In order to obtain cell-class specific synaptic proteomes, we generated three lines of mice expressing a tdTomato-synaptophysin fusion protein in PV, SST or VIP interneurons (Fig 1A, B). Following perfusion, fixation, cortical dissection, and vibratome sectioning (60 μm) we performed antibody-based proximity labeling on 6 brains from each mouse line (N=18 total, 6 per cell-class) (Fig 1C). Antibody-based proximity labeling by biotinylation was achieved by sequentially adding a primary antibody directed against the cell-type specific tdTomato:Synaptophysin antigen, an HRP-conjugated secondary antibody, followed by incubation with a lab-synthesized biotin-tyramide and hydrogen peroxide to catalyze biotinylation of tyrosine residues proximal to the tdTomato:Synaptophysin bait protein (Fig 1D). Successful biotinylation was validated by immunohistochemistry in all three interneuron cellclasses by the presence of FITC-streptavidin staining colocalized with tdTomato signal (Fig 1E). Following successful biotinylation, tissue sections were homogenized, and biotinylated proteins were enriched by streptavidin pull-down (one brain /pull down), with no pooling of samples. Proteins were then analyzed by liquid chromatography tandem mass spectrometry (LC-MSMS). To prepare samples for bottom-up shotgun proteomics, samples were reduced and alkylated, and biotinylated peptides were digested off beads with trypsin (1:20, Promega) (Fig 1F). Peptide sequences were identified based on MS2 spectra acquired in the linear ion trap, and raw data were processed using MaxQuant v1.6.0.1 and search results were filtered to a 1% false discovery rate (FDR). A total of 3853 proteins were identified across all 18 samples prior to additional filtering (Supplemental Data Table 1). The dataset was then further filtered to remove proteins identified by only 1 peptide (2258 proteins removed), and any proteins that were present in fewer than 3/6 replicates in any cell type (1014 proteins removed). The final list contains 581 high-confidence, reproducibly observed proteins (Fig 1G). 141 proteins were uniquely identified by proximity labeling in parvalbumin presynaptic terminals (PV-PL samples), 20 in somatostatin presynaptic terminals (SST-PL samples), and 30 in vasoactive intestinal peptide proximity labeled presynaptic terminals (VIP-PL samples). Seventy-five proteins were shared between PV and VIP, 76 between PV and SST, 11 between SST and VIP, and 228 were present in all three cell types. An UpSet plot (Fig 1H) displays the intersection of proximity labeled proteins across all three classes of cells (PV-PL, SST-PL, and VIP-PL). In order to determine whether the proteins identified in the three proximity labeling conditions clustered according to three cell classes, we performed unbiased and unsupervised network analysis of shared proteins. We found that the network analysis resulted in 4 clusters of proteins, one with proteins shared by all conditions and three additional clusters formed by proteins elevated in each cell class respectively (Fig 1I). This unbiased clustering approach suggests that we are able to successfully capture and preserve proteome level differences between our initial three interneuron cell classes.

To test whether our approach enriched for proteins present at synapses, we performed Gene Ontology analysis (Panther) on PV-PL, SST-PL, and VIP-PL proteomes. We found that proteins in each of the three proximity labeling conditions were significantly enriched for the GO:CC (cellular compartment) terms presynapse, synaptic vesicle membrane, and presynaptic active zone compared to crude lysate, which enriched for the non-synaptic term neuronal cell body (Fig 2A). Similarly, PV-PL, SST-PL, and VIP-PL proteomes were enriched for the GO:BP (biological process) terms synapse organization, regulation of neuronal synapse plasticity, exocytosis, and synaptic vesicle priming, thus confirming that our enrichment was for the presynaptic terminal of the neuron (Fig 2B). Examining each cell-class proteomes gene ontology networks in more detail with Cytoscape,35 we found cellular compartment gene ontology terms for presynapse, axon, and synaptic vesicle as terms shared across all 3 local proteomes (Fig 2C-E). Cytoscape analysis also revealed cellular compartment gene ontology terms that were uniquely enriched in one cell-class, such as an enrichment for cellularmembrane proteins in the SST-PL proteome(Fig 2D). Similarly, a unique enrichment for postsynaptic density and NMDA receptors was found in PV-PL proteomes (Fig 2C). KEGG molecular signaling pathways for long-term potentiation and depression, as well as several terms related to addiction, appeared in PV-PL proteomes (Fig 2F) (Supplemental Data Table 2). In contrast, SST-PL terms included several molecular signaling pathways including PI3K/Akt/mTOR and Erbb2/Ras/MAPK (Fig 2G). The VIP-PL proteins were in terms mostly shared by the other two proteomes but also contained proteins in the gastric acid secretion pathway (Fig 2H).

