Abstract
Learning and behavior activate cue-specific patterns of sparsely distributed cells and synapses called ensembles that undergo memory-encoding engram alterations. While Fos is often used to label selectively activated cell bodies and identify neuronal ensembles, there is no comparable endogenous marker to label activated synapses and identify synaptic ensembles. For the purpose of identifying candidate synaptic activity markers, we optimized a flow cytometry of synaptoneurosome (FCS) procedure for assessing protein alterations in activated synapses from male and female rats. After injecting yellow fluorescent protein (YFP)-expressing adeno-associated virus into medial prefrontal cortex (mPFC) to label terminals in nucleus accumbens (NAc) of rats, we injected 20 mg/kg cocaine in a novel context (cocaine+novelty) to activate synapses, and prepared NAc synaptoneurosomes 0–60 min following injections. For FCS, we used commercially available antibodies to label presynaptic and postsynaptic markers synaptophysin and PSD-95 as well as candidate markers of synaptic activity [activity-regulated cytoskeleton protein (Arc), CaMKII and phospho-CaMKII, ribosomal protein S6 (S6) and phospho-S6, and calcineurin and phospho-calcineurin] in YFP-labeled synaptoneurosomes. Cocaine+novelty increased the percentage of S6-positive synaptoneurosomes at 5–60 min and calcineurin-positive synaptoneurosomes at 5–10 min. Electron microscopy verified that S6 and calcineurin levels in synaptoneurosomes were increased 10 min after cocaine+novelty. Pretreatment with the anesthetic chloral hydrate blocked cocaine+novelty-induced S6 and calcineurin increases in synaptoneurosomes, and novel context exposure alone (without cocaine) increased S6, both of which indicate that these increases were due to neural activity per se. Overall, FCS can be used to study protein alterations in activated synapses coming from specifically labeled mPFC projections to NAc.
SIGNIFICANCE STATEMENT Memories are formed during learning and are stored in the brain by long-lasting molecular and cellular alterations called engrams formed within specific patterns of cue-activated neurons called neuronal ensembles. While Fos has been used to identify activated ensemble neurons and the engrams within them, we have not had a similar marker for activated synapses that can be used to identify synaptic engrams. Here we developed a procedure for high-throughput in-line analysis of flow cytometry of synaptoneurosome (FCS) and found that ribosomal S6 protein and calcineurin were increased in activated mPFC–NAc synapses. FCS can be used to study protein alterations in activated synapses within specifically labeled circuits.
Introduction
Long-lasting rewarding and aversive memories are thought to be encoded by persistent molecular and cellular alterations called engrams that are induced only in a small fraction of sparsely distributed neurons called neuronal ensembles that were selectively activated during behavior (Semon, 1904; Hebb, 1949; Sutherland and McNaughton, 2000; Cruz et al., 2013; Tonegawa et al., 2015; Whitaker and Hope, 2018; Rao-Ruiz et al., 2021). While many candidate engrams have recently been identified in cell bodies of neuronal ensembles, many of the critical engrams encoding memories are thought to be in small fractions of synapses, called synaptic ensembles, that are selectively activated during behavior (Quinlan et al., 1999; Schuman, 1999; Rosenberg et al., 2014; Benito et al., 2018; Hafner et al., 2019).
Progress has been made to identify these synaptic ensembles in situ, but it has been difficult to isolate and study them for molecular analysis in the same way that we do for neuronal ensemble cell bodies (Cruz et al., 2013, 2015). Fos and other immediate early genes have long been used as endogenous markers for identifying activated neuronal ensemble cell bodies that intermingle among the much more abundant nonactivated neuronal cell bodies, and fluorescence-activated cell sorting (FACS) has been used to sort these Fos-expressing neurons for subsequent analysis of unique molecular alterations (Guez-Barber et al., 2011; Cruz et al., 2013; Liu et al., 2014; Li et al., 2015; Rubio et al., 2015). The critical feature of FACS that allows the isolation of this very small fraction of behaviorally activated Fos-expressing neurons is “single-cell resolution” (Lyons and West, 2011; Cruz et al., 2013; Kawashima et al., 2014; Rubio et al., 2016; Whitaker and Hope, 2018; Chen et al., 2019). Unfortunately, we do not have a comparable endogenous marker to identify and isolate behaviorally activated synapses from the surrounding nonactivated synapses with single-synapse resolution.
The flow cytometry method fluorescence-activated synaptosome sorting (FASS) has recently been developed to isolate labeled synaptosomes from brain for subsequent molecular analysis (Biesemann et al., 2014; Luquet et al., 2017; Paget-Blanc et al., 2022). However, these studies examined the overall population of synapses independent of their activation state and used synaptosomes that do not have intact postsynaptic sacs containing the cytoplasmic molecules of dendritic spines, many of which have previously been proposed as key candidate engrams encoding memories (Quinlan et al., 1999; Schuman, 1999; Rosenberg et al., 2014; Benito et al., 2018; Hafner et al., 2019). To address this, we developed a flow cytometry of synaptoneurosome (FCS) procedure to identify synaptic protein alterations in activated yellow fluorescent protein (YFP)-labeled glutamatergic terminals coming from medial prefrontal cortex (mPFC) and synapsing on dendritic spines of medium spiny neurons in nucleus accumbens (NAc). Synaptoneurosomes (SNs) include both sealed presynaptic terminals and sealed postsynaptic sacs (Hollingsworth et al., 1985). We used FCS to assess several candidate synaptic activity markers 0–60 min following acute cocaine injections in a novel context, a treatment known to cause strong neural activation of mPFC and NAc brain areas (Uslaner et al., 2001; Mattson et al., 2008; Marin et al., 2009). Electron microscopy confirmed that the number of ribosomal S6 protein (S6)-containing and calcium/calmodulin-dependent serine/threonine protein phosphatase 2B (calcineurin)-containing synapses were increased 10 min following cocaine in the novel context. We then used the anesthetic chloral hydrate to block neural activity-induced synaptic activity to test whether increased levels of these two synaptic proteins are activity dependent.
Materials and Methods
Subjects
We used a total of 129 male and female Sprague Dawley and Long–Evans rats [National Institute on Drug Abuse (NIDA), Intramural Research Program (IRP) Animal Care Facility] that weighed 200–500 g at the start of the experiment. Rats were housed and maintained in the animal facility under a reverse 12 h light/dark cycle (lights off at 8:00 A.M.) with food and water freely available. Procedures followed the guidelines outlined in the Guide for the Care and Use of Laboratory Animals (eighth edition; http://grants.nih.gov/grants/olaw/Guide-for-the-Care-and-Use-of-Laboratory-Animals.pdf) and were approved by the NIDA IRP Animal Care and Use Committee.
Viral injections
For characterization experiments and Experiments 1, 2, 4, and 5, we injected AAV1-CaMKIIa::hChR2-eYFP virus [undiluted, 1013 viral genomes (vg)/ml; catalog #26969-AAV1, Addgene] bilaterally into both prelimbic (0.5 µl/side) and infralimbic (0.5 µl/side) cortices of mPFC to label presynaptic terminals from mPFC to dorsal striatum and NAc. We used channelrhodopsin 2 (ChR2) fused to YFP as a membrane anchor to ensure transport of YFP to the synaptic membranes in NAc. We injected the virus first in the infralimbic region, and then the needle was raised to inject in prelimbic region. Infralimbic coordinates were anteroposterior (AP) +2.9/3.0 mm, mediolateral (ML) ±1.5 mm (10° angle), and dorsoventral (DV) −5.0 mm; and prelimbic coordinates were AP +2.9/3.0 mm, ML ±1.2 mm (10° angle), and DV −3.5 mm, with the nose bar set at −3.3 mm (Paxinos and Watson, 2005). We used AP +2.9 mm for females and AP +3.0 mm for males because of their difference in size. For Experiment 3, we injected AAV9-CAG::ChR2-mCherry virus (diluted 10-fold to final concentration of 1012 vg/ml; catalog #100054-AAV9, Addgene) in seven rats using the same procedure described for the enhanced YFP (eYFP) virus above. After recovering from surgery, rats were placed in their home cages for 6–8 weeks to allow for robust transgene expression in the terminals.
