UNIVERSIDAD COMPLUTENSE DE MADRID FACULTAD DE CIENCIAS BIOLÓGICAS TESIS DOCTORAL GAP43: una nueva proteína interactora del receptor CB1 cannabinoide MEMORIA PARA OPTAR AL GRADO DE DOCTOR PRESENTADA POR Irene Berenice Maroto Martínez Director Manuel Guzman Pastor Madrid © Irene Berenice Maroto Martínez, 2021 UNIVERSIDAD COMPLUTENSE DE MADRID FACULTAD DE CIENCIAS BIOLÓGICAS TESIS DOCTORAL GAP43: UN NUEVA PROTEÍNA INTERACTORA DEL RECEPTOR CB1 CANNABINOIDE MEMORIA PARA OPTAR AL GRADO DE DOCTOR PRESENTADA POR IRENE BERENICE MAROTO MARTÍNEZ DIRECTOR MANUEL GUZMÁN PASTOR UNIVERSIDAD COMPLUTENSE DE MADRID FACULTAD DE CIENCIAS BIOLÓGICAS TESIS DOCTORAL GAP43: Una nueva protéına interactora del receptor CB1 cannabinoide Irene Berenice Maroto Mart́ınez Director de la Tesis Doctoral Manuel Guzmán Pastor COMPLUTENSE UNIVERSITY OF MADRID FACULTY OF BIOLOGY DOCTORAL THESIS GAP43: A new cannabinoid CB1 receptor-interacting protein Irene Berenice Maroto Mart́ınez PhD Supervisor Manuel Guzmán Pastor A los que buscan aunque no encuentren a los que avanzan aunque se pierdan No te rindas, por favor, no cedas, aunque el frio queme, aunque el sol se esconda, aun hay fuego en tu alma, aun hay vida en tus sueños porque cada dia es un comienzo nuevo. M. Benedetti De vez en cuando vale la pena salirse del camino, sumergirse en un bosque. Encontrarás cosas que nunca hab́ıas visto. Alexander Graham Bell 3 Acknowledgements Son much́ısimas todas las personas que formaron parte de la construcción de este proyecto vital que es la tesis durante los últimos cinco años, trataré de ser breve, pues al final, la tesis impregna casi todos los aspectos de una vida. Yo empecé siendo una niña y finalizo el proceso siendo adulta, he experimentado un crecimiento no solo cient́ıfico sino también de maduración personal. Empezaré agradeciendo a mi jefe Manolo, desde aquella entrevista inicial que más bien me hizo sentir como una charla amigable, hasta el d́ıa de hoy, por su paciencia y su sonrisa en todos los momentos, porque le he dado mucho la lata durante estos años con el proyecto, con las alergias e inspecciones. . . gracias por tu accesibilidad y cercańıa, por enseñarme a dar más de mı́ misma, a ser independiente y a proponer ideas cient́ıficas. Y tras él también a todo nuestro grupo de investigación, que empezó siendo grupo Huntington, donde di mis primeros pasos con Andrew a la que segúıa como a mama pato por todo el laboratorio, y con Reichel, siempre dispuesta a ayudar, a responder una duda, a acompañar y a réır, han sido mis dos hermanas mayores del labo. Evita siempre sacando las castañas del fuego y Helen, tan eficaz con esas cuantis que me salvaron y la circense Blasquis. Ahora el grupo evolucionó a ser el grupo Interactors donde tengo que agradecer a Carlos porque sin él este proyecto de tesis no habŕıa sido literalmente posible, por todo nuestro trabajo diario codo con codo, por ser mi recordadora y mi enciclopedia hasta el agotamiento, eres un fuera de serie amigo. Nuestra nueva incorporación Alba, tienes una capacidad incréıble, y nuestros nuevos técnicos mis gracias por todo el trabajo con esos malditos ratones. No me olvido de nuestro brazo del proyecto en qúımicas, el headmaster Nacho y los jóvenes cient́ıficos (cada vez menos jóvenes y más cient́ıficos), donde se inició todo este prometedor proyecto. Pero además he compartido laboratorio con el grupo de los neuros, el grupo mamas, el grupo gliomas, y el más reciente grupo de oligos, una gran familia en un espacio reducido donde la colaboración y la ayuda ha sido siempre la tónica, donde tenemos relaciones de horizontalidad y confianza con los jefes Isma, Cris, Guille y Javi, y donde se han compartido infinidad de risas y también de llantos, de relaciones, de sentimientos, de viajes, de donde salen fuertes amistades, sois muchos Alba, Juan, Esti, Samu, . . . no me olvido de ninguno, mil gracias por todo. No puedo olvidarme de mis inicios en la ciencia en la Facultad de Medicina de la Autónoma haciendo mi TFM y de donde no me he desligado nunca y he seguido colaborando, donde conoćı al amigo feo e inseparable (como le gustaba decir a Laura) que es la electrofisioloǵıa. A David, el jefe, le estoy muy agradecida por su coaching, y por creer en mi siempre, y a Laura y Jose (hola bonitos) con los que se creó una conexión casi instantánea entre risas y locura y que se han convertido en mis grandes amigos, hemos compartido y seguimos compartiendo momentos vitales en lo cient́ıfico y en lo personal. Esto me lleva a mi estancia en Nueva York. Agradezco al jefe, Pablo Castillo por la cálida primera acogida y el sentimiento de grupo que genera, y sin duda por la 4 segunda casi como a una cuarta hija (como dećıa Mónica) en momentos de pandemia mundial tan inciertos y desoladores, gran parte de mis enseres tras la estampida siguen en su garage. Gracias por la oportunidad y la motivación cient́ıfica. Y Coralie Berthoux, Coralina, colaboradora y amiga, agradezco tu enorme generosidad, tus planings en hojas de papel, las confesiones en pandemia, el trabajo hasta altas horas de la noche en el labo y la “social life” hasta altas horas de la noche en la City. También me gustaŕıa aprovechar para agradecer a todos los colaboradores de este proyecto, en especial a Estefania, nuestra extensión del grupo en Barcelona. El inicio de toda esta carrera cient́ıfica corresponde a mi familia de qúımicos, mi hermano doctor como gúıa, y mis padres, profesores de qúımica en todas sus vertientes me enseñaron a dar mis primeros pasos, pusieron en mi el granito de la curiosidad que hace al cient́ıfico y me llena de orgullo que puedan verme hoy conver- tida en doctora. Por último, no me olvido de mis amigos del alma, mi red, que han escuchado, compartido, apoyado, levantado todo este tiempo. Contribuciones espe- ciales, Mulero al diseño de portada, y, dejado a propósito para el final, el diseñador de formato de esta tesis, experto en la bibliograf́ıa, tú ya sabes quién eres (Miguel), mi alegŕıa diaria y sustento de barro, el resto lo dejo para la intimidad. 5 Contents 9 13 16 Abbreviations Resumen Abstract Introduction 18 1.1 The endocannabinoid system . . . . . . . . . . . . . . . . . . . . . . . 19 1.1.1 Cannabinoid receptors: CB1R, CB2R and others . . . . . . . . 19 1.1.2 Cannabinoids and endocannabinoids . . . . . . . . . . . . . . 22 1.1.3 Enzymes for eCB synthesis and degradation . . . . . . . . . . 23 1.2 CB1 Receptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 1.2.1 Expression in the adult brain . . . . . . . . . . . . . . . . . . 25 The hippocampal formation. Dentate gyrus and mossy cells . 27 Detailed CB1R expression on hippocampus and DG . . . . . . 30 1.2.2 Synaptic plasticity mediated by CB1R . . . . . . . . . . . . . 32 eCB-STD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 eCB-LTD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 1.2.3 Biological functions of CB1R . . . . . . . . . . . . . . . . . . . 37 Cannabinoids as therapeutic agents . . . . . . . . . . . . . . . 40 1.2.4 Context-dependent signaling by CB1R . . . . . . . . . . . . . 41 1.2.5 CB1R-interacting proteins . . . . . . . . . . . . . . . . . . . . 44 49 50 Aims Materials and Methods Results 67 4.1 Results of Aim 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.2 Results of Aim 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 101 113 Discussion Conclusions References 114 6 List of Figures 1.1 Cryo-EM Structures of the CB1R and CB2R coupled to Gi protein. . 21 1.2 Major phytocannabinoids and eCBs. . . . . . . . . . . . . . . . . . . 23 1.3 Main biosynthesis and degradation pathways in eCBs signalling. . . . 25 1.4 Expression pattern of CB1R protein and mRNA in the mice’s CNS. . 26 1.5 The hippocampal circuitry. . . . . . . . . . . . . . . . . . . . . . . . . 28 1.6 Distant and local MC-GC circuitry in ipsilateral DG. . . . . . . . . . 30 1.7 CB1R expression in hippocampus and DG. . . . . . . . . . . . . . . . 32 1.8 Molecular mechanisms underlying endocannabinoid-mediated short- and long-term synaptic plasticity. . . . . . . . . . . . . . . . . . . . . 36 1.9 Main biological functions regulated by endocannabinoid system. . . . 38 1.10 Main signaling cascades controlled by CB1R activation. . . . . . . . . 42 1.11 Schematic diagram of the human CB1R-CTD. . . . . . . . . . . . . . 46 4.12 The CTD of CB1R interacts with GAP43. . . . . . . . . . . . . . . . 68 4.13 CB1R interacts with GAP43 in neural tissue. . . . . . . . . . . . . . . 73 4.14 GAP43 phosphorylation favors its interaction with CB1R in HEK293T cells by in situ PLA detection. . . . . . . . . . . . . . . . . . . . . . . 75 4.15 GAP43 phosphorylation favors its interaction with CB1R in HEK293T cells by co-immunoprecipitation and BRET. . . . . . . . . . . . . . . 76 4.16 GAP43 reduces CB1R-mediated signaling independently of Gαi/o in HEK293T cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 4.17 GAP43 reduces CB1R-activated ROCK pathway signaling via Gαq/11 in HEK293T cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 4.18 GAP43 reduces CB1R-mediated signaling in 7-DIV primary cultured neurons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 4.19 GAP43 does not affect CB1R internalization in HEK293T cells. . . . 82 4.20 CB1R and GAP43 are present in the mouse striatum, cortex and hippocampus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4.21 CB1R and GAP43 colocalize exclusively in glutamatergic MC axons of the DG. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 4.22 CB1R and GAP43 interact in glutamatergic MC axons of the DG. . . 87 4.23 eCB-mediated CB1R function at MC-GC synapses is enhanced by PKC pharmacological inhibition. . . . . . . . . . . . . . . . . . . . . 90 4.24 Active GAP43 inhibits CB1R function at MC-GC synapses. . . . . . 92 4.25 Conditional GAP43 KO mouse generation. . . . . . . . . . . . . . . . 94 4.26 Conditional GAP43 KO mouse general characterization. . . . . . . . 97 4.27 Conditional GAP43 KO mouse learning and memory performance. . . 98 4.28 THC protects from seizure severity and progression in Glu-GAP43−/− vs WT mice after systemic KA administration. . . . . . . . . . . . . . 100 5.29 Proposed mechanism of action of GAP43 interaction on CB1R signaling.105 5.30 Proposed mechanism of action of CB1R-GAP43 complexes on eCB- STD at the MC-GC synapse. . . . . . . . . . . . . . . . . . . . . . . 109 7 List of Tables 1.1 DSI/DSE in the brain. . . . . . . . . . . . . . . . . . . . . . . . . . . 33 1.2 eCB-LTD in the brain. . . . . . . . . . . . . . . . . . . . . . . . . . . 35 1.3 Plasmids used in this Thesis. . . . . . . . . . . . . . . . . . . . . . . . 51 1.4 Primary antibodies used in this Thesis. . . . . . . . . . . . . . . . . . 58 4.5 List of potential CB1R interactors identified in proteomic-MS analyses with statistical significance. . . . . . . . . . . . . . . . . . . . . . . . 69 8 Abbreviations µHD: Mu homology domain 2-AG: 2-arachidonoylglycerol 5-HT: 5-hydroxytryptamine 5-IAF: 5-(iodoacetamido)-fluorescein A2AR: adenosine A2A receptor AA: arachidonic acid ABHD6/12: αβ-hydrolases 6 and 12 AC: adenylyl cyclase ACSF: artificial cerebral spinal fluid Anandamide: N-arachidonoylethanolamine AP3: adaptor protein 3 BDNF: brain-derived neurotrophic factor bFGF: basic fibroblast growth factor BIM-I: bisindolylmaleimide-I BRET: bioluminiscence resonance energy transfer BSA: bovine serum albumin CA: Cornu Amonis CaM: calmodulin cAMP: cyclic adenosine monophosphate CB1R: type-1 cannabinoid receptor CB2R: type-2 cannabinoid receptor CBA: chicken β-actin CBD: cannabidiol CBG: cannabigerol CBN: cannabinol CBR: cannabinoid receptors CCK: cholecystokinine CID: dissociation induced by collision CNS: central nervous system co-IP: co-immunoprecipitation CREB: cAMP responsive element-binding CRIP1a: cannabinoid receptor interacting protein-1a cryo-EM: cryoelectron microscopy CTD: C-terminal domain D2R: dopamine 2 receptor DAG: diacylglycerol DAGL: diacylglycerol lipases DAPI: 4’,6-diamidino-2-phenylindole DG: Dentate gyrus DIV: days in vitro DMEM: Dulbecco’s modified Eagle’s medium 9 DMR: dynamic mass redistribution DMSO: dimethylsulfoxide DPE: depolarization-induced potentiation of excitation DSE: depolarization-induced suppression of excitation DSI: depolarization-induced suppression of inhibition DTT: dithiothreitol EC: enthorinal cortex eCB-LTD: endocannabinoid-mediated long-term depression eCB: endocannabinoid eCB-STD: endocannabinoid -mediated short-term depression ECL: enhanced chemiluminescence ECS: endocannabinoid system EDTA: ethylenediaminetetraacetic acid EEA1: early endosome antigen 1 EGTA: ethylene glycol tetraacetic acid Epac: exchange protein directly activated by cAMP EPSC: excitatory postsynaptic current EPSP: excitatory postsynaptic potential ERK: extracellular signal-regulated kinase ESI: electrospray ionization FAAH: fatty acid amide hydrolase FAK: focal adhesion kinase FAN: neutral sphingomyelinase activation FBS: fetal bovine serum fEPSP: field EPSP FP: fluorescence polarization GAP43: growth-associated protein-43 GASP1: GPCR-associated sorting protein GC: granular cell GCL: granular cell layer GIRK: G protein-coupled inward-rectifying K+ channels GPCR: G protein-coupled receptor GPR55: G protein receptor 55 GRB2: Growth factor receptor-bound protein 2 GRK2/3: GPCR kinase 2/3 GSK3: glycogen synthase kinase 3 HBSS: Hank‘s Balanced Salt Solution HFS: high frequency stimulation HICAP: hilar commissural-associational pathway-related HIPP: hilar perforant path-associated HPA: Hypothalamic-pituitary-adrenal axis HTRF: Homogeneous time-resolved fluorescence energy transfer Hx8: helix 8 Hx9: helix 9 IDP: intrinsically disordered protein IL1: intracellular loop 1 IL2: intracellular loop 2 IL3: intracellular loop 3 10 IML: inner molecular layer IN: interneurons IP3: inositol triphsophate JAK2: Janus kinase 2 JNK: c-Jun N-terminal kinases KA: kainate LAMP1: lysosomal-associated membrane protein 1 LB: Lysogeny Broth LPP: lateral perforant pathway LTD: long-term depression LTP: long-term potentiation mAChRs: muscarinic acetylcholine receptors MAGL: monoacylglycerol lipase MC: mossy cell MEM: Minimum Essential Media mGluR-I: group I metabotropic glutamate receptors mGluR-II: group II metabotropic glutamate receptors MML: middle molecular layer MOPP: molecular layer perforant path-associated MOR: µ opiod receptor MPP: medial prefrontal pathway mTORC1: mammalian target of rapamycin complex 1 NAE: N-acylethanolamine NAPE: N-acylphosphatidylethanolamine NAPE-PLD: N-acyl-phosphatidylethanolamine-selective phospholipase D NCAMs: neural cell adhesion molecules NGF: nerve growth factor NHERF: sodium hydrogen exchanger regulatory factor nLC/MS-MS: nano-scale liquid chromatographic tandem mass spectrometry NMII: non-muscle myosin II NMDAR: N-methyl-D-aspartate receptors OEA: oleoylethanolamide OML: outer molecular layer PA: phosphatidic acid. PBS: phosphate buffered saline PEA: palmitoylethanolamide PFA: paraformaldehyde solution PI(4,5)P2: phosphatidilinositol-4,5-bisphosphate PKA: protein kinase A PKC: protein kinase C PLA: proximity ligation assay PLCβ: phospholipase Cβ PMSF: phenylmethylsulfonyl fluoride - PMSF PP: perforant pathway PPAR: peroxisome proliferator-activated receptor PPD: paired-pulse depression PPF: paired-pulse facilitation PPR: paired-pulse ratio 11 PTH1R: parathyroid hormone receptor PV: parvalbumin PVDF: polyvinylidene difluoride PVT: Polyvinyltoluene rAAV: Recombinant adeno-associated vector ROCK: Rho-associated coiled-coil containing protein kinase RRP: ready releasable pool RT: Room Temperature SC: Schaffer Collateral SDS: dodecyl sulfate SGIP1: SRC Homology 3-containing GRB2-like protein 3-interacting protein 1 SPA: scintillation proximity assay STD: short-term depression STDP: spike-timing dependent plasticity STP: short-term potentiation Syn: synaptophysin TBS: theta burst stimulation TBS-T: Tris buffered saline-tween THC: ∆9-tetrahydrocannabinol TLE: temporal lobe epilepsy TRPV1: transient receptor potential cation channel subfamily V member I VGCC: voltage-gated calcium channel VTA: ventral tegmental area WIN: WIN-55,212-2 WT: wild type 12 Resumen GAP43: una nueva protéına interactora del receptor CB1 cannabinoide La planta del cáñamo (Cannabis sativa L.) se ha utilizado en medicina desde hace al menos cincuenta siglos. Sin embargo, la estructura qúımica de sus componentes activos, los cannabinoides (∆9-tetrahydrocannabinol - THC and cannabidiol - CBD) no fue dilucidada hasta la década de 1960. Tres décadas después, dos receptores de cannabinoides acoplados a protéınas G fueron caracterizados: el de tipo 1 (CB1R), que es especialmente abundante en las áreas del sistema nervioso central implicadas en el control de la actividad motora, aprendizaje y memoria o emociones; y el de tipo 2 (CB2R), que se expresa preferentemente en el sistema inmune. Estos receptores son activados por ligandos endógenos, los endocannabinoides (eCBs). Concreta- mente a través de la activación de CB1R, los cannabinoides, tanto endógenos como exógenos, ejercen importantes efectos neuromoduladores en nuestro cerebro. Niveles de expresión de CB1R particularmente altos se pueden encontrar en la formación hipocampal, que alberga circuitos altamente interconectados implicados en la con- solidación de la memoria. El giro dentado (DG) es la primera estación de procesami- ento de la información. Los dos tipos celulares glutamatérgicos más importantes del DG, las células granulares (GCs) y las mossy cells (MCs), establecen un circuito re- currente excitatorio GC-MC-GC que regula la tranferencia de información al resto de estructuras hipocampales. Este circuito recurrente esta implicado no solo en consolidacion de memoria sino tambien en la epilepsia. CB1R esta localizado prin- cipalmente en los terminales presinápticos, donde regula la liberacion de neurotrans- misores. De esta manera, CB1R media formas de plasticidad a corto plazo conocidas como supresion de la inhibición o de la excitación inducida por depolarizacion (DSI y DSE respectivamente), aśı como depresion a largo plazo (LTD), en