Our Gene Ontology results above showed surprising enrichment in PV-PL for several terms related to the post-synapse and more specifically to glutamate receptor signaling. In order to obtain a more detailed assessment of this enrichment in glutamatergic postsynaptic mechanisms, we turned to the expert-curated synapse protein localization tool SynGO (syngoportal.org). While all three interneuron classes had proteins present from both pre- and postsynaptic compartments (Fig 3A-C), PV-PL and SST-PL samples showed a significantly higher ratio (Fig 3D) of post-synaptic proteins relative to presynaptic proteins compared to VIPPL [ANOVA; F(2,15)=5.77, p=0.0138; Tukey PV-PL vs SST-PL, p=0.023, SST-PL vs VIP-PL, p=0.0228]. As there were over 40 proteins in PV-PL contained within postsynaptic densities (Fig 3A), we examined these in more detail (Fig 3E). PV-PL samples included several proteins associated with NMDAR signaling: Cnksr2, Dlg2, Grin2B, Iqsec2, Lrrc4b, Lrrc7,and Rtn4 (Fig 3E). Baiap2 and Dlg2 were identified in all three cell classes; however, co-occurrence was most frequent in PV-PL cells (Fig 3F, G). As Grin2B and Grin2A were unique to PV-PL, we compared the spectral counts of Grin1, Grin2A, and Grin2B across cell classes for each biological replicate. We found that total Grin1 spectra identified was greatest for PV-PL, identified in 5 out of 6 replicates (Fig 3H). Furthermore, Grin2A spectral counts were only identified in PV-PL samples, and only a single biological replicate of SST-PL contained Grin2B spectra (Fig 3I,J). Together these patterns indicate that NMDA receptor subunits and associated proteins are enriched in the PV-PL samples.

In order to further assess the potential localization of NMDARs to PV presynaptic terminals we used a synaptotagmin-2 (Syt2) primary antibody for both proximity labeling and STED superresolution microscopy. Syt2 is a member of the synaptotagmin family of synaptic proteins which facilitates calcium-dependent fusion of synaptic vesicles and presynaptic membranes37 (Fig 4A), and it has been shown to localize to PV synaptic terminals in cortex and hippocampus38. We show by STED superresolution microscopy that Syt2 is present in PV-Cre:tdTomatosynaptophysin terminals (Fig 4B). Proximity labeling experiments using a Syt2 primary antibody in sections from three mice (C57BL/6) (Fig 4C) further showed that the resulting SYT2-PL proteome is similar to the PV-PL proteome (Fig 4D). As with PV-PL proteomes, SYT-PL proteomes are significantly enriched for the GO:CC terms presynaptic active zone, exocytic vesicle membrane, synaptic vesicle membrane, and post-synaptic specialization (Fig 4E). Furthermore, we investigated the localization of Grin1 relative to Syt2 by STED. Multiple terminals show signal for both Syt2 and Grin1, consistent with our observation of Grin1's presence in the PV-PL pre-synaptic proteome (Fig 4F). The fluorescent signal of Grin1 and Syt2 within terminals was more completely overlapping (Fig 4F) than the pattern observed for Syt2 and tdTomato-synaptophysin within terminals (Fig 4B), suggesting that Grin1 may be in close proximity to the point of synaptic vesicle and presynaptic membrane fusion in PV cell synapses.

