histochemistry. Histochem. Cell Biol. 106, 1–8. Schikorski, T., 2010. Pre-embedding immunogold localization of antigens in mammalian brain slices. Methods Mol. Biol. 657, 133–144. Schmitz, F., 2009. The making of synaptic ribbons: how they are built and what they do. Neuroscientist 15, 611–624. Schmitz, F., Bechmann, M., Drenckhahn, D., 1996. Purification of synaptic ribbons, structural components of the photoreceptor active zone complex. J. Neurosci. 16, 7109–7116. Schmitz, F., Königstorfer, A., Südhof, T.C., 2000. RIBEYE, a component of synaptic ribbons: a protein's journey through evolution provides insights into synaptic ribbon function. Neuron 28, 857–872. Slot, J.W., Geuze, H.J., 2007. Cryosectioning and immunolabeling. Nat. Protoc. 2, 2480–2491. Souslova, E.A., Mironova, K.E., Deyev, S.M., 2016. Application of genetically encoded photosensitization miniSOG: from correlative light microscopy to immunophotosensitizing. J. Biophotonics http://dx.doi.org/10.1002/jcbio.201600120. Sterling, P., 2013. Some principles of retinal design: the Proctor lecture. Invest. Ophthalmol. Vis. Sci. 54, 2267–2275. Takizawa, T., Robinson, J.M., 2000. FluoroNanogold is a bifunctional immunoprobe for correlative fluorescence and electron microscopy. J. Histochem. Cytochem. 48, 481–486. Takizawa, T., Robinson, J.M., 2003. Ultrathin cryosections: an important tool for immunofluorescence and correlative microscopy. J. Histochem. Cytochem. 51, 707–714. Takizawa, T., Powell, R.D., Hainfeld, J.F., Robinson, J.M., 2015. FluoroNanogold: an important probe for correlative microscopy. J. Chem. Biol. 8, 129–142. Tokuyasu, K.T., 1973. A technique for ultracryotomy of cell suspensions and tissues. J. Cell Biol. 57, 551–565. Tokuyasu, K.T., 1980. Immunocytochemistry on ultrathin frozen sections. Histochem. J. 12, 381–403. Tokuyasu, K.T., Singer, S.J., 1976. Improved procedures for immunoferritin labeling of ultrathin frozen sections. J. Ultrastruct. Res. 71, 281–287. Van de Plas, P., Leunissen, J.L., 1993. Ultrasmall gold probes: characteristics and use in immunocytochemical studies. Methods Cell Biol. 37, 241–257.

16

S. Suiwal et al. / Journal of Immunological Methods 444 (2017) 7–16

Van den Pol, A., 1985. Dual ultrastructural localization of two neurotransmitter-related antigens: colloidal gold-labeled neurophysin-immunoreactive supraoptic neurons receive peroxidase-labeled glutamate decarboxylase or gold-labeled GABA-immunoreactive synapses. J. Neurosci. 5, 2940–2954. Van Rijnsoever, C., Oorschot, V., Klumperman, J., 2008. Correlative light-electron microscopy (CLEM) combining live-cell imaging and immunolabeling of ultrathin cryosections. Nat. Methods 5, 973–980. Wahl, S., Katiyar, R., Schmitz, F., 2013. A local periactive zone endocytic machinery at photoreceptor synapses in close vicinity to synaptic ribbons. J. Neurosci. 33, 10278–10300. Wahl, S., Magupalli, V.G., Dembla, M., Katiyar, R., Schwarz, K., Köblitz, L., Alpadi, K., Krause, E., Rettig, J., Sung, C.H., Goldberg, A.F., Schmitz, F., 2016. The disease protein Tulp1 is essential for periactive zone endocytosis in photoreceptor ribbon synapses. J. Neurosci. 36, 2473–2493.

Watanabe, S., Jorgensen, E.M., 2012. Visualizing proteins in electron micrographs at nanometer resolution. Methods Cell Biol. 111. http://dx.doi.org/10.1016/B978-0-12416026-2.00015-7. Watanabe, S., Punge, A., Hollopeter, G., Willig, K.I., Hobson, R.J., Davis, M.W., Hell, S.W., Jorgensen, E.M., 2011. Protein localization in electron micrographs using fluorescence nanoscopy. Nat. Methods 8, 80–84. Watanabe, S., Lehmann, M., Hujber, E., Fetter, R.D., Richards, J., Söhl-Kielczynski, B., Felies, A., Rosenmund, C., Schmoranzer, J., Jorgensen, E.M., 2014. Nanometer resolution fluorescence microscopy in cultured cells. Methods Mol. Biol. 1117, 503–526. Zenisek, D., Horst, N.K., Merrifield, C., Sterling, P., Matthews, G., 2004. Visualizing synaptic ribbons in the living cell. J. Neurosci. 24, 9752–9759.