sinapsis tanto inhibitorias como excitatorias. El mecanismo de acción canónico de CB1R implica el acoplamiento a proteinas G heterotriméricas, principalmente al subtipo Gαi/o. Sin embargo, existen diferencias llamativas en las cascadas de señalizacion desencade- nadas por CB1R en función del tipo celular, la situación patofisiológica concreta u otros factores dependientes del contexto que se desconocen por el momento. Estu- dios recientes sugieren que factores moleculares espećıficos de tipo celular pueden facilitar, impedir o sesgar/desviar la señalización de CB1R. Concretamente, diver- sos estudios han mostrado interacciones entre protéınas intracelulares y el dominio citoplasmático C -terminal de CB1R como posibles moduladoras de la acción del receptor. Sin embargo, estos estudios no han desvelado ninguna implicación patofi- siológica de estas interacciones. Por tanto, la evaluación de la relevancia fisiológica y del potencial terapéutico de las distintas poblaciones del CB1R se ve impedida, al menos en parte, por la falta de conocimiento de la posibles interactores intracelulares expresados en un contexto celular espećıfico. Con estos antecedentes, el objetivo global de esta tesis doctoral es descubrir y caracterizar en detalle nuevos interactores intracelulares de CB1R que tengan lugar 13 de manera espacio-temporal en el cerebro. Tras un análisis inicial de detección por proteómica, decidimos centrarnos en la protéına GAP43 como posible interactor. Este objetivo global se puede dividir en dos objetivos espećıficos: Objetivo 1: Identificar y validar GAP43 como nuevo interactor de CB1R, ca- racterizando para ello la interacción CB1R-GAP43 y sus consecuencias sobre la señalización del receptor in vitro. Objetivo 2: Localizar los posibles complejos CB1R-GAP43 en el cerebro de ratón, con especial atención a distintas poblaciones neuronales, y estudiar las pos- ibles consecuencias funcionales de la interacción sobre la modulación de la trans- misión sináptica y las funciones fisiopatológicas que media el receptor. Para abordar el Objetivo 1, expresamos el dominio C -terminal del receptor purificado en una columna de cromatograf́ıa de sefarosa, la cual se expuso a un ho- mogeneizado de cerebro completo. Las protéınas unidas fueron eluidas y digeridas y se identificaron por espectrometŕıa de masas (nLC/MS-MS). Esto nos permitió detectar una lista de ∼50 interactores potenciales del receptor, de los cuales se se- leccionó la protéına GAP43/neuromodulina en base a la significatividad estad́ıstica y a su conocido papel en el sistema nervioso. GAP43 es una protéına principalmente presináptica y asociada al citoesqueleto. Puede interaccionar con otras protéınas como calmodulina, F-actina, SNAP25 y rabaptina-5. La actividad de GAP43 de- pende mayoritariamente de la fosforilacion en su residuo S41 por PKC. Es un factor clave en el crecimiento y gúıa axonal durante el desarrollo. En el cerebro adulto mod- ula procesos de endocitosis, reciclaje de protéınas de membrana, plasticidad y apren- dizaje. Para validar bioqúımicamente la interacción CB1R-GAP43, llevamos a cabo un gran número de experimentos en la ĺınea celular HEK293T, expresando CB1R y GAP43 [nativa o la versión activa (S41D) o inactiva (S41A) de la protéına]. Aśı, observamos por tres técnicas distintas de interacción protéına-protéına que CB1R interacciona con GAP43 principalmente en su forma fosforilada en S41. Además, ob- servamos que GAP43 ejerce un bloqueo parcial en el acoplamiento a proteinas G y la señalizacion mediada por el receptor, especificamente sobre la via de ROCK/cofilina desencadenada por Gαq11, mientras que las rutas de AMPc/PKA y ERK, canonica- mente dirigidas por acoplamiento a Gαi/o, no se vieron afectadas. Además, corrob- oramos estos hallazgos en cultivos primarios de neuronas hipocampales. Con respecto al Objetivo 2, detectamos complejos GAP43-CB1R en tejido de cerebro de ratón, selectivamente en las terminales glutamatérgias de las MCs que se sitúan en la capa molecular interna del DG, empleando para ello ensayos de li- gadura de proximidad in situ en ratones knockout condicionales selectivos y sus correspondientes rescates. Estos resultados sugieren que GAP43 podŕıa afectar a la funcionalidad de la población de CB1R localizada en las MCs. Dos aproximaciones complementarias, como son la inhibición farmacológica de la activación de GAP43 por PKC y la inoculación de virus recombinantes adenoasociados que expresan la forma activa (S41D) o inactiva (S41A) de GAP43 en MCs, nos permitieron determ- inar que la forma activa de GAP43 disminuye la DSE en las sinapsis MC-GC. Esta observación indica que GAP43 modula negativamente la actividad de CB1R en di- chas sinapsis. Por último, generamos ratones knockout condicionales con deleciones de GAP43 espećıficas en neuronas glutamatérgicas o GABAergicas. El estudio de 14 conductas dependientes de hipocampo y de epilepsia inducida por kainato en estos ratones mostraron que la deleción de GAP43 en neuronas glutamatérgicas conlleva pérdida de memoria y modula las propiedades anticonvulsivantes adscritas a CB1R. En suma, como conclusiones, podemos destacar que GAP43, principalmente en su forma fosforilada, interacciona con el dominio C -terminal de CB1R y reduce su señalización. En el cerebro de ratón, esta interacción se observa selectivamente en terminales de las MCs del DG, donde reduce la plasticidad a corto plazo mediada por el receptor. Además, GAP43 localizado en neuronas glutamatérgicas puede modular la actividad antiepiléptica de CB1R, en la cual las MCs desempeñan un papel fundamental. 15 Abstract GAP43: A new cannabinoid CB1 receptor-interacting protein The hemp plant (Cannabis sativa L.) has been used in medicine for at least fifty centuries. However, the chemical structure of its specific active components, the cannabinoids (∆9-tetrahydrocannabinol - THC and cannabidiol - CBD), was not elucidated until the early 1960s. Afterwards, two specific G protein-coupled can- nabinoid receptors were identified: CB1R, which is especially abundant in areas of the central nervous system (CNS) involved in the control of motor behaviour, learning and memory, or emotions; and CB2R, which is preferentially expressed in the immune system. These receptors are activated by endogenous ligands, the endocannabinoids (eCBs). By engaging CB1R in particular, both endogenous and exogenous cannabinoids exert pleiotropic, neuromodulatory effects on our brain. Particularly high levels of CB1R occur in the hippocampal formation, which shows a highly organized intrinsic circuit with the main purpose of memory consolidation. The dentate gyrus (DG) is the first step in the processing of information. The two principal glutamatergic cells in the DG, granule cells (GCs) and hilar mossy cells (MCs), establish an associative GC-MC-GC excitatory circuit that gates informa- tion transfer within the hippocampus. This circuit has been implicated not only in memory but also in epilepsy. CB1R is located mainly on neuron terminals to regulate synaptic function. This is the principal mode by which eCBs mediates short-term forms of plasticity known as depolarization-induced suppression of inhibition (DSI) or excitation (DSE), as well as long-term depression (eCB-LTD), at both excitatory and inhibitory synapses. The canonical mechanism of CB1R action involves its coupling to heterotrimeric G proteins, especially Gαi/o family members. However, there are striking differences in the signaling events triggered by CB1R among various cell types, pathophysiological situations and other context-related elements that remain unexplained. Recent find- ings suggest that cell type-intrinsic molecular factors may facilitate, impede and/or bias CB1R action. Specifically, a number of studies have dealt with the possible interaction of intracellular proteins with the cytoplasmatic domain of CB1R. How- ever, these studies have not shown the potential pathophysiological relevance of these putative interactions. Hence, the assessment of the physiological and thera- peutic impact of distinct CB1R pools is still hampered by the lack of knowledge of possible context-specific CB1R intracellular interactors. Based on this background, the general aim of the present Doctoral Thesis is to discover and characterize in detail new and spatio-temporally specific CB1R intracellular interactor(s). Upon a starting proteomic analysis, we decided to focus our efforts on the protein GAP43. The aforementioned general aim of this Doctoral Thesis can therefore be divided into two specific aims: 16 Aim 1: To identify and validate GAP43 as a new interactor of CB1R, by char- acterizing CB1R-GAP43 interaction and its signaling consequences in vitro. Aim 2: To map the CB1R-GAP43 complexes in the mouse brain in a cell- population specific manner, and to assess their functional consequences in CB1R- associated synaptic transmission and pathophysiology. To address Aim 1, we applied purified lectin-CB1R-C -terminal domain onto a Seph- arose chromatography column, which was challenged to a whole-brain homogenate. Then, bound and digested proteins were subjected to nLC/MS-MS proteomics ana- lysis, which allowed us to detect ∼50 potential CB1R-interacting proteins. Based on their statistical significance and known neurobiological roles, we decided to focus on growth-associated protein 43 (GAP43/neuromodulin). GAP43 is a neural-specific cytoskeleton-associated protein that is mainly located in the presynaptic terminal. It can interact with a number of other proteins as calmodulin, F-actin, SNAP25 or rabaptin-5. GAP43-mediated effects rely on PKC phosphorylation of its S41 residue to fulfill its functions. GAP43 is especially expressed during development, orches- trating axonal outgrowth and pathfinding. In the adult brain, GAP43 fine-tunes endocytosis, recycling of membrane proteins, synaptic plasticity and learning. We performed several experiments to validate CB1R-GAP43 interaction in HEK293T cells expressing CB1R and GAP43 [WT or two mutant versions: active (S41D) or inactive (S41A)]. We observed, by three different protein-protein interac- tion approaches, that CB1R interacts with GAP43 in its S41-phosphorylated form. Moreover, GAP43 exerted a partial blockade on CB1R-evoked G-protein coupling and signaling, especially the Gαq/11 protein-driven ROCK/cofilin cascade, while the canonical Gαi/o protein-driven cAMP/PKA and ERK pathways remained unaltered. In addition, we extrapolated these findings to primary hippocampal neuron cultures. Regarding Aim 2, in brain tissue, we detected GAP43-CB1R complexes select- ively in glutamatergic axon terminals of the MCs, located at the inner molecu- lar layer (IML) of the DG, as shown by in situ PLA in conditional knockout and rescue mouse models. These findings indicate that GAP43 could affect a pool of CB1R molecules specifically located on MC terminals. By using two complementary approaches on MCs, namely pharmacological inhibition of PKC-mediated GAP43 activation and adeno-associated virus (AAV)-mediated intracranial delivery of act- ive (S41D) or inactive (S41A) GAP43, we could determine that the active form of GAP43 leads to a decrease in DSE at MC-GC synapses. This observation strongly suggests that activated GAP43 negatively modulates CB1R-mediated regulation at the MC-GC synapse. In addition, we generated conditional GAP43-deficient mice, specifically in glutamatergic and GABAergic neurons. We tested hippocampal- dependent behaviors and kainate-induced epilepsy on these mouse models, and found that glutamatergic-neuron GAP43 deletion leads to memory impairment and fine- tunes CB1R-mediated anti-seizure activity. In sum, as conclusions, GAP43, mainly in its S41-phosphorylated form, in- teracts with CB1R-CTD and decreases cannabinoid-evoked signaling. In addition, GAP43-CB1R interaction occurs selectively in glutamatergic terminals of MCs, and decreases short-term cannabinoid-mediated plasticity. Moreover, glutamatergic- neuron GAP43 fine-tunes CB1R-mediated anti-seizure activity, in which MCs are highly relevant. 17 Introduction The hemp plant or marijuana (Cannabis sativa L.) has been used since ancient times for achievement of religious ecstasy as well as in medicine. The earliest record of the use of cannabis by humans comes from Ancient China, where it was prescribed as a herbal remedy. In Assyria, about 800 b.c., they called it gan-zi-gun-nu (the drug that takes away the mind) when used religiously or azallu when being prescribed. In many countries, hemp was grown also for obtaining durable fibres. Our present-day society follows a long tradition of recreational, industrial and medical cannabis use. C. sativa and its related subspecies are to date the only ones among the whole plant kingdom that produce significant amounts of phytocannabinoids. The most relevant ones, both in abundance and in pharmacological action, are ∆9-tetrahydrocannabinol (THC) and cannabidiol (CBD). The seminal work by Ro- ger Adams and Raphael Mechoulam led to the elucidation of their chemical structure in the early 1950-1960s upon their isolation from the plant. Around 30 years later, their first molecular targets, the cannabinoid receptors, were discovered in the mam- malian brain. Thus, the receptor stimulated by THC responsible for its psychoactive effects was designated as type-1 cannabinoid receptor. Specific receptors should ex- ist in our body because they are biologically activated by endogenous molecules, and therefore there was a strong reason to look for endogenous cannabinoids from brain and periphery and to find out the physiology of this signalling system. Of note, the first endocannabinoid to be discovered in 1992 was anandamide, named from the Sanskrit word ananda, which means “joy, bliss, delight” (Mechoulam et al., 2014). Results obtained from animal models and human studies over the past two decades greatly increased our understanding of the molecular mechanism by which cannabin- oids engage their receptors. They are expressed in virtually every single synapse in the brain and other cell types in the body. By activating their receptors, cannabin- oids cause a myriad of physiological responses, including feelings of well-being or psychosis, impaired memory and cognitive processing or learning, impaired motor function, as well as antinociceptive, antiemetic, antispastic and sleep-promoting ef- fects (Koppel et al., 2014). The brain is a very complex organ, actually working as an “organ of organs”. It is a “machine” that is continuously changing with different cell elements undergoing opposite and redundant functions that combine and synchronize to build in the cir- cuitry that governs our elaborated behaviors and feelings. Therefore, the complexity coming out from medical cannabinoid studies would raise from the unwanted effects of cannabinoids acting in the brain where and how we do not want them to act. This is why the necessity to look for a regional specificity of the whole cannabinoid system appears. Previous studies have provided scientific support for targeting the endocannabinoid signalling system to treat different devastating diseases, including neurodegenerative and chronic inflammatory diseases. Raphael Mechoulam, in a re- cent review (2014), asks: “Is the endocannabinoid system going to bring a revolution in therapy? This might be the case as investigators are now able to target cell-specific 18 synthetic enzyme pathways, allosteric modulators and receptor-associated proteins”. After nearly five decades of cannabinoid research and in spite of the social stigma and psychoactive side effects, the medical use of cannabis extracts is approved by twenty-two European countries. Some other countries, such as Argentina, Australia, Canada, Chile, Colombia, Ecuador, Israel, Lebanon, Malawi, New Zealand, Thail- and, Uruguay, Zambia, Zimbabwe and the United States (33 states and the District of Columbia) have also regulated the medical use of marijuana. This encourages researchers to continue seeking answers to the benefits and risks of marijuana use in patients. Cannabis uses and legalization is a hot topic nowadays and cell-type and patient-specific medicine is a prominent challenge. 1.1 The endocannabinoid system The endocannabionid system (ECS) is a cell signaling system classically defined as the ensemble of 1) two 7-transmembrane-domain and G protein-coupled recept- ors (GPCRs): type-1 cannabinoid receptor (CB1R) and type-2 cannabinoid re- ceptor (CB2R); 2) the most studied endogenous ligands or “endocannabinoids” N-arachidonoylethanolamine (anandamide or AEA) and 2-arachidonoylglycerol (2- AG); 3) the enzymes responsible for these endocannabinoids’ biosynthesis [i.e., N- acyl- phosphatidylethanola-mine-selective phospholipase D (NAPE-PLD) and diacylgly- cerol lipases (DAGL) α and β, for anandamide and 2-AG, respectively] and hydro- lytic inactivation [i.e., fatty acid amide hydrolase (FAAH) and monoacylglycerol lipase (MAGL), for anandamide and 2-AG, respectively]; and 4) the proteins for cellular uptake, transport and bioconversion of endocannabinoids (Marzo and Pis- citelli, 2015). 