Primary antibody: The polyclonal rabbit Arc antibody (Synaptic Systems, Göttingen, Germany, Cat#156003, RRID: AB_887694) was raised against recombinant protein corresponding to AA 1–396 from mouse Arc (UniProt Id: Q9WV31), used at 1:400 for electron microscopy (EM), 1:500 for light microscopy (LM) and 1:600 for western blot (WB). The specificity of this Arc antibody has been verified by immunocytochemistry of dissociated hippocampal neuron cultures prepared from wild type (WT) and Arc knockout (KO) littermates (Niere et al., 2012), and has been characterized in our lab previously (Egbenya et al., 2023). Monoclonal actin antibody (Millipore, Germany, Cat# MAB1501, RRID:AB_2223041), raised against amino acids 357–375 at the C-terminus of actin of human origin, used at 1:20000 for WB.

Secondary antibodies: Goat anti-rabbit coupled to 10 nm gold particle (Abcam, Cambridge, United Kingdom, IgG H&L antibody, Cat#AB27234, Lot#GR73433 and Lot#GR104536-1, RRID: AB_954427) used at 1:40 for EM. Biotinylated donkey anti-rabbit antibody (GE healthcare, United Kingdom, lot#9554939) was used at 1:100 for LM. For WB, a monoclonal anti-rabbit IgG alkaline phosphatase antibody produced in mouse (Sigma, United States, lot#083M4782V) was used at 1:10,000.

For the WB experiments, three Wistar rats weighing 250–300 g (Scanbur, Nittedal, Norway) were used for the regional brain experiments. Three Wistar rats weighing 250–300 g (Scanbur, Nittedal, Norway) were used for the LM experiments. Four Sprague–Dawley rats (Harlan Sprague–Dawley Inc., Indiana, United States) weighing 120–240 g were used for the EM experiments. Experimental protocols were approved by the Institutional Animal Care and Use Committee and conform to the National Institute of Health guidelines for the care and use of animals, as well as international laws on protection of laboratory animals, with the approval of a local bioethical committee and under the supervision of a veterinary commission for animal care and comfort. The animals were treated in accordance with the guidelines of the Norwegian Committee on Animal Experimentation (Norwegian Animal Welfare Act and European Communities Council, Directive of 24 November 1986-86/609/EEC). Every effort was made to minimize the number of animals used and their sufferings. This article does not contain any studies with human participants performed by any of the authors.

Whole rat brain vibratome sections (50 μm) from three rats were immunostained with anti-Arc antibody using the labeled streptavidin-biotin method to optimize antibody concentration and labeling conditions. The sections were incubated in 1 M ethanolamine in sodium phosphate buffer (0.1 M, pH 7.4) for 30 min. After washing 3 × 1 min in sodium phosphate buffer, the sections were incubated in blocking buffer (10% normal goat serum in sodium phosphate buffer) for 1 h prior to primary antibody incubation overnight (diluted in blocking buffer at concentration 1:100). The following day, the sections were rinsed in sodium phosphate buffer for 3 × 5 min, and then incubated in 1% normal goat serum in sodium phosphate buffer for 20 min. The sections were incubated in biotinylated secondary antibody diluted in the same buffer for 1 h at room temperature. Washing 3 × 5 min with sodium phosphate buffer followed. Then the sections were incubated with streptavidin-biotinylated horse-radish peroxidase (HRP) diluted in sodium phosphate buffer with 1% normal goat serum for 1 h, before washing in sodium phosphate buffer for 5 × 10 min. As the final steps, the sections were incubated in 0.05% diaminobenzidine (DAB) in sodium phosphate buffer for 5 min, and then incubated in 0.01% H₂O₂ and 0.05% DAB diluted in buffer for 6 min, before final washing in sodium phosphate buffer 3 × 5 min. Sections were mounted on glass slides with glycerol gelatin.

