Cell Type-Specific Tandem Affinity Purification of the Mouse

Sep 5, 2016 - Phone: +49 6131 39 25912. ... We used a combination of viral-mediated cell type-specific expression of a tagged CB1 fusion protein (CB1-...
0 downloads 0 Views 5MB Size
Subscriber access provided by Northern Illinois University

Article

Cell type-specific tandem affinity purification of the mouse hippocampal CB1 receptor-associated proteome Tobias Mattheus, Katharina Kukla, Tina Zimmermann, Stefan Tenzer, and Beat Lutz J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00339 • Publication Date (Web): 05 Sep 2016 Downloaded from http://pubs.acs.org on September 7, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Cell type-specific tandem affinity purification of the mouse hippocampal CB1 receptorassociated proteome Tobias Mattheus1, Katharina Kukla1, Tina Zimmermann1, Stefan Tenzer2, Beat Lutz1* 1

Institute of Physiological Chemistry, University Medical Center of the Johannes Gutenberg-

University Mainz, Duesbergweg 6, 55128 Mainz, Germany 2

Institute for Immunology, University Medical Center of the Johannes Gutenberg-University

Mainz, Langenbeckstrasse 1, 55131 Mainz, Germany

* Corresponding author Email: [email protected] Phone: +49 6131 39 25912

Author contributions B.L. conceived and guided the study, T.M., K.K., and T.Z. carried out the biochemical and animal experiments, S.T. performed the mass spectrometry analysis and T.M the integrative network analysis; T.M. wrote, and S.T. and B.L. edited the manuscript.

Abstract G protein coupled receptors (GPCR’s) exert their effects through multiprotein signaling complexes. The cannabinoid receptor type 1 (CB1) is among the most abundant GPCR’s in the mammalian brain and involved in a plethora of physiological functions. We used a combination of viral-mediated cell type-specific expression of a tagged CB1 fusion protein (CB1-SF), tandem affinity purification (TAP) and proteomics on hippocampal mouse tissue to analyze the composition and differences of CB1 protein complexes in glutamatergic neurons 1

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

and in GABAergic interneurons. Purified proteins underwent tryptic digestion and were identified using deep-coverage data-independent acquisition with ion mobility separationenhanced mass spectroscopy, leading to the identification of 951 proteins specifically enriched in glutamatergic and GABAergic CB1-SF TAP samples as compared to controls. Gene Ontology and protein network analyses showed an enrichment of single proteins and functional clusters of proteins involved in already well described domains of CB1 functions. Supported by this consistent dataset we could confirm already known CB1 interactors, reveal new potentially interacting proteins and differences in cell type-specific signaling properties of CB1, thereby providing the foundation for further functional studies on differential CB1 signaling.

Keywords Cannabinoid receptor type 1, CB1, ECS, endocannabinoid system, protein interactions, proteomics, GPCR, signaling, TAP, tandem affinity purification

Introduction The cannabinoid receptor type 1 (CB1) is among the most abundant G protein coupled receptors (GPCRs) in the mammalian brain with the highest expression levels in various neuronal subtypes in the basal ganglia, substantia nigra, globus pallidus, cerebellum and hippocampus 1. It is mainly coupled to Gα proteins of the Gi/o-family, which inhibit adenylyl cyclase and affect the activity state of mitogen-activated protein kinases (MAPK) and several types of calcium and potassium channels 2-5. In the hippocampus, CB1 is expressed in glutamatergic principal and GABAergic interneurons, which exert antagonizing effects on the overall neuronal circuitry output 6. The main action of the mainly presynaptically localized CB1, is the attenuation of neurotransmitter release 7. CB1 is activated through binding of its ligands, the endocannabinoids, which are 2

ACS Paragon Plus Environment

Page 2 of 48

Page 3 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

synthesized on demand at the postsynaptic site and travel retrogradely through the synaptic cleft and bind the receptors. Despite the fact that in the hippocampus CB1 density is remarkably higher in GABAergic interneurons than in glutamatergic neurons, previous studies have shown that coupling to G protein signaling is several fold stronger in glutamatergic neurons that in GABAergic interneurons, which indicates cell type-specific differences of signaling complex composition or properties 8. Consistent with the high abundance in the limbic system, numerous effects in learning and memory, anxiety and depression, and the reward system have been described

6, 9, 10

.

Furthermore, the endocannabinoid system (ECS) plays an important role in the developing brain in axonal outgrowth and guidance and in the adult brain in modulating synaptic strength and adult neurogenesis

11

. Besides its fundamental roles in neurons, CB1 is also expressed

in astrocytes, where it is activated upon release of endocannabinoids produced by surrounding neurons in consequence of their depolarization. In the hippocampus, activation of astrocytic CB1 leads to an increase in astrocytic intracellular calcium levels, which can then potentiate synaptic transmission at the tripartite synapse

12, 13

. The involvement of the

ECS in all of the afore-mentioned fields are well studied, but still face a substantial lack of understanding of the highly complex intracellular signaling at the molecular level. While several CB1 interacting proteins have been described

4, 14

, the majority of proteins

assembling with CB1 in a multiprotein complex are still unknown. Several techniques are available for the identification of protein interactions. Coimmunoprecipitation (CoIP) using antibodies against a bait protein and subsequent screening for co-purified interaction proteins by Western blot is the standard assay in this field, but only allows for directed screening towards expected proteins. Proximity Ligation Assays (PLA) or Fluorescence Resonance Energy Transfer (FRET) represent alternative methods but like performing CoIP only candidate proteins can be tested. Yeast two-hybrid assays, however, allow for high throughput screens but are performed within a highly artificial cellular environment. Tandem Affinity Purification (TAP) is a method originally developed by Rigaut et al. in 1999 and continuously improved since then, which allows for high throughput 3

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 48

screening, whereby interactions are detected within a natural cellular environment 15. The two step purification process furthermore minimizes false positive results. This method has recently been applied for other GPCRs, i.e. the A2A Adenosine receptor MT2 melatonin receptors

16

, and the MT1 and

17

, but not yet for CB1. For this receptor, only yeast two-hybrid

screens have been performed

18, 19

, and a number of investigations leading to the discovery

of several single functionally interacting GPCRs 20-22 and signaling proteins 14. Recently, the first proteomic approach to identify CB1-interacting proteins has been published and led to new insights on how CB1 signaling alters the actin cytoskeleton and subsequently growth cone morphology and the remodeling of synaptic spines via the WAVE1 complex in the mouse cortex 23. In the present study, we took advantage of adeno-associated virus (AAV) vectors, Creexpressing mouse lines and the TAP technique using a StrepII/FLAG-tagged CB1 24 to purify CB1 signaling complexes. We revealed differences between CB1 protein complex compositions in glutamatergic neurons versus GABAergic interneurons in the hippocampal synaptosomal preparations. Using a deep-coverage proteomic approach and further enrichment analyses, we could identify a plethora of co-purified proteins with minor differences in cell subtype specific CB1 protein complexes. These differences in cell typespecific enrichment of a number of proteins require further investigations and might be responsible for differences in CB1 signaling in the respective neuronal subtype. The identification of individual potentially interesting interactors now paves the way for further investigations on their functional relevance in CB1 signaling.

Experimental Procedures Animals Male and female mice at an age of 2 – 6 months were used and maintained under standard conditions with food and water ad libitum. All experimental procedures were approved by the

4

ACS Paragon Plus Environment

Page 5 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Committee on Animal Health and Care of the local government. Conditional Glu-CB1-KOmice and GABA-CB1-KO-mice were obtained as described previously

25

and used for viral-

mediated Cre recombinase-dependent Glu- or GABA-specific overexpression of CB1-SF.

DNA constructs All DNA constructs used in this study are based on either pcDNA3 mammalian expression or pAM AAV expression vector backbones. pcDNA3 vectors contain a CMV (cytomegalovirus) promoter and a bovine growth hormone polyadenylation signal (BGHpA) to drive expression of a recombinant protein in mammalian cells. In order to produce AAVs containing the CDS and regulatory sequences to drive expression of a recombinant protein after infection, pAM vectors were used, containing a CAG promoter (CMV early enhancer element, promoter, first exon and the first intron of the chicken beta-actin gene and the splice acceptor of the rabbit beta-globin gene), woodchuck hepatitis virus posttranscriptional regulatory element (WPRE), AAV2 inverted terminal repeats (ITRs) and a bovine growth hormone polyadenylation signal (BGHpA). A StrepII/FLAG-tag (SF-tag)

24

cDNA sequence was synthesized by Eurofins

Genomics, Ebersberg (Germany), taking into account the codon usage of Mus musculus. Endonuclease restriction sites were added by PCR and subcloned into a mouse CB1 cDNA containing pcDNA3 expression plasmid generated by our lab at the 3’ end of the CB1 cDNA sequence to obtain a pcDNA3/CMV-CB1-SF-bGHpA (pcDNA3-CB1-SF) plasmid. For the generation of AAV plasmids the CB1-SF cDNA sequence was cut out using specific restriction enzymes and ligated into a linearized pAM AAV expression plasmid as described above to generate a pAM/CAG-CB1-SF-WPRE-bGHpA (pAM-CB1-SF). For a Cre recombinase-dependent expression of CB1-SF we generated a pAM plasmid containing a 340 bp transcriptional Stop cassette composed of a herpes simplex virus thymidin kinase pA signal sequence and a pA terminator from pGL3 (Promega, Madison, USA) flanked by loxP sites, which was synthesized by a commercial provider (Epochbiolabs, Missouri City, USA). This floxed stop cassette was transferred downstream the CAG promoter to obtain a

5

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

pAM/CAG-Stop-WPRE-bGHpA used as a basis for fusing sequentially the SF-tag downstream the floxed stop cassette and finally the Mus musculus CB1 cDNA sequence between the floxed stop cassette and the StrepII/FLAG-tag generating a pAM/CAG-StopCB1-SF-WPRE-bGHpA (pAM-Stop-CB1-SF).