Synaptoneurosome preparation
Rats were rapidly decapitated using a guillotine. All of the following steps for preparing synaptoneurosomes were performed on ice. We washed each brain in PBS to remove blood cells and used razor blades to cut a 3 mm total thickness of NAc (2.52–0.60 mm from bregma) in a metal brain matrix. We placed the brain slices on a glass plate on ice to dissect NAc core from both hemispheres from one to two slices. We used a Nightsea lamp with a Royal Blue color Light Head (excitation, 440–460 nm; emission filter, 500 nm longpass; catalog #SFA-LFS-RB, Electron Microscopy Sciences) to confirm YFP-labeled mPFC processes in the NAc core before dissecting the region of interest. The lamp was at least 6 inches above the brain section with a light intensity maximum of 0.36 mW/mm2. The brain and brain sections were kept on ice for at least 2 min before the blue light was turned on. The ice-cold conditions prevent significant activation of signaling enzymes and transport mechanisms necessary for S6 or calcineurin translocation.
Each sample was obtained from one rat. The tissue was placed into a Teflon-glass tissue grinder tube with 2 ml of ice-cold KREB-sucrose solution [Krebs-Henseleit Buffer Modified; catalog #K3753, Sigma-Aldrich) and 0.32 m sucrose] plus EDTA-free protease inhibitors (Complete; catalog #11836170001, Roche) and phosphatase inhibitors (PHOSS-RO, PhosSTOP, Roche) and then homogenized with five to six up and down strokes of a motor-driven pestle attached to a Wheaton overhead stirrer (∼1500 rpm; catalog #WHE-903475, DWK Life Sciences). The homogenate was then transferred to a 10 ml polycarbonate tube and centrifuged at 1000 × g for 10 min at 4°C, with maximum acceleration and deceleration (Avanti J-26 XPI Centrifuge, JA-20 rotor, Beckman Coulter). The supernatant (S1) was transferred to a new 10 ml polycarbonate tube on ice and centrifuged at 13,800 × g for 20 min. After removing the supernatant, the pellet (P2) was resuspended in 1.8 ml of KREBS-sucrose solution and layered on to a Percoll gradient (2 ml/layer of 23%, 15%, 10%, and 3%) and centrifuged at 35,500 × g for 7 min at maximum acceleration and 2 min of deceleration. We used a Pasteur pipette to collect the synaptoneurosome-enriched layer, identified as a white viscous band between the 10% and 15% Percoll layers. We added KREBS-sucrose solution to the collected synaptoneurosomes to a final volume of 10 ml and centrifuged the sample at 32,800 × g for 40 min at maximum acceleration and deceleration. After discarding most of the supernatant, the synaptoneurosome pellet that is floating in the bottom of the tube was resuspended in ∼1 ml of KREBS-sucrose solution, transferred to a 1.5 ml Eppendorf tube, and centrifuged at 15,000 × g at 4°C for 10 min to produce the P3 pellet.
To fix the synaptoneurosomes, the P3 pellet was resuspended in 1 ml of 0.1% paraformaldehyde (PFA) in PBS solution. The tubes were vortexed briefly and then rotated end over end at 4°C for 15 min. The tubes were centrifuged again at 15,000 × g at 4°C for 30 min, and then the pellets were resuspended in 1 ml of 0.2% Tween 20 in PBS solution. The tubes were vortexed then rotated end over end at 4°C for 30 min. The tubes were centrifuged at 15,000 × g at 4°C for 7 min, and the pellets were resuspended in 500 μl of 5% DMSO in KREBS-sucrose solution. Finally, pellets were split into 125 μl samples for slow freezing (−20°C overnight, then −80°C within a Styrofoam box) and stored at −80°C.
Immunolabeling for flow cytometry
After thawing fixed P3 pellets on ice, we split one of the 125 μl samples into multiple tubes (up to 10) for immunolabeling with different antibodies. A primary antibody or a conjugated primary antibody was added to each of these tubes, and the final volume was brought to 350 μl with PBS on ice. We used the primary antibodies: calcium/calmodulin-dependent protein kinase type II subunit α (CaMKIIα; dilution, 1:200; 6G9; catalog #MA1-048, Thermo Fisher Scientific), phospho-Thr286-CaMKIIα (dilution, 1:200; 22B1; catalog #MA1-047, Thermo Fisher Scientific), calcineurin A (dilution, 1:500; catalog #ab109412, Abcam), and phospho-Ser469-calcineurin (dilution, 1:100; a gift from Karen Perez Del Arce, Broad Institute, Cambridge, MA). We used the following conjugated primary antibodies: synaptophysin-AF647 (dilution, 1:1000; Catalog #ab196166, Abcam), postsynaptic density-95 (PSD-95)-AF594 (dilution, 1:25; catalog #sc-32290 AF594, Santa Cruz Biotechnology), Arc-AF647 (dilution, 1:1000; C-7; catalog #sc-17839 AF647, Santa Cruz Biotechnology), S6 (54D2)-AF647 (dilution, 1:50; catalog #5548S, Cell Signaling Technology), and phospho-Ser235/236-S6 ribosomal protein-PE-Cy7 (D57.2.2E; dilution, 1:500; catalog #34411S, Cell Signaling Technology). We incubated samples for 30 min with the primary antibodies, or 45 min with the conjugated primary antibodies, using an end-over-end mixer at 4°C at ∼20 rpm. After spinning down and washing the synaptoneurosome pellets in PBS, we added the following secondary antibodies: anti-mouse AF647 (dilution, 1:1000; catalog #MA1-10 405, Thermo Fisher Scientific) and anti-rabbit AF647 (dilution, 1:1000; catalog #08–6199, Thermo Fisher Scientific). We incubated the samples for 15 min using an end-over-end mixer at 4°C. All samples were washed with 1 ml of PBS and centrifuged at 15,000 × g at 4°C for 7 min. Finally, samples were resuspended in 100 μl of PBS and filtered through a 40 μm strainer into a FACS tube (catalog # 352235, Falcon) for flow cytometry.
Flow cytometry
The FCS analysis procedure shown here is a modification of our previous FACS procedure used to isolate and study neurons (Liu et al., 2014). Since synaptoneurosomes are 10–20 times smaller than neuronal cell bodies, we optimized the detection of events closer to the lower size limit for flow cytometers (0.2 µm). The initial pilot and characterization experiments (Fig. 1A) used a FACS Aria II Fusion Flow Cytometer (BD Biosciences). Putative synaptoneurosomes were first defined by forward scatter (FSC) and side scatter (SSC) using a light scatter plot, and their sizes were estimated using fluorescent (Alexa Fluor 488) size calibration beads (220, 440, 880, and 1350 nm beads; catalog #NFPPS-52-4K, Spherotech). The synaptoneurosome gate was defined by FSC and SSC properties, while the YFP-specific gate for synaptoneurosomes projecting from the mPFC to NAc was further defined by YFP-positive expression detected in the fluorescein isothiocyanate (FITC) channel and side scatter. For Experiments 2, 4, and 5, we switched to a FACSAria Fusion SORP Flow Cytometer (BD Biosciences) equipped with UV 355 nm, Violet 405 nm, Blue 488 nm, YG 561 nm, and Red 637 nm lasers for all of the following experiments (see Figs. 3, 5, 8, 11). Fluorescent immunolabels were detected in the allophycocyanin (APC), phycoerythrin (PE), or PE-CF594 channels. Samples without antibody or only primary antibody were used to set the thresholds for each marker. Data were analyzed using FCS Express 6 Flow Research Edition for Windows. Gating strategies using a density or contour-type plot were compared to determine the optimal procedure.