Three rats were decapitated and the brains were removed. Brain regions of interest were dissected and submerged in ice-cold Hepes-buffered sucrose (0,32 M sucrose, 4 mM Hepes, pH 7.4) containing protease and phosphatase inhibitor cocktails. Corresponding brain regions from the different rats were combined and homogenized in a Hepes buffer with a glass-Teflon homogenizer (10–15 strokes) and centrifuged (800–1,000 g, 10 min, 4°C). 20 μg of protein extracts (brain region materials) were run on a 4–20% SDS-acrylamide gel (Bio-Rad Laboratories Inc., United States, Cat# 345-0007) at 200 V and 400 mA for 50–55 min followed by blotting for 30 min at 25 V, 1 A. The different protein extract amounts were used to optimize band intensities for quantification. Prior to this, the samples were mixed with loading buffer (62.5 mM Tris–HCl, 10% glycerol, 2% SDS, 5% 2-mercaptoethanol, 0.025% bromophenol blue). The membrane was blocked with 5% skim-milk solution for 60 min followed by an overnight incubation with primary antibody (in 2.5% skim-milk solution). Secondary antibody incubation (in 1.25% skim-milk solution), washing (4 × 15 min using TBS-T buffer) and signal detection using ECF substrate (GE Healthcare, United Kingdom, Cat#1067873) were performed. BioRad Scanner (Bio-Rad Laboratories, Inc., United States) was used to visualize the fluorescence signals. Three blots were used for the semiquantitative WB analyses. We utilized photoshop to create a rectangular marker designed to tightly enclose the bands on the WB membrane images. Values “mean” and “pixels” were then transferred to an Excel sheet. The intensity of each band was calculated by taking the mean times the pixels divided by 1,000. The background was measured and substracted from mean band intensity. Band intensity of the blot was normalized to loading controls (actin). The regional WB were performed to visualize brain regional differences in expression of Arc, not for statistical comparison.

Four rats were used for EM. We optimized conditions for ultrastructural visualization of vesicles with two different fixation protocols. Two of the rats were deeply anaesthetized with Equithesin (0.4 mL/100 g, i.p.) and transcardially perfused with a flush of 30 mL of 2% Dextran-T70 in 0.1 M phosphate buffer and thereafter with 1.0 L of a fixative made up of glutaraldehyde (0.5%) and formaldehyde (4.0%) in the same buffer. Two of the other rats were deeply anesthetized with an i.p. injection of a mixture of ketamine (100–200 mg/kg) and xylaxine (10–20 mg/kg) and sacrificed by transcardial perfusion with 30 mL 4% Dextran-T70 in 0.1 M sodium phosphate buffer (pH 7.4) followed by a mixture of 0.1% glutaraldehyde and 4.0% formaldehyde in the same buffer. The fixed brains were left in situ overnight at 4°C, and then dissected out from the skulls and stored in the above-mentioned fixatives after diluting 1: 10 in 0.1 M phosphate buffer at 4°C. No differences between the two protocols were observed, so all four brains were used in the study.

Small (0.5–1.0 mm) blocks dissected from the CA1 area of the hippocampus were freeze-substituted, sectioned, and immunolabeled as described previously (Hussain et al., 2019). After sectioning and mounting on nickel grids, the sections were first incubated in TBST (20 mL 0.05 M Tris–HCl, 180 mL dH2O with 1.62 g NaCl, 0.02 g Triton X-100) with glycine for 10 min. Then they were incubated in blocking buffer (2% human serum albumin in TBST) for 10 min, before incubation overnight in primary antibody diluted in blocking buffer. Next day, the sections were rinsed 3 × 10 min in TBST, before incubating in blocking buffer for 10 min. They were then incubated for 2 h in secondary antibody coupled to 10 nm colloidal gold in blocking buffer. Subsequently, the grids were briefly washed with distilled water six times. After drying, they were incubated in drops of 1% uranyl acetate for 1.5 min, then washed three times with distilled water, and incubated in drops of 0.3% lead citrate for 1.5 min, and finally washed with distilled water and dried.