Cell culture HEK293 cells were kept in 6 well plates to prepare cell lysates, or grown on poly-L-lysine coated coverslips (if used for immunocytochemistry) in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (growth medium) at 37°C in 5% CO2. Transfection with pcDNA3-CB1-SF, pcDNA3-SF or pAM-CB1-SF plasmid was performed using the calcium phosphate precipitation method 26 2 h after changing the growth medium to Improved Minimum Essential Medium supplemented with 5% fetal bovine serum. 16 h after the transfection, the medium was replaced with DMEM, and cells kept for another 24 h until they were used for further analysis.

Agonist and antagonist treatments of cells HEK293 cells were treated with the CB1 agonist CP55,940 (Sigma Aldrich) or antagonist SR141716 (NIMH Chemical Synthesis and Drug Supply Program) pre-diluted in DMSO 48 h after transfection by replacing the medium with growth medium containing 40 µM of agonist or antagonist (final concentration 0.1% DMSO). Cells were directly fixated afterwards and used for immunocytochemistry analysis.

AAV injection Chimeric AAV serotype 1/2 vectors were produced using the pAM-Stop-CB1-SF (AAV-StopCB1-SF) and the genomic titers determined using an Applied Biosystems ABI 7500 real time

6

ACS Paragon Plus Environment

Page 6 of 48

Page 7 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

PCR cycler as described before

27

. The AAV-Stop-CB1-SF was delivered to anaesthetized

adult Glu-CB1-KO mice or GABA-CB1-KO mice bilaterally into the dorsal (-3.1 mm AP,-3.0 mm ML,-3.5 mm DV from bregma) and ventral (-2.0 mm AP,-2.0 mm ML,-2.0 mm DV from bregma) hippocampus using a microprocessor controlled mini-pump (World Precision Instruments) with 34xG beveled needles (World Precision Instruments) in a stereotaxic frame (Kopf Instruments), injecting 1µl per injection site at a rate of 200nl/min.

Synaptosome preparation The hippocampi of 5 mice were dissected and homogenized in 4 ml Buffer A (0.32 M sucrose, 5 mM HEPES pH 7.4, 1 mM EGTA, HaltTM Protease/Phosphatase inhibitor (Thermo Scientific)) using a glass Potter homogenizer. After centrifugation for 10 min with 1000 x g at 4°C the supernatant was transferred to a Falcon tube and the pellet resuspended and homogenized again. The supernatant after centrifugation of the second homogenate was then combined with the first one and centrifuged for 15 min with 12,000 x g at 4°C. After discarding the supernatant, the pellet was resuspended in 2 ml Buffer A and centrifuged again for 20 min with 12,000 x g at 4°C. The supernatant was then discarded and the pellet resuspended in 400 µl Buffer B (0.32 M sucrose, 5 mM Tris pH 8.1, 1 mM EGTA, HaltTM Protease/Phosphatase inhibitor (Thermo Scientific)). After loading the resuspended pellet onto a sucrose gradient (1.2 M, 1 M, 0.85 M in Buffer B) it was centrifuged for 2 h with 85,000 x g at 4°C and the synaptosome containing phase between 1.2 M and 1 M sucrose collected 28.

Tandem Affinity Purification Synaptosome preparations were transferred to 5 ml of ice-cold Lysis Buffer (30 mM Tris pH 7.4, 150 mM NaCl, 0.3% BrijO10, HaltTM Protease/Phosphatase inhibitor (Thermo Scientific)) and incubated on a rocker at 4°C overnight. Next day, the synaptosome debris was

7

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

centrifuged for 10 min with 5,000 x g at 4°C, and the supernatant was filtered through a 0.22 µm Millex GP sterile filter (Merck Millipore). The lysate was then incubated with 150 µl of Streptactin Superflow Beads (IBA) on a rocker for 1 h at 4°C. After centrifugation for 30 sec with 1,000 x g at 4°C, the supernatant was discarded and the bead mixture transferred to Illustra MicroSpin Columns (GE Healthcare). The beads were then washed 3x with 500 µl Lysis Buffer, and the proteins were eluted with 700 µl of 3X desthiobiotin (IBA) by incubating for 10 min in an overhead rotor at 4°C. The eluate was transferred to fresh microspin columns together with 50 µl of anti-FLAG M2 Agarose Beads (Sigma Aldrich) and incubated for 1 h at 4°C in an overhead rotor. The column was then washed 3x with 500 µl Lysis Buffer and afterwards the protein complexes eluted from the beads by incubation with 250 µl of 2X FLAG Peptide (Sigma Aldrich) for 20 min in an overhead rotor at 4°C. The final samples were then collected by centrifugation of the eluate for 5 sec with 2,000 x g at 4°C into fresh Eppendorf tubes.

Western blot Cell lysates (supernatant of homogenized HEK293 cells in HaltTM Protease/Phosphatase inhibitor (Thermo Scientific) containing lysis buffer (10 mM Tris pH 7.4, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100 centrifuged at 20,000 x g for 10 min at 4°C), hippocampal lysates (supernatant of homogenized hippocampal tissue in HaltTM Protease/Phosphatase inhibitor (Thermo Scientific) containing RIPA buffer (10 mM Tris pH 7.4, 140 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 0.1% Triton X-100, 0.1% SDS, 0.1 g sodium-deoxycholate centrifuged at 20,000 x g for 30 min at 4°C), synaptosome preparations or TAP eluates (as described above) were prepared. The protein concentration of hippocampal lysates and synaptosome preparations were determined using the BCATM protein assay kit (Pierce), and 20 µg of protein was heated to 95°C for 5 min in Laemmli reducing sample buffer. The proteins were resolved by 12% SDS-PAGE and electroblotted onto a nitrocellulose membrane. After blocking in 5% nonfat dry milk in TBS containing 0.1% Tween 20 (TBS-T) for 45 min at room

8

ACS Paragon Plus Environment

Page 8 of 48

Page 9 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

temperature, the membrane was incubated with primary antibodies: rabbit anti-CB1 [1:500] (Frontier Sciences, Hokkaido, Japan), rabbit anti-FLAG [1:500] (Sigma Aldrich) or rabbit antiActin [1:10,000] (Merck Millipore) at 4°C overnight. After washing with TBS-T, the membranes were incubated in horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG secondary antibodies [1:5,000] (Dianova, Hamburg, Germany) in 5% nonfat dry milk in TBS-T for 1 h at room temperature. After washing with TBS-T, Western blot was performed by incubating the membrane with the Amersham ECL Prime Western Blotting Detection Reagent (GE Healthcare Life Sciences) for 5 min, followed by an analysis with the FUSIONSL chemiluminescence imaging system (Peqlab).

Immunocytochemistry on cells Coverslips with adherent cells were washed two times with PBS and incubated in 4% paraformaldehyde (Sigma Aldrich) for 15 min for fixation. Afterwards, cells were permeabilized in PBS containing 0.1% Tween 20 (PBS-T) for 5 min and blocked in 4% normal donkey serum (NDS) for 30 min to reduce non-specific background. Primary antibody incubation was performed at 4°C overnight diluted in 4% NDS using rabbit anti-CB1 [1:1,000] (Frontier Sciences). On the next day, coverslips were washed three times in PBS-T for 10 min and incubated with fluorescence labeled secondary antibody Alexa546 goat IgG [1:1,000] (Invitrogen) in 4% NDS for 1 h in the dark. After three washing steps with PBS-T nuclei were counterstained by incubation in Draq5 (BioStatus) for 30 min. Coverslips were mounted on glass slides with Mowiol mounting medium, and fluorescence was visualized using a Zeiss LSM 710 confocal microscope (Zeiss).

Immunohistochemistry on brain sections 40 µm free-floating coronal cryosections of perfused AAV-injected Glu-CB1-KO and GABACB1-KO mice, and non-injected Glu-CB1-wildtype (wt) and GABA-CB1-wt littermates were

9

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 48

permeabilized in PBS containing 0.2% Triton X-100 (PBS-TX) and then blocked in 4% normal goat serum (NGS) in PBS-TX for 15 min at room temperature. Primary antibody incubation was performed at 4°C overnight in 4% NGS using the following primary antibodies: mouse anti-FLAG [1:500] (Sigma Aldrich), guinea pig anti-VGlut1 [1:500] (Merck Millipore), guinea pig anti-VGAT [1:100] (Merck Millipore). Next day, sections were washed with PBS-TX and incubated with Alexa488- or Alexa546-conjugated secondary antibodies goat IgG [1:1,000] (Invitrogen) in 4% NGS for 1 h in the dark. After washing with PBS-TX, the slices were counterstained with the nuclear dye 4’,6-diamidino-2-phenylindole (DAPI) for 5 min and washed in PBS. Sections were transferred onto glass slides and coverslipped with Mowiol mounting medium. Fluorescence was visualized with a Leica TCS SP5 Confocal Microscope (Leica).

Protein Preparation for MS TAP Eluates were frozen at -80°, lyophilized to dryness and solubilized in 200 µl 7 M urea, 2 M thiourea and 2% CHAPS. Proteins were digested with sequencing grade trypsin (Trypsin Gold, Promega) using a modified Filter-aided sample preparation (FASP)

29

. After FASP

digest, resulting tryptic peptides were concentrated to 20 µl by lyophilization and 5 µl of 100 fmol/µl MassPrep Enolase Digestion Standard (Waters) were added to each sample and transferred into an Autosample Vial.

UPLC-MS Analysis For nanoUPLC-MS analysis, 2 µl were used per injection. Samples were analyzed in three technical replicates. Tryptic peptides were separated by reversed-phase nanoUPLC in direct injection mode on a Waters nanoAcquity System equipped with a C18 HSS-T3 75µm x 250 mm column using a gradient from 4% to 40% B over 90 min as described before

29

. Buffer A

was 0.1% formic acid in water + 3% DMSO. Buffer B was 0.1% formic acid in acetonitrile +

10

ACS Paragon Plus Environment

Page 11 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

3% DMSO. The column was coupled to a nanoelectrospray Source on a Waters Synapt G2S mass spectrometer operated in ion-mobility enhanced, data-independent acquisition mode as described previously 30.