Electron microscopy of synaptoneurosomes
Electron microscopy was used to assess the structure of our synaptoneurosome preparations using a modification of our previous procedure (Zhang and Morales, 2019). The unsorted P3 fraction was used to demonstrate sealed presynaptic and postsynaptic compartments in our synaptoneurosomes (Fig. 1D). Another P3 fraction was sorted in the FACS Aria II Fusion Flow Cytometer to examine synaptoneurosome structure in presorted and postsorted samples based on YFP expression from the same preparation (Fig. 2B). Synaptoneurosome samples were centrifuged in a 1.5 ml Eppendorf centrifuge tube at 15,000 × g at 4°C for 7 min. The supernatant was discarded, and we added 1 ml of fixative (2.5% glutaraldehyde, 2% paraformaldehyde, 2 mm CaCl2 in 0.1 m sodium cacodylate buffer) and left the Eppendorf tube at 4°C overnight. We centrifuged the tube at 15,000 × g at 4°C for 7 min and removed the fixative, and rinsed the pellets twice with 1 ml of 0.1 m sodium cacodylate buffer for 5 min. We then postfixed the pellets with 2% osmium tetroxide in 0.1 m sodium cacodylate buffer for 1 h, removed the osmium solution, and rinsed the pellets twice with double-distilled water (ddH2O) for 5 min. We then added 0.5 ml of 3% agarose dissolved in ddH2O at 37°C to the Eppendorf tube and centrifuged the tube for 1 min to push the pellet into the agarose. We left the agarose in the tube on ice for 5–10 min to solidify the agarose, and then removed and cut it into 1 mm pieces on ice and transferred them into glass vials, where they were rinsed five times with cold ddH2O for 5 min each. We labeled the samples with 4% uranyl acetate for 40 min on ice and rinsed them five times with cold ddH2O for 5 min each. We dehydrated the samples with ice-cold graded ethanol as follows: 5 min for each sample in 30%, 50%, 70%, and 90%, 10 min for each sample in 100%, 100%, and 100%, and 2 × 5 min for each sample in ice-cold acetone. The samples were infiltrated first with a 1:1 mixture of acetone and Durcupan ACM epoxy resin for 90 min on a rotator, with pure Durcupan ACM epoxy resin overnight on a rotator, and then with freshly made resin for 90 min on the rotator the next day. We embedded the samples in beam capsules with freshly made resin and polymerized the resin at 60°C for 48 h. Sections of 60 nm were cut from the resin block with an ultramicrotome UC7 (Leica Microsystems) using a diamond knife (Diatome). The sections were collected on formvar-coated single-slot grids and counterstained with Reynold's lead citrate solution. Sections were examined and photographed using a transmission electron microscope (Tecnai G2 12, Thermo Fisher Scientific) equipped with the OneView digital micrograph camera (Gatan).
Immunohistochemical electron microscopy of brain sections
We perfused rats with 1000 U/ml heparin solution followed by fixative solution [4% PFA, 0.15% glutaraldehyde, 15% picric acid solution in 0.1 m phosphate buffer (PB), pH 7.4] and kept the brains in the same fixative solution at 4°C for another 2 h. We placed the brains in 2% PFA fixative solution and postfixed the brains at 4°C overnight. After rinsing in 0.1 m PB, coronal serial sections (50 µm) were cut with a vibratome (catalog #VT1000S, Leica Microsystems). The sections were rinsed and incubated with 1% sodium borohydride to inactivate free aldehyde groups, rinsed, and then incubated in blocking solution. Sections were then incubated with the primary antibodies rabbit anti-DsRed (1:1000; catalog #632496, Takara) and mouse anti-S6 (1:50; catalog #23175, Cell Signaling Technology); or with mouse anti-mCherry (1:1000; catalog #632543, Takara) and rabbit anti-calcineurin (1:100; catalog #ab109412, Abcam), diluted in 1% normal goat serum, 4% BSA, 0.02% saponin in PB at 4°C overnight. Sections were rinsed and incubated overnight at 4°C in the corresponding secondary antibodies, as follows: biotinylated goat-anti-Rb for DsRed detection and Nanogold-Fab′ goat anti-mouse IgG (1:100; catalog #2002, Nanoprobes) for S6 detection; or biotinylated goat-anti-mouse for mCherry detection and Nanogold-Fab′ goat anti-rabbit IgG (1:100; catalog #2004, Nanoprobes) for calcineurin detection. Sections were rinsed in PB, and then sections were incubated in avidin-biotinylated horseradish peroxidase complex in PB for 2 h at room temperature and washed. Next, sections were postfixed with 1.5% glutaraldehyde for 10 min and rinsed in PB and double-distilled water, followed by silver enhancement of the gold particles with the Silver Enhancement Kit (catalog #2012, Nanoprobes) for 7 min at room temperature. Next, peroxidase activity was detected with 0.025%vdiaminobenzidine (DAB) and 0.003% H2O2 in PB for 5–10 min. Sections were rinsed with PB and fixed with 0.5% osmium tetroxide in PB for 25 min, washed in PB followed by ddH2O water, and then contrasted in freshly prepared 1% uranyl acetate for 35 min. Sections were dehydrated through a series of graded alcohols and with propylene oxide. Afterward, they were flat embedded in Durcupan ACM epoxy resin (catalog #14040, Electron Microscopy Sciences). Resin-embedded sections were polymerized at 60°C for 2 d. Sections of 60 nm were cut from the outer surface of the tissue with an ultramicrotome UC7 (Leica Microsystems) using a diamond knife (Diatome). The sections were collected on formvar-coated single-slot grids and counterstained with Reynold's lead citrate. Sections were examined and photographed using a Tecnai G2 12 transmission electron microscope (Thermo Fisher Scientific) equipped with the OneView digital micrograph camera (Gatan).
Quantitative ultrastructural analysis
Serial ultrathin sections were analyzed with the observer blind to the experimental conditions. Synaptic contacts were classified according to their morphology and immunolabel and were photographed at a magnification of 6800–13,000×. The morphologic criteria used for identification and classification of cellular components or type of synapse observed in these thin sections were previously described in the study by Zhang et al. (2015). Pictures were adjusted to match contrast and brightness by using Photoshop (Adobe Systems). This experiment was successfully repeated three times.
Fluorescent histochemistry of injection site and fiber distribution
For histologic visualization of viral expression of YFP, we perfused rats with 4% PFA in PBS, pH 7.4, and kept the brains in the same fixative solution at 4°C for another 2 h. Coronal serial sections (40 µm) were cut with a cryostat (catalog #CM3050S, Leica Microsystems), and floating sections were immunostained for the dopaminergic marker tyrosine hydroxylase using a primary mouse antibody (1:500; catalog #AB318, Millipore) and a secondary antibody Alexa Fluor 594 donkey anti-mouse F(ab')2 (1:200; catalog #715–586-151, Jackson ImmunoResearch). Brain sections containing mPFC or NAc were mounted and coverslipped using Vectashield Vibrance with DAPI (catalog #H-1800, Vector Laboratories). Injection sites and fiber distribution were imaged in a slide scanner (model VS120, Olympus USA). The following three fluorophores were imaged: (1) DAPI as the marker for cytoarchitecture; (2) YFP as the marker of viral vector expression; and (3) AF-594 as the fluorophore to identify TH (tyrosine hydroxylase) fibers. Mosaic images of the mPFC and striatum were acquired at 10× magnification with a resolution of 645.33 nm/pixel.