1.1.1 Cannabinoid receptors: CB1R, CB2R and others CB1R was the first discovered receptor for the phytocannabinoid THC. It was cloned from rat cerebral cortex (Matsuda et al., 1990), human brain and testis (Gérard et al., 1991), and mouse brain (Chakravarthy et al., 2008). A high conservation between the genes encoding the human, rat, mouse, fish and bovine cannabinoid receptors (CBRs), among others, suggests an important biological function. The coding region of the CB1R gene (Cnr1) is intronless. This means that the expression of the Cnr1 gene will have one major RNA processing event, accelerating its protein expression (Onaivi et al., 1999). Nevertheless, the presence of splice isoforms both in humans and mice (Ruehle et al., 2017), coming from 5-UTR introns of the gene, and possible post-translational modifications, demonstrates that CB1R can appear in different forms already at transcriptional and translational levels, with potential signaling differences (Bagher et al., 2013; Oddi et al., 2017; Straiker et al., 2012). Shortly after, a second receptor, CB2R, was identified and cloned in rat (Munro et al., 1993) and later in mouse (Shire et al., 1996). The CB2R gene (Cnr2) in mouse presents three exons which encode two transcripts (mCB2A and mCB2B). The mCB2A transcript contains exon 1 and exon 3, and the mCB2B transcript contains exon 2 and exon 3. In contrast, Cnr2 includes 3 exons in the rat that can be spliced to four rCB2R transcript isoforms - CB2A, CB2B, CB2C and CB2D, and 19 each displays different expression in the brain and peripheral tissues (Jordan and Xi, 2019). CB1R and CB2R belong to the Class A GPCR superfamily, composed by seven transmembrane helixes separated by three intracellular loops (named IL1, IL2 and IL3) and three extracellular loops facing the matrix plus an extracellular glycosylated N -terminal domain and an intracellular C -terminal domain (CTD) ori- ented towards the cytoplasm (Figure 1.1). They share an overall homology of 44% and 68% in their transmembrane domains. Both are highly conserved across species and share common ligands. However, even if they share their ability to bind both phytocannabinoids and endocannabinoids, binding affinity and specificity is differ- ent for each receptor. There is no significant homology between the intracellular CTD of both receptors and they present a distinct tendency to select intracellular interacting partners, which gives each CBR type a unique signaling profile. Analyses of CBR’s crystal structure have uncovered the differences between CB1R and CB2R in activation, ligand recognition and G protein coupling. Two antagonist/inverse agonist-bound (Hua et al., 2016, 2017; Shao et al., 2016) and one agonist-bound (Hua et al., 2017) crystal structures of CB1R have recently revealed both the inactive and the active conformations of the receptor, respectively. For CB2R, high-resolution cryoelectron microscopy (cryo-EM) structure of the agonist-bound form of the re- ceptor in complex with a heterotrimeric G protein was defined (Figure 1.1). The 3D structure study reveals the agonist-driven mechanism of receptor activation in- volving the structural arrangement of critical residues (Xing et al., 2020). Ligand binding within the orthostetic pocket in the extracellular site triggers a number of positional changes of many of the transmembrane helices and intracellular domains and stabilizes the receptor in an active state to facilitate nucleotide exchange in the G protein (Kumar et al., 2019). Both CBRs have as well allosteric binding sites for endogenous and synthetic ligands that are different from the orthosteric site. The binding of allosteric modu- lators typically leads to a conformational change of the receptor, which affects the affinity and/or efficacy of the orthosteric ligands, thereby fine-tuning their biological activities. There have been more studies of allosteric modulators for CB1R than for CB2R (Price et al., 2005; Piazza et al., 2017). For instance, the endogenous anti- inflammatory lipid lipoxin A4 may be an allosteric enhancer of CB1R signaling in the brain (Pamplona et al., 2012), as well as CBD, that acts as a negative allosteric modulator (Tham et al., 2019). On the contrary, the neurosteroid pregnenolone is an allosteric signal-specific inhibitor of CB1R, able to protect the brain from excessive cannabinoid intoxication (Vallée et al., 2014). Recently cholesterol was described as an endogenous “allosteric” modulator of CB1R (Hua et al., 2020). However, only a few allosteric modulators of CB2R have been identified so far and they have micro- molar activity. For example, trans-β-caryophyllene and dihydro-gambogic acid are the two reported CB2R negative modulators. In addition, gambogic acid analogs and pepcan-12 (RVD-hemopressin, an endogenous peptide) had been reported as positive modulators (Pandey et al., 2020; Petrucci et al., 2017). CB1R is located primarily in the central nervous system (CNS), controlling neur- otransmitter release, but is also expressed in peripheral tissues like heart, uterus, testis, liver and small intestine (Zou and Kumar, 2018). CB2R is expressed mainly by immune cells (for example, B and T lymphocytes, and macrophages). When activ- ated, CB2R can affect the release of chemical messengers, in this case the secretion of 20 Figure 1.1: Cryo-EM Structures of the CB1R and CB2R coupled to Gi protein. A. Cryo-EM density maps of the CB1R-Gi and CB2R-Gi complexes, with colored subunits. upon activation with agonists AM841 and AM10233 respectively. Color code for the proteins is as follows: CB1R, orange; CB2R, green; Gαi in CB1R, slate; Gαi in CB2R, yellow; Gβ, cyan; Gγ, magenta. B and C. Cryo-EM structures of CB1R-Gi (B) and CB2R-Gi (C) complexes using same color code as in (A), with AM841 and AM10233 shown as yellow and orange sticks, respectively (Hua et al., 2020). cytokines by immune cells, and they can modulate immune cell trafficking (Pertwee et al., 2010). However, it is now accepted that CB2R is expressed in multiple tis- sues: hematopoietic system, endocrine pancreas, adipose tissue and also in nervous system, mainly in microglial cells, although it has also been reported in astrocytes (Sagredo et al., 2009), neural progenitors (Palazuelos et al., 2006), peripheral ter- minals and some restricted populations of neurons in low amounts (Stempel et al., 2016). Apart from the two main cannabinoid receptors, CB1R and CB2R, a third GPCR named G-protein receptor 55 (GPR55), which shares 14% sequence homology with the CBRs, was proposed to interact with some cannabinoid compounds; it can be targeted by exogenous cannabinoids (e.g., THC as agonist and CBD as ant- agonist) and endogenous cannabinoids (e.g., N-palmitoylethanolamine as agonist) (Ross, 2009; Ryberg et al., 2007) and mediate part of their effects, at least in some pharmacological studies in vitro. GPR55 can be involved in tumor development (Pérez-Gómez et al., 2013) and in energy balance (Simcocks et al., 2014). CBD 21 can exert some of its functions as anticonvulsivant and anti-tumoral by binding to GPR55 to block its activity (Devinsky et al., 2014). Additionally, some cannabinoids (anandamide, and related fatty acylethanolamides and N-arachidonoyldopamines) can often bind non-GPCR receptors as vainilloid receptor/transient receptor po- tential cation channel subfamily V member I (TRPV1), a ligand-dependent ion channel typically responding to capsaicin, with effects in synaptic plasticity and pain regulation (Bisogno et al., 2001; Zygmunt et al., 1999), or nuclear receptors as the peroxisome proliferator-activated receptor (PPAR) family members, with neuroprotective effects (O’Sullivan, 2007). 1.1.2 Cannabinoids and endocannabinoids The endocannabinoids (eCBs) are lipid molecules derived from arachidonic acid and isolated from brain and peripheral tissues that include amides and esters of long-chain polyunsaturated fatty acids. They act as “THC mimetics”, thus enga- ging CBRs and controlling different biological processes. Anandamide or AEA, an amide derivative of arachidonic acid, and 2-AG, an ester derivative of arachidonic acid, are to date the best established eCBs and the most important in terms of bio- logical functions (Figure 1.2). The first to be identified was AEA in porcine brain. Its structure was established by mass spectrometry, NMR spectroscopy and by its chemical synthesis (Devane et al., 1992). Using the same techniques that were used to isolate anandamide, it was possible to isolate 2-AG for the first time from canine intestines (Mechoulam et al., 1995) and brain (Sugiura et al., 1995). AEA and 2-AG show some differences in activity. Thus, 2-AG is a full agonist for both receptors, with high efficacy and moderate-to-low affinity. Furthermore, it is increasingly more accepted that it represents the main ligand of presynaptic CB1R in the control of synaptic activity and plasticity (Katona and Freund, 2008). Meanwhile, AEA is an intermediate-efficacy, high-affinity partial agonist for CB1R, and almost inactive at CB2R. It can bind presynaptic CB1R, but is also involved in other eCB-mediated forms of plasticity, for example upon TRPV1 (Zygmunt et al., 1999) or PPAR-γ (Bouaboula et al., 2005) binding. eCBs are not stored in intracellular compartments but are generated by “on demand” synthesis and cleavage of cellular membrane lipid precursors, which are hydrophobic molecules with polyunsaturated fatty acyl chains. This is a remarkable feature of the ECS that makes a difference from almost the rest of neurotransmitter- based cell communication systems and implies the constant bioavailability of the substrates and the permanence of metabolic enzymes in an active state. The pro- cess occurs often in response to increased concentrations of intracellular calcium. eCBs are then released and function as retrograde synaptic messengers by acting on presynaptic CB1R, that can in turn prevent the development of excessive neuronal synaptic activity in the CNS and thereby contribute to the maintenance of neural tissue homeostasis (Mechoulam et al., 2014; Marzo et al., 2015). Besides these two compounds, other fatty acyl ethanolamides such as oleoyle- thanolamide (OEA), palmitoylethanolamide (PEA), virodhamine and N- arachido- noyldopamine have been identified as “ endocannabinoid-like” compounds. They are present in considerable amounts in several organs. They might bind to CBRs, probably to a non-orthosteric site, although they have not been univocally proven to exert CBR-evoked biological actions. They might also potentiate the activity of 22 AEA or 2-AG by inhibiting their degradation (the so-called “entourage effect”). Another class of cannabinoids is constituted by the phytocannabinoids, namely the naturally-occurring compounds coming from C. sativa. Among the more than a hundred different active components from the plant, we may recognize for example THC, CBD (Figure 1.2), cannabinol (CBN), ∆8-THC and cannabigerol (CBG), to name just a few, with distinct affinities for CBRs. The binding affinity of THC for CB1R is similar to that of 2-AG (tens of nM range). THC causes a typical tetrad of effects in animals, the so-called “cannabinoid tetrad”, which includes antinocicep- tion, catalepsia, hypomotility and hypothermia (Zou and Kumar, 2018; Mechoulam et al., 2014). Figure 1.2: Major phytocannabinoids and eCBs. Chemical structures of the phytocannabinoids ∆9-THC and CBD, and the two endogenous cannabinoids anandamide and 2-AG (Mechoulam and Parker, 2013). Last, it is worth mentioning that there is a plethora of synthetic cannabinoids engineered for research and medical purposes. Some of them were used in this Thesis. WIN-55212-2 (WIN) is an aminoalkylindole (Howlett et al., 2004) widely used in vitro because of its high affinity and potency (Breivogel et al., 1998). Some other synthetic agonists are HU-210 and CP55,940 (dual agonists with very high affinity and potency at CBRs). SR141716 (rimonabant) and AM251 are CB1R-selective antagonists/inverse agonists (Pertwee et al., 2010). 1.1.3 Enzymes for eCB synthesis and degradation AEA and 2-AG are synthesized differently in the various producing tissues. AEA synthesis pathway starts with the N-acylation of phosphatidyletanolamine at cel- lular membranes by a Ca2+-dependent N-acyltransferase (through the transfer of the corresponding acyl chain from the sn-1 position of phospholipids to the amine group of phosphatidylethanolamine), thus resulting in the synthesis of N- acylphos- phatidylethanolamine (NAPE). This is the Ca2+-sensitive and rate-limiting step of AEA production. Subsequently, N-acylethanolamine (NAE) is released upon NAPE hydrolysis by NAPE-PLD, an enzyme that therefore generates AEA and phos- phatidic acid (PA). AEA synthesis can occur as well in two or three steps through alternative routes (Maccarrone et al., 2007). In contrast, 2-AG and other 2-acylglycerols are produced in one step from the hydrolysis of diacylglycerols (DAGs) by either of two DAGLs, DAGLα or DAGLβ, 23 although most if not all 2-AG mediating synaptic transmission in the adult brain is generated by DAGLα (Tanimura et al., 2010; Murataeva et al., 2014). DAGs are produced in most cases from the hydrolysis of phosphoinositides by phospholipase Cβ (PLCβ), in a Ca2+-sensitive and rate-limiting step. After release into the intracellular space, due to their uncharged hydrophobic nature, eCBs are unable to diffuse freely like other neurotransmitters. eCBs are taken up by cells through a mechanism that is not yet fully elucidated, but likely to involve simple diffusion driven by concentration gradients generated from enzymatic degradation, membrane carriers like fatty acid-binding proteins, and/or possibly en- docytosis involving caveolae/lipid rafts. Once eCBs are taken up by the cells, they can be degraded through hydrolysis and/or oxidation to terminate their activity. AEA is degraded mainly by FAAH into free arachidonic acid (AA) and ethanolam- ine, whereas 2-AG is mostly hydrolyzed by MAGL into AA and glycerol; several other enzymes could be involved as well in these processes, for example the serine hy- drolases αβ-hydrolases (ABHD) 6 and 12. Oxidation of both AEA and 2-AG could involve cyclooxygenase-2 and several lipoxygenases. In addition, eCB-degradation products are themselves precursors of bioactive lipid mediators, such as prostagland- ins, prostamides and other eicosanoids, thus making their metabolic regulation and function be tightly regulated. These enzymes are therefore “degrading” for eCBs and “biosynthetic” for other mediators (Nomura et al., 2011). The location of enzymes engaged in the synthesis and degradation of eCBs is closely linked to the purpose of eCB signaling (Figure 1.3). As CB1R is presynaptic, the retrograde nature of cannabinoid signalling is made possible by the postsynaptic localization of DAGLα, which has a major role in 2-AG production in the brain (Yoshida, 2006; Jung et al., 2007). DAGL is found on the plasma membrane of the perisynaptic domain, where is functionally coupled to PLCβ and metabotropic re- ceptors. MAGL, on the contrary, resides in the presynaptic axon membrane, where it co-localizes with CB1R. From the presynaptic localization at various terminals, MAGL limits spatial and temporal extents of 2-AG, which is released from post- synaptic neurons to the extracellular space but also contributes to degradation of constitutively produced 2-AG and prevention of its accumulation around presyn- aptic terminals. Thus MAGL activity determines basal eCB tone and modulates retrograde eCB signaling (Hashimotodani et al., 2007). As these MAGL-expressing elements cover a fairly large surface area, 2-AG diffusion from activated to neigh- bouring synapses may be restricted. FAAH is expressed mainly in postsynaptic dendrites and soma, in membranes of intracellular organelles (for example, to ac- tivate postsynaptic TRPV1). This suggests that AEA and 2-AG signalling may subserve functional roles that are spatially segregated at least at the stage of their metabolism (Gulyas et al., 2004). A complete identification of the different synthesis and degradation pathways will provide novel targets for drugs to manipulate brain eCB levels. Given the big amount of elements and players in this system including eCBs-like mediators, and their several receptors and metabolic enzymes, a new term has lately emerged for assembling all of them: the “endocannabinoidome” (Marzo and Piscitelli, 2015). 24 Figure 1.3: Main biosynthesis and degradation pathways in eCBs signalling. The subcellular distribution in neurons of enzymes regulating the levels of eCBs is shown. The biosynthesis of AEA occurs through the action of NAPE-PLD, which is located in intracellular membranes both pre- and postsynaptically. AEA is degraded by FAAH, which is located postsynaptically. 2-AG is synthesized by DAGLα, located postsynaptically from the hydrolysis of DAGs in the plasma membrane. 2-AG is degraded by MAGL, which is presynaptic. These enzymes also regulate the levels of the eCB-related mediators. ER, endoplasmic reticulum; PIP2, phosphatidylinositol-4.5-bisphosphate; PE, phosphatidyletanolamine VGCCs, voltage-gated calcium channels. 