Data processing Resulting raw data files were processed by ProteinLynxGlobalServer (PLGS, v3.0.2), and database search was performed against the mouse UniProt Reference Proteome database supplemented with common contaminants (trypsin, bovine serum albumin, human keratins, etc) as described before

30

. Data post-processing and TOP3-based label-free quantification

were performed in the ISOQuant software 29.

Bioinformatics A network of all identified proteins with a molar ratio of at least 1% relative to the bait protein was generated using the STRING database for known and predicted protein-protein interactions (string-db.org)

31

with required medium confidence scores (0.4). Clustering

analysis of the network based on the confidence scores of protein interactions was performed using Cytoscape

32

with ClusterMaker2 plugin

33

using the Markov Cluster

Algorithm (MCL). Additional Gene Ontology (GO) annotation was performed using DAVID (https://david.ncifcrf.gov) 34 or BiNGO Cytoscape plugin 35.

Results Expression of StrepII/FLAG-tagged CB1 in HEK293 cells Since the tagging of CB1 may disturb the transport of the protein to the plasma membrane, the correct subcellular distribution of StrepII/FLAG (SF)-tagged CB1 (CB1-SF) was evaluated. The localization of endogenous CB1 was shown to be activity-dependent 11

ACS Paragon Plus Environment

36

, thus

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 48

agonist and antagonist treatment was performed to induce or prevent internalization of CB1SF. HEK293 cells were transfected with pcDNA3-CB1-SF for the expression of CB1-SF. Mock-transfected cells without DNA were used as the negative control. After two days, the cells were treated with either the CB1 agonist CP55,940 [40 µM] to induce an activation and subsequent internalization of CB1, or the antagonist SR141716 [40 µM] to force membrane localization of CB1. Immunocytological analyses were then performed to determine the subcellular localization of the CB1-SF protein. HEK293 cells treated with the CB1 agonist CP55,940 mainly showed fluorescent signal for CB1 in internalized cytoplasmic vesicles (Fig. 1A), while cells treated with the CB1 antagonist SR141716 displayed a localization of the CB1 fusion protein predominantly in the cell membrane (Fig. 1B), indicating an accurate biological trafficking of the tagged receptors

37

.

No signal was detectable in the mock-transfected cells (Fig. 1C).

Specific expression of CB1-SF in hippocampal glutamatergic and GABAergic neurons To selectively target glutamatergic or GABAergic subtypes of neurons in the hippocampus, we combined an AAV vector containing a transcriptional stop sequence flanked by loxP sites (floxed Stop cassette) 38 between the CAG promoter region and the coding sequence of a Cterminal StrepII-FLAG-tagged CB1 (AAV-Stop-CB1-SF) together with neuronal subtypespecific Cre recombinase-expressing conditional CB1 knockout mouse lines (Tab. S1). The AAV was delivered via bilateral stereotactic injection into the hippocampi of either Glu-CB1KO mice, which showed a restricted expression of Cre recombinase under control of the NEX promoter in dorsal telencephalic glutamatergic neurons, or GABA-CB1-KO mice, which express Cre recombinase under control of the dlx5/6 promoter only in forebrain GABAergic interneurons

25

. Furthermore, in both mouse lines, the genomic CB1 locus was modified

whereby the CB1 open reading frame containing sequence was flanked by loxP sites. The expression of Cre recombinase therefore determines both the deletion of the endogenous CB1 (through floxed CB1 CDS) and the virally induced expression of the SF-tagged CB1 12

ACS Paragon Plus Environment

Page 13 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

(through a floxed Stop cassette). We used this strategy in order to express CB1-SF in neurons which lacked the endogenous CB1, avoiding that putative interactions of proteins with CB1-SF are not competed by the endogenous CB1 (Fig. 2A). For an in vivo expression analysis, we performed stereotactic injections of AAV-Stop-CB1-SF in the hippocampi of adult Glu-CB1-KO mice, or GABA-CB1-KO mice. 3 weeks after injection, we performed Western blot and immunohistological analysis. In the Western blot analysis of hippocampal lysates, we compared the endogenous and recombinant CB1 expression levels of AAV-Stop-CB1-SF-injected Glu-CB1-KO mice (AAV-Glu-CB1-SF) and GABA-CB1-KO mice (AAV-GABA-CB1-SF), non-injected Glu-CB1-KO and GABA-CB1-KO mice and C57BL/6N mice (Fig. 2B). Using a CB1-specific antibody, we could clearly distinguish between the endogenously expressed CB1 (~53 kDa) and the recombinant CB1SF (~64 kDa). The highest band intensity of endogenous CB1 was visible in the C57BL/6N mice. Decreased CB1 expression was observed in the Glu-CB1-KO mice and even less in the GABA-CB1-KO mice, in accordance with a significantly higher amount of CB1 in GABAergic interneurons although the higher number of glutamatergic principal neurons in the hippocampus 8. Densitometric quantification of Western blot band intensities showed approximately 3x higher protein leveles of CB1-SF in glutamatergic neurons as compared to endogenous CB1, and vice versa a slightly lower level of CB1-SF as compared to endogenous CB1 in GABAergic neurons (Fig. S1). It has to be considered that the transgenic CAG promoter is driving expression with the same intensity in both neuronal populations, whereas the endogenous CB1 expression in glutamatergic and GABAergic neurons differs to a very high degree 8. Therefore, the CB1-SF band intensity is higher in the AAV-Glu-CB1-SF mice as compared to the AAV-GABA-CB1-SF mice because of the higher number of glutamatergic neurons relative to GABAergic interneurons. Most importantly, the overall band intensities of the recombinant CB1 are comparable to those of the endogenous CB1 without vast differences that might occur using AAV-mediated overexpression under the CAG promoter. After stripping the membrane and re-incubating with a FLAG specific antibody, we

13

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

confirmed the 64 kDa band visible as the recombinant CB1-SF, again showing a stronger signal in the AAV-Glu-CB1-SF mice as compared to the AAV-GABA-CB1-SF mice (Fig. 2B). FLAG immunoreactivity was determined immunohistologically with a co-staining for the vesicular glutamate transporter (VGlut1) as a pre-synaptic marker for glutamatergic neurons in AAV-Glu-CB1-SF mice, and with vesicular GABA transporter (VGAT) as a pre-synaptic marker for GABAergic neurons in AAV-GABA-CB1-SF mice, respectively (Fig. 2C). In AAVGlu-CB1-SF mice, we observed a strong FLAG antibody signal in the pyramidal neurons of the stratum oriens and stratum radiatum of the CA1–CA3 region, as well in the mossy fibers that project from the dentate gyrus (DG) to CA3 but not in the granule cells of the DG, as it was described for NEX gene activity in mice after P10 39. In AAV-GABA-CB1-SF mice, FLAG staining was in general less intense consistent with the results obtained in the Western blot analysis. The strongest immunoreactivity was observed in the stratum moleculare and the stratum pyramidale of CA3, CA2 and lateral CA1. Consistent with this observation, Dlx5/6 was shown to be active in neocortical parvalbumin positive interneurons, which in the hippocampus can be subdivided into basket, axo-axonic, bistratified and oriens-lacunosum moleculare (O-LM) cells that innervate pyramidal cells on different subcellular regions respectively

40

. The overlap with VGAT immunoreactivity confirms the presynaptic

localization of CB1-SF on GABAergic interneurons.

Analysis of cell type-specific tandem affinity purified CB1 receptor complexes of glutamatergic and GABAergic neurons and enrichment analyses After the isolation of hippocampi from AAV-Glu-CB1-SF-, AAV-GABA-CB1-SF- and nonAAV-injected control mice, synaptosomal fractions including mitochondria as evaluated by western blot (data not shown) were prepared, and were then used as source material for three independent TAPs. Western blot analysis of the final eluates using a CB1-specific antibody showed specific bands of ~64 kDa in the AAV-Glu-CB1-SF-TAP, as well as in AAVGABA-CB1-SF-TAP, but not in the control-TAP (Fig. 3B). 14

ACS Paragon Plus Environment

Page 14 of 48

Page 15 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Using a deep-coverage proteomics approach

29

, we identified and quantified 951 specifically

enriched proteins in the AAV-Glu-CB1-SF-TAP and AAV-GABA-CB1-SF-TAP as compared to the control-TAP sample as a negative control (Tab. S2). In order to achieve a sufficient amount of protein in the samples for the mass spectrometry, 4 AAV-Glu-CB1-SF-TAPs and 8 AAV-GABA-CB1-SF-TAPs from independent experiments were pooled, respectively. Although all proteins in the dataset (Tab. S2) were detected as high confidence interactors, a STRING network analysis was performed with only the proteins with a relative molar ratio of at least 1% as compared to the bait CB1-SF protein for the following reason. STRING does not consider the abundance of the detected proteins, and CRIP1a, a previously described interactor of CB1

18, 41, 42

, was identified in our dataset as a CB1 interactor with a relative

molar ratio of close to 1% in both neuronal subtypes (Glu: 0.009, GABA: 0.008). Among these 571 proteins, the majority showed no enrichment in glutamatergic or GABAergic CB1 complexes. Specificity was defined as a ≥ 2x enrichment of a protein in one cell type as compared to the other one. 33 proteins in the AAV-Glu-CB1-SF-TAP and 56 proteins in the AAV-GABA-CB1-SF-TAP fulfilled this criterion (Fig. 3A and Tab. 1). Western blot analyses of another three TAP samples of AAV-Glu-CB1-SF-, AAV-GABA-CB1SF- and control hippocampal synaptosome preparations confirmed the presence of several selected proteins in the respective samples according to the acquired MS data (Fig. 3C). The ratio of the band intensity of an identified interactor protein to the band intensity of CB1 in a particular sample can be considered as the equivalent of the calculated relative abundance of the protein in the label-free quantification MS approach. Despite using different antibodies for each protein in the Western blot analysis, we compared the ratio of this ratio of Western blot band intensities for an interactor protein in the AAV-Glu-CB1-SF-TAP to the AAV-GABACB1-SF-TAP to the ratio of the calculated abundances of the same protein in the AAV-GluCB1-SF-TAP and AAV-GABA-CB1-SF-TAP in the label-free MS quantification. We found a high correlation (r2 value = 0.9809) (Fig. S2), thereby confirming the label-free MS quantification by Western blot analyses.

