Male Wistar rats and C57BL/6 mice (Charles River, Calco, Lecco, Italy) of about 3 months of age were kept in animal house in a temperature-controlled room under a 12-h light-dark regimen (lights on at 6 AM) with food and water ad libitum. The animals were anesthetized by intra-peritoneal injection of chloral hydrate (40 mg/100 g body weight) and decapitated with a guillotine. The brain was removed, and brain cortex and cerebellum were quickly dissected and used immediately for synaptosomal preparation.

Synaptosomal fractions were prepared as previously described (Eyman et al., 2007). Briefly, cerebral cortex (Cx), and cerebellum (Cb) were homogenized in nine volumes of cold isotonic medium (HM: 0.32 M sucrose, 10 mM Tris-HCl, pH 7.4), using a Dounce homogenizer. After centrifugation of the homogenate (2,200 g, 1 min, 4°C), the sediment was resuspended in the same volume of HM and centrifuged under the same conditions to yield a washed sediment containing nuclei, cell debris, and other particulates (P1 fraction). The two supernatant fractions were mixed and centrifuged at a higher speed (22,000 g, 4 min, 4°C), to obtain a second sediment that was resuspended in the same volume of HM and centrifuged as described above. The washed sediment contained free mitochondria, synaptosomes, myelin, and microsomal fragments (P2 fraction). An aliquot of the P2 fraction (1 ml), with a protein concentration of 3.5 mg/ml was layered over a discontinuous gradient of 5% and 13% Ficoll dissolved in HM (2 ml each) and centrifuged at 45,000 g, 45 min, 4°C. The purified synaptosomes were recovered at the interface between the two Ficoll layers, diluted with nine volumes HM and sedimented by centrifugation at 22,000 g, 20 min, 4°C. The sediment was homogenized in HM and used for subsequent analyses.

Aliquots of homogenate and synaptosomes were lysed in RIPA buffer (50 mM Tris-HCl pH 8.8, 150 mM NaCl, 0.1% SDS, 0.5% NP-40, 0.5% DOC; protease and phosphatase inhibitor cocktail, Sigma-Aldrich) and centrifuged in Eppendorf 5415C microcentrifuge at 14,000 rpm, 5 min, 4°C. Proteins (5 μg/lane, in Sample Buffer: 60 mM Tris-HCl pH 6.8, 10% Glycerol, 2% SDS, 100 mM DTT 0.1% bromophenol blue) were separated in 10%–15% SDS-polyacrylamide gel and transferred to PVDF membranes (Merck-Millipore). Western blot analysis was performed as previously reported (Chun et al., 2004; Volpicelli et al., 2019). Briefly, filters were blocked for 2 h at RT in 5% (w/v) non-fat milk in Tris-buffered saline Tween-20 (TBST; 0.1% Tween, 150 mM NaCl, 10 mM Tris-HCl, pH 7.5) and probed with the following antibodies: anti-CSTB (1:4,000, ABIN271833 Antibodies), anti-Synaptophysin (SYP; 1:400,000, AB9272, Merck-Millipore), anti-β-actin (1:2,000, 612656, BD Biosciences), and anti-SOD-1 (1:1,000, ADI-SOD-100, Enzo Life Sciences). After several washings in TBST, membranes were incubated with secondary antibody against rabbit (1:20,000, A0545, Sigma-Aldrich) or mouse (1:20,000, NA931, GE Healthcare) linked to horseradish peroxidase. Signals were visualized by chemiluminescence (ECL, Millipore) on autoradiographic film (Fujifilm X-Ray Film).

The synaptosomal fraction was incubated in Ringer medium (90 mM NaCl, 3 mM KCl, 2 mM MgCl₂, 1 mM CaCl₂, 1 mM glucose, 100 mM sucrose, 30 mM Tris-HCl, pH 7.5) in presence or absence (negative control) of 50 μM L-Homopropargylglycine (HPG, Invitrogen). After 2 h incubation at 37°C the reaction was stopped by cooling the samples on ice. Synaptosomes were then collected by centrifugation at 190,000× g, 20 min, 4°C, and the pellet was resuspended in an appropriate volume of a buffer containing 1% SDS and 50 mM Tris-HCl pH 8. The protein concentration was determined by Pierce BCA (Thermo Scientific). Aliquots of 50 μL containing 200 μg of proteins were processed for Click-iT reaction, following the manufacturer’s instruction (Invitrogen). Briefly, 100 μL of Component A containing biotin azide (40 μM) and 10 μL of deionized water were added to the aliquot. After vortexing, the sample, 10 μL of Component B (CuSO₄), 10 μL of reaction buffer 1 and 20 μL of reaction buffer 2 were added. All the samples are gently mixed on Roto-Torque for 20 min at RT. Proteins were precipitated as a pellet by methanol/chloroform precipitation and store at −20°C until use. Newly synthesized proteins were isolated using Dynabeads M-280 Streptavidin (Invitrogen). Aliquots of 50 μL of Dynabeads were washed in PBS pH 7.4. The protein pellet was resuspended in PBS containing 0.02% Tween 20, added to the beads and mixed gently on Roto-Torque for 30 min. The samples were exposed to DynaMags for few minutes to separate newly synthetized proteins, bound to Dynabeads, from the rest of the reaction mixture. After four washing using PBS pH 7.4 Dynabeads were boiled in 0.1% SDS for 5 min at 95°C and the newly-synthetized proteins were collected in sample buffer for western blot analysis.