1.2 CB1 Receptor 1.2.1 Expression in the adult brain CB1R is one of the most abundant G-protein coupled receptor/metabotropic re- ceptor in the mammalian CNS. Due to its predominant presynaptic location, there is often a discrepancy between the distribution of CB1R mRNA and protein in neur- onal populations across brain areas, especially when those are projecting neurons (Figure 1.4). CB1R protein in the murine brain is highly detected by immun- ostaining or binding of synthetic radioactively-labeled ligands in the hippocampus, the olfactory bulb, the main target nuclei of the striatum (i.e. globus pallidus, entopeduncular nucleus, substantia nigra pars reticulata) and cerebellar molecular layer. Moderate levels are noted in the cerebral cortex (higher in the frontal, parietal and cingulated areas than in other cortical areas), septum, basolateral amygdala, ventromedial hypothalamus, some nuclei of the brainstem (interpeduncular nucleus, parabrachial nucleus, nucleus of solitary tract) and spinal dorsal horn. The thal- amus, other nuclei in the brainstem and spinal ventral horn have a low receptor expression. These overall mapping properties are preserved across mammals (Kano et al., 2009; Herkenham et al., 1990, 1991) Regarding CB1R mRNA, two distinct patterns of CBR expression are distin- guished. There is a uniform labeling (resulting from CB1R expression in principal cells) found in the striatum, thalamus, hypothalamus, cerebellum and lower brain- stem, whereas a non-uniform signal is observed in the hippocampus, cerebral cortex 25 Figure 1.4: Expression pattern of CB1R protein and mRNA in the mice’s CNS. (A-D) Overall distribution in sagittal (A, D) and coronal (B, C) brain sections of wild-type (A-C) and CB1R-knockout (CB1-KO) (D) mice immunolabeled with a polyclonal antibody against mouse CB1R. CB1R immunoreactivity is highest along striatal output pathways, including the substantia nigra pars reticulata (SNR), globus pallidus (GP) and entopeduncular nucleus (EP) and cerebellar cortex (Cb). High levels are also observed in the hippocampus (Hi), dentate gyrus (DG) and cerebral cortex, caudate putamen (CPu) and ventromedial hypothalamus (VMH) among others. (E) CB1R immunolabeling in the spinal cord. (F) In addition, CB1R immunoreactivity also shows laminar patterns in the hippocampus and dentate gyrus. (G,H) Representative images showing CB1R in situ hybridization in a sagittal (H) or coronal (G) section of adult mouse brain. Note a slightly different pattern distribution comparing to protein. Image credit: Allen Institute & (Kano et al., 2009) and amygdala (Kano et al., 2009). In recent years, it has been demonstrated that CB1R can be distributed in many cell types and intracellular compartments, but its pattern highly differs from one cell type to another. It is mainly expressed in virtually every neuron type: (most subtypes of) GABAergic and glutamatergic neurons -the former show much higher expression compared to the later (Marsicano and Lutz, 1999), serotoninergic neurons (Häring et al., 2015) and cholinergic neurons (Nýıri et al., 2005) -cells in the medial septum projecting to hippocampus in mice and rats, (Degroot et al., 2006), but also it has been detected in noradrenergic neurons (Oropeza et al., 2007)- in frontal cortex, and dopaminergic neurons (Lau and Schloss, 2008). The development of cell- type specific conditional CB1R knockout models has extensively contributed to these findings. In addition, it is expressed in other neural cells, though at lower levels, such as astrocytes (Sánchez et al., 1998), where it modulates neuron-astrocyte com- munication or tripartite synapses (Navarrete and Araque, 2008), microglia (Stella, 2010), oligodendrocytes (Molina-Holgado et al., 2002) and adult neural progenitors (Jin et al., 2004; Aguado et al., 2005). At a subcellular level it is mostly located at the presynaptic plasma membrane (Katona, 2006; Schlicker and Kathmann, 2001; Hu and Mackie, 2015), where it inhibits neurotransmitter release. However, there are some reports that show CB1R 26 in the postsynaptic zone, where may be involved in self-inhibition in a subset of pyramidal neurons and interneurons in the neocortex (Marinelli et al., 2009; Bacci et al., 2004). Additionally, part of the receptor pool is present on intracellular structures, mostly endosomes, due to its constitutive recycling. Moreover, a new location for the receptor has been described in the last years: the outer membrane of the mitochondria (Bénard et al., 2012), where cannabinoid compounds are able to modulate respiration and therefore neuronal activity, thus interfering with memory (Hebert-Chatelain et al., 2016) and social interaction (Jimenez-Blasco et al., 2020) processes. The hippocampal formation. Dentate gyrus and mossy cells The hippocampal formation comprises three main regions: subiculum, dentate gyrus (DG) and Cornu Amonis (CA) or hippocampus proper, which can be di- vided into three different subfields attending to morphology and function: CA1, CA2 and CA3. In some cases, a CA4 field containing cells within the hilus of the DG is considered. Subiculum is directly followed by enthorinal cortex (EC) and parahipocampal gyrus. One of the unique features of the hippocampal formation is that many of its connections are unidirectional and that it presents highly stratified lamina. This led archetypically to a highly organized connectivity with the main purpose of long-term episodic memory consolidation. This anatomical organization has influenced the conceptualization of hippocampal information processing. From a classical view, the hippocampus intrinsic trisynaptic circuitry stands as follows (Figure 1.5): DG receives information directly from layer II of EC (where information from olfactory bulb, amigdala, or prelimibic, visual, auditory and gust- atory cortex converges) via the so-called perforant pathway (PP). Some PP afferents can also synapse onto CA3 and CA1 pyramidal cells (Doller and Weight, 1982; Yeckel and Berger, 1990). The output from DG is the so-called mossy fibers that impinge in CA3. CA3 pyramidal cells give rise to collateralized axons that terminate within CA3 itself but also their projections provide the major input to CA1, the so-called Schaffer Collaterals (SC). Finally, CA1 pyramidal cells project to subiculum or to cortex. Axons from cells in subiculum and CA1 form the efferences back to deeper layers (V-VI) of the EC, which will connect with other cortex associative areas to close the circuit. Efferent fibers from subiculum and CA1 pyramidal cells can travel as well by fimbria/fornix to subcortical areas: mammillary nucleus, hypothalamus, thalamus, accumbens, etc. Additionally, there is also an associational/commisural pathway with SC fibers and others which connects one hippocampus to the contralat- eral one (Groen and Wyss, 1990; Amaral and Witter, 1989; Haines and Mihailoff, 2017). The hippocampal formation shows basically a three-layer cytoarchitecture. CA regions are formed by a central pyramidal layer composed of pyramidal neuron somas. They are termed “pyramidal” due to the conical shape of the soma from which two basal and one apical dendrite emerge. They are glutamatergic projec- tion neurons. Above this pyramidal layer, there is a molecular layer divided in stratum lucidum (only in CA3, harboring mossy fiber afferents), stratum radiatum and stratum lacunosum-moleculare harbouring apical dendrites of pyramidal cells and PP, SC and commissural afferents. Below the pyramidal layer the stratum ori- ens is located, with the basal dendrites and axons from pyramidal cells and some recurrent collaterals. Different types of inhibitory interneurons are present in every 27 Figure 1.5: The hippocampal circuitry. The DG receives most of its inputs from layer II of the Enthorinal Cortex (EC) through the MPP and LPP, respectively. The axons of the GCs (in orange over red) extend towards the pyramidal cells of the CA3 region, forming the mossy fibers. The projections of the CA3 pyramidal cells form the Schaffer collaterals (SC), which establish synapses with dendrites of the CA1 pyramidal cells (in orange over blue). Finally, CA1 neurons complete the hippocampal circuitry by projecting their fibres to the deep layers of the EC. CA1 also receives direct input from layer III of the EC through the temporoammonic (TA) pathway (Pinar et al., 2017). layer. They are commonly classified by the location of their cell body and axon, and the specificity of their terminal for a sublayer of the hippocampus (Freund and Buzsáki, 1996). The DG presents a three-layer structure as well, with an anatomical organiz- ation similar in rodents and primates. There is a relatively cell-free layer called the molecular layer. It is divided in three fractions: the outer molecular layer (OML), the middle molecular layer (MML) and the inner molecular layer (IML). The major input towards DG, the PP, can be divided into two separated pathways, the lateral perforant pathway (LPP), which terminates in the OML, and the medial prefrontal pathway (MPP), which reaches the MML and originates from the lateral and the medial enthorinal areas, respectively. They differ in histochemical prop- erties and convey different sources of information. The LPP transmits non-spatial information as objects and odors and the MPP is in charge of spatial information as places (Deshmukh and Knierim, 2011). Next to IML there is the principal cell layer or granule cell layer (GCL), which is made up of densely packed granule cells oriented in a stereotypical manner. The dendrite branches extend throughout the molecular layer and the distal tips of the dendritic tree end just at the hippocampal fissure or at the ventricular surface. Last, one can find the polymorphic cell layer of the hilus that contains the granule cell axons, which are called mossy fibers (that present large irregular axonal terminals) and some other cell types (Amaral et al., 2007). The major glutamatergic cell type of the DG is the previously mentioned granule cell (GC), whose axons shape the output of processed information in the DG to the next structure. Most GCs are located in the GCL, but there are small subsets in the IML (known as semilunar GCs) and hilus (known as ectopic GCs). To note, DG harbors one of the two niches of neural stem cells in the adult brain located in the subgranular zone. These subgranular zone progenitors divide throughout life and migrate primarily to the GCL, where they become GCs and integrate into the DG circuitry in a similar manner to GCs born in early life (Kempermann et al., 2015). There are several types of GABAergic interneurons (IN) in the DG. The 28 most studied one is the basket cell, with a pyramidal soma located in the interface between GCL and hilus or within the GCL. They have a principal aspiny apical dendrite directed into the molecular layer and several basal dendrites into the poly- morphic cell layer. Part of them is parvalbumin (PV)-positive, very resistant to injury, and their axons form pericellular plexus surrounding GC bodies. Another part is cholecystokinine (CCK)-positive, which establish synapses with GC periso- matically and in their proximal dendrites of IML. In the molecular layer one can find the soma of molecular layer perforant path-associated (MOPP) cells, with axons dir- ected towards the outer two thirds of the molecular layer and chandelier/axo-axonic cells, whose axons are directed to the initial segment of GC axon (PV+). In the hilus one can find hilar perforant path-associated (HIPP) cells, which connect to GC dendrites in the outer two-thirds of the ML and are somatostatin-positive and hilar commissural-associational pathway-related (HICAP) cells that extend their axon through the GCL and branch profusely in the IML (Amaral et al., 2007; Houser, 2007). However, the DG is different from other hippocampal and brain areas due to the presence of another additional type of glutamatergic cell with an “interneuron” function, the so-called mossy cells (MCs). This cell type is probably what Ramón y Cajal referred to as the “stellate or triangular” cells, but received its current name from Amaral (1978). Their triangular multipolar cell bodies are located in the hilus, where it remains most of their dendritic branches, covered by large complex spines like thorny excrescences with a mossy appearance. MC axons target specifically the IML where they primarily innervate GCs both ipsi and contralaterally by commis- sural/associational projections. Because of their proximity to the GC soma, MC-GC synapses are in an ideal position to influence the activity of GCs. Moreover, MCs not only contact GCs locally (same lamella) but also project widely along the longit- udinal axis of the hippocampus, both dorsally and ventrally/septally and temporally from the point of origin (Amaral et al., 2007; Buckmaster et al., 1996). The pro- jection to the molecular layer at the septotemporal level of origin is weak but gets increasingly stronger at distance from the cells of origin (Figure 1.6) (Scharfman, 2016; Scharfman and Myers, 2013; Hashimotodani et al., 2017). At the same time, MCs receive inputs from collaterals of GC axons. Thereby an intrinsic or associative positive-feedback loop is established, receiving powerful input from a relatively small number of GCs and providing highly distributed excitatory output to a large number of GCs (Amaral et al., 2007; Buckmaster and Schwartzk- roin, 1994; Buckmaster et al., 1996; Scharfman and Myers, 2013). In addition to this recurrent excitatory circuit, MCs also contact INs (mostly basket cells), which me- diate feed-forward inhibition onto GCs (Larimer and Strowbridge, 2008; Scharfman, 1995). Therefore, the net effect of MCs on GC activity is difficult to predict. There is an excitatory/inhibitory balance orchestrated by MCs that depends on specific spatiotemporal context cues to reinforce excitation or inhibition (Hashimotodani et al., 2017; Scharfman, 2016). Other afferences to DGs are glutamatergic supra- mammillary projections located just superficially to the GC layer, cholinergic inputs from septal nuclei to GC layer and IML, noradrenergic fibers from locus coeruleus and ventral tegmental area to hilus, and serotoninergic fibers from rafe nuclei to INs of the hilus. The DG is functionally heterogeneous along its axis; the dorsal/septal limb is primarily involved in spatial memory, while the ventral/temporal area is associated 29 Figure 1.6: Distant and local Mossy cell-Granular cell (MC-GC) circuitry in ipsilateral DG. Layer structure of DG is depicted on the left. The axon of a single ventral MC is illustrated schematically. Far from the soma, the distant ipsilateral or distant contralateral branches of the axon project primarily to the IML, on spines of GCs. In addition, the MC axon extends hilar, OML and MML collaterals. Near the soma, the local ipsilateral branches of the mossy cell axon make synapses in the hilus (HIL) where they contact Gabaergic interneuron (IN). MCs also make local projections to the IML, but these are not as numerous as those to distant sites. SGZ, subgranular zone (Scharfman, 2016). with emotional memory or anxiety (Fanselow and Dong, 2010; Strange et al., 2014). GCs and MCs are a key component of the computation performed by the DG. Thus, the projection of MCs could modulate GC activity throughout the hippocampus, linking functionally diverse areas (Scharfman and Myers, 2013). Specifically, MCs, via their communication with GCs, are known for contributing to various forms of learning and memory. This includes associative memory in a way that they allow to physically separate subsets of GCs to be associated with one another along the axis, recall of memory sequences (Lisman et al., 2005) or pattern separation, i.e. the ability to separate patterns of afferent input that contains elements that are identical or overlapping. It allows similar experiences to be stored in different subsets of CA3 pyramidal cells, thus facilitating accurate memory retrieval. For two patterns of input from the EC, separation could occur if the shared (identical) element of one pattern was coincident with MC input to GCs but not coincident with the other pattern. MCs are also involved in novelty recognition or signaling changes in the environment. They can contribute to this function by exciting GCs when a novel sensory input is processed in the lateral EC and is sent to the DG by the LPP (Scharfman, 2016). On the other hand, the recurrent excitatory loop between MCs and GCs predispose MCs to excitotoxic death. They are highly vulnerable to injury (Buckmaster and Schwartzkroin, 1994). MCs have been linked to temporal lobe epilepsy (TLE) onset, one of the most common forms of epilepsy in adults in which seizures affect the temporal lobe of the cortex, either by becoming overactive and driving GC firing (Ratzliff et al., 2002; d. H. Ratzliff, 2004), or by dying and reducing the magnitude of feed-forward inhibition since a premature loss of MCs has been observed in TLE (Sloviter, 1991). Detailed CB1R expression on hippocampus and DG Particularly high levels of CB1R are shown on the hippocampus. The highest levels are found on GABAergic neurons. At the hippocampal pyramidal cell layer, CB1R is found on large CKK-positive basket and SC-associated cells. A prominent expression 30 of CB1R in the IML of DG, particularly on presynaptic terminals often clustered around principal cell somata and proximal dendrites of CCK-positive INs, is evident (Marsicano and Lutz, 1999; Katona et al., 1999). Little or no CB1R was found on PV-positive hippocampal INs and calretinin-positive INs (Katona et al., 1999; Soltesz et al., 2015). Regarding glutamatergic neurons, CB1R protein is present at low levels in pyr- amidal neurons in CA1 and CA3 (Katona, 2006; Marsicano et al., 2003). Meanwhile, CB1R is absent from GCs of the DG, although low levels of CB1R were found in a subset of progenitor cells in the subgranular zone, which regulates proliferation and differentiation of these cells (Aguado et al., 2005; Galve-Roperh et al., 2007). However, with no doubt, among glutamatergic neurons, the highest CB1R levels are present in IML-located MC terminals of DG. MC axons express uniquely high levels of CB1R. This was elegantly shown in Monory et al. (2006), where they generated selective conditional CB1R knockout mouse models targeting GABAergic neurons, in which it is possible to note the remaining immunohistochemical labeling for CB1R in the IML (Figure 1.7). This labeling is reduced in selective CB1R knockout mice targeting glutamatergic cells. The LPP and MPP glutamatergic terminals express as well low but functional levels of CB1R (Wang et al., 2016; Peñasco et al., 2019). Additionally, CB1R immunoreactivity was also identified in the majority of hippo- campal cholinergic nerve terminals, controlling acetylcholine release (Degroot et al., 2006). The basic features of CB1R expression in the DG described so far appear to be largely conserved in gray mouse lemur, primate, and human brain (Ong and Mackie, 1999; Katona et al., 2000; Harkany et al., 2005) and along development (Kawamura, 2006; Hu and Mackie, 2015; Frazier, 2007). In the hippocampus, eCB synthesis and degradation enzymes show as well specific expression patterns. DAGLα, with a transmembrane topology, is expressed by hippocampal principal cells and is strikingly concentrated in dendritic spine heads within a perisynaptic annulus encircling the postsynaptic density of excitatory syn- apses. The expression level of DAGL mRNA is very high in CA1 and CA3 pyramidal neurons, and moderate to high in the GCs of the DG. DAGL protein is highly ex- pressed in the IML, indicating that it is targeted to neural processes of GCs after synthesis in the cytoplasm. Meanwhile, weakly labeled cells were scattered in the hilus of the DG, corresponding to MCs or GABAergic INs. INs from other areas or glial cells are not labeled (Katona, 2006). Differently, NAPE-PLD mRNA is expressed at the highest levels in GCs of the DG. NAPE-PLD protein is highly ex- pressed in presynaptic terminals of mossy fibers projecting to CA3 pyramidal cells, although mossy fibers do not express CB1R. Within mossy fibers, NAPE-PLD is localized predominantly on the endoplasmic reticulum, an intracellular Ca2+ store (Egertová et al., 2008; Kano et al., 2009). MAGL appears to be expressed in axon terminals. Stronger MGL immunoreactivity is found in basket cells around prin- cipal neurons and hilar neurons, and again in mossy fibers. Additionally, terminals of Schaffer collaterals innervating CA1 pyramidal cells express MAGL, while ter- minals of CA1 pyramidal cells innervating CA1 interneurons are devoid of MAGL. Interestingly, MAGL is expressed not only in CCK+/CB1R+ basket cell terminals, but also in PV+/CB1R- basket cell terminals (Kano et al., 2009). The pyramidal cells of the hippocampus show the strongest FAAH immunoreactivity. It is select- ive to somatodendritic elements near intracellular Ca2+ stores of principal neurons, but not of INs (Tsou et al., 1998; Gulyas et al., 2004). This expression pattern is 31 complementary to that of CB1R and MAGL. Figure 1.7: CB1R expression in hippocampus and DG. Immunohistochemical images of CB1R expression in wild-type (A–C), glutamatergic-CB1R knockout (D–F), GABA-CB1R knockout (G–I) and complete CB1R knockout mice (J–L).