Our gene ontology analysis of PV-PL proteomes indicated enrichment for several pathways related to addiction (Fig 2F; nicotine, amphetamine, and cocaine), and we performed an additional gene ontology analysis using Qiagen Ingenuity Pathway Analysis (IPAtm) to investigate whether additional potentially enriched pharmacological pathways are present at PV terminals. In addition to synaptogenesis and calcium signaling pathways, IPA analysis revealed that SYT2-PL proteomes contained proteins implicated in opioid signaling pathways (Fig 5A). Presynaptic effects of mu-opioid receptor activation on PV interneuron transmitter release have been shown previously39, and we hypothesized that mu-opioid receptor activation by the partial agonist buprenorphine could alter SYT2-PL proteomes. Our rationale was that pharmacological activation of presynaptic opioid receptors on PV cells would activate a signaling cascade that could alter presynaptic protein localization which would be detectable by both proximity labeling and STED. We administered 0.05 mg/ml of the mu-opioid high-affinity partial agonist buprenorphine or saline to C57BL/6 mice at postnatal day 30 (Fig 5B, n=3 per condition). Biotinylation in molecular proximity to Syt2 was confirmed by double labeling for Syt2 and FITCStreptavidin (Fig 5C). Proteomic analysis of Syt2-PL revealed after filtering a total of 70 proteins unique in buprenorphine-treated mouse cortices, 167 proteins in cortices from saline-injected controls, and 128 shard between them. (Fig 5D) (Supplemental Table 4). SYNGO analysis for the synaptic compartments of proteins identified in the two conditions indicated a decrease in postsynaptic proteins in the buprenorphine condition (Fig 5E). Consistent with this analysis, LC/MS data also showed a decrease in the average precursor intensities for Grin1 and Dlg1 in the buprenorphine-treated samples (Fig 5F). Using STED microscopy and double labeling for Syt2 and NMDAR1 we found colocalization of both fluorescent signals in the presynaptic terminals of saline-treated mice (Fig 5G). Fluorescence intensity versus position plots indicate that SYT2 and NMDAR1 intensity peaks are in overlapping positions in X-Y and X-Z projections in saline-treated controls (Fig 5 H,I upper panels). In contrast, in buprenorphine-treated samples, fluorescence intensity peaks are shifted relative to each other in both projections (Fig 5 H,I lower panels). The STED microscopy results support the proximity labeling results and indicate that mu-opioid receptor activation by buprenorphine alters the proximity of SYT2 and NMDAR1 in PV synaptic terminals.

All mice were housed in compliance with the Institute for Animal Care and Use Committee at the University of Connecticut. In order to target specific GABAergic interneuron cell types, VIPIRES-Cre (VIPtm1(cre)Zjh/J JAX stock no 010908), SST-IRES-Cre (Ssttm2.1(cre)Zjh/J, JAX stock no 013044) and PV-Cre (Pvalbtm1(cre)Arbr/J, JAX stock no 017320) were used. VIP-IRES-Cre, SSTIRES-cre, and PV-Cre mice were crossed to a tdTomato-Synaptophysin (Gt(ROSA)26Sortm34.1(CAG-Syp/tdTomato)Hzs/J, JAX stock no 012570) reporter mouse. VIP:tdTomatoSynaptophysin, SST:tdTomato-Synaptophysin, and PV:tdTomato-Synaptophysin mice were used for the proximity labeling experiments

Two cohorts of 6 C57BL/6 male mice were subcutaneously injected with 0.05 mg/ml buprenorphine or saline control. Mice were kept in their home cages for 4 hours post-injection and then transcardially perfused with 4% paraformaldehyde (PFA) and brains were allowed to postfix for 2 hours in PFA. The cortex was then microdissected and sliced at 60μm before performing proximity labeling as described above with SYT2 as the primary antibody.

Mice were transcardially perfused with fresh, ice-cold 2% (PFA). Brains were post-fixed in 2% PFA for 2 hours before removal of the cortex through microdissection and vibratome sectioning at 60µm. Slices were collected in 5ml tubes and proximity labeling was performed on freefloating slices.

Slices were permeabilized with phosphate buffered saline with 0.1% Triton-X (PBS-T) before endogenous peroxidases were quenched with 0.5% hydrogen peroxide (H2O2) for 10 minutes. After washing, slices were blocked for 2 hours in blocking buffer (1% BSA in PBS + 0.1% Tween-20). A primary antibody (rabbit anti-RFP, Rockland 1:500, mouse anti-synaptotagmin 2, Zebrafish International, 1:300, Rabbit PSD-95, 1:2000) was added with an overnight incubation followed by a 4 hour incubation in HRP-conjugated secondary antibody (1:300). Following antibody labeling, samples were incubated for 15 minutes in reaction buffer (100mM borate with 2% dextran sulfate and 0.1% Tween-20) with biotin-tyramide at a ratio of 1:500 (buffer:biotintyramide). Biotin-tyramide was synthesized in-house (see Organic Synthesis of BiotinTyramide). Biotinylation was initiated through the addition of 0.003% H2O2 and the reaction was allowed to proceed for 1 minute before being quenched with 500mM sodium ascorbate. Samples were rapidly washed in PBST to remove any additional biotin-tyramide before immunohistochemical validation.