Synaptosomes were incubated in Ringer medium or in depolarizing medium (43 mM NaCl, 50 mM KCl, 2 mM MgCl₂, 1 mM CaCl₂, 1 mM glucose, 100 mM sucrose, 30 mM Tris-HCl, pH 7.5) to detect depolarization-induced secretion. The reaction was initiated by adding the synaptosomal fraction (300 μg/ml) to the incubation medium. Immediately after that, an aliquot of the incubation medium was transferred to ice to stop the reaction (sample at time 0). After 2 h incubation at 37°C, the reaction was stopped by cooling the sample on ice. Synaptosomes were collected by centrifugation at 190,000 g, 20 min, 4°C. Secreted proteins, present in the supernatant, were concentrated by Amicon Centrifugal Filter Devices with a cut-off of 10 kDa (Merck-Millipore) and the concentration was determined by Bradford assay (BIORAD). The synaptosomal proteins, retained in the pellet, were resuspended in RIPA Buffer. Aliquots of all samples in the sample buffer were separated in SDS-PAGE as previously described and visualized by Coomassie gel staining or Western blot; for Coomassie staining gel was placed in Coomassie solution (10% Acetic acid, 25% Isopropyl alcohol, 0.25% Coomassie brilliant blue R-250 SIGMA).

Proteins identified by Mass Spectrometry were assigned to Gene Ontology (GO) biological process groups (Annotation data set: GO biological process complete release 20190202) using the PANTHER Overrepresentation Test (release 20190308; Mi et al., 2017) with default parameters and in reference to the Rattus norvegicus genes as background.

Induced pluripotent stem cells reprogrammed from human newborn foreskin fibroblasts (CRL-2522, ATCC; O’Neill et al., 2018; Klaus et al., 2019) were used to generate cerebral organoids as previously described (Lancaster et al., 2013; Lancaster and Knoblich, 2014). Organoids were kept in 10 cm dishes on an orbital shaker at 37°C, 5% CO₂ and ambient oxygen level with medium changes every 3–4 days. Organoids were analyzed at 35 days, 60 days and 70 days after plating. For synaptosomal fraction purification, a pool of 20–40 organoids was collected by centrifugation (500× g for 10 min). Organoids were resuspended in HM and homogenized in a Dounce homogenizer with nine volumes of HM. The P2 crude synaptosomal fraction was prepared as described above. Homogenate and P2 fraction, resuspended in the sample buffer, were processed for western blot analysis as previously described. For immunostaining 16 μm sections of organoids were made using a cryotome. Immunostainings were performed as described previously (Cappello et al., 2013). Nuclei were visualized using 0.1 μg/ml 4,6-diamidino-2-phenylindole (DAPI, Sigma Aldrich). SYP antibody (AB9272, Merck-Millipore), doublecortin (DCX) antibody (AB2253, Millipore), and CSTB antibody (ABIN223204, Antibodies) were incubated at the dilution of 1:1,000, 1:1,000 and 1:400 respectively. Immunostained sections were analyzed using Leica laser-scanning microscopes.

All the statistical analyses were performed using GraphPad Prism 7 software. Data were expressed as mean ± SEM. Differences among groups were compared by ANOVA or t-test as indicated in the Figure legends. Differences were considered statistically significant at p < 0.05.

We isolated synaptosomal fractions from a homogenate of both cerebral cortex and cerebellum of rats as previously described (Eyman et al., 2007). By western blot analysis, we first assessed the distribution of a typical cytoskeletal protein, β-actin, in the homogenate and in synaptosomes. As shown in Figures 1A,B, β-actin was slightly less abundant in the synaptosomal fractions of both brain regions in comparison with its levels in the homogenates. By contrast, SYP, a well-known presynaptic protein, was significantly enriched in the synaptosomes of the brain cortex (Figure 1A) and cerebellum (Figure 1B). The differential distribution of these two proteins confirms that the synaptosomal fraction is a subcellular compartment representing the synaptic region of the neuron. When the distribution of CSTB in the synaptic compartment was examined (Figures 1A,B), it was evident that CSTB was present in rat synaptosomal fractions although it was more abundant in the homogenate, in keeping with its well-known cytosolic localization. The presence of CSTB in the synaptic region was also confirmed in the mouse cerebral cortex where the ratio of CSTB in synaptosomes vs. homogenate was even higher than in the rat (Figure 1C). Altogether, these results clearly indicate the synaptic localization of CSTB, suggesting its involvement in synaptic plasticity.