(B, E, H, and K) Higher magnification images of the areas enclosed in the square in (A), (D), (G), and (J), respectively. (C, F, I, and L) Detail of the CA3 hippocampal region. GC, granule cell layer; Hil, hilar region of dentate gyrus; LMol, (stratum lacunosum-molecularis); Mol, (stratum molecularis); Or, (stratum oriens); Pyr, pyramidal cell layer of hippocampus; Rad, (stratum radiatum). Asterisks indicate the IML. Bar, 100 m (A, C, D, F, G, I, J, and L); 25 µm (B, E, H, and K) (Monory et al., 2006). 1.2.2 Synaptic plasticity mediated by CB1R The main function ascribed to the ECS is the modulation of synaptic transmission and plasticity from presynaptic terminals. Synaptic plasticity refers to the activity- dependent modification of the strength and/or efficacy of synaptic transmission at preexisting synapses, and plays a central role in the capacity of the brain to incor- porate transient experiences into persistent memories. Diverse types of plasticity have been defined. Synapses can experience short-term plasticity (from milliseconds to minutes) and long-term plasticity (from hours to days), involving either a po- tentation (short-term potentiation or STP. and long-term potentiation or LTP) or a depression (short-term depression or STD, and long-term depression or LTD) of transmission. The synaptic potentials produced by the postsynaptic neuron upon neurotransmitter reception are classified as excitatory (EPSPs-excitatory postsyn- aptic potentials) o inhibitory (IPSPs-inhibitory postsynaptic potentials) due to their capacity to increase or decrease the action potential burst probability, respectively (Citri and Malenka, 2008). eCBs are produced and released from postsynaptic neurons phasically in an activity-dependent manner, or sometimes tonically under 32 basal conditions. The released eCBs diffuse in retrograde direction, activate pre- synaptic CB1Rs and suppress transmitter release either transiently (eCB-STD) or persistently (eCB-LTD) at both excitatory and inhibitory synapses in the CNS, in agreement with the ubiquitous expression of CB1Rs (Kano et al., 2009). eCB-STD eCB-mediated short-term changes in synaptic transmission (tens of seconds) encom- pass depolarization-induced suppression of excitation (DSE) and inhibition (DSI), depending on whether eCBs target glutamatergic or GABAergic terminals, respect- ively. DSI and DSE were first described and are well characterized in cerebellum (Purkinje cells) and hippocampus (CA1 pyramidal cells) (Llano et al., 1991; Pitler and Alger, 1992; Wilson and Nicoll, 2001). Thus, researchers noticed that a brief depolarization of principal neurons triggers a transient suppression of GABAergic synaptic input via eCBs released as retrograde messengers. Likewise, a DSI-like phe- nomenon at excitatory synapses was discovered in cerebellar Purkinje cells (Kreitzer and Regehr, 2001). Since then, DSI and DSE have been observed at synapses in many brain regions. These include inhibitory and excitatory synapses in the basolat- eral amygdala (Zhu, 2005), DG of the hippocampus (Isokawa and Alger, 2005; Chiu and Castillo, 2008), neocortex (Trettel and Levine, 2003; Bodor, 2005; Fortin and Levine, 2006), striatum and substantia nigra (Narushima et al., 2006; Yanovsky et al., 2003) or hypothalamus (Hentges, 2005), to mention some. Of particular in- terest for this Thesis is the DSE that occurs at the MC-GC synapse (Chiu and Castillo, 2008). This DSE was shown to be enhanced by agonists of cholinergic and group I metabotropic glutamate receptors (mGluR-I) and to be independent from DAGL activity. Moreover, FAAH blockade does not increase DSE either at the MC-GC synapse. In this same hippocampal domain, DSI at IN-GC and IN-MC synapses has been observed (Hofmann et al., 2006) (Table 1.1). Brain area Postsynaptic neuron Input/STD Dependence Independence Enhancement Hippocampus CA1 DSI CB1R, Ca2+, Gi (pre) G(post),PKA, RIM1α,PLC,DAGL AChR, mAChR, mGluR-I DSE CB1R CA3 DSI Ca2+ mGluR-II CCK-IN DSI CB1R GC DSI (CCK-IN) CB1R, Ca2+, Ca2+ store DSE (MC) CB1R, Ca2+ DAGL AChR, mGluR-I MC DSI CB1R, Ca2+ mAChR Cerebellum PC DSI CB1R, Ca2+, DAGL,CaMKII PLC, mGluR, GABAb mGluR-I DSE(CF) CB1R, Ca2+, DAGL mGluR, GABAb, A1 mGluR-I DSE(PF) CB1R, Ca2+, DAGL mGluR, DAGL GABAb, A1 BC DSE(PF) CB1R SC DSE(PF) CB1R GC No DSE(PF) Striatum MSN(Str) DSI CB1R, DAGL mGluR-I, mAChR(M1), mGluR-I No DSE MSN (SNr) DSI CB1R Cortex L2/3 PyC DSI CB1R, Ca2+ L5 PyC DSI DSE CB1R, Ca2+ mGluR-I Amigdala (BLA) Principal cell DSI CB1R Isolated cell DSI CB1R, Ca2+ Table 1.1: DSI/DSE in the brain. To note, DSI/E is not observed at all synapses where presynaptic CB1Rs are found. For example, in the dorsal striatum eCBs-dependent DSE is not observed at glutamatergic synapses known to express CB1R and other forms of eCB-dependent plasticity (Gerdeman et al., 2002; Kreitzer, 2005). CCK-IN, CCK-positive interneuron; I, inhibitory; E, excitatory; MCF, pre, presynaptic; post, postsynaptic; M1, muscarinic receptor 1; PC,purkinje cell BC, basket cell; SC, stellate cell; GC, Golgi cell; CF, climbing fiber; PF, parallel fiber; CaMKII, Ca2+/calmodulin-dependent protein kinase II; A1, A1 adenosine receptor. SNr, substantia nigra pars reticulata; MSN, medium spiny neuron; Str, striatum PyC, pyramidal cell; BLA, basolateral amygdala (Kano et al., 2009; Heifets and Castillo, 2009). Two major pathways cause eCB release during STD. The typical induction pro- tocol for triggering DSE and DSI is a direct postsynaptic neuronal depolarization 33 (or action potential generation), causing an influx of Ca2+ into the postsynaptic cell through voltage-gated calcium channel (VGCC) activation, mostly L-type channels. This increase in intracellular Ca2+ concentration may be amplified by recruitment of Ca2+ release from intracellular stores. The Ca2+ dependence of eCB release dur- ing DSI/E has been demonstrated in all brain structures, and it is necessary and sufficient for DSI/E induction. A second pathway for postsynaptic eCB production can be induced by other patterns of synaptic stimulation of excitatory afferents that trigger the activation of postsynaptic metabotropic receptors. Glutamate release onto postsynaptic mGluR-I, but also muscarinic acetylcholine receptors (mAChRs) and other metabotropic receptors (CCK-receptors, 5-hydroxytryptamine (5-HT)2 receptors) that couple to Gαq/11 proteins, can generate 2-AG by activating PLCβ and, subsequently, DAGL. Activation of these metabotropic receptors is sufficient to trigger eCB release for example in hippocampal neurons (Ohno-Shosaku et al., 2002; Fukudome et al., 2004). Nonetheless, although each pathway can usually be triggered independently of the other (Chevaleyre and Castillo, 2003), most likely a cooperative mechanism occurs in which Ca2+ influx through VGCCs and down- stream signaling from mGluR-I activation converge on the same metabolic pathway to mobilize 2-AG (Ohno-Shosaku et al., 2002) (Figure 1.8A). PLCβ is believed to act as a coincidence detector integrating two types of postsynaptic activation: metabotropic receptor activation and Ca2+ increase (Hashimotodani et al., 2005). Therefore, cannabinoid production is a complex and tightly regulated, activity- dependent process that integrates both postsynaptic depolarization and synaptic input to shape the strength of synaptic transmission. The newly synthesized eCB traverses the synaptic cleft and binds to presyn- aptic CB1Rs. The duration of DSI and DSE, less than 1 minute, appears to be directly related to the lifetime of the lipid in the synaptic cleft (Wilson and Nicoll, 2001). CB1R activation inhibits N-type Ca2+ channels by direct interaction with liberated G protein βγ subunits(Mackie and Hille, 1992; Qin et al., 1997) (Figure 1.8A). eCBs can also activate presynaptic voltage-dependent K+ channels and G protein-coupled inward-rectifying K+ channels (GIRK) by direct G protein subunit interaction (Daniel and Crepel, 2001; Daniel et al., 2004; Varma et al., 2002). Ad- ditionally, extracellular signal-regulated kinase (ERK)-mediated control of synaptic output by modulation of Munc-18, which is part of the vesicle release machinery, has been reported in hippocampus (Schmitz et al., 2016). Therefore, cannabinoid signaling in the presynaptic terminal can vary from one synapse to another or within the same terminal, and various transduction pathways may be recruited following CB1R activation for blunting membrane depolarization and exocytosis (Chevaleyre et al., 2006; Kano et al., 2009; Lovinger, 2008; Araque et al., 2017). eCB-LTD eCBs mediate as well long-term synaptic plasticity. Activation of CB1R and sub- sequent long-term reduction of transmitter release determines generally a presyn- aptic LTD. The first evidence implicating eCB signaling in LTD emerged at excit- atory synapses in the dorsal striatum (Gerdeman et al., 2002) and concurrently in the nucleus accumbens (Robbe et al., 2002). Since then, eCB-LTD has been repor- ted in several other brain structures such as the amygdala (Marsicano et al., 2002), hippocampus (Chevaleyre and Castillo, 2003), neocortex, (Sjöström et al., 2003), cerebellum (Soler-Llavina and Sabatini, 2006), ventral tegmental area (VTA) (Pan 34 et al., 2008), and brain stem (Tzounopoulos et al., 2007) (Table 1.2). There are some exceptions in which eCBs can mediate LTP by stimulation of astrocyte–neuron sig- naling in CA1 (Araque and Navarrete, 2010; Gómez-Gonzalo et al., 2015). Also an atypical LTP in the DG (Wang et al., 2016, 2018) was recently described. Brain area Postsynaptic neuron Input Induction protocol Hippocampus CA1 I HFS, TBS E (immature) HFS Cerebellum Stellate IN E 4 sets of 25 stimuli at 30 Hz every 0.33 Hz Purkinje E(PF) STDP (PF+CF stimulation) Dorsal Striatum MSN (Str) E LFS, STDP Cortex L2/3(somatosensory) E STDP (postsynaptic bursts) L5 PyC (visual) E STDP and LFS L2/3 (visual) E TBS L5/6(prefrontal) E Moderate 10 Hz stimulation for 10 min Amigdala (BLA) Principal cell E LFS Ventral tegmental area DA I Moderate 10 Hz stimulation for 5 min Table 1.2: eCB-LTD in the brain. L: Layer; E, excitatory; I, inhibitory; HFS, high-frequency stimulation (100 Hz); LFS, low-frequency stimulation (1 Hz); STDP, spike timing–dependent plasticity; TBS, theta burst stimulation. BLA, basolateral amygdala; MSN, medium spiny neuron; PyC, pyramidal cell; PF, parallel fiber; E, excitatory; I, inhibitory (Kano et al., 2009; Heifets and Castillo, 2009). For inducing eCB-LTD, a few minutes of CB1R stimulation are needed. There- fore, eCB-LTD is generated by different activation patterns allowing for associativity between synaptic events and long-lasting production of eCBs. The protocols used can be afferent-only repetitive stimulation protocols (high frequency stimulation - HFS, theta burst stimulation - TBS, etc) or pairing presynaptic stimulations with postsynaptic depolarizations yielding protocols of spike-timing dependent plasti- city (STDP). These protocols result in similar mechanisms for eCB release to those described for eCB-STD: a Ca2+ rise through L-type and T-type VGCCs, N- methyl-D-aspartate receptors (NMDARs) and intracellular stores and/or activation of mGluR-I (or other Gαq/11-GPCRs) on the postsynaptic compartment, acting syn- ergistically to trigger commonly a relatively longer-lasting eCB mobilization (Katona and Freund, 2012; Chevaleyre and Castillo, 2003). Released eCBs activate CB1R on the presynapsis of either the original afferent (homosynaptic eCB-LTD) or nearby afferents (heterosynaptic eCB-LTD) (Figure 1.8B). Once eCBs are released, eCB- LTD induction requires CB1R activation plus coincident presynaptic activity of the target afferent. Silent afferents do not undergo eCB-LTD. This provides a mech- anism of synapse specificity such that only active fibers undergo eCB-LTD. Also, restricting the amount and lateral extent of eCB release contributes to this synapse specificity of plasticity. However, once established, maintenance of eCB-LTD does not require continued CB1R activation. Accordingly, washing in a CB1R antagonist after eCB-LTD is established fails to reverse plasticity in all synapses where this manipulation has been tested (Heifets and Castillo, 2009; Chevaleyre and Castillo, 2003; Sjöström et al., 2003; Ronesi, 2004). The signaling pathways downstream CB1R activation that mediate transi- ent versus long-lasting depression differ due to the increased time requirement for eCB-LTD induction. Most evidence suggests that CB1R-dependent inhibition of the cyclic adenosine monophosphate (cAMP)/ protein kinase A (PKA) signaling pathway, through its coupling to Gαi/o, is a critical step in LTD at both excitat- ory and inhibitory synapses. This cascade triggers PKA-mediated changes in the component of the release machinery Rab3B/RIM1α and possibly several other pre- 35 Figure 1.8: Molecular mechanisms underlying endocannabinoid-mediated short- and long-term syn- aptic plasticity. A, eCBS-STD. Postsynaptic activity triggers Ca2+ influx via VGCCs. Ca2+ promotes DAGLα- mediated eCB production by an unknown mechanism. Presynaptic activity can also lead to eCB mobilization by activating postsynaptic I-mGluRs. PLCβ can now act as a coincidence detector (highlighted) integrating pre- and postsynaptic activity. 2-AG release which retrogradely targets presynaptic CB1Rs to reduce neurotransmitter release via βγ subunits and VGCCs activity. B, eCB-mediated excitatory and inhibitory LTD. Patterned presyn- aptic stimulation releases glutamate which activates postsynaptic mGluRs coupled to PLCβ and DAGLα. 2-AG homosynaptically targets CB1Rs localized to excitatory terminals and heterosynaptically at inhibitory terminals. A Gαi/o-dependent reduction in AC/PKA activity suppresses transmitter release. At inhibitory synapses together with calcineurin (CaN), a shift on the phosphorylation status of vesicle-associated targets (T) upstream Rab3B or RIM1α is depicted (Castillo et al., 2012). synaptic proteins involved in exocytosis of neurotransmitter (Castillo et al., 2012). Also, some other effectors downstream CB1R/PKA have been identified as Ca2+ influx via P/Q-type (Mato et al., 2008) and T-type (Qian et al., 2017; Ross et al., 2008). Alternative CB1R signaling cascades can involve actin cytoskeleton modu- lations (McFadden et al., 2018) and button structural changes. Protein synthesis is known to be required in some forms of eCB-LTD (Younts et al., 2016; Monday et al., 2020). While 2-AG seems to be the major eCB required for activity-dependent retro- grade signaling, the relative contribution of the main eCBs to synaptic transmis- sion is still debated. For STD is not completely clear. Evidence suggests that metabotropic receptor–induced STD is mediated by 2-AG that is synthesized via DAGL, whereas depolarization-induced STD is mediated either by another eCB or by a pool of 2-AG that is synthesized via a DAGL-independent pathway (Bisogno et al., 1999). Regarding the type of eCB mediating LTD, evidence exists for suppres- sion mediated by 2-AG in cerebellar Purkinje cells, CA1 pyramidal cells and VTA dopaminergic cells, and suppression mediated by AEA in basolateral amygdala and neocortex. However, AEA can participate in LTD at a slower rate than 2-AG by acting through CB1R (eCB-LTD) and through postsynaptic TRPV1 (AEA-TRPV1- LTD) in an autocrine manner. AEA-TRPV1-LTD is present at both glutamater- gic and GABAergic synapses (Chávez et al., 2010; Ohno-Shosaku and Kano, 2014). Stimulation of AEA production appears to be triggered through a pathway involving mGluR-I-activated PKA in amygdala (Azad, 2004) or, alternatively, involving PLC and dephosphorylation (Liu et al., 2006). Recent findings suggest that 2-AG and AEA can be recruited differentially from the same postsynaptic neuron depending on the type of presynaptic activity (Lerner and Kreitzer, 2012; Puente et al., 2011). 