15

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 48

To gain insights into the functional properties of the identified interactors we performed a functional protein network analysis using the STRING online tool was imported into Cytoscape

32

31

. The resulting network

, with every node representing a protein and confidence

values of the protein interactions, calculated by STRING, represented as edges between the proteins. Using the ClusterMaker2 plugin, proteins were clustered into functional groups based on the interaction confidence values using the Markov Cluster Algorithm (MCL) and an inflation value of 1.8. Proteins that could not be embedded in the network by the STRING algorithm were discarded and finally, a network containing 483 proteins in 7 main clusters containing at least 15 proteins remained. Within these clusters, interaction proteins were colored in green (Glu-specific) or red (GABA-specific) to highlight neuron-subtype specific enrichment of proteins in these functional clusters, and color saturation displaying their abundance (Fig. 4 and Tab. 2) (for a close-up view of individual clusters including the visualization of intracluster edges displaying the interaction confidence values between nodes, see Fig. S3A-G). For reasons of clarity, the interaction intercluster edges between the protein nodes were illustrated in a separate network with transparency adjustment of the edges displaying the confidence values calculated by the STRING algorithm (Fig. S4). The whole network and the main clusters were analyzed in respect of Gene Ontology (GO) term enrichment. A visualized gene annotation of the entirety of all proteins within the network using the BiNGO plugin for Cytoscape and reduced GOSlim_GOA subset of GO terms has been performed to gain comprehensive insights into the distribution and function of the purified proteins (Fig. 5). Enrichment of the proteins was tested against the synaptosomal proteome, which was assessed by the same MS approach used for the detection of proteins in the TAP samples by measuring three independent samples of synaptosome preparations of AAV-Glu-CB1-SF, AAV-GABA-CB1-SF, and non-injected control mice, respectively, which served as input controls for the TAPs (Tab. S3). The cellular component (CC) ontology shows a significant enrichment of proteins annotated to the membrane. For the biological process (BP) ontology, proteins are significantly enriched in several cellular processes, in particular cell communication, multicellular organismal development, secretion and regulation

16

ACS Paragon Plus Environment

Page 17 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

of biological processes in general, as well as catabolic process and behavior. The molecular function (MF) ontology shows enrichment for proteins involved in protein binding, structural molecule activity, receptor activity, activity of hydrolases, and ion channel and transporter activity. Aiming for a better understanding of how the identified individual Glu- or GABA-specific proteins could affect the overall output of CB1 signaling in the neuronal subtypes, GO analyses using GO_FAT categories of the single main clusters were performed using the DAVID Functional Annotation Tool and compared to the GO_FAT analysis of the whole network. The two annotations with the highest significance for each category (CC, BP, MF) of the single clusters were compared to the p-values of the same annotations for the whole network, and, if also significant for the whole network GO analysis, summarized for each cluster (Tab. 3). An accumulation of Glu- or GABA-specific proteins in one cluster, such as the GABA-specific proteins in Cluster 5 (Fig. 4 and Tab. 2), which is annotated generally to potassium channel activity, might hence suggest the possibility that these cell-type specific interacting proteins might be responsible for differentially regulating CB1 signaling output in one particular cell type by acting through these functional clusters.

Discussion In the present work, we took advantage of a TAP approach for the identification of co-purified proteins using deep coverage mass spectrometry, and revealed the composition of the CB1 multiprotein complex in mouse hippocampal glutamatergic neurons and GABAergic interneurons. The mere number of 951 identified proteins together with the amount of scaffolding proteins and cytoskeletal elements suggests that not only CB1 with directly interacting proteins was isolated but also higher order multiprotein structures and potentially indirectly interacting proteins. It has also to be considered that membrane proteins and receptors undergo a complex and dynamic biogenesis, and are trafficked through different subcellular 17

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 48

compartments. In consequence, it was shown that they have a large and dynamic interactome 43. In order to assess the relevance of the proteins identified in the TAP samples, in a first step, the relative amount of purified CB1, and already known CB1-interacting proteins were screened. In both TAP samples of Glu-CB1-SF-C and GABA-CB1-SF-C, CB1 is the most abundant protein, followed by the G protein subunit Gαo with a relative molar ratio of 0.428 (Glu) and 0.474 (GABA) (from here, molar ratios of a protein to CB1 are presented in brackets after the respective protein), respectively. Since CB1 is a Gαi/o-coupled GPCR 44, with the G protein as a directly interacting protein, this depicts a first and very important proof for a valuable dataset. Furthermore, various Gβ and Gγ subunits could also be identified, with Gβ1 (Glu: 0.272, GABA: 0.219) and Gγ3 (Glu: 0.033, GABA: 0.021) being the most abundant variants. Other G proteins, which were shown to presumably couple to CB1, were also identified. These are the inhibitory G protein subunits Gαi1 (Glu: 0.216, GABA: 0.174), Gαi2 (Glu: 0.127, GABA: 0.124), and Gαi3/k (Glu: 0.050, GABA: 0.035), and members of other G protein classes, which are Gαs (Glu: 0.019, GABA: 0.018)

45

, Gαq (Glu: 0.054, GABA:

0.037), and Gα11 (Glu: 0.020, GABA: 0.017) 46. Consequently, also downstream effectors of CB1-G protein-mediated activation were identified, namely AC1 (Glu: 0.008, GABA: 0.011)

47

and several PKA subunits, with type II

subunit α being the most abundant (Glu: 0.013, GABA: 0.045). The identification of VGCC subunits β4 (Glu: 0.004, GABA: 0.006) and γ8 (Glu: 0.085, GABA: 0.062) is consistent with the description of CB1-mediated inhibition of neurotransmitter release via G proteindependent inhibition of presynaptic VGCCs. Furthermore, the CB1-mediated activation of the ERK signaling cascade is reflected by the identification of MAPK1 (Glu: 0.007, GABA: 0.006) and MAPK3 (Glu: 0.003, GABA: 0.003) 48. Other identified proteins in the TAP samples that have already been shown to be functionally involved in CB1-mediated actions are GSK3β (Glu: 0.010, GABA: 0.012) 0.072, GABA: 0.080) GABA: 0.037)

19

51

, Crip1a (Glu: 0.009, GABA: 0.008)

, PP2A (Glu: 0.017, GABA: 0.027)

18, 41, 42

49, 50

, Rab3B (Glu:

, 14-3-3β (Glu: 0.046,

52

, and Src family non-receptor tyrosine

kinase Fyn (Glu: 0.024, GABA: 0.022) 3. Other CB1-interacting proteins, such as WAVE1, 18

ACS Paragon Plus Environment

Page 19 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

NCKAP1, CYFP2, Rac1 and ABI2, were previously shown to mediate the actions of cannabinoids on the actin cytoskeleton in growth cones of cortical neurons

23

. Of these

proteins, only CYFP2 (Glu: 0.014, GABA: 0.011) and Rac1 (Glu: 0.158, GABA: 0.137) were co-purified in the TAP, but not WAVE1, NCKAP1 and ABI2. This may be a result of the preceding synaptosome preparations, which served as input for the TAPs, since in the work of Njoo et al. whole cell extracts were used. Another difference between both studies are the investigated brain areas. The interactions of proteins with CB1 in the work of Njoo et al. occurred in the mouse cortex, and therefore do not necessarily need to interact with CB1 in other brain areas, such as the hippocampus, which was investigated in this work. Considering the GO annotation analysis of the whole network containing 571 proteins with a molar ratio of > 1% to CB1, the entirety of the proteins and their GO annotations reflect known functions of CB1 and actions of this GPCR in different multiprotein signaling subsets. The significant enrichment of membranous proteins in the CC ontology reflects the role of CB1 as a membrane protein and associated signaling proteins in or at the membrane where they exert their actions. The BP ontology terms also reflect already described functions of CB1, such as cell communication or secretion reflecting the modulation of neurotransmitter release. Other terms, such as regulation of biological processes, catabolic process, or behavior also relate to described functions of CB1 but are rather vague. The MF terms, on the contrary, are specific enough to allow for a comparison with known properties of CB1. Gi/o proteins, which are the major mediators of CB1 signaling, regulate ion channel and transporter activity

4

. Protein binding and signal transducer/receptor activity are main

characteristics of GPCR signaling cascades as well as the catalytic activity of hydrolases. Enrichment in structural molecule activity indicates enrichment of further scaffolding and cytoskeletal proteins (Fig. 5). Clustering of the proteins embedded in the network led to functional groups with consistent GO annotations (Tab. 3). Cluster 1, which contains CB1 itself and its closest predicted interactors based on the calculated confidence values of the STRING database, showed

19

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

enrichment in ontology terms, which can be directly related to GPCR signaling such as GTPase activity, heterotrimeric G-protein complex, plasma membrane, intracellular signaling cascade, or G-protein coupled receptor protein signaling pathway. The next largest clusters 2 and 3 still show enrichment in nucleotide binding annotations and therefore direct signaling cascade properties regarding the MF ontology, with an additional enrichment of proteins involved in protein transport in synaptic and vesicular vesicles in cluster 3. Cluster 4 consists of proteins mainly located in mitochondria with an enrichment in GO terms related to transmembrane movement of substances. Although it has become clear recently that there is a mitochondrial location of CB1 and that it is playing a regulatory role in the respiratory chain it has been still a matter of debate among the scientific community 53-55. The co-purification of mitochondrial proteins involved in these processes using the TAP approach therefore contributes further evidence to the presence of CB1 in mitochondria. The enrichment of proteins annotated to potassium channel activity in Cluster 5 represents a more detailed view of the enrichment of proteins with ion transmembrane transporter activity and channel activity of the GO_Slim annotation of the whole network (Fig. 5). Potassium channel regulation is a fundamental property of Gi/o protein signaling in neurons, which can therefore be easily related to CB1 signaling. All but one protein in cluster 6 are keratins without any enrichment in GO terms and can be most likely disregarded regarding CB1 signaling. The proteins in Cluster 7 generally play roles in symporter and phosphatase activity in the plasma membrane, related to membranous signaling protein complexes. Taken together, the integrative network and GO analyses confirmed that the entirety of purified proteins in the TAPs reflect known functions of CB1 and thereby revealing a high degree of relevance of this dataset. In the context of these findings, with the concomitant identification of already known CB1-interacting proteins, all the identified yet unknown potential interactors in the dataset of detected proteins are interesting candidates with a potential functional relevance in CB1 signaling. In neurons, CB1 is considered to be located and to act generally in the presynaptic compartment, thereby decreasing neurotransmitter release after being activated by 20