In view of the crucial role played by synaptic protein synthesis in brain plasticity, we tested if CSTB was locally synthesized in the synaptosomal fraction. To this end, we performed metabolic labeling of newly synthesized proteins using Click-iT^TM L-HPG as a precursor (Best, 2009). Newly synthesized proteins incorporating HPG were isolated from the background proteins by biotinylation and subsequent purification with streptavidin beads. In Western blot analysis, CSTB was identified as one of the newly synthesized (HPG-labeled) proteins in the synaptosomes of rat cerebral cortex (Figure 2). The CSTB band was negligible in the negative control, i.e., the reaction mixture of the synaptosomal fraction incubated without HPG. These data demonstrated, for the first time, that CSTB is, at least in part, synthesized in the synaptic region of the rat cerebral cortex.

To investigate the role of CSTB in synaptic physiology, we tested its possible secretion from synaptosomes. Thus, the synaptosomal fraction was incubated for 2 h in physiological extracellular solution (Ringer medium). At the end of the incubation period, the incubation medium was collected, and its contents were concentrated using Amicon filters with a cut-off size of 10 kDa. As judged by the Coomassie blue staining, only few proteins were found secreted in the medium, with a pattern (Figure 3A, lane 5) that was strikingly different from that of synaptosomal proteins (Figure 3A, lane 2). Among the secreted proteins, we detected CSTB by Western blot analysis, while SYP, a very abundant protein of synaptic vesicles, was completely absent. Likewise, also β-actin was not present in the incubation medium (Figure 3B, lane 5).

As a next step, we checked if CSTB expression level was modulated by depolarization. Thus, synaptosomes were incubated for 2 h in a depolarizing medium containing high concentration of K⁺ (50 mM, see “Materials and Methods” section). The CSTB, as well as SYP expression levels in synaptosomes did not change in the two incubation conditions (Figure 3B, lanes 2 and 3, and Figure 3C). Interestingly, in depolarizing condition, the amount of secreted CSTB significantly increased compared with the incubation in physiological extracellular solution (Figure 3B, lanes 5 and 6 respectively, and Figure 3D), indicating that the secretion of this protein is regulated by depolarization. SYP, as well as β-actin were not detectable in the incubation medium in both experimental conditions in any of the biological replicates (Figure 3B, lanes 5 and 6).

In order to identify the proteins secreted from synaptosomes and to survey the possible presence of known CSTB partners in the secreted medium, we performed a mass-spectrometry analysis of the secreted proteins after 2 h incubation of the synaptosomes either in Ringer medium (sample-R), or in the depolarizing medium (sample-D) by use of nanoLC-Tandem high resolution MS/MS. Using the filtering criteria indicated in “Materials and Methods” section, we detected 253 proteins in sample R and 160 in sample D, with 125 common proteins (Figure 4A; Supplementary Table S1). Among the proteins present in both samples, we identified some of the known partners of CSTB, such as Spectrin beta chain, Neurofilament light polypeptide (Di Giaimo et al., 2002) and SOD1 (Ulbrich et al., 2014), while we did not find any protease target of CSTB such as cathepsin B. The presence of SOD1 and the absence of cathepsin B in the secreted samples was validated by Western blot analysis (Figure 4B). As to the same set of common proteins, the results of GO analysis showed a strong enrichment for the terms related to the synaptic activity (Figure 4C), providing additional hint toward the involvement of CSTB in neuronal plasticity.

To extend our study to humans, we took advantage of the 3D human multicellular model of brain cortical tissue, namely cerebral organoids (Lancaster et al., 2013; Lancaster and Knoblich, 2014; Camp et al., 2015). We first used immunohistochemical analyses to characterize the organoids. The ventricle like structures in the organoids were detected by immunostaining with beta-catenin (CTNNB) that labels adherent junctions allowing the identification of the apical belt, and with DCX that detects young neurons in the cortical plate-like zone (Figures 5A,B,D). To identify synaptic regions, 70 days old human brain organoids were stained with the presynaptic marker SYP (Figure 5C). As shown in Figures 5D,E,E′, the colocalization of DCX and SYP indicates that neurons in brain organoids developed in our laboratory express presynaptic markers to a certain extent, already at 70 days. In addition, we observed the localization of CSTB in some of the SYP positive-synaptic regions (Figures 5F,G). We prepared crude synaptosomal fraction (P2) from 35, 60 and 70 days old human brain organoids (pool of 20–40 organoids for each time point), and we analyzed the distribution of SYP, β-actin and CSTB in synaptosomes and in the corresponding homogenates (Figure 5H). SYP is highly enriched in the crude synaptosomal fraction compared to the homogenate, as expected for a pre-synaptic protein. On the other hand, as expected for a ubiquitous cytoskeletal protein, β-actin was equally distributed in the two fractions throughout the developmental stages of the organoids. These data are comparable to the ones obtained from rodent brains fractionation (Figure 1), indicating that synaptosomal fraction prepared from human brain organoids is a good model to study in vitro the synaptic region in a human model. CSTB has a different subcellular distribution than SYP and β-actin, being more abundant in homogenate than in synaptosomes (Figure 5H). These data indicate that in human brain organoids, as in rodent brains, CSTB is mainly a cytosolic protein, although it is partially localized also in the synaptic region. Interestingly, CSTB is almost undetectable in synaptosomal fraction from 35 days old organoids, while its level increases in synaptosomes at later stages, suggesting that the CSTB localization in the presynaptic terminals progressively increases with neuronal maturation.