36 1.2.3 Biological functions of CB1R By modulating both excitatory and inhibitory synaptic strength, eCBs can regulate a large number of brain functions, including memory, cognition, emotions, feeding behaviours, pain and motor behavior (Figure 1.9). Deregulation of the ECS has been implicated, for example, in neuropsychiatric conditions, such as depression, stress or anxiety (Lutz et al., 2015; Mechoulam and Parker, 2013; Breivogel and Sim-Selley, 2009). Activity-dependent changes in synaptic efficacy play an important role in learn- ing and memory formation. The functional consequences of those changes depend, among other factors, on their direction (potentiation/depression), duration, timing, magnitude, and the type of synapse (excitatory/inhibitory) in which they occur. In this regard, CB1R occupies a position of great potential influence because it can mediate short- and long-term plasticity at both excitatory and inhibitory synapses, and in multiple locations such as hippocampus and amygdala (Chevaleyre et al., 2006). It is well known that cannabinoid agonists can interfere with long-term memory consolidation causing amnesic-like effects. CB1R inhibition can enhance memory (Puighermanal et al., 2012). But specific manipulations that elevate endo- genous cannabinoids do not consistently produce such impairments, and there are differences reported between systemic and localized administration of cannabinoid agonists. Neuronal network oscillations in the theta and gamma (30–80 Hz) ranges are integral to cognitive processes and behaviours. Exocannabinoids disrupt the finely tuned firing of neuronal ensembles and network oscillations. Amnesic effects can result from a generalized, circuit-independent activation of CB1R in the hippo- campus and other brain areas (Piomelli, 2003), mostly due to an imbalance between excitatory and inhibitory transmission (Busquets-Garcia et al., 2015). Alternatively, these cannabinoid-mediated impairments on memory have been ascribed to specific pools of CB1Rs (Han et al., 2012; Viñals et al., 2015). Conversely, some eCB- mediated forms of plasticity show a cognitive-enhancing action. CB1R is required for memory extinction of aversively motivated learning in auditory fear conditioning tests (Marsicano et al., 2002). It is also known to facilitate forms of hippocampus- dependent associative learning by inducing DSI or LTD of inhibitory transmission, thereby facilitating subsequent induction of LTP at excitatory inputs required for memory acquisition (Chevaleyre and Castillo, 2004; Carlson et al., 2002) a process called metaplasticity. eCB-STD might regulate the pattern and timing of neuronal activity in discrete circuits. It can alter and/or equalize synaptic efficacy of incoming afferent information across dendrites to influence behaviour. DSI can be induced in vitro by levels of neural activity that could also be encountered in vivo (Melis, 2004). In this manner, eCBs can regulate hippocampal gamma oscillations or hippocampal cell assemblies that work as an engram (Wilson, 2002; Piomelli, 2003; Freund et al., 2003) to synchronize network rhythms that are essential for cognitive processes and normal brain function. In conclusion, CB1R is believed to be transiently activated at distinct synapses that have been stimulated beyond a certain threshold to fine- tune those specific networks. Although exogenous cannabinoids generally disrupt physiological network rhythms, eCBs have complex, and sometimes opposing mod- ulatory effects on neuronal circuit behaviour. eCB-dependent forms of plasticity show physiologically a role of maintenance of balance in memory processes. Another function of CB1R is the homeostatic control of stress and anxiety by acting as a gatekeeper. Tonic eCB signaling constrains hypothalamic–pituitary–adrenal 37 Figure 1.9: Main physiological functions regulated by endocannabinoid system. CB1R is involved in the regulation of a multitude of behavioural functions, from which the best-established are shown in the schematic figure. Purple intensity correlates with CB1R protein expression levels (Flores et al., 2013). (HPA) axis activity, ultimately habituating the stress response and restoring homeo- stasis, thus avoiding deleterious consequences (Morena et al., 2016; Mechoulam and Parker, 2013). CB1R influences as well affective states in physiological and pathological conditions. Exogenous cannabinoids influence anxiety-like behavior in a biphasic manner, with low and high doses of THC exerting anxiolytic and anxio- genic states, respectively, in both animals and humans. It has been reported that the anxiolytic effect of cannabinoids at low doses depends on the presence of CB1R on cortical glutamatergic neurons, whereas the anxiogenic effect of higher doses is mediated by CB1R on forebrain GABAergic neurons (Rey et al., 2012; Ruehle et al., 2013; Lutz et al., 2015). Both STD and LTD are involved in sensory processing in somatosensory cortex (Sjöström et al., 2003; Maglio et al., 2018). CB1R signaling in the forebrain and sympathetic neurons controls thermogenesis and energy balance by triggering energy storage. Its blockade alleviates obesity-related metabolic disorders (Quarta et al., 2010; Piazza et al., 2017). However, CB1R activation can regulate feeding behavior again in a biphasic manner: ventral striatal CB1Rs in GABAer- gic neurons exert a hypophagic action. Conversely, CB1Rs modulating excitatory transmission (in olfactory afferences) mediate the well-known orexigenic effects of cannabinoids (Bellocchio et al., 2010; Soria-Gómez et al., 2014). Receptor activation is also involved in sleep regulation as a sleep-inducer (Mechoulam et al., 1997). Ad- ditionally, it can control nociception as well (Cravatt and Lichtman, 2004). CB1R plays a role in cerebellar-based forms of learning and memory including pavlovian conditioning and adaptation of reflexes and in striatal-dependent motor behavior (Kishimoto, 2006). It facilitates reward-based learning of a motor sequence by con- ferring the flexibility with which animals can switch between strategies (Tanigami et al., 2019). eCBs are considered as well bidirectional (depression vs. potentiation of synaptic strength) neuromodulators of striatal functions (Cui et al., 2016). Both, brain aging and neurodegenerative diseases are associated to drastic changes in the ECS. CB1R activation has been shown to exert neuroprotection in a vari- ety of in vitro and in vivo models of neurodegeneration as global ischemia, multiple sclerosis, traumatic brain injury, Huntington’s disease, etc (Kano et al., 2009). CB1R 38 activation can act against neuronal death by activating cell survival signalling cas- cades such as brain-derived neurotrophic factor (BDNF) transcription or ERK/c-fos, enhancing brain microcirculation, and controlling microglial function. Independ- ently of CB1R engagement, various cannabinoids with phenolic structure exert a CB1R-independent neuroprotective effect owing to their intrinsic antioxidant prop- erties (Sinor et al., 2000). Nevertheless, it is mainly CB1R that plays an intrinsic protective role in suppressing pathologic neuronal excitability, and so a key manner for CB1R to safeguard the CNS against excitotoxicity is the suppression of glutama- tergic transmission (Chiarlone et al., 2014). The role of eCBs and CB1R in brain hyperexcitability, seizures, and epilepsy has also been investigated. Epilepsy is a neurological disorder characterized by a pre- disposition to recurrent, unprovoked seizures and imbalanced synaptic input, which may cause an excessive neuronal activity that transform normal brain synchron- ized activity and eventually leads to neuronal death and synaptic reorganization (Karlócai et al., 2011). Febrile seizures are the most common seizures during child- hood. Epileptic syndromes can be partial (occurring within localized brain regions) or generalized (occurring in the entire forebrain) (Lutz, 2004). To model epilepsy, chemo-convulsants (for example, kainic acid and pilocarpine) are injected into ro- dents; subsequently, intermittent or continuous acute seizures lasting for tens of minutes to a few hours occur and are followed by a seizure-free period, which is then interrupted by the emergence of spontaneous chronic epileptic seizures by days and weeks later. The seizure activity induces permanent neuronal plasticity changes, a process called epileptogenesis, resulting in neuronal hyperexcitability and a recurrent seizure activity. Such neuronal insults evoke eCB release to exert an overall pro- tective effect against over-excitation (Panikashvili et al., 2001; Wettschureck et al., 2006; Marsicano et al., 2003), whereas perturbations of the ECS lead to reduced seizure thresholds and increased incidence of epileptic seizures (Wettschureck et al., 2006). Epilepsy can substantially modify the ECS. This system can both alter and be altered by epileptiform activity in a wide range of in vitro and in vivo models of epilepsy. A temporal pattern of changes in eCB levels and CB1R expression has been observed in epilepsy models. In the pilocarpine model in rats, CB1Rs are markedly downregulated throughout the hippocampus in the acute phase (which is shortly after the initiating insult) but upregulated in the chronic phase of the disorder (Falenski et al., 2009; Wallace et al., 2003). Also, both 2-AG and AEA levels are increased in the acute phases of pilocarpine- and kainate (KA)-induced seizures (Wallace et al., 2003; Marsicano et al., 2003). Importantly, such modifications are cell type-specific. In experimental fever-induced seizures model, CB1R mediated retrograde signalling increased selectively at inhibitory, but not excitatory synapses that showed spontaneous seizures. This was not due to sprouting of CB1R-positive axons but to an increase in CB1R abundance (Chen et al., 2003, 2007). Similar results were obtained in pilocarpine-induced epilepsy in rats, where a downregulation of CB1Rs in the IML of the DG was reported and it was not entirely due to MC loss (Wallace et al., 2003; Falenski et al., 2007, 2009). In samples of patients with chronic epilepsy, a decrease in CB1R expression ratio in the IML specific to glutamatergic synapses was reported. This was the first evidence of a potential impairment of the ECS in patients with epilepsy, and implicates that decreases in CB1R expression at glutamatergic synapses may be involved in the etiology of epilepsy (Ludanyi et al., 39 2008). Finally, data from cell type-specific knockout mice indicate that the selective deletion of CB1R on excitatory cells, but not on GABAergic INs, worsens the KA- induced model of acute seizures (Monory et al., 2006; Soltesz et al., 2015; Augustin and Lovinger, 2018). The eCB-mediated negative feedback control system seems to be functionally compromised in chronic ongoing epilepsy, either because of alterations in eCB syn- thesis pathways or because of decreased CB1R expression on glutamatergic termin- als, and is therefore unable to prevent the generation of seizures. CB1R agonism gen- erally exerts anti-convulsant, antiepileptic and anti-epileptogenic effects in animal models of epilepsy, presumably by decreasing glutamatergic transmission (Wallace et al., 2002, 2003; Marsicano et al., 2003; Monory et al., 2006) but with some excep- tions (Whalley et al., 2019). CB1R antagonists can be proconvulsant, but exhibit anti-epileptogenic effects (to prevent the long-term increase in seizure susceptibil- ity) if employed during a precise time window (Echegoyen et al., 2009; Keeler and Reifler, 1967). This might be caused by a CB1R dependent disinhibition ascribed to GABAergic terminals (Lutz, 2004). Cannabinoids as therapeutic agents The potential therapeutic applications of cannabinoids currently constitutes a widely debated issue with ample scientific and clinical implications (Pertwee, 2012). The content of THC and the ratio of THC to CBD (which limits THC psychoactive effects) are two key factors that define the efficacy and safety of cannabis prepar- ations. A variety of formulations has been used, with differing amounts of THC and CBD and routes of administration: for example, pills, mucosal sprays, and vaporized or smoked cannabis. Nowadays, it is permitted to prescribe capsules of THC (dronabinol; Marinol®) or nabilone (Cesamet®), a synthetic THC analogue, to inhibit nausea and vomiting, and to attenuate appetite loss and cach- exia/energy expenditure, in cancer or AIDS patients. Another cannabinoid-based medicine, an oro-mucosal spray composed of a cannabis plant extract enriched in THC and CBD at equimolar amounts (nabiximols; Sativex®), is currently ap- proved in Canada for the management of multiple sclerosis-associated neuropathic pain and opioid-resistant cancer pain, and in many countries worldwide (including Spain) for the treatment of multiple sclerosis-associated spasticity, painful spasms and bladder dysfunction. An additional cannabinoid preparation (CBD; Epidi- olex®) has gained recent approval by the FDA and EMA to palliate seizures associated to some pediatric refractory epileptic syndromes (O’Connell et al., 2017). A number of clinical studies with cannabinoids in other diseases are currently in pro- gress (Koppel et al., 2014). CB1R antagonists (e.g., rimonabant) showed promise as pharmacotherapeutics for obesity and addiction, but most clinical trials were ter- minated in 2008 over concerns of depression and suicide (Breivogel and Sim-Selley, 2009). For epilepsy, some clinical cases showed that THC but mainly CBD may have anti-epileptic effects in some individuals, used as an adjunctive medicine to prevent epileptic seizures (Lutz, 2004). With a high CBD/THC ratio, medical cannabis for deleterious epilepsies such as Dravet syndrome has been successful, apparently without major side effects (Maa and Figi, 2014). However, safe and side effect-free future cannabinoid-based medications are likely to be those that can be developed to target cannabinoid receptors and related signalling molecules with high cell-type, temporal and spatial selectivity (Soltesz et al., 2015). 40 1.2.4 Context-dependent signaling by CB1R The complex and even biphasic fine-tuning of physiological events and the wide range of behavioral consequences of CB1R activity in the brain can be explained by the fact that CB1R cell signaling is highly dependent on the context. CB1R couples to heterotrimeric G proteins to modulate multiple downstream signaling pathways (Figure 1.10). The canonical signaling pathway involves the ac- tivation of Gαi/o family members (Gαi1, Gαi2, Gαi3 and Go1 and Go2) and the sub- sequent dissociation into the respective αi/o and βγ subunits (Howlett and Fleming, 1984; Howlett et al., 1986). Activation of the αi/o subunit inhibits adenylyl cyclase (AC), decreases cAMP levels and hence reduces the activity of cAMP-dependent cytoplasmic proteins, especially PKA (Howlett, 1985) and, in some instances, the exchange protein directly activated by cAMP (Epac) (Ramı́rez-Franco et al., 2014). PKA regulates in turn other kinases and transcription factors (Demuth and Molle- man, 2006; Bosier et al., 2010), for example focal adhesion kinase (FAK) and Raf-1. This can lead also to the activation of ERK (Davis et al., 2003) and the concerted regulation of nuclear transcription factors and gene expression involved in synaptic plasticity, cell survival and differentiation. Further intracellular effects of this limb include modulation of small GTPases and their effectors as Rho-associated coiled- coil containing protein kinase (ROCK), net dephosphorylation of channels and ac- tivation of the A-type K+ currents to hyperpolarize the membrane and cause a depression in neurotransmitter release (Childers and Deadwyler, 1996; Deadwyler et al., 1993, 1995). On the other hand, the βγ dimer modulates as well different sig- naling pathways. βγ stimulates PLC and PLA2, as well as calcium influx (Hunter et al., 1986), inhibits N-type and P/Q type VGCCs (Pan et al., 1996; Twitchell et al., 1997) and stimulates GIRK channels (Vásquez et al., 2003). It also mod- ulates MAPK superfamily, including ERK (Bouaboula et al., 1995; Galve-Roperh et al., 2002), p38 MAPK (Liu et al., 2000; Rueda et al., 2000), and c-Jun N-terminal kinases (JNK) (Rueda et al., 2000; Derkinderen et al., 2001). This modulation is triggered for example by fosfatidilinositol-3-kinase (PI3K)-dependent phosphoryla- tion and activation of Akt (del Pulgar et al., 2000). PI3K/Akt can alternatively activate glycogen synthase kinase 3 (GSK3) (Ozaita et al., 2007) or the well-known mammalian target of rapamycin complex 1 (mTORC1) to regulate protein trans- lation (Puighermanal et al., 2009; Younts et al., 2016) and autophagy (Blázquez et al., 2020). Cannabinoids also regulate other Gαi/o protein-independent path- ways involved in the control of cell proliferation and survival, including activation of ceramide-generation pathways, in which CB1R activation leads to ceramide gener- ation through sphingomyelin hydrolysis or synthesis de novo (Galve-Roperh et al., 2000), the former via the adaptor protein factor associated with neutral sphingomy- elinase activation (FAN) (Sánchez et al., 2001). Prolonged exposure to receptor agonists leads to a decreased ability of CB1R to further activate effectors’ pathways, namely receptor desensitization, and to a reduction in the number of receptor molecules on the cell surface, namely receptor internalization (Lu and Mackie, 2016). After activation, CB1R is endocytosed from the cell surface into endosomes to facilitate ligand detachment. CB1R desensit- ization involves GPCR kinase 2/3 (GRK2/3)-mediated phosphorylation of mostly two serine residues on the receptor cytoplasmic domains (S426 and S430), which recruits β-arrestins 1/2. However, phosphorylation of a combination of several serine and threonine residues in the C-terminal domain that modulate arrestin re- 41 Figure 1.10: Main signaling cascades controlled by CB1R activation. Several pathways activated by CB1R signaling are depicted. To denote some major examples, CB1R can inhibit voltage-dependent Ca2+ channels (VDCC) or activate GIRKs through βγ subunits. β-arrestins participate in CB1R internalization and desensit- ization, while GPCR-associated sorting protein 1 (GASP1) directs this receptor toward lysosomal degradation. The canonical activation of MAPKs (ERK, JNK, p38) by βγ subunits and inhibition of AC and protein PKA by Gαi/o are shown. Alternatively, CB1R can activate RhoA/ROCK cascade siganlling by Gαi/o and mTORC1 signaling by βγ-dependent recruitment of PI3K-Akt. cruitment can be required as well, and the studies that report