Four rats were used for the quantitative postembedding electron microscopy. From each animal, three ultrathin sections were immunolabeled. Twenty-five synaptic profiles from each section were selected for the quantitative analysis. A total of 291 synaptic profiles were quantified. From each image of a synapse, we examined 4 regions of interest (ROIs). Similar selections were made for the myelin profiles in separate images. Five myelin profiles from each section were selected for the quantitative analysis. A total of 56 images of myelin were used for the quantification. Electron micrographs were obtained at random from the middle layer of the stratum radiatum of the dorsal CA1 region of the hippocampus, midway between the stratum radiatum and stratum lacunosum. Asymmetric synapses in this location are well-defined glutamatergic synapses. Immunolabeling was quantified as number of gold particles/μm of membrane length in asymmetric synapses (from Schaffer collaterals) and as number of gold particles/μm² in the intracellular compartments. Specific plasma membrane and cytoplasmic compartments were defined as ROIs and used for quantification. These included the presynaptic cytoplasm (PreCy), the active zone (AZ), the postsynaptic cytoplasm (PoCy) and the postsynaptic density (PSD). Images of synaptic profiles clearly showing the ROI were acquired for the quantitative analysis. The following total numbers of synaptic profiles were analyzed for the different ROI: PreCy = 291, PoCy = 291, AZ = 291, PSD = 291. Arc immunonegative synaptic profiles were also included in the quantitative analyses. For the quantitative analysis of percentage of synapses exhibiting MVBSL, EVs, and Arc immunolabeling in different synaptic compartments, we utilized a set of 190 synaptic profiles with well-preserved ultrastructure. Excitatory synapses were identified by the presence of two closely aligned membranes with a synaptic cleft, a prominent postsynaptic density and circular synaptic oriented vesicles at the presynaptic side. An in-house extension to analysis connected with SPSS (SPSS Inc., Chicago, IL, United States) was used to evaluate the gold particle labeling of the specific plasma membrane and cytoplasmic compartments. The software calculated area particle density (number per unit area) over cytoplasmic compartments and linear particle density (number per unit length of curve) over membrane domains. In the latter case, it measured the distance from each particle-center to the membrane and included only those particles which were within an operator-defined distance from the curve segment. For general plasma membranes, the inclusion distance was symmetric between −21 nm and + 21 nm (negative value signifying an intracellular location). The inclusion distance was defined as the distance between the epitope and the center of the gold particle, corresponding to the radius of the particle (5 nm) and the length of the interposed primary antibody (8 nm) and length of the secondary antibody (8 nm). Data for particles were collected in ASCII files as flat tables and exported to SPSS for further statistical and graphical analysis.

Quantitative data from EM were analyzed with SPSS by performing a Mann–Whitney non-parametric test, as all criteria for this test were met and the data were not normally distributed. The criterion for statistical significance was set to p < 0.05 as standard.

To determine the expression of Arc in different brain regions in the central nervous system (CNS), we performed western blotting and immunostaining for light microscopy. The anti-Arc antibody gave a single immunoreactive band at the expected molecular weight. Arc was expressed in different brain regions (Figure 1), including the brain stem, cortex, cerebellum, thalamus, and the hippocampus. The highest band intensities of Arc, normalized against actin, were observed in the cortex and the hippocampus (Figures 1A,B). Concentrations seem to be lower in the brain stem, thalamus and the cerebellum (Figures 1A,B). These immunoblotting results correspond to the immunoreactivity observed for immunostained sections using the labeled streptavidin-biotin method for light microscopy (Figure 1C). Most prominent labeling intensities were observed in the cortex and the hippocampus (Figure 1C). While the cerebellar cortex showed prominent immunoreactivity, the underlying white matter showed only weak staining, explaining the low band intensity for the cerebellar region as a whole.

The specificity of the Arc antibody has previously been verified by immunocytochemistry of dissociated hippocampal neuron cultures prepared from WT and Arc KO littermates (Niere et al., 2012), and we have also characterized it in a recent study (Egbenya et al., 2023). However, considerable effort was employed to verify the specificity of the anti-Arc antibody used in our quantitative immunogold electron microscopical studies. As a negative control, we quantitated the immunogold labeling over myelin sheaths around axons and compared to levels in the postsynaptic cytoplasm (Figure 2). The postsynaptic cytoplasm has a well-established expression of Arc, as shown consistently in our electron micrographs (Figure 2A). However, the myelin sheaths were virtually devoid of immunogold particles (Figure 2B), though the axon profiles inside the sheaths were typically labeled, in spite of the fact that these sheaths contain high concentrations of other “sticky” proteins (Morell, 1999). Besides neurons, Arc expression has also been shown in astrocytes, but not in oligodendrocytes (Rodriguez et al., 2005; Schochet et al., 2005; Uhlen et al., 2015). We chose here to use the myelin sheaths as bona fide knock-out tissue, with the additional advantage that this tissue had been subject to the exact same fixation and other treatments as the synaptic structures investigated. Quantitative analyses of synaptic and myelin profiles revealed a highly significant difference in gold particle densities between the postsynaptic cytoplasm (55.2 gold particles/μm² ± 3.7) and the myelin sheaths (10.4 gold particles/μm² ± 2.2) (n = 291 in the PoCy group, n = 56 in the myelin group, p < 0.0001, Mann–Whitney U test) (Figure 2C). As additional control measurements, sections were incubated with primary antibody coupled with subsequent incubation with a secondary antibody from an inappropriate species, revealing no labeling at all. Also, no labeling was observed when sections were incubated in secondary antibody only, omitting primary antibody incubation. We concluded that the Arc antibody is specific when used under the present electron microscopical conditions.