ACS Paragon Plus Environment

Page 20 of 48

Page 21 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

endocannabinoids produced at the postsynaptic site. However, the identified interactors cover both presynaptically and postsynaptically located proteins. It has been shown that there are compartment-selective actions of CB1 in polarized cells where the receptor stably accumulates in the axonal plasma membrane after going through cycles of endocytosis and recycling in the somatodendritic compartment, involving specific intracellular pathways before being delivered to axons

56

. Whereas a CB1-dependent strong decrease of PKA activity can

be observed in the nerve terminals, somatodendritic CB1 is constitutively activated by locally produced 2-AG, thereby constitutively inhibiting the cAMP/PKA pathway and ultimately leading to its relocation to the axonal compartment

57

. The identification of interacting

proteins that are predominantly located in the postsynapse therefore does not exclude that they are bound in CB1 protein complexes in vivo, considering the activity-dependent cycles of endocytosis, recycling and transport CB1 is undergoing in neurons after its translation. The conditional Cre-expressing mouse lines drive expression under the NEX transcriptional regulatory elements in order to target dorsal telencephalic glutamatergic neurons

39

or the

Dlx5/6 enhancer elements in order to target all forebrain GABAergic interneurons

25

.

Expression of the SF-tagged CB1 under control of the NEX regulatory elements spatially resembles endogenous CB1 expression in hippocampal glutamatergic neurons, including the lack of expression in dendate gyrus granule neurons. For the GABAergic expression, the Dlx5/6 enhancer elements drive expression in all GABAergic interneurons, while CB1 is endogenously only expressed in cholecystokinin-positive (CCK+) interneurons

58, 59

. This

might lead to the detection of false positive interactions of CB1 with proteins in other interneuron subtypes (in particular in parvalbumin-positive neurons). However, this was unavoidable in our experimental approach due to limitations in the availability of Cre-driver mouse lines. The identification of CB1-interacting proteins is a first step to understand signaling and all proteins have the potential to play a role in the CB1 signaling pathway. Therefore, this study provides a list of potentially interesting target proteins whose effects have to be further investigated. 21

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Cell type-specific isolated proteins provide a basis for new hypotheses regarding differences in CB1-mediated manifestations of glutamatergic or GABAergic signaling output. These proteins could play important roles on their own by their functional properties, especially if they have a high abundance and are annotated in clusters with functions directly related to GPCR signaling. Additionally, enrichment of proteins specific for only one of the neuronal subtypes in one cluster hints at differences regarding cell type-specific CB1 signaling through the function of this cluster. In the CB1-containing Cluster 1, we observed an even distribution of AAV-Glu-CB1-SF- and AAV-GABA-CB1-SF-specific proteins, which have to be considered as candidates to change CB1 signaling in the respective cell type by their own functional properties. On the other hand, Cluster 5 clearly shows enrichment of GABA-specific proteins, which might indicate that potassium channel activity in general is more strongly influenced by CB1 in GABAergic interneurons. Taken together, we generated a powerful experimental set up for a screening of CB1interacting proteins in a spatially well-defined area, in only a subset of neurons by combining AAV-mediated overexpression of the SF-tagged CB1, TAP, and deep-coverage MS. After establishing this purification technique, this approach can be extended to other brain areas of interest by changing the injection sites of the AAV, and also to other neuronal subtypes, and glial cells (e.g., astrocytes), using different Cre-driver mouse lines. Our work outlined a comprehensive view on a large number of up to now unidentified and potentially interesting CB1-interacting proteins and functional groups of proteins in the mouse hippocampus. Although the majority of proteins were found in both neuronal subtypes, a number of neuron subtype-specific interactors may pave the way to put forward new hypotheses of how differential CB1 signaling in glutamatergic and GABAergic neurons in the hippocampus may emerge. Based on these findings, further functional studies need to be performed to reveal the functional relevance of interesting target proteins, direct protein interactions and distinct signaling pathways. Thus, we provide an extensive basis for the generation of new perspectives for endocannabinoid signaling, which is of steadily growing importance in

22

ACS Paragon Plus Environment

Page 22 of 48

Page 23 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

understanding physiological functions and potential dysregulation in pathophysiology and their clinical implications.

23

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Supporting Information Table S1. Generation of mouse model to specifically express CB1-SF in hippocampal glutamatergic and GABAergic neurons. Overview of conditional Cre recombinase expressing CB1f/f mice used in this study. Expression of Cre recombinase in a specific neuronal subtype leads to the simultaneous excision of the endogenous floxed CB1 CDS and excision of the Stop cassette of the delivered AAV, thus, AAV-mediated expression of CB1-SF.

Table S2. Full list of proteins identified in TAP samples. All proteins identified in the AAVGlu-CB1-SF-TAP and AAV-GABA-CB1-SF-TAP. Specificity was defined as a ≥ 2x enrichment of the abundance of a protein in one cell type compared to the other and indicated with green (AAV-Glu-CB1-SF-TAP) or red (AAV-GABA-CB1-SF-TAP) background. Description, detailed description of the identified protein including official gene symbol. Accession, UniProtKB accession number of the protein. Molar ratio vs CB1 GABA, molar ratio of the identified protein in the AAV-GABA-CB1-SF-TAP sample relative to CB1 in the AAV-GABA-CB1-SF-TAP sample. Molar ratio vs CB1 Glu, molar ratio of the identified protein in the AAV-Glu-CB1-SF-TAP sample relative to CB1 in the AAV-Glu-CB1-SF-TAP sample. Average molar ratio vs CB1, average of molar ratio of the identified protein relative to CB1 in the AAV-Glu-CB1-SF-TAP and AAV-GABA-CB1-SF-TAP samples. Log2 ratio GABA vs Glu, ratio of abundance relative to CB1 of identified protein in AAV-GABA-CB1-SF-TAP vs AAVGlu-CB1-SF-TAP sample on a logarithmic scale. Specificity of a protein for a cell type is defined as the log2 value exceeds or equals the absolute value of 1.

Table S3. Proteins identified in synaptosome samples. The synaptosomal proteome assessed by three individual synaptosome samples of AAV-Glu-CB1-SF- (Lysate NEX), AAV-GABA-CB1-SF- (Lysate DLX), and non-injected control mice (Lysate NC) by the same label-free quantification approach used for the TAP samples. Description, detailed 24

ACS Paragon Plus Environment

Page 24 of 48

Page 25 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

description of the identified protein including official gene symbol. Accession, UniProtKB accession number of the protein. ppm, parts per million of identified protein.

Figure S1. Densitometric quantification of the expression of recombinant CB1-SF as compared to endogenous CB1 (eCB1). Densitometric quantification of Western blot band intensities of AAV-Stop-CB1-SF-injected Glu-CB1-KO mice (Glu-CB1-SF + GABA-eCB1) and AAV-Stop-CB1-SF-injected GABA-CB1-KO mice (GABA-CB1-SF + Glu-eCB1) (n=3). Expression of CB1-SF is approximately 3x higher as compared to eCB1 in glutamatergic neurons, whereas in GABAergic interneurons the expression of CB1-SF is only slightly lower than eCB1. Importantly, the overall expression levels of the recombinant CB1 are comparable to those of the endogenous CB1 without a pronounced overexpression that might occur using AAV vectors.

Figure S2. Comparison of the abundance of proteins obtained by MS and band intensities of proteins in the Western blot analysis (Fig. 3). Glu vs GABA ratio of the ratio of the WB band intensities of a protein to CB1 is compared to the Glu vs GABA ratio of abundance of the protein calculated in the label-free quantification MS approach. Despite the strong background in the WB samples, densitometry could be performed successfully and the ratio of a protein in the Glu-CB1-SF-C-TAP to the same protein in the GABA-CB1-SF-CTAP as compared to the ratio of abundances of the same protein in the MS approach showed a good linear correlation (r2 value = 0.9809), thereby confirming the quantitative MS data.

Figures S3A-G. Close-ups of single clusters 1-7 (A-G respectively) of network analysis of proteins identified in the TAP samples (Fig. 4). Network of functionally related proteins embedded in a cluster including cell type specificity of single proteins illustrated with color

25

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

code (yellow, CB1; brown, no specificity for either cell type; green, specific for AAV-Glu-CB1SF; red, specific for AAV-GABA-CB1-SF). Abundance of single proteins by their relative molar ratio to CB1 is illustrated with a transparency gradient. Protein-protein interaction confidence values calculated by the STRING algorithm illustrated with lines between the proteins and a transparency gradient representing the confidence values.

Figure S4. Interaction of proteins between different clusters. Network with functionally related proteins embedded in clusters as shown in Fig. 4 including protein-protein interaction confidence values calculated by the STRING algorithm illustrated with dashed lines between the intercluster proteins and a transparency gradient representing the confidence values. Pronounced protein interactions are visible throughout the whole network.

Acknowledgments This study was supported by the German Research Foundation by Research Unit FOR926 (Subproject SP3) and the Forschungszentrum für Immuntherapie (FZI) of the Johannes Gutenberg University Mainz. We thank Ruth Jelinek, Andrea Conrad and Ruben Spohrer for excellent technical assistance.