Experiments
Experiment 1: characterization and validation of FCS procedure.
For fluorescent microscopy, we used three rats for histologic visualization of mPFC projections to striatum (Fig. 1B). For electron microscopy of synaptoneurosomes (Fig. 1D), we used a total of four rats. Samples were processed as described above for synaptoneurosome preparation, but NAc tissue used for electron microscopy was pooled from two rats each. We used three rats for flow cytometry detection of striatal YFP-tagged striatal synaptoneurosomes. As described above for Viral injections, AAV1-CaMKIIa::hChR2-eYFP was unilaterally injected into mPFC (Fig. 1C). After YFP expression and translocation to the striatum, we collected and processed striatal tissue, containing both dorsomedial striatum (DMS) and NAc, separately from the ipsilateral and contralateral hemispheres.
Experiment 2: FCS analysis of time course following cocaine + novelty stimulus.
We used a total of 56 rats (n = 7–9/group) in a one-factor between-subjects design with Time (0, 5, 10, 30, and 60 min) as the main factor. We injected AAV1-CaMKIIa::hChR2-eYFP into infralimbic and prelimbic regions as described in viral injections. On test day, 6–8 weeks later, rats were injected intraperitoneally with 20 mg/kg cocaine (Cocaine-HCl, NIDA Pharmacy) and were immediately placed in a novel context with bedding and toys. Brains were extracted 0, 5, 10, 30, and 60 min after cocaine injections. After hand dissecting the YFP-positive NAc region, we prepared synaptoneurosomes as described above. Levels of the candidate synaptic activity markers Arc, CaMKII and phospho-CaMKII, S6 and phospho-S6, and calcineurin and phospho-calcineurin were assessed in-flight during FCS within the synaptoneurosome gate and within the YFP-positive gate.
Experiment 3: immunohistochemical electron microscopy validation of synaptic activity markers.
We used a total of seven rats (n = 3/group and 4/group) in a one-factor between-subjects design with Time (0, 10 min) as the main factor. We injected AAV9-CAG::ChR2-mCherry into infralimbic and prelimbic regions as described in the subsection Viral injections. On test day, 6–8 weeks later, rats were injected with 20 mg/kg cocaine and immediately placed in a novel context with bedding and toys. Brains were extracted 0 and 10 min after cocaine injections. Levels of the candidate synaptic activity markers S6 and calcineurin in mCherry-labeled synapses were assessed as described in the subsection Immunohistochemical electron microscopy of brain sections.
Experiment 4: assessing neural activity dependence using chloral hydrate inactivation.
We used a total of 26 rats (6–7 rats/group) in a full factorial 2 × 2 between-subjects design with the factors Anesthetic (chloral hydrate, no chloral hydrate) and Drug (cocaine, no cocaine). We injected AAV1-CaMKIIa::hChR2-eYFP into infralimbic and prelimbic regions as described in the subsection Viral injections. On test day, 6–8 weeks later, we injected rats with the anesthetic chloral hydrate (400 mg/kg, i.p.) or no anesthetic, and then 10 min later we injected 20 mg/kg cocaine and immediately placed the rats in a novel context with bedding and toys. Brains were extracted 0 and 10 min after cocaine injections. After hand dissecting the YFP-positive NAc region, we prepared synaptoneurosomes as described above. Levels of the candidate synaptic activity markers S6 and calcineurin were assessed in-flight during FCS for the synaptoneurosome gate (which includes YFP-negative and YFP-positive gates) and for the YFP-positive gate only.
Experiment 5: distinguishing the separate effects of injections, cocaine, and context.
We used a total of 30 rats (6 rats/group) in a full factorial 2 × 2 between-subjects design with the factors Drug (cocaine, saline) and Context (novelty, home cage). We injected AAV1-CaMKIIa::hChR2-eYFP into infralimbic and prelimbic regions as described in the subsection Viral injections. On test day, 6–8 weeks later, we injected rats with either saline or 20 mg/kg cocaine in their home cage or in a novel context. All brains were extracted 10 min after injections. We also included a group of naive rats for baseline comparison. After hand dissecting the YFP-positive NAc region, we prepared synaptoneurosomes as described above. Levels of the candidate synaptic activity markers S6 and calcineurin were assessed in-flight during FCS for the YFP-positive gate only.
Statistical analysis
We used GraphPad Prism 9 for Windows for all statistical analyses. Flow cytometry data for Experiment 2 (Cocaine time course) were analyzed using one-way ANOVA followed by Dunnett's multiple-comparisons post hoc tests (Tables 1, 2). Data for Experiment 3 (see subsection Immunohistochemical electron microscopy validation) were analyzed using t tests (see Table 6). Flow cytometry data for Experiment 4 were analyzed using two-way ANOVA followed by Holm–Sidak multiple-comparisons post hoc tests (see Tables 7, 8). Flow cytometry data for Experiment 5 were analyzed using both one-way ANOVA followed by Dunnett's multiple-comparisons post hoc tests and two-way ANOVA followed by Holm–Sidak multiple-comparisons post hoc tests (see Tables 9, 10). Results are expressed as the mean ± SEM, and p values < 0.05 were considered statistically significant.
Statistical analysis for FCS data from YFP-positive gate (PFC to NAc synaptoneurosomes)
Statistical analysis for FCS data from synaptoneurosome gates (all synaptoneurosomes)
Results
Experiment 1: characterization and validation of FCS procedure
Experimental timeline for Experiment 1 is shown in Figure 1A. YFP was strongly expressed in mPFC cell bodies and transported to terminals in DMS and NAc (Fig. 1B, left). Following differential centrifugation, P2 pellets from striatum were centrifuged through a Percoll gradient, and synaptoneurosomes (including YFP-positive and YFP-negative terminals) were collected from a green fluorescent band (SN fraction) enriched for synaptoneurosomes (Fig. 1B, right). Electron microscopy of the SN fraction from NAc confirmed selective enrichment of ∼500–1000 nm synaptoneurosomes containing both sealed presynaptic and postsynaptic terminals (Fig. 1D).
Characterization and validation of FCS procedure. A, Experimental timeline for Experiment 1: characterization and validation of FCS procedure. Rats received unilateral injections of AAV1-CaMKIIa::hChR2-eYFP into prelimbic (PL) and infralimbic (IL) regions of the mPFC to tag presynaptic terminals from mPFC to striatum. Brains were processed for either synaptoneurosome or histology preparations followed by flow cytometry and electron microscopy. B, Left, Representative coronal brain sections shown YFP-labeled fluorescent cells (green) in the mPFC sending fibers to DMS and NAc. Middle, Tube with fluorescent P3 fraction enriched with synaptoneurosomes, visible as a green band under blue light between 10% and 15% Percoll gradients. C, Density plot from FCS indicating YFP-labeled presynaptic terminals from a rat that received a unilateral injection of AAV1-CaMKIIa::hChR2-eYFP. Strong YFP-positive expression (FITC channel) was found in striatum of the ipsilateral hemisphere (ipsi), but not in the contralateral hemisphere (contra). YFP-expressing events were not detected in striatum of uninjected brains (no virus). D, Representative electron microscopy image of a synaptoneurosome, showing intact sealed presynaptic and postsynaptic elements. Green arrows indicate some of the hundreds of round-shaped vesicles within the presynaptic terminal. White arrows indicate dense black structures called post-synaptic densities (PSD) within the postsynaptic compartment. Bar in D is 200 nm. See Figure 2 for more information.