We found that both presynaptic terminals and postsynaptic spines in Schaffer collateral synapses in the stratum radiatum of the CA1 region in the rat hippocampus displayed Arc immunogold labeling (Figure 3A). The labeling of Arc often appeared as clusters of gold particles associated with about 2–5 vesicles in the cytoplasmic regions of interest (Figure 3B). The presynaptic vesicles were about the size of regular synaptic vesicles, i.e., about 40 nm in diameter. In the presynaptic cytoplasm, the immunogold particles were present over some synaptic vesicles (Figure 3C), i.e., the gold particles were localized either close to the vesicle membrane or within the vesicle lumen. The protein was also present over the plasma membrane, predominantly over the active zone (Figure 3D). In the postsynaptic spine, several gold particles were seen in the postsynaptic cytoplasm (Figure 3E). The highest density of Arc in the spines was typically observed over the postsynaptic density (PSD) (Figure 3F). Some vesicles in or close to the PSD were Arc positive. Quantitative analysis of Arc immunogold labeling density in synaptic ROIs (Figure 4A) significantly showed 25% higher concentration of the protein in the postsynaptic cytoplasm (55.2 gold particles/μm² ± 3.7) compared to the presynaptic cytoplasm (44.0 gold particles/μm² ± 2.9) (n = 291 in both groups, p = 0.029, Mann–Whitney U test) (Figures 4B). Similarly, the concentration of Arc in the PSD (10.3 gold particles/μm ± 0.7) was significantly 26% higher than in the presynaptic active zone (7.6 gold particles/μm ± 0.6) (n = 291 in both groups, p = 0.006, Mann–Whitney U test) (Figure 4C). Quantification of Arc immunogold labeling of synapses revealed that Arc is expressed exclusively in the presynaptic terminal in 21% of synapses and solely in the postsynaptic spine in 10% of synapses. The majority of synapses (62%) exhibited simultaneous presence of Arc in both the terminal and the spine. Approximately 7% of synapses showed no immunoreactivity for Arc (Figure 4D). Quantitative analysis of Arc immunogold labeling in the presynaptic terminals revealed that 63% of terminals displayed Arc immunogold labeling exclusively in the presynaptic cytoplasm, while 2% exhibited labeling specifically at the active zone. In 18% of terminals, Arc was localized in both the presynaptic cytoplasm and at the active zone, whereas 17% of terminals showed no immunoreactivity for the protein (Figure 4E). Furthermore, within each synaptic terminal, there was an average presence of approximately 40 synaptic vesicles. Among these presynaptic vesicles, a fraction of 6% exhibited the presence of Arc protein. Quantification of immunogold labeling in the postsynaptic spines demonstrated that Arc is present particularly in the postsynaptic cytoplasm in 56% of spines, while 3% showed specific labeling at the postsynaptic density. Furthermore, in 14% of spines, Arc was localized in both the postsynaptic cytoplasm and at the postsynaptic density. However, 27% of spines displayed no immunoreactivity for Arc protein at all (Figure 4F).

Extracellular vesicles (EVs) with an approximate diameter of 26 nm were detected in the synaptic cleft or close to the synapse (Figure 5). Some of these vesicles were labeled for Arc. Unlike the clusters of gold particles associated with synaptic vesicles in the cytoplasm, only single gold particles were observed attached to these extrasynaptic vesicles. EVs were detected in 10% of the synapses. Approximately 6% of synapses showed Arc-positive EVs, while 4% of synapses exhibited Arc-negative EVs. Overall, 63% of the synaptic EVs contained Arc, while the remaining 37% were immunonegative for the protein (Table 1). We also found examples of clusters of vesicles, some of which were enclosed by an outer membrane within the cytoplasm. These round structures resembled multivesicular bodies and were classified as multivesicular body-like structures (MVBLS). Some of these MVBLS were labeled with several gold particles (Figure 6). The MVBLS were observed both in the presynaptic terminals (Figure 6A) and postsynaptic spines (Figure 6B). Approximately 4% of presynaptic terminals contained MVBLS, while 5% of postsynaptic spines displayed MVBLS. Interestingly, within both the presynaptic terminals and the postsynaptic spines, there was an equal distribution, with 50% of the MVBLS being positive for Arc and the remaining 50% being negative for Arc (Table 2).

In the dendrites immediately adjacent to the spines, Arc was present in the cytoplasm attached to single vesicles (Figure 7A) or MVBLS. Labeling of dendritic membrane was rarely observed. Arc was clearly present also in the axonal cytoplasm (Figure 7B). The protein was not, however, observed in myelin sheaths surrounding the axons. Occasionally, isolated gold particles were noted within the cytoplasm of astrocytes (Figure 7C). This observation was interpreted as incidental background labeling. Notably, Arc was not detected at the astrocytic membrane.