26

ACS Paragon Plus Environment

Page 26 of 48

Page 27 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Figures/Tables Figure 1

Figure 1. Immunocytochemical detection of CB1 in HEK293 cells after treatment with CB1 agonist CP55,940 or antagonist SR141716. A, HEK293 cells transfected with pcDNA3-CB1-SF treated with CP55.940. The majority of CB1-SF fusion protein is internalized. B, HEK293 cells transfected with pcDNA3-CB1-SF and treated with the CB1 antagonist SR141716. The majority of CB1-SF fusion protein is located in the plasma 27

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

membrane. C, Untreated HEK293 cells transfected with pcDNA3-CB1-SF. The CB1-SF fusion protein is primarily located in the plasma membrane, but with a proportion internalized, representing an intermediate phenotype of agonist (A) and antagonist (B) treated cells. D, Non-transfected HEK293 cells; no CB1 staining detectable. Red: CB1; blue: Draq5-stained cell nuclei. Scale bar: 20 µm.

28

ACS Paragon Plus Environment

Page 28 of 48

Page 29 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Figure 2

Figure 2. Cell type-specific AAV-mediated overexpression of CB1-SF. A, Cell typespecific knockout of endogenous CB1 and simultaneous expression of recombinant CB1-SF. 29

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Glu-CB1-KO mice express Cre recombinase under control of the NEX promoter and GABACB1-KO mice under control of the dlx5/6 promoter, respectively. Subsequently, expression of Cre recombinase leads to the excision of the endogenously floxed CB1 CDS and, after stereotactic delivery, the floxed Stop cassette of AAV-Stop-CB1-SF in glutamatergic neurons or GABAergic interneurons, respectively. Thus, a concomitant expression of recombinant CB1-SF and knockout of endogenous CB1 is obtained specifically in the respective cell type. B, Western blot analysis of endogenous CB1 and recombinant CB1-SF in hippocampal homogenates of AAV-Glu-CB1-SF, AAV-GABA-CB1-SF, Glu-CB1-KO, GABA-CB1-KO and wild-type C57BL/6N mice. AAV-Glu-CB1-SF and AAV-GABA-CB1-SF show expression of recombinant CB1-SF protein detected with anti-FLAG antibody at ~64 kDa. The same bands are visible using anti-CB1 antibody showing the recombinant CB1-SF together with endogenous CB1 expressed at varying levels in the different mouse lines at ~53 kDa whereas the CB1-SF-specific signal appears at ~64 kDa using anti-CB1 antibody in accordance to the anti-FLAG signal. Highest band intensity of endogenous CB1 was visible in wild-type C57BL/6N mice. Decreased CB1 expression was observed in the Glu-CB1-KO mice and even less in the GABA-CB1-KO mice due to the significantly higher loss of CB1 in GABAergic interneurons compared to glutamatergic neurons. In contrast to the endogenous CB1 expression levels, the recombinant AAV-mediated CB1-SF levels are inverted because of the higher number of glutamatergic principal cells compared to GABAergic interneurons and AAV-mediated CB1-SF expression under control of the same CAG promoter. Unspecific background bands at ~62 kDa and ~73 kDa appear above and below the CB1-SF-specific band in the anti-CB1 blot. Actin is shown for comparison of total protein amount used for each sample. One lane of the blot was removed as indicated by discontinued lines. C, SFtagged CB1 expression in the hippocampus of AAV-Glu-CB1-SF and AAV-GABA-CB1-SF mice. AAV-Stop-CB1-SF was bilaterally injected in the hippocampus of Glu-CB1-KO mice and GABA-CB1-KO mice, respectively, to obtain cell type-specific expression of CB1-SF in glutamatergic neurons (AAV-Glu-CB1-SF) and GABAergic interneurons (AAV-GABA-CB1SF), respectively. FLAG immunoreactivity and presynaptic location was determined together

30

ACS Paragon Plus Environment

Page 30 of 48

Page 31 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

with anti-VGlut1 antibody as a pre-synaptic marker for glutamatergic neurons in AAV-GluCB1-SF mice, and with anti-VGAT antibody as a presynaptic marker for GABAergic neurons in AAV-GABA-CB1-SF mice. AAV-Glu-CB1-SF mice showed a strong FLAG immunostaining in the glutamatergic neurons of the stratum oriens and stratum radiatum of the CA1 – CA3 region, as well in the mossy fibers that project from the dentate gyrus (DG) to CA3. In AAVGABA-CB1-SF mice, FLAG staining is less intense as compared to AAV-Glu-CB1-SF due to the lower amount of GABAergic interneurons as compared to glutamatergic neurons. The strongest immunoreactivity is visible in the stratum moleculare and the stratum pyramidale of CA3, CA2 and lateral CA1. Higher magnification of CA1 in AAV-Glu-CB1-SF and DG in AAVGABA-CB1-SF (areas indicated by white boxes in overview) showing co-expression of FLAG and VGlut1 or VGAT respectively as indicated by white arrows. Green: FLAG; red: Vglut1 (left: AAV-Glu-CB1-SF), VGAT (right: AAV-GABA-CB1-SF); blue: DAPI-stained cell nuclei. StrO, stratum oriens; Pyr, pyramidal cell layer; StrR, stratum radiatum; StrL, stratum lacunosum; StrM, stratum moleculare; GC, granule cell layer; Hil, hilus. Scale bar: 500 µm.

31

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3

Figure 3. Identification of proteins in hippocampal TAP eluates from AAV-Glu-CB1-SF, AAV-GABA-CB1-SF and non-injected control mice. A, Dot plot of identified proteins in TAP samples with a molar ratio of at least 1% to CB1. 571 proteins represented as squares with a molar ratio of ≥ 1% to CB1 protein were co-purified altogether in the AAV-Glu-CB1-SF-

32

ACS Paragon Plus Environment

Page 32 of 48

Page 33 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

TAP and AAV-GABA-CB1-SF-TAP. x-axis, molar ratio to CB1; y-axis, logarithmic ratio of abundance of identified protein relative to CB1 in AAV-Glu-CB1-SF to abundance of identified protein relative to CB1 in AAV-GABA-CB1-SF. Specificity is defined as ≥ 2x enrichment of an identified protein in one cell type as compared to the other (brown, not specific for either cell type; green, specific for AAV-Glu-CB1-SF-TAP; red, specific for AAVGABA-CB1-SF-TAP; yellow, CB1). B, Western blot analysis of TAP eluates from AAV-GluCB1-SF, AAV-GABA-CB1-SF and non-injected control mice. TAP eluates of hippocampal synaptosome preparations showing the CB1-SF-specific band at ~64 kDa in AAV-Glu-CB1SF-TAP and AAV-GABA-CB1-SF-TAP samples, but not in control-TAP sample. Figure is assembled from three different blots as indicated by discontinued lines. C, Western blot analysis of TAP eluates from AAV-Glu-CB1-SF, AAV-GABA-CB1-SF and non-injected control mice. Syntaxin (Syntaxin), guanine nucleotide-binding protein G(z) subunit alpha (Gz), vesicular inhibitory amino acid transporter (VGAT), and receptor-type tyrosine-protein phosphatase σ (PTPRS) were identified in the AAV-Glu-CB1-SF-TAP and AAV-GABA-CB1SF-TAP in the MS-based identification approach and were confirmed using specific antibodies against the respective proteins by Western blot analysis. Synaptosomal preparations, taken as input controls, showed specific bands for each protein at the respective molecular weights. TAP eluates showed specific bands only in the AAV-Glu-CB1SF-TAP and AAV-GABA-CB1-SF-TAP but not in the control-TAP for the respective proteins with band intensities apparently in accordance with their abundance calculated by the labelfree quantification MS approach. All bands shown derive from the same membrane. Image acquisition had to be adjusted separately for synaptosomes and TAP samples due to huge differences of protein concentrations in both preparations but is consistent within these groups. Molar ratio to CB1 in MS data for comparison: Syntaxin, 0.209 (AAV-Glu-CB1-SF), 0.254 (AAV-GABA-CB1-SF); Gz, 0.141 (AAV-Glu-CB1-SF), 0.101 (AAV-GABA-CB1-SF); VGAT, 0.005 (AAV-Glu-CB1-SF), 0.107 (AAV-GABA-CB1-SF); PTPRS, 0.122 (AAV-GluCB1-SF), 0.079 (AAV-GABA-CB1-SF).

33

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4

Figure 4. Network analysis of proteins identified in the TAP samples. Functional protein network analysis based on all proteins identified in the TAP samples with a molar ratio of at 34

ACS Paragon Plus Environment

Page 34 of 48

Page 35 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

least 1% to CB1 using STRING. Of 571 proteins, the analysis of protein-protein interactions derived from the STRING database lead to the generation of a network containing 483 embedded proteins displayed with their respective official gene symbol using a medium threshold for protein-protein interaction confidence value (0.400). Proteins were clustered based on the interaction confidence values using the Markov Cluster Algorithm (MCL) with an inflation value of 1.8. 7 main clusters containing at least 15 proteins were obtained. For a close-up of single clusters including visualization of intracluster protein interaction confidence values see Fig. S3A-G. Network with functionally related proteins embedded in clusters including cell type specificity of single proteins illustrated with color code (yellow, CB1; brown, no specificity for either cell type; green, specific for AAV-Glu-CB1-SF; red, specific for AAV-GABA-CB1-SF). Abundance of single proteins by their relative molar ratio to CB1 is illustrated with a transparency gradient.

35

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5

36

ACS Paragon Plus Environment

Page 36 of 48

Page 37 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Figure 5. Gene Ontology (GO) analysis of proteins identified in the TAP samples. GO term enrichment using the reduced GOSlim_GOA subset of GO terms based on all proteins identified in the TAP samples with a molar ratio of at least 1% to CB1 against the synaptosomal proteome as background set using the BiNGO plugin for Cytoscape. Visualization of enrichment of GO terms is based on the amount of proteins annotated in a GO term illustrated with the size of the circles and significance of the GO term enrichment visualized with a color gradient ranging from yellow (p = 0.05) to orange (p ≤ 5E-7). Analysis was separately performed for biological process (top), molecular function (middle) and cellular component (bottom).