After biotin labeling, a subset of slices were taken for immunohistochemical validation that biotinylation occurred. Slices were labeled with FITC-streptavidin (Invitrogen, 1:1000), DAPI (1:1000), and an mouse Alexa-568 conjugated secondary antibody (Thermo Fisher, 1:1000). Slices were labeled, mounted, and viewed on a Leica SP8 confocal microscope. The reaction was considered successful if FITC labeling corresponded to the location of the original primary antibody. If the labeling was successful, the remaining free-floating slices were prepared for mass spectrometry.

Biotin-tyramide was synthesized by dissolving 10mg of NHS-Biotin (Thermo Fisher) in dimethylformamide (DMF, Sigma) and allowed to react at room temperature with 10mg tyramine-HCl (Sigma, T90352) dissolved in DMF and triethylamine (Sigma) for 2 hours in the dark. The reaction was stopped through the addition of 100% ethanol. 1 volume of ethyl acetate (Sigma) was then added. Following the reaction, DMF was removed through a liquid-liquid extraction using repeated (3-5) extractions of 5x volume of water, and the biotin-tyramide product was removed and stored at 4C until use.

Slices were homogenized in a 1ml glass Teflon homogenizer. A final concentration of 3% sodium deoxysulfate (Sigma) and 2% sodium deoxycholate (Sigma) were added and samples were heated at 99°C for 1 hour. Samples were centrifuged at 13,000xg for 5 minutes. 150ul of the supernatant was saved for pre-enrichment analysis (<input= LCMS samples), and the remaining supernatant was added to 150µl of Streptavidin M-280 Dynabeads. After incubation at room temperature for 2 hours, the supernatant was removed and saved for analysis (<unbound= LCMS samples). Beads were washed 5x in PBS, and finally resuspended in 300ul of 0.1M ammonium bicarbonate (Sigma).

Bead-bound samples were reduced in 5 mM dithiothreitol (Thermo Fisher) in 0.1 M ammonium bicarbonate at 23°C for 1.5 h followed by alkylation using 10mM iodoacetamide (Sigma) in 0.1 M ammonium bicarbonate for 45 min at 23°C, covered from light. Samples were proteolyzed with porcine modified sequencing grade trypsin (Promega) at a ratio of 1:20 enzyme:protein at 37°C for 16 hr. Following proteolysis, the peptide-containing supernatant was removed and digestion was quenched with concentrated formic acid to yield a final pH of 3.0. Peptides were desalted on Pierce C18 Peptide Desalting Spin Columns according to manufacturer’s specifications. After desalting, peptides were dried to completion, resuspended in 0.1% formic acid (Pierce) in water, and quantified on a UV-VIS Spectrophotometer using A280/A260 (1 Abs = 1 mg/mL). Injection amounts for mass spectrometry analysis were normalized across all samples based on A280 readings.