We characterized particles in the SN fraction using flow cytometry. Alexa Fluor 488-sized beads of 220, 440, 880, and 1350 nm were used in forward/side scatter light plots to determine the approximate size of events in the SN fraction (Fig. 2A). Although the different material and shape of the beads make size comparisons less accurate, nearly all events were between the 880 and 1350 nm beads, which overlapped with the size range ∼500–1000 nm that we observed for synaptoneurosomes using electron microscopy (Figs. 1D, 2A). We tentatively defined a gate around this cluster of larger events (between the 880 and 1350 nm beads) with >102 side scatter values as the synaptoneurosome gate. This synaptoneurosome gate was confirmed when we found nearly all YFP-labeled events within this gate. We speculate that the cluster of smaller events (lower forward scatter values) detected by the flow cytometer with >102 side scatter values corresponds to the debris seen in electron micrographs. To confirm that YFP labeling in flow cytometry analysis was specific for the mPFC to striatal terminals, we unilaterally injected YFP-expressing adeno-associated virus (AAV; or no injection for control) into one hemisphere and assessed YFP-labeled synaptoneurosomes from striatum, which included DMS and NAc (Fig. 1C). Most events on the ipsilateral side were YFP-positive, while a much smaller number were YFP-positive on the contralateral side, likely because of transport via crossing fibers. No events were above the YFP-positive threshold in the No injection control samples. Thus YFP-labeled events in flow cytometry indicate mPFC-striatal synaptoneurosomes. Electron microscopy of the synaptoneurosome fraction after FASS confirmed that synaptoneurosomes maintained both sealed presynaptic and postsynaptic sacs during flow cytometry (Fig. 2B).
Synaptoneurosome validation. Size of synaptoneurosomes. A, Logarithmic density plot of forward (x-axis, cell size) and side (y-axis, granularity) scatter properties of all events in the synaptoneurosome fraction. The ranges of four different AF488-sized beads are displayed, as follows: 220, 440, 880, and 1350 nm. The synaptoneurosome gate was determined based on the size and YFP expression of synaptoneurosomes. B, Electron microscopy images of synaptoneurosomes before sorting (top) and after sorting (bottom) in a FACS Aria II Fusion Flow Cytometer. White arrows point to PSDs, which is characteristic of asymmetric/excitatory synapses.
Experiment 2a: flow cytometry characterization of presynaptic and postsynaptic markers in synaptoneurosomes
The experimental timeline for Experiment 2 is shown in Figure 3A. We assessed expression of the following three proteins: the presynaptic protein synaptophysin and the postsynaptic protein PSD-95 using specific antibodies and detected using Alexa Fluor 647 in APC channel and PE-Cy7 in PE-CF594 channel, respectively, and YFP expression in mPFC-NAc terminals from the SN fraction at each time point. The synaptoneurosome gate (defined above) contains ∼95% of all events in the forward/side light scatter plot (Fig. 3B, left). A majority of events (60–80%) in the synaptoneurosome gate were double-labeled for both synaptophysin and PSD-95 in samples taken from all five time points (Fig. 3C, left, bottom). When assessing YFP expression from the synaptoneurosome gate, two sets of events were identified as YFP-positive or YFP-negative based on fluorescence in the FITC channel (Fig. 3B, right). YFP-positive events were 44 ± 3% of all events. Nearly all (97–100%) of events in this YFP-positive gate were double-labeled for both synaptophysin and PSD-95 (Figs. 3C, right, bottom, 4A). The contribution of aggregates to the dual labeling was low at 4–20% of all events (Fig. 4B). Overall, nearly all events in the YFP-positive gate and most events in the synaptoneurosome gate contained both presynaptic and postsynaptic sacs.
FCS for assessing presynaptic and postsynaptic markers in synaptoneurosomes. A, Experimental timeline for Experiment 2: time course experiment. Rats received bilateral injections of AAV1-CaMKIIa::hChR2-eYFP into prelimbic (PL) and infralimbic (IL) regions of the mPFC to tag presynaptic terminals from mPFC to NAc. Six to 8 weeks later, rats were injected with 20 mg/kg cocaine and immediately placed into a novel environment. Synaptoneurosomes were isolated from NAc 0, 5, 10, 30, or 60 min later. B, Gating strategy for flow cytometry of NAc synaptoneurosomes. Logarithmic density plots showing the synaptoneurosome gate based on forward (x-axis, cell size) and side (y-axis, granularity) scatter properties (left); the YFP-positive gate and YFP-negative gate based on YFP fluorescence (x-axis), detected in FITC channel, relative to side scatter (y-axis; right). C, Logarithmic density plots from the same synaptoneurosome sample showing events double labeled for synaptophysin and PSD-95 in the top right quadrants for events taken from the synaptoneurosome gate (left) or from the YFP-positive gate (right) shown in B. Bottom, Graph showing the percentage of synaptophysin and PSD-95 double-labeled events from the synaptoneurosome and YFP-positive gates 0–60 min following cocaine injections. Data came from a total of 50,000 YFP-positive events for each sample (n = 5–7 samples/time point). See Figure 6 for more information.
Validation of synaptophysin and PSD-95. A, Representative logarithmic density plots of an aggregate control experiment using the same sample but split into the following 4 conditions: (1) synaptophysin (top left) or PSD-95 (top right) antibodies were incubated separately; (2) synaptophysin and PSD-95 antibodies were each incubated with one of the split samples and then the two labeled split samples were combined (bottom left); and (3) synaptophysin and PSD-95 antibodies were incubated together (bottom right). B, Immunoreactivity of each condition was assessed and plotted as bar graphs. LL, Lower left quadrant; UL, upper left quadrant; LR, lower right quadrant; UR, upper right quadrant.
Experiment 2b: time course for candidate markers of activated mPFC–NAc synapses
We used flow cytometry to assess altered protein and phosphorylation levels of candidate synaptic activity markers obtained from rats 0–60 min following cocaine injections in a novel environment (Fig. 5) using the same NAc samples previously used for PSD-95 and synaptophysin in Figure 3A. We selected enzymes thought to be activated during neural activity in presynaptic and postsynaptic compartments, including calcineurin (Goto et al., 1986; Gomez et al., 2002; Dell'Acqua et al., 2006) and CaMKII (Shen and Meyer, 1999; Dosemeci et al., 2001, 2002), as well as proteins known to translocate to dendritic spines during neural activity, including Arc (Steward and Schuman, 2001; Kim et al., 2012; Okuno et al., 2012) and S6 (Steward and Schuman, 2001; Kim et al., 2005; Rangaraju et al., 2017). Synaptoneurosomes were immunolabeled for Arc, CaMKII, phospho-CaMKII, S6, phospho-S6 protein, calcineurin, and phospho-calcineurin, and were detected using Alexa Fluor 647 in APC channel or PE-Cy7 in the PE-Cy7 channel. Levels are represented as the fold change relative to 0 min in either the mPFC-NAc synaptoneurosomes (YFP-positive gate detected in the FITC channel) or all synaptoneurosomes in the synaptoneurosome gate. We then compared levels at each time point (5–60 min) to 0 min using Dunnett's post hoc analyses after significant one-way ANOVAs (Tables 1, 2). In the YFP-positive gate, Arc and CaMKII protein levels were not significantly altered. S6 was increased at all time points (5–60 min), while calcineurin levels were transiently increased at 5–10 min, but not 30–60 min (Fig. 5A, left column). Phosphorylated forms of CaMKII, S6, and calcineurin were not altered. In the synaptoneurosome gate, protein or phosphorylation levels were not significantly altered, although S6 levels trended toward an increase (p < 0.1) at 30 min, while phospho-calcineurin levels (p < 0.01) decreased at 10 min following the cocaine+novelty stimulus (Fig. 5A, right column). Overall, the responses in the synaptoneurosome gate were blunted compared with the responses in the YFP-positive gate specific for the mPFC-NAc terminals. Levels for each protein are also represented as the percentage of YFP-positive or of all synaptoneurosomes in Figure 6. Baseline levels were similar for all proteins in the YFP-positive synaptoneurosome gate (10–15%; Fig. 6A) and for all proteins in the synaptoneurosome gate (7–9%; Fig. 6B). Statistical analyses of data in Figure 6 are provided in Tables 3, 4, and 5.