37

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 1

38

ACS Paragon Plus Environment

Page 38 of 48

Page 39 of 48

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 39

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 1. AAV-Glu-CB1-SF-TAP- and AAV-GABA-CB1-SF-TAP-specific proteins. All cell type-specific proteins identified in the AAV-Glu-CB1-SF-TAP and AAV-GABA-CB1-SF-TAP with a relative molar ratio of at least 1% to CB1. Specificity was defined as a ≥ 2x enrichment of the abundance of a protein in one cell type compared to the other. gene name, official gene symbol of the respective protein; molar ratio to CB1, molar ratio of the protein relative to CB1 either in the AAV-Glu-CB1-SF-TAP or in the AAV-GABA-CB1-SF-TAP respectively; UniProt entry, UniProtKB/Swiss-Prot entry name of the protein; description, detailed description of the identified protein. a, AAV-Glu-CB1-SF-TAP-specific proteins. b, AAVGABA-CB1-SF-TAP-specific proteins.

40

ACS Paragon Plus Environment

Page 40 of 48

Page 41 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Table 2

Table 2. AAV-Glu-CB1-SF-TAP- and AAV-GABA-CB1-SF-TAP-specific proteins in the 7 main clusters after network analysis of the identified proteins in the TAP samples using the STRING database and MCL clustering. For each of the main clusters AAV-GluCB1-SF-TAP-specific proteins are highlighted in green and sorted from top to bottom based on their relative molar ratios to CB1 which are displayed in the column to the right next to 41

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

their official gene symbols. AAV-GABA-CB1-SF-TAP-specific proteins are highlighted in red with their associated molar ratios to CB1 in the column to their right.

42

ACS Paragon Plus Environment

Page 42 of 48

Page 43 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Table 3.

Table 3. Gene Ontology (GO) analysis of the 7 main clusters obtained by the network analysis of the identified proteins in the TAP samples using the STRING database and 43

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

MCL clustering. GO analyses using GO_FAT categories for molecular function (MF), cellular component (CC) and biological process (BP) of all proteins embedded in the network and for its single main clusters were performed using the DAVID Functional Annotation Tool. The two GO term annotations with the highest significance for each category (CC, BP, MF) of the single clusters were summarized for each cluster if the GO terms were also significant for the whole network GO analysis. Cluster/Category, analyzed cluster and category for MF, CC or BP; GO term, enrichment of proteins in respective cluster and category for GO term; p value, significance for GO term displayed by p value (Benjamini corrected) calculated for GO analysis of all proteins in the network.

References (1)

Marsicano, G.; Kuner, R., Anatomical distribution of receptor, ligands and enzymes in the brain and in the spinal cord: circuitries and neurochemistry. In Cannabinoids and the brain, Köfalvi, A., Ed. Springer: New York, 2008; pp 161-201. (2) Bouaboula, M.; Poinot-Chazel, C.; Bourrie, B.; Canat, X.; Calandra, B.; Rinaldi-Carmona, M.; Le Fur, G.; Casellas, P., Activation of mitogen-activated protein kinases by stimulation of the central cannabinoid receptor CB1. Biochem J 1995, 312 ( Pt 2), 637-41. (3) Derkinderen, P.; Valjent, E.; Toutant, M.; Corvol, J. C.; Enslen, H.; Ledent, C.; Trzaskos, J.; Caboche, J.; Girault, J. A., Regulation of extracellular signal-regulated kinase by cannabinoids in hippocampus. J Neurosci 2003, 23 (6), 2371-82. (4) Turu, G.; Hunyady, L., Signal transduction of the CB1 cannabinoid receptor. J Mol Endocrinol 2010, 44 (2), 75-85. (5) Dalton, G. D.; Bass, C. E.; Van Horn, C. G.; Howlett, A. C., Signal transduction via cannabinoid receptors. CNS Neurol Disord Drug Targets 2009, 8 (6), 422-31. (6) Lutz, B.; Marsicano, G.; Maldonado, R.; Hillard, C. J., The endocannabinoid system in guarding against fear, anxiety and stress. Nat Rev Neurosci 2015, 16 (12), 705-18. (7) Soltesz, I.; Alger, B. E.; Kano, M.; Lee, S. H.; Lovinger, D. M.; Ohno-Shosaku, T.; Watanabe, M., Weeding out bad waves: towards selective cannabinoid circuit control in epilepsy. Nat Rev Neurosci 2015, 16 (5), 264-77. (8) Steindel, F.; Lerner, R.; Haring, M.; Ruehle, S.; Marsicano, G.; Lutz, B.; Monory, K., Neuron-type specific cannabinoid-mediated G protein signalling in mouse hippocampus. J Neurochem 2013, 124 (6), 795-807. (9) Mechoulam, R.; Parker, L. A., The endocannabinoid system and the brain. Annu Rev Psychol 2013, 64, 21-47. (10) Parsons, L. H.; Hurd, Y. L., Endocannabinoid signalling in reward and addiction. Nat Rev Neurosci 2015, 16 (10), 579-94. 44

ACS Paragon Plus Environment

Page 44 of 48

Page 45 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

(11) Maccarrone, M.; Guzman, M.; Mackie, K.; Doherty, P.; Harkany, T., Programming of neural cells by (endo)cannabinoids: from physiological rules to emerging therapies. Nat Rev Neurosci 2014, 15 (12), 786-801. (12) Navarrete, M.; Araque, A., Endocannabinoids potentiate synaptic transmission through stimulation of astrocytes. Neuron 2010, 68 (1), 113-26. (13) Metna-Laurent, M.; Marsicano, G., Rising stars: modulation of brain functions by astroglial type-1 cannabinoid receptors. Glia 2015, 63 (3), 353-64. (14) Smith, T. H.; Sim-Selley, L. J.; Selley, D. E., Cannabinoid CB1 receptor-interacting proteins: novel targets for central nervous system drug discovery? Br J Pharmacol 2010, 160 (3), 454-66. (15) Rigaut, G.; Shevchenko, A.; Rutz, B.; Wilm, M.; Mann, M.; Seraphin, B., A generic protein purification method for protein complex characterization and proteome exploration. Nat Biotechnol 1999, 17 (10), 1030-2. (16) Bergmayr, C.; Thurner, P.; Keuerleber, S.; Kudlacek, O.; Nanoff, C.; Freissmuth, M.; Gruber, C. W., Recruitment of a cytoplasmic chaperone relay by the A2A adenosine receptor. J Biol Chem 2013, 288 (40), 28831-44. (17) Daulat, A. M.; Maurice, P.; Froment, C.; Guillaume, J. L.; Broussard, C.; Monsarrat, B.; Delagrange, P.; Jockers, R., Purification and identification of G protein-coupled receptor protein complexes under native conditions. Mol Cell Proteomics 2007, 6 (5), 835-44. (18) Niehaus, J. L.; Liu, Y.; Wallis, K. T.; Egertova, M.; Bhartur, S. G.; Mukhopadhyay, S.; Shi, S.; He, H.; Selley, D. E.; Howlett, A. C.; Elphick, M. R.; Lewis, D. L., CB1 cannabinoid receptor activity is modulated by the cannabinoid receptor interacting protein CRIP 1a. Mol Pharmacol 2007, 72 (6), 1557-66. (19) Jung, H. W.; Park, I.; Ghil, S., Cannabinoid receptor activation inhibits cell cycle progression by modulating 14-3-3beta. Cell Mol Biol Lett 2014, 19 (3), 347-60. (20) Hudson, B. D.; Hebert, T. E.; Kelly, M. E., Physical and functional interaction between CB1 cannabinoid receptors and beta2-adrenoceptors. Br J Pharmacol 2010, 160 (3), 627-42. (21) Ferre, S.; Lluis, C.; Justinova, Z.; Quiroz, C.; Orru, M.; Navarro, G.; Canela, E. I.; Franco, R.; Goldberg, S. R., Adenosine-cannabinoid receptor interactions. Implications for striatal function. Br J Pharmacol 2010, 160 (3), 443-53. (22) Fatahi, Z.; Assar, N.; Mahmoudi, D.; Pahlevani, P.; Moradi, M.; Haghparast, A., Functional interaction between the orexin-1 and CB1 receptors within the nucleus accumbens in the conditioned place preference induced by the lateral hypothalamus stimulation. Pharmacol Biochem Behav 2015, 132, 42-48. (23) Njoo, C.; Agarwal, N.; Lutz, B.; Kuner, R., The Cannabinoid Receptor CB1 Interacts with the WAVE1 Complex and Plays a Role in Actin Dynamics and Structural Plasticity in Neurons. PLoS Biol 2015, 13 (10), e1002286. (24) Gloeckner, C. J.; Boldt, K.; Schumacher, A.; Roepman, R.; Ueffing, M., A novel tandem affinity purification strategy for the efficient isolation and characterisation of native protein complexes. Proteomics 2007, 7 (23), 4228-34. (25) Monory, K.; Massa, F.; Egertova, M.; Eder, M.; Blaudzun, H.; Westenbroek, R.; Kelsch, W.; Jacob, W.; Marsch, R.; Ekker, M.; Long, J.; Rubenstein, J. L.; Goebbels, S.; Nave, K. A.; During, M.; Klugmann, M.; Wolfel, B.; Dodt, H. U.; Zieglgansberger, W.; Wotjak, C. T.; Mackie, K.; Elphick, M. R.; Marsicano, G.; Lutz, B., The endocannabinoid system controls key epileptogenic circuits in the hippocampus. Neuron 2006, 51 (4), 455-66. (26) Chen, C.; Okayama, H., High-efficiency transformation of mammalian cells by plasmid DNA. Mol Cell Biol 1987, 7 (8), 2745-52. (27) During, M. J.; Young, D.; Baer, K.; Lawlor, P.; Klugmann, M., Development and optimization of adeno-associated virus vector transfer into the central nervous system. Methods Mol Med 2003, 76, 221-36. (28) Carlin, R. K.; Grab, D. J.; Cohen, R. S.; Siekevitz, P., Isolation and characterization of postsynaptic densities from various brain regions: enrichment of different types of postsynaptic densities. J Cell Biol 1980, 86 (3), 831-45.