Peptide samples were subjected to UPLC-MS/MS using a Thermo Scientific Ultimate 3000 RSLCnano ultra-high performance liquid chromatography (UPLC) system coupled to either a Thermo Scientific Q Exactive HF or Orbitrap Eclipse Tribrid mass spectrometer. For Q Exactive HF experiments, desalted peptides resuspended in Solvent A (0.1% formic acid in water) were injected onto a nanoEase M/Z Peptide BEH C18 column (1.M , 75M x 250mm, Waters Corporation) and separated by a 300 nL/min reversed-phase UPLC gradient of 4-30% Solvent B (0.1% formic acid in acetonitrile) over 90 min, 30% to 90% B in 20 min, 90% B hold for 10 min, and return to initial conditions over 2 min with an additional 18 min of column re-equilibration at 4% Solvent B with a total runtime of 150 min. Peptides were ionized directly into the mass spectrometers using positive mode ESI. Q Exactive HF-specific MS parameters included 60,000 resolution, AGC target of 1e6, maximum injection time (MIT) of 60ms, and a mass range of 3001,800 m/z. Data-dependent higher-energy C-trap dissociation (HCD) MS/MS scan acquisition parameters included 15,000 resolution, AGC target of 1e5, MIT of 40 ms, loop count of 15, isolation window of 2.0 m/z, dynamic exclusion <on= with a window of 30 sec, normalized collision energy of 27, and charge exclusion <on= for all unassigned, +1, and > +8 charged species. For experiments acquired using the Orbitrap Eclipse Tribrid, either the 150 min UPLC gradient described above was used, or a shortened version with a total runtime of 90 min using identical flow rates and UPLC column was used. Orbitrap Eclipse Tribrid MS parameters were as follows: resolution of 60,000, Standard AGC target, 60 ms MIT, mass range of 300-1800 m/z, and 30% RF lens. Data-dependent MS/MS spectra were acquired in the linear ion trap (IT) using collision-induced dissociation (CID) and the following parameters: Dynamic exclusion <on= with a window of 30 sec, charge state inclusion of +2 to +8 charge states only, exclude after n times set to 1, intensity threshold of 5.0e3, cycle time of 3 sec, isolation window of 1.6 m/z, 35% CID collision energy and 10 ms CID activation time, <Rapid= IT scan rate, <Normal= mass range, <Standard= AGC target, and <Dynamic= MIT.

Peptide identification and label-free quantitation via MaxQuant and data analysis using Scaffold 5 Peptides were identified and quantified for all experiments with the Andromeda search engine embedded in MaxQuant v1.6.0.1 (Cox, J., and Mann, M, 2008). Raw data were searched against the UniProt Mus Musculus reference proteome (UP000000589, Accessed 12/12/2017 and updated 05/16/2017) with the fluorescent protein TdTomato sequence appended, plus the MaxQuant contaminants database. Oxidation(M), acetylation(protein N-term), Gln->pyro-Glu, and a Biotin-Phenol(Y) custom modification of 361.1460 Da were set as variable modifications. Carbamidomethylation(C) was set as a fixed modification. Protease specificity was set to trypsin/P with a maximum of 2 missed cleavages. LFQ quantification was enabled and minimum number of amino acids/peptide set to 5. All results were filtered to a 1% false discovery rate at both the peptide and protein levels based on a target/decoy method; all other parameters were left as default values. Peptide/protein identifications by Andromeda and quantification data by MaxQuant were subsequently visualized and analyzed in Scaffold v5.0.1 (Proteome Software) with a minimum of 2 peptides required for identification and average precursor intensity (API) values used for quantification. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD036293 and reviewer access as follows: Username: reviewer_pxd036293@ebi.ac.uk Password: 9uS9FYaz=

Scaffold data were imported into MATLAB (version 2021a) for filtering and analysis. Contaminants were removed and final protein list was obtained by removing proteins present in fewer than 3/6 replicates. API values for remaining proteins were averaged across replicates. A background list of proteins known to promiscuously bind to M-280 streptavidin dynabeads was compared to each sample’s protein list and removed from each sample when identified. Protein lists were analyzed using Gene Ontology (Panther, Geneontology.org), Cytoscape, SynGo (SyngoPortal.org), and IPA Ingenuitytm. Cytoscape analysis was performed using the ClueGO plugin35 using integrated GO terms. Panther Gene Ontology was run using the Mus musculus background with a Fisher’s Exact test and Bonferroni correction for multiple hypothesis testing.

SynGO results were compared to a background of brain-expressed genes. All SynGO and gene 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 ontology lists were reported in descending order of enrichment and display the top results obtained from the analysis. Network analysis (Fig 1I) was generated using differentially expressed proteins with the greatest fold changes (top 10%, defined by the absolute value of expression ratios) using a force-directed layout36 and inverse edge weighting over 1000 iterations.