Time courses for candidate synaptic activity markers in NAc synaptoneurosomes. Graphs show the fold change expression at each time point, 0, 5, 10, 30, and 60 min after cocaine+novelty, relative to levels at the 0 min time point. Candidate synaptic markers were Arc, CaMKII, S6, calcineurin, and their corresponding activated phosphorylated forms (pCaMKII, pS6, pCalcineurin). Total protein levels for each marker are indicated by closed circles, while phosphorylated protein levels are indicated by open circles. A, Fold change values of flow cytometry data from only the YFP-tagged mPFC–NAc synaptoneurosomes (YFP-positive gate). B, Fold change values of flow cytometry data from all brain inputs to NAc synaptoneurosomes (synaptoneurosome gate). Asterisks indicate significant differences (p < 0.05) relative to 0 min; n = 6–8 samples for each time point for each protein. See Figure 9 and Table 1–Table 5 for more information.
Statistical analysis for FCS data from YFP-positive gate (PFC–NAc synaptoneurosomes) using the density plot method
Statistical analysis for FCS data from synaptoneurosome gate (all synaptoneurosomes) using the density plot method
Statistical analysis for FCS data using density plot threshold method
Percentage raw data of time courses for candidate synaptic activity markers in NAc synaptoneurosomes. Graphs represent the percentage of YFP-positive or all synaptoneurosomes that were labeled for each protein at each time point, 0, 5, 10, 30, and 60 min after cocaine+novelty. Candidate synaptic markers were Arc, CaMKII α, S6, calcineurin, and their corresponding activated phosphorylated forms (pCaMKII, pS6, pCalcineurin). Total protein levels for each marker are indicated by closed circles, while phosphorylated protein levels are indicated by open circles. A, Percentage of YFP-tagged synaptoneurosomes from only mPFC inputs to NAc (YFP-positive gate) that were labeled for each protein. B, Percentage of all synaptoneurosomes from all brain inputs to NAc synaptoneurosomes (synaptoneurosome gate) that were labeled for each protein. Asterisks indicate significant differences (p < 0.05); n = 6–8 samples for each time point for each protein.
Experiment 3: electron microscopy of S6 and calcineurin proteins in mPFC–NAc synapses
We used electron microscopy to confirm increased expression of S6 and calcineurin in mPFC–NAc synapses 10 min following cocaine injections in a novel environment (Fig. 7). The 10 min time point was chosen because the flow cytometry data in Figure 5 indicate that both S6 and calcineurin were increased in YFP-positive mPFC-NAc terminals at this time. Representative electron micrographs for S6-positive and S6-negative synapses are shown in Figure 7A and indicate that S6 was found only in the postsynaptic compartment. We counted the number of S6-positive synapses that contained at least one immunogold-labeled particle versus S6-negative synapses, only in synapses that contained both postsynaptic densities and DAB-labeled mCherry (instead of YFP) transported from the mPFC to NAc. Quantitative analysis shown in the graph at the right indicates that the number of S6-positive synapses increased from 0 to 10 min, similar to our S6 flow cytometry data in Figure 5. Representative electron micrographs for calcineurin-positive and calcineurin-negative synapses are shown in Figure 7B and indicate that calcineurin was found in both the presynaptic and/or postsynaptic compartments. We counted the number of calcineurin-negative versus calcineurin-positive synapses separately for presynaptic versus postsynaptic calcineurin expression. Negative or positive synapses were analyzed independently using an unpaired t test (Table 6, details). Quantitative analysis shown in the graph indicates that the number of calcineurin-positive postsynaptic terminals did not increase significantly from 0 to 10 min, while the number of calcineurin-positive presynaptic terminals increased from 0 min (10.33 ± 3.33) to 10 min (37.00 ± 6.5).
Electron microscopy of immunolabeled S6 and calcineurin proteins in mPFC–NAc synapses. In Experiment 3, rats received bilateral injections of AAV1-CAG::hChR2-mCherry into mPFC. mPFC terminals in NAc core were immunolabeled with anti-mCherry or anti-DsRed antibodies and revealed with DAB. S6 and calcineurin were immunolabeled with S6 and calcineurin primary antibodies and Nanogold-conjugated secondary antibodies. A, B, Left panels, Electron microscopy images of mCherry-positive synapses with presynaptic (red contour) and postsynaptic (yellow contour) specializations delineated by dashed lines. The dark DAB-induced staining in the presynaptic terminal (pre) indicates the presence of mCherry. Yellow arrowheads point to high-density Nanogold particles that are associated with S6 (A) or calcineurin (B) immunolabeling. The white arrow points to the PSD in the postsynaptic ending (post). Left images show examples of S6-positive (A) or calcineurin-positive (B) synapses. Right images show examples of S6-negative (A) or calcineurin-negative (B) synapses. Bars in A and B are 200 nm. Bar graphs indicate the number of postsynaptic endings that are immunoreactive (IR) for S6 (A) and the number of postsynaptic or presynaptic endings immunoreactive for calcineurin (B). Asterisks indicate significant differences (p < 0.05) relative to 0 min; n = 3 rats/time point. The number of synapses counted per rat: 70–139 synapses for rats in the 0 min group; 62–160 synapses for rats in the 10 min group. See Table 6 for more information.
Statistical analysis for immunohistochemical data from NAc core brain sections
Experiment 4: effect of chloral hydrate on cocaine+novelty-induced S6 and calcineurin levels
Experimental timeline for Experiment 4 is shown in Figure 8A. To confirm that S6 and calcineurin levels are markers of synaptic activity, we injected the anesthetic chloral hydrate 10 min before the cocaine+novelty stimulus to reduce synaptic activity. We compared levels using Holm–Sidak post hoc comparisons after significant two-way ANOVAs with cocaine and chloral hydrate as factors (Tables 7, 8, details). In unanesthetized rats, cocaine increased S6 (Fig. 8B) and calcineurin (Fig. 8C) levels, similar to our results in Experiment 2b for the YFP-positive mPFC–NAc synaptoneurosomes (Figs. 5, 6). However, chloral hydrate anesthesia blocked the cocaine+novelty-induced increase of S6 and calcineurin levels (Fig. 8B,C). Similar effect of chloral hydrate was observed in the synaptoneurosome gate containing all (YFP-positive and YFP-negative) synaptoneurosomes for S6 (Fig. 9A) and calcineurin (Fig. 9A) levels, without altering expression of the other candidate proteins (Fig. 10). It should be noted that S6-positive and calcineurin-positive activated synaptoneurosomes make up ∼8–12% of all synaptoneurosomes (Fig. 9). Overall, the results for all synaptoneurosomes (Fig. 9) were similar to those for YFP-positive synaptoneurosomes (Fig. 8) but blunted in comparison.
Chloral hydrate blocks cocaine+novelty-induced expression of S6 and calcineurin synaptic activity markers in synaptoneurosomes. A, Experimental timeline for Experiment 4: chloral hydrate inactivation experiment. Rats received bilateral injections of AAV1-CaMKIIa::hChR2-eYFP into mPFC to tag presynaptic terminals from mPFC to NAc. On test day, 6–8 weeks later, rats were injected with 400 mg/kg chloral hydrate or no injection and kept in their home cages for 10 min. Rats were then injected with 20 mg/kg cocaine and immediately placed in a novel context for 10 min. Synaptoneurosomes were isolated from NAc 10 min after cocaine injections or removed from their home cages without an injection (0 min). B, C, Bar graphs indicate S6-positive events (B) or calcineurin-positive events (C) as a percentage of all YFP-tagged mPFC to NAc synaptoneurosomes (YFP-positive gate). Asterisks indicate significant differences (p < 0.05); n = 5–7 samples for each group. See Figures 9 and 10 and Tables 7 and 8 for more information.