45

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(29) Distler, U.; Kuharev, J.; Navarro, P.; Levin, Y.; Schild, H.; Tenzer, S., Drift time-specific collision energies enable deep-coverage data-independent acquisition proteomics. Nat Methods 2014, 11 (2), 167-70. (30) Distler, U.; Schmeisser, M. J.; Pelosi, A.; Reim, D.; Kuharev, J.; Weiczner, R.; Baumgart, J.; Boeckers, T. M.; Nitsch, R.; Vogt, J.; Tenzer, S., In-depth protein profiling of the postsynaptic density from mouse hippocampus using data-independent acquisition proteomics. Proteomics 2014, 14 (21-22), 2607-13. (31) Jensen, L. J.; Kuhn, M.; Stark, M.; Chaffron, S.; Creevey, C.; Muller, J.; Doerks, T.; Julien, P.; Roth, A.; Simonovic, M.; Bork, P.; von Mering, C., STRING 8--a global view on proteins and their functional interactions in 630 organisms. Nucleic Acids Res 2009, 37 (Database issue), D412-6. (32) Cline, M. S.; Smoot, M.; Cerami, E.; Kuchinsky, A.; Landys, N.; Workman, C.; Christmas, R.; Avila-Campilo, I.; Creech, M.; Gross, B.; Hanspers, K.; Isserlin, R.; Kelley, R.; Killcoyne, S.; Lotia, S.; Maere, S.; Morris, J.; Ono, K.; Pavlovic, V.; Pico, A. R.; Vailaya, A.; Wang, P. L.; Adler, A.; Conklin, B. R.; Hood, L.; Kuiper, M.; Sander, C.; Schmulevich, I.; Schwikowski, B.; Warner, G. J.; Ideker, T.; Bader, G. D., Integration of biological networks and gene expression data using Cytoscape. Nat Protoc 2007, 2 (10), 2366-82. (33) Morris, J. H.; Apeltsin, L.; Newman, A. M.; Baumbach, J.; Wittkop, T.; Su, G.; Bader, G. D.; Ferrin, T. E., clusterMaker: a multi-algorithm clustering plugin for Cytoscape. BMC Bioinformatics 2011, 12, 436. (34) Huang da, W.; Sherman, B. T.; Lempicki, R. A., Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009, 4 (1), 44-57. (35) Maere, S.; Heymans, K.; Kuiper, M., BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics 2005, 21 (16), 3448-9. (36) Hsieh, C.; Brown, S.; Derleth, C.; Mackie, K., Internalization and recycling of the CB1 cannabinoid receptor. J Neurochem 1999, 73 (2), 493-501. (37) Leterrier, C.; Bonnard, D.; Carrel, D.; Rossier, J.; Lenkei, Z., Constitutive endocytic cycle of the CB1 cannabinoid receptor. J Biol Chem 2004, 279 (34), 36013-21. (38) Guggenhuber, S.; Monory, K.; Lutz, B.; Klugmann, M., AAV vector-mediated overexpression of CB1 cannabinoid receptor in pyramidal neurons of the hippocampus protects against seizureinduced excitoxicity. PLoS One 2010, 5 (12), e15707. (39) Goebbels, S.; Bormuth, I.; Bode, U.; Hermanson, O.; Schwab, M. H.; Nave, K. A., Genetic targeting of principal neurons in neocortex and hippocampus of NEX-Cre mice. Genesis 2006, 44 (12), 611-21. (40) Klausberger, T.; Marton, L. F.; O'Neill, J.; Huck, J. H.; Dalezios, Y.; Fuentealba, P.; Suen, W. Y.; Papp, E.; Kaneko, T.; Watanabe, M.; Csicsvari, J.; Somogyi, P., Complementary roles of cholecystokinin- and parvalbumin-expressing GABAergic neurons in hippocampal network oscillations. J Neurosci 2005, 25 (42), 9782-93. (41) Guggenhuber, S.; Alpar, A.; Chen, R.; Schmitz, N.; Wickert, M.; Mattheus, T.; Harasta, A. E.; Purrio, M.; Kaiser, N.; Elphick, M. R.; Monory, K.; Kilb, W.; Luhmann, H. J.; Harkany, T.; Lutz, B.; Klugmann, M., Cannabinoid receptor-interacting protein Crip1a modulates CB1 receptor signaling in mouse hippocampus. Brain Struct Funct 2016, 221 (4), 2061-74. (42) Blume, L. C.; Eldeeb, K.; Bass, C. E.; Selley, D. E.; Howlett, A. C., Cannabinoid receptor interacting protein (CRIP1a) attenuates CB1R signaling in neuronal cells. Cell Signal 2015, 27 (3), 716-26. (43) Pankow, S.; Bamberger, C.; Calzolari, D.; Martinez-Bartolome, S.; Lavallee-Adam, M.; Balch, W. E.; Yates, J. R., 3rd, F508 CFTR interactome remodelling promotes rescue of cystic fibrosis. Nature 2015, 528 (7583), 510-6. (44) Howlett, A. C.; Qualy, J. M.; Khachatrian, L. L., Involvement of Gi in the inhibition of adenylate cyclase by cannabimimetic drugs. Mol Pharmacol 1986, 29 (3), 307-13. (45) Glass, M.; Felder, C. C., Concurrent stimulation of cannabinoid CB1 and dopamine D2 receptors augments cAMP accumulation in striatal neurons: evidence for a Gs linkage to the CB1 receptor. J Neurosci 1997, 17 (14), 5327-33.

46

ACS Paragon Plus Environment

Page 46 of 48

Page 47 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

(46) Lauckner, J. E.; Hille, B.; Mackie, K., The cannabinoid agonist WIN55,212-2 increases intracellular calcium via CB1 receptor coupling to Gq/11 G proteins. Proc Natl Acad Sci U S A 2005, 102 (52), 19144-9. (47) Rhee, M. H.; Bayewitch, M.; Avidor-Reiss, T.; Levy, R.; Vogel, Z., Cannabinoid receptor activation differentially regulates the various adenylyl cyclase isozymes. J Neurochem 1998, 71 (4), 1525-34. (48) Kano, M.; Ohno-Shosaku, T.; Hashimotodani, Y.; Uchigashima, M.; Watanabe, M., Endocannabinoid-mediated control of synaptic transmission. Physiol Rev 2009, 89 (1), 309-80. (49) Gomez del Pulgar, T.; Velasco, G.; Guzman, M., The CB1 cannabinoid receptor is coupled to the activation of protein kinase B/Akt. Biochem J 2000, 347 (Pt 2), 369-73. (50) Ozaita, A.; Puighermanal, E.; Maldonado, R., Regulation of PI3K/Akt/GSK-3 pathway by cannabinoids in the brain. J Neurochem 2007, 102 (4), 1105-14. (51) Tsetsenis, T.; Younts, T. J.; Chiu, C. Q.; Kaeser, P. S.; Castillo, P. E.; Sudhof, T. C., Rab3B protein is required for long-term depression of hippocampal inhibitory synapses and for normal reversal learning. Proc Natl Acad Sci U S A 2011, 108 (34), 14300-5. (52) Dalton, G. D.; Howlett, A. C., Cannabinoid CB1 receptors transactivate multiple receptor tyrosine kinases and regulate serine/threonine kinases to activate ERK in neuronal cells. Br J Pharmacol 2012, 165 (8), 2497-511. (53) Benard, G.; Massa, F.; Puente, N.; Lourenco, J.; Bellocchio, L.; Soria-Gomez, E.; Matias, I.; Delamarre, A.; Metna-Laurent, M.; Cannich, A.; Hebert-Chatelain, E.; Mulle, C.; OrtegaGutierrez, S.; Martin-Fontecha, M.; Klugmann, M.; Guggenhuber, S.; Lutz, B.; Gertsch, J.; Chaouloff, F.; Lopez-Rodriguez, M. L.; Grandes, P.; Rossignol, R.; Marsicano, G., Mitochondrial CB(1) receptors regulate neuronal energy metabolism. Nat Neurosci 2012, 15 (4), 558-64. (54) Hebert-Chatelain, E.; Reguero, L.; Puente, N.; Lutz, B.; Chaouloff, F.; Rossignol, R.; Piazza, P. V.; Benard, G.; Grandes, P.; Marsicano, G., Cannabinoid control of brain bioenergetics: Exploring the subcellular localization of the CB1 receptor. Mol Metab 2014, 3 (4), 495-504. (55) Fisar, Z.; Singh, N.; Hroudova, J., Cannabinoid-induced changes in respiration of brain mitochondria. Toxicol Lett 2014, 231 (1), 62-71. (56) Leterrier, C.; Laine, J.; Darmon, M.; Boudin, H.; Rossier, J.; Lenkei, Z., Constitutive activation drives compartment-selective endocytosis and axonal targeting of type 1 cannabinoid receptors. J Neurosci 2006, 26 (12), 3141-53. (57) Ladarre, D.; Roland, A. B.; Biedzinski, S.; Ricobaraza, A.; Lenkei, Z., Polarized cellular patterns of endocannabinoid production and detection shape cannabinoid signaling in neurons. Front Cell Neurosci 2014, 8, 426. (58) Katona, I.; Sperlagh, B.; Sik, A.; Kafalvi, A.; Vizi, E. S.; Mackie, K.; Freund, T. F., Presynaptically located CB1 cannabinoid receptors regulate GABA release from axon terminals of specific hippocampal interneurons. J Neurosci 1999, 19 (11), 4544-58. (59) Marsicano, G.; Lutz, B., Expression of the cannabinoid receptor CB1 in distinct neuronal subpopulations in the adult mouse forebrain. Eur J Neurosci 1999, 11 (12), 4213-25.

47

ACS Paragon Plus Environment

Journal of Proteome Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For TOC only 85x47mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 48 of 48