STED images were collected on an Abberior Instruments Expert Line STED system (Göttingen, Germany) using an Olympus 100x/1.40 UPLANSAPO oil immersion lens. STAR RED was excited with a pulsed 640 nm laser, and STAR ORANGE was excited using a 561 nm pulsed laser. Both dyes were imaged with a pulsed 775 nm depletion laser. STED images were deconvolved to a maximum of 20 iterations and a 0.01 threshold using Huygens software (SVI, Hilversum, Netherlands). 1. Ascoli, G. A. et al. Petilla terminology: Nomenclature of features of GABAergic interneurons of the cerebral cortex. Nat. Rev. Neurosci. 9, 5573568 (2008). Interneuron Precursors is Prescient of their Mature Function. Cereb. Cortex 18, 230632317 (2008). 3. Paul, A. et al. Transcriptional Architecture of Synaptic Communication Delineates

GABAergic Neuron Identity. Cell 171, 522-539.e20 (2017). 4. Galarreta, M. & Hestrin, S. Electrical and chemical synapses among parvalbumin fastspiking GABAergic interneurons in adult mouse neocortex. Proc. Natl. Acad. Sci. U. S. A. 99, 12438312443 (2002). 5. Barthó, P. et al. Characterization of neocortical principal cells and interneurons by network interactions and extracellular features. J. Neurophysiol. 92, 6003608 (2004). 6. Gentet, L. J. et al. Unique functional properties of somatostatin-expressing GABAergic neurons in mouse barrel cortex. Nat. Neurosci. 15, 6073612 (2012). cellular design to microcircuit function. Science 345, (2014). 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 (2014). proximity labeling reveals activity-dependent nuclear proteome dynamics in the striatum. Nat. Commun. 12, 4855 (2021).

Curr. Opin. Chem. Biol. 48, 137 (2019).

Cell 166, 1295-1307.e21 (2016). Parvalbumin-Expressing Interneurons Is Decreased in the Prefrontal Cortex in Autism.

Cereb. Cortex 28, 6903690 (2018).

Prefrontal Cortex-Dependent Behavior and Psychiatric Disorders. Front. Neural Circuits 12, 18. Meldrum, B. S. Neurotransmission in Epilepsy. Epilepsia 36, 30335 (1995).

37 (2018).

a022764 (2016).

17. Khazipov, R. GABAergic Synchronization in Epilepsy. Cold Spring Harb. Perspect. Med. 6, 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 With a Decreased Seizure Threshold and a Progressive Loss of Parvalbumin-Positive

Neurosci. 17, 389433906 (1997). containing interneurons by VIP-immunopositive interneurons in the primary somatosensory cortex of the adult rat. Eur. J. Neurosci. 25, 232932340 (2007). 25. Pi, H.-J. et al. Cortical interneurons that specialize in disinhibitory control. Nature 503, 5213 524 (2013). dependent biotinylation to characterize protein complexes and map organelles. Curr. Opin.

Chem. Biol. 48, 44354 (2019). 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 124 (2021).

Methods 15, 1273133 (2018).

Mouse Cerebral Cortex. Neuron 91, 122831243 (2016).

Neurosci. 36, 1833215 (2013). ontology and pathway annotation networks. Bioinformatics 25, 109131093 (2009).

Pract. Exp. 21, 112931164 (1991). receptors in hippocampal parvalbumin interneurons. eLife 10, e69746 (2021). Methyl-D-Aspartate Receptors Enhance Parvalbumin Cell3Mediated Inhibition of Pyramidal Cells in Mouse Prefrontal Cortex. Biol. Psychiatry 84, 4603470 (2018). 41. Paquet, M. & Smith, Y. Presynaptic NMDA receptor subunit immunoreactivity in GABAergic terminals in rat brain. J. Comp. Neurol. 423, 3303347 (2000). 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 Depression by Subunit-Preferring NMDA Receptor Antagonists in Mouse Barrel Cortex.

Cereb. Cortex 19, 295932969 (2009). Newly Appreciated Roles in Cortical Synaptic Function and Plasticity. The Neuroscientist 14, 6093625 (2008). identity of NMDA receptors at synaptic and non-synaptic inputs in cerebellar molecular layer interneurons. Front. Synaptic Neurosci. 7, (2015). health and disease. Neuropharmacology 193, 108615 (2021). Georgi Tushev, Nicole Fürst, Julian D. Langer, Erin M. Schuman,The proteomic landscape of synaptic diversity across brain regions and cell types, Cell 186 (2023)

E.

V P T S S P I

V bioRxiv preprint doi: https://doi.org/10.1101/2024.01.03.574066; this version posted January 4, 2024. The copyright holder for this preprint (which was not certified by peer Brev.iew) is the author/funder, who hDasgranted bioRxiv a license to display the preprint in perpetuity. It is made .

available under aCC-BY-NC-ND 4.0 International license.

BiotinTyramide

H2O2 Synaptic

Vesicle