Statistical analysis for FCS data from YFP-positive gate (PFC–NAc synaptoneurosomes) using contour plot threshold method
Statistical analysis for FCS data from synaptoneurosomes gate (all synaptoneurosomes) using the contour plot threshold method
Chloral hydrate effect on S6 and calcineurin expression in all NAc synaptoneurosomes. A, B, Bar graphs indicate S6-positive events (A) or calcineurin-positive events (B) as a percentage of all NAc synaptoneurosomes (containing YFP-positive and YFP-negative populations). Asterisks indicate significant differences (p < 0.05); n = 5–7 samples for each group.
Chloral hydrate does not affect protein expression of Arc, CaMKII, pCaMKII, pS6, and pCalcineurin in synaptoneurosomes. A, B, Bar graphs indicate positive events as a percentage of all YFP-tagged mPFC–NAc synaptoneurosomes (YFP-positive gate; A) or as a percentage of all NAc synaptoneurosomes (synaptoneurosome gate; B). Not significant differences; n = 5–7 samples for each group.
Experiment 5: effect of injections, cocaine and context on S6 and calcineurin levels
Experimental timeline for Experiment 5 is shown in Figure 11A. To evaluate the separate contributions of injection procedure, cocaine, and novel context exposure, we injected four groups of rats with saline or cocaine in their home cage or a novel context, along with an additional group of naive rats, and assessed S6 and calcineurin levels 10 min later. For S6 and calcineurin, one-way ANOVAs followed by Dunnett's multiple-comparisons tests with the naive group (Table 9, details) indicated no effect of the handling/injection procedure in the saline plus home cage group, but increased levels in the cocaine plus novel context group (Fig. 11B,C), which is similar to that seen for the previous comparison of the 0 min group (equivalent to the naive group here) versus the 10 min group in Experiment 2b. Relative to the naive group, the saline+novelty context increased S6 levels while cocaine plus home cage had no significant effect. Neither novel context nor cocaine alone increased calcineurin levels. For S6 and calcineurin, two-way ANOVAs (not including the naive group) indicated significant main effects of cocaine and novel context, but no interaction between them (Table 10, details). Relative to saline injections in the home cage, Holm–Sidak post hoc comparisons indicated that novelty exposure alone, but not cocaine alone, increased S6 levels, whereas neither novelty exposure nor cocaine alone increased calcineurin levels.
Effect of novel context versus cocaine on S6 and calcineurin synaptic expression. A, Experimental timeline for Experiment 5: novel context versus cocaine experiment. Rats received bilateral injections of AAV1-CaMKIIa::hChR2-eYFP into mPFC to tag presynaptic terminals from mPFC to NAc. On test day, 6–8 weeks later, rats were injected with either saline or 20 mg/kg cocaine in two different contexts: home cage or novel context. Synaptoneurosomes were isolated from NAc 10 min after injections. Naive rats were taken directly from their home cages without an injection. B, C, Bar graphs indicate S6-positive events (B) or calcineurin-positive events (C) as fold change values relative to those from naive rats. Asterisks indicate significant differences (p < 0.05) relative to naive group; n = 6 samples for each group. Data came from a total of 100,000 YFP-positive events for each sample. See Tables 9 and 10 for more information.
Statistical analysis for FCS data from YFP-positive gate (PFC–NAc synaptoneurosomes) using contour plot threshold method
Statistical analysis for FCS data from YFP-positive gate (PFC–NAc synaptoneurosomes) using the contour plot threshold method
Discussion
We developed an optimized FCS protocol to identify protein alterations in activated mPFC–NAc synapses with single-synapse resolution. FCS and electron microscopy confirmed that synaptoneurosomes contained sealed presynaptic and postsynaptic compartments. Anterograde labeling with YFP allowed FCS analysis of synapses coming specifically from mPFC axons onto dendritic spines of NAc neurons. Using FCS analysis, we found increased numbers of synapses expressing S6 and calcineurin at 5–60 and 5–10 min, respectively, following activation with cocaine in a novel context (cocaine+novelty), and electron microscopy results confirmed these findings. Pretreatment with the anesthetic chloral hydrate, known to inhibit glutamate-mediated neural activity without altering cocaine-induced dopamine in striatum, blocked cocaine+novelty-induced increases of S6 and calcineurin, indicating a dependence on neural activity rather than cocaine-induced “metabolic” mechanisms. This was further supported when novel context exposure alone increased S6 levels (but not calcineurin levels). We propose that S6 and calcineurin can be used with FCS to study synaptic protein alterations induced preferentially in synapses activated during behavior and memory reactivation, similar to how Fos has been used with FACS to study engram alterations in activated neuronal ensemble cell bodies (Guez-Barber et al., 2011; Fanous et al., 2013; Liu et al., 2014; Cruz et al., 2015; Li et al., 2015; Rubio et al., 2015).
Neural activity during behavior or memory recall is thought to activate information-specific patterns of synapses interspersed among a much larger percentage of nonactivated synapses and produce long-lasting memory-encoding alterations called engrams selectively in these activated synapses (Semon, 1904; Hebb, 1949; Sutherland and McNaughton, 2000; Tonegawa et al., 2018; Whitaker and Hope, 2018; Rao-Ruiz et al., 2021). To identify engram alterations in this small percentage of selectively activated synapses, one needs to identify these activated synapses and then use a quantitative method for analyzing protein alterations with single-synapse resolution to distinguish between the intermingling activated and nonactivated synapses. Immunohistochemical methods detect the presence or absence of a protein with single-synapse resolution, but they are generally poor for detecting different levels of these proteins. In contrast, Western blots and proteomic methods have been used to assess different levels of proteins in homogenates of synaptosomes and postsynaptic densities (Carlin et al., 1980; Li et al., 2005; Schrimpf et al., 2005; Villasana et al., 2006), but they require large amounts of tissue and do not distinguish between different types of synapses intermingled in tissue.
Flow cytometry approaches such as our FCS procedure use small amount of tissue and can distinguish and analyze different types of synapses in tissue with single-synapse resolution. FASS has previously been used to isolate synaptosomes before postsorting quantitation of molecular alterations in synaptosomes (Wolf et al., 1989; Wolf and Kapatos, 1989a, b; Biesemann et al., 2014; Paget-Blanc et al., 2022). Later, levels of fluorescently labeled presynaptic and postsynaptic proteins in synaptosomes were quantitatively assessed in-line as they passed through the flow cell (Gylys et al., 2004a, b; Hafner et al., 2019; Castillo-Ocampo et al., 2021), similar to that in our analysis procedure. However, in addition to not distinguishing activated versus nonactivated synapses, these previous studies analyzed synaptosomes that do not have the sealed postsynaptic sacs containing cytosolic proteins (Whittaker, 1988) that we obtained using synaptoneurosomes in our FCS procedure (Hollingsworth et al., 1985). These postsynaptic cytosolic proteins are thought to be important for memory consolidation and engram formation, including ribosomes for local translation, endosomal vesicles for glutamatergic receptor recycling, and other signaling molecules (Quinlan et al., 1999; Schuman, 1999; Rosenberg et al., 2014; Benito et al., 2018; Hafner et al., 2019). Indeed, one of the most significant activity-dependent alterations in our study was increased S6 (and likely the rest of the ribosome) in the cytosol of dendritic spines, which would not have been detected in the previous FASS procedures that used synaptosomes missing the sealed postsynaptic sacs. Freezing the synaptoneurosome samples in our procedure has the added advantage of allowing us to collect samples for FCS later, similar to that for FACS of cells from frozen brain tissue (Rubio et al., 2016). For these reasons, we estimate up to 40 synaptic proteins from a single rat NAc tissue punch can be analyzed using our protocol. This is particularly important for molecular and cellular studies using very limited amounts of tissue obtained from complex behavioral experiments that require many weeks to months of training (Liu et al., 2014; Li et al., 2015; Rubio et al., 2015).