a direct interaction are scarce. Binding of β-arrestins prevents the receptor from G protein coupling, dampens receptor-evoked signalling, recruits components of the endocytic machinery to initiate receptor endocytosis and triggers chlatrin-mediated receptor internaliz- ation (Howlett et al., 2010). The pattern and rate of endocytosis of CB1R varies according to the cell type or membrane subdomain. For example, differential endo- cytosis rates between axons and the somatodendritic compartment contributes to the axonally-polarized surface expression of CB1R, although there are data support- ing that CB1R is established early in the secretory pathway by directed trafficking of CB1R-containing vesicles to axons (Fletcher-Jones et al., 2019). β-arrestins 1 and 2 are expressed ubiquitously and have two major roles, one as negative regulators of heterotrimeric G protein signaling during receptor desens- itization and internalization, and another as signaling scaffolds. CB1R-β-arrestin complexes can work as transduction platforms by triggering signaling pathways such those reliant on MAPKs or cAMP responsive element-binding (CREB) (Delgado- Peraza et al., 2016). CB1R signaling can conceivably change over time in three different waves (Nogueras-Ortiz and Yudowski, 2016). The first wave is transient (few minutes) and involves heterotrimeric G proteins. The second wave is mediated by β-arrestins. Finally, the third wave takes place in the intracellular compartments and can occur either by G proteins or β-arrestins. However, β-arrestins 1 and 2 play different roles in signalling and internalization, respectively. β-arrestin 2 may not have a primary role in ERK signaling but be critical for receptor internalization (Delgado-Peraza et al., 2016). Moreover, the roles of both arrestins in CB1R desens- itization and tolerance are highly dependent on the brain region and agonist type (Nguyen et al., 2012; Breivogel and Vaghela, 2015; Henderson-Redmond et al., 2020). Both GPCRs and β-arrestins can be modified post-translationally in several ways, 42 which regulates their stability and activity. One such well-studied modification is ubiquitination (Rajagopal and Shenoy, 2018). The desensitizing action of β-arrestins on CB1R can occur in the short term, over minutes, and is primarily associated with β-arrestins preventing G protein in- teraction with the GPCR; the receptor is then mostly recycled back to the surface. But, also, receptor can traffic for degradation in lysosomes. This is a longer-term de- sensitization (over hours to days), referred to as downregulation, which is a key mo- lecular process unlerlying cannabinoid tolerance (Rajagopal and Shenoy, 2018). Tolerance to the effects of cannabinoids causes a rapid attenuation of behavioural re- sponsiveness. It decreases the efficacy of cannabinoid-based medicines progressively, especially upon use over long time periods. Cannabinoid tolerance is mainly attrib- uted to pharmacodynamic changes, especially a decrease in the number of total and functionally-coupled CB1R molecules on the cell surface, with a minor pharmacokin- etic component caused by enhanced cannabinoid biotransformation and excretion (Grotenhermen, 2003; Panlilio et al., 2015). That being said, it is certain that the CB1R seems to have only a few “intrinsic” signaling properties, but its effects largely “emerge” from specific temporal and spa- tial constraints largely depending on the “context” (cell type, subcellular location, cellular functional state, nature and dose of the ligand, etc.) where it is activated. Hence, CB1R signaling is highly pleiotropic and we cannot assign the same cascades or effects to the receptor at the different locations where it is present. This issue is discussed here below. First, signaling potency is not an intrinsic property of the CB1R, but it de- pends on the cell type or subcellular compartment where it is expressed. For ex- ample, expression levels are relatively low in hypothalamic regions (Wittmann et al., 2007) but higher activity of cannabinoid-dependent signaling occurs. This is ex- plained by the fact that CB1R potency (number of G proteins activated per CB1R) is highest in hypothalamus, implying that stimulation of fewer CB1R receptors is sufficient to achieve biological effects (Breivogel et al., 1997). Likewise, hippocam- pal GABAergic INs contain higher levels of CB1R than glutamatergic neurons. But there is much higher coupling to Gαi/o protein-dependent signaling of CB1R in glutamatergic neurons than in neighboring GABAergic INs (Steindel et al., 2013). And indeed, this may account for the dose-dependently biphasic actions of can- nabinoid receptor agonists, for example in anxiety. However, the specific cellular factors that determine increased coupling efficiency are not known. Perhaps some intracellular proteins serve to tether Gαi/o protein subunits in the vicinity of the receptor and increase signaling potency. CB1R interactions with non-Gαi/o proteins have been shown in particular con- texts (Diez-Alarcia et al., 2016). Thus, the receptor is capable of signalling in brain cells through Gαs, which stimulates AC and PKA (Howlett et al., 2010) in some contexts like dopamine 2 receptor (D2R) and CB1R coexpression in a subpopulation of neurons (Glass and Felder, 1997). Blockade of Gαi/o in the globus pallidus can switch the effect of CB1R toward activation of Gαs and potentiation of neurotrans- mission (Caballero-Florán et al., 2016). It also can signal through Gαq/11 (Lauckner et al., 2005) which canonically stimulates the PLC cascade, for example when ex- pressed in astrocytes (Navarrete and Araque, 2010). Alternatively, CB1R can couple to Gα12/13 (Roland et al., 2014), which regulates small Rho family GTPases, for ex- ample when it is expressed in axonal growth cones (Roland et al., 2014). 43 Some other factors may account for the differential G protein signaling of the re- ceptor. Biased agonism for CB1R signaling according to ligand type exists. CB1R can bind a wide range of structurally different endogenous or exogenous ligands that favor a specific conformation of the receptor. Consequently, they might preferen- tially direct the receptor signaling toward a specific pathway (Diez-Alarcia et al., 2016; Priestley et al., 2017; Laprairie et al., 2014). CB1R can as well transactivate other receptors as tyrosine kinase receptors (Dalton and Howlett, 2012). Addition- ally, CB1R have been shown to homo- and heteromerize (Mackie, 2005; Callén et al., 2012) with other GPCRs - e.g., D2R (Kearn et al., 2005); µ opiod receptor - MOR (Rios et al., 2006); adenosine A2A receptor - A2AR (Moreno et al., 2018). Finally, the presence of specific cytoplasmic CB1R-interacting proteins, which will be extensively described below as it is central to this Thesis, could play a role in the cell-specific modulation of cannabinoid signaling. Several studies indicate that interaction between CB1R and intracellular proteins is established with the CB1R-CTD (Figure 1.11). CTD studies help identifying structural motifs involved in activities and direct interactions with the receptor. The whole CB1R-CTD encompasses 73 residues in humans delimitated from R400 to L472. It shows two cysteine residues (C415 and C416 in rat) which, if palmitoylated, may act as membrane anchors. Additionally, CTD shows 11 S and 5 T phosphorylat- able residues recently described as a “bar code” (Booth et al., 2019) required for protein binding modulation in signaling processes and internalization (as for β- arrestin recruitment) (Zhou et al., 2017). The proximal part of the CTD presents a highly conserved NPXXY motif at the C -end of the TM7 considered to be critical for receptor activation. Next, a conserved short α-helix named helix 8 (Hx8) en- compassing residues S401 to F412 appears. It is observable in the crystal structure of the receptor disposed in parallel to the cellular membrane with a hydrophobic face of the helix oriented toward the membrane and linking perpendicular to TM7. It is the most commonly studied region. H8 domain plays a role in docking specific G-protein subunits and in the transition between conformational states. It contrib- utes to receptor stabilization during biosynthesis, proper trafficking and folding of the receptor, homodimerization, endoplasmic reticulum exit, cell surface expression, G protein coupling, β-arresting binding, and internalization (Mukhopadhyay et al., 2000; Anavi-Goffer et al., 2007). An additional helix, Hx9, in the CTD has been identified. It is located towards the end of the CTD, between central and distal parts of the CTD and encompassing residues A440-M461. It is as well perpendicu- lar to the TM7 bundle and lying on the membrane surface. This complex has been postulated to comprise a Gαq-binding site (Fletcher-Jones et al., 2019). Residues carboxyl-terminal to H9 were found to be unstructured (Stadel et al., 2011). Distal and central parts of the CTD flanking the helices contain the multiple S and T phosphorylation sites referred above (Booth et al., 2019). 1.2.5 CB1R-interacting proteins GPCRs commonly present receptor-selective partners (apart from G proteins, GRKs, arrestins and other GPCRs) that mediate receptor signaling, modulate it through G proteins and/or influence receptor pharmacology. Additionally, they can direct post-endocytic trafficking of GPCRs to proteolytic and/or lysosomal degradation or back to the plasma membrane, or anchor GPCRs in particular subcellular loc- 44 ations. Since many of these receptor-selective partners exhibit restricted patterns of tissue expression, these interactions can help to explain many examples of cell- specific fine-tuning of GPCR functional activity. To mediate GPCR signaling, the interaction with the receptor should be regulated by agonist stimulation to initiate a response. Examples of this phenomenon are the interaction of the non-receptor tyrosine kinase Janus kinase 2 (JAK2) with the angiotensin AT1 receptor, or the interaction between the sodium hydrogen exchanger regulatory factor (NHERF) scaffold proteins and the parathyroid hormone receptor (PTH1R) leading to a pref- erential enhancement of the downstream Gαq/11 signaling. When PTH1R is in a different cell type or a separate cellular compartment in which NHERF proteins are absent, PTH1R instead preferentially signals through Gαs. In contrast, a number of other GPCR-interacting proteins associate with receptors in an agonist-independent manner. Such proteins can increase the speed and efficiency of GPCR signaling by acting as scaffolds to tether downstream effectors in close proximity to the receptor. Examples include a variety of PDZ scaffolds, such Homer proteins, which associate with mGluR-I and intracellular Inositol triphosphate (IP3) receptors in postsynapses to increase the efficiency of mGluR-stimulated calcium signaling. Conversely, pro- tein partners can decrease the intensity, efficiency and/or kinetics of GPCR signaling by disrupting association with G proteins or, in some cases, recruiting negative reg- ulators of GPCR signaling like arrestins, calmodulin in some cases, or periplakin with the MOR (Ritter and Hall, 2009). Several studies have aimed to define the CB1R interactome, namely all the interactions or associations that the receptor establishes with other proteins at a certain moment in the cell. Some specific CB1R-interacting proteins have been identified that bind to particular regions of the receptor CTD (Smith et al., 2010). They can greatly determine differential eCB signaling in the brain. As previously mentioned, the protein FAN was defined as an adaptor protein for CB1R to activate neutral sphingomyelinase, which leads to acute ceramide accumulation in the cell (Sánchez et al., 2001). The putative CB1R-binding site ascribed to FAN is located in the CB1R-CTD, from residues 431 to 435 (Stadel et al., 2011) (Figure 1.11). CB1R can interact as well with GPCR-associated sorting protein (GASP1). The region of interaction proposed is the last 14 residues of the CTD, although it is still not clear; binding to the conserved F408/R409 motif in H8 has been described too. GASP1 interaction promotes the endocytic targeting of agonist-internalized CB1R to lysosomes, and it has been implicated in tolerance to cannabinoids (Martini et al., 2010). Adaptor protein 3 (AP3) interacts with CB1R and modulates CB1R intra- cellular trafficking via endosome-derived synaptic vesicles and divert CB1R away from plasma membrane localization (Howlett et al., 2010). The best example of CB1R-interacting proteins is the extensively studied can- nabinoid receptor interacting protein-1a (CRIP1a), a protein of 164 amino acids in humans that is able to bind the CTD of CB1R. There are two alternative spli- cing forms for Cnrip1a, yielding two different mRNA molecules, one that originates CRIP1a (164 amino acids) and the other CRIP1b (128 amino acids), being CRIP1b a specific isoform for primates (Fletcher-Jones et al., 2019). CRIP1a was sought and discovered when deletion of the distal C-terminal amino acids augmented the CB1R-evoked constitutive inhibition of the voltage-dependent Ca2+ current in su- perior cervical ganglion neurons, thus suggesting that the distal CTD induced an inhibitory function. By reasoning that the CTD binds to a regulatory protein, a 45 yeast two-hybrid screening was performed with the last C-terminal 55 residues of CB1R (from 418 to 472) as bait. CRIP1a was identified as a key CB1R-associated protein. Co-immunoprecipitation of CRIP1a with CB1R demonstrated that these two proteins form a complex in solubilized rat brain membranes. Further deletion mapping studies demonstrated that the distal C-terminal nine amino acids of the CB1R constituted the minimum domain required for CRIP1a binding (Niehaus et al., 2007), five aminoacids 467, 468,469, 472, 473 being essential (Mascia et al., 2017) (Figure 1.11). However, residues from central regions of the CTD, and maybe also in IL3, play a role in CRIP1a binding too. The primary structure of CRIP1a has been identified and found to be highly conserved across species. However, there appears to be no amino acid sequence homology with other known proteins, making hard to model the interaction with CB1R-CTD, although it contains a palmitoylation site that may facilitate anchoring to the membrane and association with CB1R (Booth et al., 2019). Figure 1.11: Schematic diagram of the human CB1R-CTD. The relevant portion of TM7 is shown. The NPXXY motif, helix 8, and helix 9 are highlighted in cyan, purple, and green, respectively and central and distal regions within CTD are indicated. Two domains implicated in desensitization and internalization are indicated. A putative palmitoylated cysteine (depicted by a black zigzag line) at position 415 aids Helix 8 association with the inner leaflet of the plasma membrane. Phosphorylation sites are indicated by red filled circles involved in desensitization in central region and desensitization in distal region. Also CTD regions of interaction with β- arrestins, FAN, GASP1 and CRIP1 are indicated by lines in fuschia, orange, grey, and blue respectively. The exact binding site of SGIP1 is unknown, but it binds downstream of A419 (Stadel et al., 2011). Subsequent studies in cell culture models have demonstrated that CRIP1a inhib- its both constitutive and agonist-stimulated CB1R-mediated G-protein activity and its signalling cascades (cAMP/PKA and ERK) (Blume et al., 2015), and that it fine- tunes agonist-dependent membrane trafficking of the receptor. CRIP1a co-localizes with CB1R at the plasma membrane, traffics to the same subcellular compartments (Niehaus et al., 2007) and attenuates agonist-induced internalization of cell surface 46 CB1R, without affecting total CB1R labeling. It can compete with β-arrestin for the same region in CB1R-CTD and so diminishes β-arrestin-mediated CB1R internaliz- ation (Fletcher-Jones et al., 2019; Mascia et al., 2017; Blume et al., 2016, 2017). The neuroanatomical distribution of CRIP1a and CB1R was studied in the cerebellum (Smith et al., 2015) and hippocampal formation, where both proteins are highly co- expressed in pyramidal neurons and interneurons of the CA1 region (Guggenhuber et al., 2015). The later study is the only one to date that proposes some physiological role for the CB1R-CRIP1a interaction in vivo; however, in contrast with previous findings, it strikingly shows that CRIP1a enhances cannabinoid-induced G protein activation and depression of excitatory currents in CA1 pyramidal neurons (Gug- genhuber et al., 2015). Finally, Src homology 3-domain growth factor receptor-bound 2 (GRB2)-like (endophilin) interacting protein 1 (SGIP1) was identified as a new CB1R part- ner. SGIP1 is specifically expressed in the mouse CNS. It was initially identified in a screen for CNS regulators of energy balance and obesity, and it was found upregulated in the hypothalamic regions of an obese mouse line. SGIP1a suppres- sion reduced body weight in obese mice via reduced food intake. In line with this observation, SGIP1 was associated with measures of obesity such as fat mass in humans (Tehran et al., 2019). SGIP1 mRNA and CB1R mRNA are highly coincid- ent, but expression levels differ among brain regions (Allen Brain Atlas). SGIP1α is a splice variant of SGIP1 expressed as well in CNS with two small insertions of 18 and 20 amino acids. Later on, this protein was shown to be associated to the CTD of CB1R. It was discovered in a yeast two-hybrid experiment using CB1R-CTD as bait. SGIP contains a µ-homology domain (µHD) known to establish protein- protein interactions. The last 100 carboxi-terminal amino acids in this domain were responsible for CB1R interaction. SGIP1 can interact in this domain with scaffold protein Eps15, which is involved in chlatrin machinery, or with synaptotagmin, to sort neurotransmitter vesicles, by different regions of the domain. In fact, it is pos- tulated that SGIP1 µHD interaction with CB1R might displace synaptotagmin but simultaneously bind Eps15 and CB1R-CTD. The N-terminal domain of SGIP har- bors a region for binding to membrane phospholipids (Lee et al., 2019; Tehran et al., 2019). Since its identification, SGIP1 has been shown to bind liposomes and inter- act with a number of adaptor proteins with roles in endocytosis. In co-transfected HEK293T cells, SGIP1 increases stability of the cell surface CB1R pool and impedes endocytosis of activated and non-activated CB1R. β-Arrestin-associated signaling is changed profoundly, since SGIP triggers an increase in β-arrestin 2 association most likely as a consequence of the prevention of chlatrin-mediated receptor internaliza- tion. This would decrease agonist-mediated ERK phosphorylation without altering Gαi/o and Gαq/11 activation. Therefore, SGIP influences CB1R signaling in a biased manner. SGIP1 and CB1R co-localize on presynaptic portions of afferent neurites of primary cortical neurons. There, SGIP1 modulation of CB1R function may ex- plain the dual internalization behavior in distinct neuronal compartments, in which a rapid CB1R internalization is characteristic for the pool of CB1Rs located in the neuronal soma, while the CB1Rs in axons, and especially at synapses, are resistant to internalization, presumably by SGIP1 association (Hájková et al., 2016). However the physiological consequences of this interaction remain enigmatic yet. The identification of specific proteins associated to CB1R is a key step to under- 47 stand the precise signalling events triggered by different receptor agonists. Moreover, by finely tuning the spatial and temporal resolution of signaling, certain CB1R inter- actors could dramatically affect the ability of the receptor to transduce extracellular stimuli into changes in cellular physiology. Understanding this issue constitutes the core goal of this Thesis. 