Acute behavioral activation in a novel context with or without cocaine increased the percentage of S6-positive synaptoneurosomes from 5 min to at least 60 min after activation. Increased S6 levels in postsynaptic terminals correlates with previous observations of ribosome translocation to the head of dendritic spines (Steward and Schuman, 2001; Kim et al., 2005; Rangaraju et al., 2017). We assume that increased expression of postsynaptic proteins, such as S6, in our FCS and electron microscopy assays are because of translocation of the proteins from below the spine or the lower part of the spine to the spine head. Functionally, such translocated ribosomes, which include S6, are thought to be part of the preinitiation translation complex leading to synthesis of new proteins in active synapses (Roux et al., 2007). Many mRNAs such as CaMKII α subunit (Ouyang et al., 1999), the immediate early gene Arc/Arg 3.1 (Steward et al., 1998), and BDNF (Tongiorgi et al., 1997) are also transported to dendritic spines in an activity-dependent manner where they can be translated by S6-containing ribosomes in the activated synapses.
Acute behavioral activation in a novel context increased the percentage of calcineurin-positive synaptoneurosomes from 5 to 10 min after activation. While calcineurin is present in both presynaptic and postsynaptic endings (Goto et al., 1986), our electron microscopy analysis indicated that calcineurin was increased significantly only in the presynaptic ending. Unaltered calcineurin levels in the postsynaptic endings reduce the effectiveness of calcineurin as an overall marker of activated synapses, at least in NAc neurons, but it remains a useful marker of activated presynaptic terminals. Functionally, calcineurin is the cytosolic catalytic subunit of calcium-activated phosphatase 2B shown to play many different roles in signal transduction (Screaton et al., 2004). The large increase of presynaptic calcineurin-positive terminals in our study is likely because of translocation from some unspecified location in the nearby axon to the presynaptic ending in activated synapses where it can affect signal transduction selectively within activated presynaptic terminals.
To confirm that increased ribosomal S6-positive and calcineurin-positive synapses are because of increased synaptic activity, we used an approach similar to that used in our previous study to demonstrate that Fos is a marker of neural activity and not because of an increase in metabolic activity (Kreuter et al., 2004). In this study, the anesthetic chloral hydrate blocked Fos expression following cocaine injections in a novel context. Chloral hydrate also decreased glutamate-mediated neural activity without affecting cocaine-induced dopamine in the striatum, and the addition of the glutamate receptor agonists AMPA and NMDA partially recovered cocaine+novelty-induced Fos expression in the presence of chloral hydrate. The key point was that cocaine-induced dopamine (and related metabolic signaling) by itself produced little to no Fos expression, while glutamate-mediated novel context-induced neural activity increased Fos expression. Based on similar results for chloral hydrate blockade of S6 and calcineurin increases in the current study, we hypothesize that cocaine+novelty increased the percentage of S6-containing and calcineurin-containing synaptoneurosomes because of glutamate-mediated excitation of selectively activated mPFC–NAc synapses and not because of some metabolic consequence of cocaine-induced dopamine. Furthermore, novel context exposure without cocaine (saline+novelty) in Experiment 5 was sufficient to increase S6 (but not calcineurin) levels in synaptoneurosomes from the mPFC to the NAc pathway.
Surprisingly, we did not observe altered Arc or CaMKII levels in our FCS assay. Arc/Arg3.1 gene is an immediate early gene that is rapidly upregulated after strong synaptic activity. Newly transcribed mRNA (including Arc mRNA) in specialized ribonucleoprotein complexes can be transported from nuclei and stored at the base of dendritic spines (Steward and Schuman, 2001; Moga et al., 2004; Kim et al., 2005; Chowdhury et al., 2006) and then later translocated and translated in the dendritic spine depending on activity state (Kim et al., 2012; Okuno et al., 2012). Regarding CaMKII, activation of NMDA glutamate receptors can activate CaMKII through phosphorylation of Thr286 and translocation to the postsynaptic density (Shen and Meyer, 1999; Dosemeci et al., 2001, 2002). The lack of changes in Arc and CaMKII proteins in our FCS analysis may be due to these proteins translocating locally within the spine, which would not be detectable by our FCS procedure.
Methodological considerations: It is important to note that S6 and calcineurin levels are nonzero in our control groups. Basal levels of S6-positive and calcineurin-positive synaptoneurosomes were 10–17% of YFP-labeled mPFC synaptoneurosomes and 7–9% of all synaptoneurosomes, and likely due to normal ongoing brain activity before and during our manipulations. Cocaine+novelty-induced neural activity significantly increased these numbers approximately twofold in YFP-labeled mPFC synaptoneurosomes without significant increases in the All synaptoneurosomes pool (Fig. 6). Therefore, while we cannot conclude that every S6-positive and calcineurin-positive synaptoneurosome was recently activated, there was an enrichment of activated synapses in this pool of S6-positive and calcineurin-positive synaptoneurosomes. Similar twofold increases over significant basal levels of Fos have been observed in mPFC and used to identify activated neuronal ensembles in this brain region (Cruz et al., 2013; DeNardo et al., 2019; Warren et al., 2019). We then used these increased Fos levels with FACS to isolate and enrich for activated neuronal ensemble cell bodies (Guez-Barber et al., 2011; Fanous et al., 2013; Cruz et al., 2013; Rubio et al., 2015; Whitaker and Hope, 2018). Using this method, we and others identified unique molecular alterations induced preferentially in activated ensembles that we proposed as candidate engrams in the formation and maintenance of memories. We propose that it is similarly possible to use FCS to isolate and enrich for activated synapses (increased S6-positive and calcineurin-positive synaptoneurosomes) to identify candidate synaptic engrams.
In summary, our optimized FCS protocol can be used to study protein alterations in synapses of specifically labeled circuits using commercial antibodies. Furthermore, labeling for S6 and calcineurin can be used with FCS to enrich for activated synapses and identify other molecular alterations regulated preferentially in these activated synapses, without the need for genetic manipulations (Gobbo et al., 2017). Of course, future studies are required to confirm that increases of S6 and calcineurin are also induced in synapses coming from different afferents in different brain areas following different experiences. One could also inject AAVs encoding two or more fluorophores into different inputs to the target region using the same AAV tagging strategy used here and in the study by Castillo-Ocampo et al. (2021) to simultaneously analyze different sets of molecular alterations in activated synapses coming from different afferent inputs. We propose that our FCS approach with S6 and calcineurin labels can be used in future studies to identify molecular alterations (and possibly engrams) in synaptic ensembles that were selectively activated during learning or memory recall.
Footnotes
This research was supported by the Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health. We thank Karen Perez de Arce (Broad Institute, Cambridge, MA) for the gift of phospho-calcineurin antibody and scientific input regarding calcineurin. We also thank Rong Ye and Kevin Yu of the Confocal and Electron Microscopy Core, National Institute on Drug Abuse Intramural Research Program, for electron microscopic images used in this study.
The authors declare no competing financial interests.
- Correspondence should be addressed to Bruce T. Hope at bhope{at}intra.nida.nih.gov