48 Aims As explained in the Introduction, there are striking differences in the signalling events triggered by CB1R in distinct environments, among various cell types or under different physiopathological situations that, in turn, can execute context-related al- terations in synaptic transmission and patterns of neuronal activity. However, these differences still remain largely unexplained. CB1R-evoked action can be modulated in different manners, one of them being conceivably its association to intracellular proteins through its large cytoplasmic CTD. Studies aimed at discovering CB1R- interacting partners are scarce and have been mostly unable to proceed beyond the mere biochemical demonstration of the protein-protein interaction and, thus, have not shown yet any key physiopathological impact of the respective putative inter- actions. Hence, unfortunately, the assessment of the physiological relevance and therapeutic potential of distinct CB1R pools is still hampered, at least in part, by the lack of knowledge of possible context-specific CB1R intracellular interactors. It would be thus extremely interesting to test whether specific interactions with still unknown cytoplasmic proteins in the presynaptic terminal might alter CB1R function, possibly in a cell-type-dependent manner, under different physiological or pathological conditions. Based on this background, the general aim of the present Doctoral Thesis is to discover and characterize in detail new and spatio-temporally specific CB1R intracellular interactor(s). Upon the starting proteomic analysis, we decided to focus our efforts on the protein GAP43. The aforementioned general aim of this Doctoral Thesis can therefore be divided into two specific aims: Aim 1: To identify and validate GAP43 as a new interactor of CB1R, by char- acterizing CB1R-GAP43 interaction and its signaling consequences in vitro. Aim 2: To map the CB1R-GAP43 complexes in the mouse brain in a cell- population specific manner, and to assess their functional consequences in CB1R- associated synaptic transmission and pathophysiology. 49 Materials and Methods Animals All the experimental designs and procedures used were performed in accordance with the guidelines and with the approval of the Animal Welfare Committee of Uni- versidad Complutense de Madrid and Comunidad de Madrid, and in accordance with the directives of the European Commission (Directive 2010/63/EU). For ex- periments conducted in Universidad Autónoma de Madrid, all animal procedures were approved by the Universidad Autónoma de Madrid Ethical Committee of An- imal Welfare and conformed to Spanish and European guidelines for the protection of experimental animals (Directive 2010/63/EU). For experiments performed in Al- bert Einstein College of Medicine, experimental procedures adhered to NIH and Albert Einstein College of Medicine Institutional Animal Care and Use Committee guidelines. Adequate measures were taken to minimize pain and discomfort of the animals throughout the experiments. All efforts were made to minimize the number of animals used. Mice were maintained in standard conditions, keeping littermates grouped in breeding cages, at a constant temperature (20±2ºC) on a 12-h light/dark cycle with food and water ad libitum. Besides wild-type C57BL/6N mice (bred in-home), conditional mutant mice generated by the Cre-loxP technology were used as well. Mice that condition- ally express Cre recombinase under the control of the transcription factor Nex1 (Nex -Cre) or the transcription factor Dlx5/6 (Dlx -Cre) were used. By crossing these mice with CB1R-floxed (CB1Rfl/fl) mice we generated conditional CB1R- deficient mice in which the CB1R gene (Cnr1 ) is primarily absent from neocortical glutamatergic neurons from the dorsal telencephalon (Cnr1 floxed/floxed;Nex1−Cre; herein referred to as Glu-CB1R−/− mice), or from GABAergic neurons from dorsal telencephalon (Cnr1 floxed/floxed;Dlx5/6−Cre; herein referred as GABA-CB1R−/− mice). We used as well systemically-null CB1R−/− mice, in which the CB1R gene was deleted from all body cells (Cnr1 floxed/floxed;EIIa−Cre; herein referred to as CB1R−/− mice). CB1R−/−, Nex -CB1R−/− and Dlx -CB1R−/− colony founding mice were originally provided by Prof. Beat Lutz (Johannes Gut- tenberg University,Mainz, Germany) and Dr. Giovanni Marsicano (INSERM, Bor- deaux, France). The generation and genotyping of CB1R−/−, Nex -CB1R−/−, Dlx - CB1R−/− and CB1Rfl/fl littermates has been reported elsewhere and was performed accordingly (Monory et al., 2007). On the other hand, CB1R expression was select- ively rescued in dorsal telencephalic glutamatergic neurons (herein referred to as Glu-CB1R-RS mice) and GABAergic neurons (herein referred to as GABA-CB1R- RS) by expressing Cre under the regulatory elements of the Nex1 or Dlx5/6 gene respectively. As a control, an EIIa-Cre-mediated, global CB1R expression-rescue in a CB1R-null background was conducted (herein referred to as CB1R-RS mice) (Ruehle et al., 2013). As for GAP43 animal models, we purchased B6Dnk;B6Brd;B6N-Tyrc-Brd 50 Gap43tm1a(EUCOMM)Wtsi/WtsiBiat mice from EMMA Mouse Repository. This is a Gap43 knockdown model that, upon crossing with mice carrying a constitutive ex- pression of Flp recombinase under the control of the general promoter ACTB (The Jackson Laboratory; kindly provided by Dr. Rui Benedito, CNIC, Madrid), al- lows a rescued, “conditional ready” floxed allele (as Gap43 exon 2 is flanked by loxP sites.) Subsequent crossing with the aforementioned Nex1 or Dlx5/6 -Cre-expressing mice results in the corresponding (herein referred to as Glu-GAP43−/− and GABA- GAP43−/−) conditional knockout mice, respectively. Plasmids All plasmids used in this Thesis are listed in Table 1.3 and were validated by Sanger sequencing before use. The recombinant expression of human(h)CB1R C -terminal domain (CTD; amino acids 408-472) for liquid chromatography-coupled mass spectrometry was achieved upon cloning this fragment into the expression vector pKLSLt (kindly given by Dr. José M. Mancheño; IQFR, CSIC, Madrid, Spain). This plasmid contains a lectin from the fungus Laetiporus sulphurous, followed by the cutting sequence of TEV protease (-ENLYFQIG-) and a multiple cloning site. It presents kanamycin resistance and its expression is controlled by the IPTG-inducible lactose operon. Expression and purification of hGAP43 and hCB1R-CTD (400-472) for fluorescence polarization assays was performed by cloning the gene into the pBH4 plasmid, a pET-28 derivative previously used in our laboratory, with an N -terminal 6xHis tag (Merino-Gracia et al., 2017). This vector allows protein expression in bacteria under the control of the lactose operon and contains an ampicillin resistance gene. Plasmid name Resistance Application pKLSLt KanR Recombinant expression in E. coli (IPTG dependent) pKLSLt-CB1R (408-472) KanR Recombinant expression in E. coli (IPTG dependent) pBH4-GAP43 AmpR Recombinant expression in E. coli (IPTG dependent) pBH4-CB1R-CTD (400-472) AmpR Recombinant expression in E. coli (IPTG dependent) pcDNA3.1-mCB1R-myc AmpR Mammalian cell expression pEGFP-C2 KanR Mammalian cell expression pEGFP-hGAP43 KanR Mammalian cell expression pEGFP-hGAP43 S41A KanR Mammalian cell expression pEGFP-hGAP43 S41D KanR Mammalian cell expression pEGFP-hCRIP1a KanR Mammalian cell expression pEGFP-hGAP43 (5-68) KanR Mammalian cell expression pEGFP-hGAP43 (60-238) KanR Mammalian cell expression pcDNA3.1 AmpR Mammalian cell expression pcDNA3.1-hGAP43 AmpR Mammalian cell expression pcDNA3.1-hGAP43 S41A AmpR Mammalian cell expression pcDNA3.1-hGAP43 S41D AmpR Mammalian cell expression pRLuc-N1 KanR Mammalian cell expression pRLuc-CB1R KanR Mammalian cell expression pRLuc-CB1R (1-458) KanR Mammalian cell expression pAM-CBA-HA-CB1R AmpR Mammalian cell expression pAM-CBA-GAP43 S41D-CFP AmpR AAV production / CFP-fused pAM-CBA-GAP43 S41A-CFP AmpR AAV production / CFP-fused Table 1.3: Plasmids used in this Thesis. The pEGFP-C2 plasmid was purchased from ClonTech (Mountain View, CA, USA). 51 This vector harbors a multiple cloning site after green fluorescent protein (GFP), which allows creating a fusion protein with GFP at the N -terminal end. It has a kanamycin resistance gene. Using standard cloning procedures, we generated pEGFP-hGAP43 and pEGFP-hCRIP1a. cDNA from hGAP43 and hCRIP1a was acquired from DNAsu (Clone ID HsCD00042260) or kindly donated by Dr. Allyn C. Howlett (Wake Forest School of Medicine, Winston-Salem, NC, USA), respectively. pEGFP-hGAP43-S41D and pEGFP-hGAP43-S41A were constructed by Quickchange mutagenesis by using pEGFP-hGAP43 as template. The pcDNA3.1 (+) plasmid was purchased from Invitrogen. This vector con- tains a multiple cloning site after the cytomegalovirus promoter. It has an ampi- cillin resistance gen. pcDNA3.1-CB1R-myc, pAM-CBA-HA-CB1R, as well as ad- enoviral backbone vectors were kindly provided by Dr. Luigi Bellocchio (Neuro- centre Magendie, INSERM, Burdeaux, France), and pRLucN1 by Dr. Estefańıa Moreno (Department of Biochemistry and Molecular Biology, University of Bar- celona, Spain). pcDNA-hGAP43, pcDNA-hGAP43 S41D, pcDNA-hGAP43-S41A, pcDNA-hCRIP1a, pRLuc-CB1R and pRLuc-CB1R (1-458) were built by PCR and restriction cloning. Finally, for recombinant adeno-associated virus (rAAV) ex- pression vectors, pAM-CBA-GAP43-S41D-CFP and pAM-CBA-GAP43-S41A-CFP were subcloned by using standard molecular cloning techniques with a minimal chicken β-actin (CBA) promoter for generalized expression and fused to CFP re- porter. Expression and purification of recombinant proteins Protein expression was based on bacterial systems that express T7 phage poly- merase. Briefly, E.Coli BL21DE3 strain competent bacteria (produced in-home) were transformed by heat shock with pBH4 plasmids encoding hGAP43 and CB1R- CTD. Cells were incubated with 1 ng of the plasmid for 30 min on ice, placed at 37º for 45 s, and maintained for 2 more min on ice. Then, 1 mL of LB (Lysogeny Broth: 1% w/v tryptone, 0.5% w/v yeast extract, and 10 g/L NaCl, pH 7.0) was added to let bacteria grow, and, after 1 hour, 100 µL of the suspension was seeded on a LB-agar (15 g/L) plate with the selection antibiotic (kanamycin or ampicil- lin). An initial pre-inoculum of an individual colony was cultured overnight in LB media with the appropriate antibiotic under constant agitation (230 rpm) at 37ºC for 6 hours. In all cases, bacteria were inoculated after in 2xYT media (1.6% w/v tryptone, 1% w/v yeast extract, and 5 g/L NaCl, pH 7.0) with the selection anti- biotic at 37ºc and continuous shaking (230 rpm); when the optical density at 580 nm reached values around 1 (exponential growth phase), 0.5 mM isopropyl 1-thio-β- D-galactopyranoside (IPTG) was added, and the temperature was lowered to 20ºC to disfavor protein aggregation. For protein purification, bacteria were pelleted by centrifugation at 7000 rpm for 15 min at room temperature (RT) and resuspended in ice-cold lysis buffer (100 mM Tris, pH 7.0, 100 mM NaCl, 10 mM imidazole, and 0.2 mg/mL lysozyme, supplemented with 1x protease inhibitors cocktail containing 1 µg/mL aprotinin, 1 µg/mL leupeptin, and 200 µM PMSF), followed by 3 cycles of sonication on ice. The lysate was subsequently clarified by centrifugation at 10,000 rpm, 30 min, 4ºC, and filtered through porous paper. Recombinant His6-tagged proteins were sequentially purified on a nickel-nitrilotriacetic acid affinity column previously equilibrated with 50 mM Tris, pH 7.0, 100 mM NaCl, and 10 mM im- 52 idazole. Extensive washing was conducted with 50 mM Tris, pH 7.0, 100 mM NaCl, and 25 mM imidazole. For elution of purified proteins, 250 mM imidazole in 50 mM Tris, pH 7.0, 100 mM NaCl, supplemented with the aforementioned protease inhibitors was used. Protein purity was confirmed by sodium PAGE-dodecyl sulfate (SDS), and Coomassie Brilliant Blue or Silver Blue staining. Protein concentration was estimated by measuring absorbance at 280 nm. Pure protein solutions were con- centrated by centrifugation in Centricon® tubes (Millipore, Burlington, MA, USA). Affinity-based proteomics Purification of CB1R-CTD bound to lectin was performed similarly as explained in the previous section. However, in this case, the bacterial lysate expressing lectin- CB1R-CTD plasmid was loaded onto a Sepharose 4B column, previously equilibrated with 100 mM Tris, 25 mM NaCl, pH 7.0. Lectin can bind to sepharose and, thereby, anchor and expose CB1R-CTD (amino acids 408-472). Purified empty vector har- bouring a lectin was loaded as well in an additional sepharose column. A sheep whole brain was homogenized by mechanical disaggregation in RIPA buffer (50 mM Tris-HCl, 150 mM NaCl, 1% v/v Triton X-100, 0.1% w/v deoxycholic acid, 0.1% w/v SDS, pH 7.35). To eliminate possible unspecific interactions of the homogenate with the sepharose resin, the soluble fraction of the homogenate was first loaded onto a Sepharose 4B column after extensively washing the column with 100 mM Tris, 200 mM NaCl, pH 7.0. The soluble fraction that eluted through the column was then collected and loaded onto the Sepharose 4B column saturated with CB1R- CTD bound to lectin or lectin alone. After washing with RIPA and TBS Buffer (50 mM Tris-HCl, 150 mM NaCl, pH 7.0), the bound fraction was eluted with 200 mM lactose, and protein quantity was estimated by measuring absorbance at 280 nm. The fractions with the highest amounts of protein were subjected to nLC/MS-MS proteomics analysis. Briefly, the samples were loaded on a 12% acrylamide gel and a denaturing elec- trophoresis was carried out. Once samples had reached the resolving gel, electro- phoresis was stopped and the gel was died with Coomassie Colloidal Blue overnight. After fading, the region of the gel containing the sample was cut just above the recombinant protein, this piece being divided in smaller fractions that were there- after digested with trypsin. The resulting peptidic fragments were retained in an Acclaim PepMap 100 precolumn (Thermo Scientific, Waltham, MA, USA) and then eluted in an Acclaim PepMap 100 C18 column, 25 cm long, 75 µm internal diameter and 3 µm particle size (Thermo Scientific, Waltham, MA, USA). The peptides were separated in a gradient for 110 min (90 min 0-35% of Buffer B; 10 min 45%-95% Buffer B; 9 min 95% Buffer B; and 1 minute 10% Buffer B; Buffer B containing 0.1% formic acid in acetonitrile) at 250 nL/min in a nanoEasy nLC 1000 (Proxeon) coupled to an ionic source with nanoelectrospray (Thermo Scientific, Waltham, MA, USA) for electrospray ionization (ESI). Mass spectra were acquired in an LTQ–Orbitrap Velos mass spectrometer (Ther- mo Scientific, Waltham, MA, USA) working in positive mode, which measures the mass-to-charge ratio (m/z ) of ionized particles and detects the relative number of ions at each m/z ratio. Mass spectra corresponding to a full screening (m/z 400- 2000) were obtained with a resolution of 500,000 to 60,000 (m/z=200) and the 15 most intense ions from each screening were selected for fragmentation by cleavage 53 at peptide bonds via collision with a gaseous matrix by dissociation induced by collision (CID) with the energy of collision normalized to 35 %. Ions with unique charge or no charge were discarded. A dynamic exclusion of 45 s duration was done. The masses of the fragments were then determined by ion trap to define the amino acid sequence of peptides. Spectrum files (*.raw) were challenged to data bases and Uniprot of sheep (Ovis aries) (23112 sequences) and other mammals (mammalian) (1184488 sequences) by using the software Proteome Discoverer (version 1.4.1.14, Thermo Scientific, Waltham, MA, USA) and the searching tool Sequest. In the searching criteria, carbamide methylation, cysteine nitrosylation and methionine oxidation were es- tablished as dynamic modifications. Tolerance to precursor selection and product ions was fixed at 10 ppm and 0.5 Da, respectively. Peptide identification was valid- ated by the Percolator algorithm using q ≤ 0,01 (q-value is the p-value, additionally adjusted to the False Discovery Rate, which is the probability of a signal showing as significant, even though it is not.) Fluorescence polarization CB1R-CTD was labeled with 5-(iodoacetamido)-fluorescein (5-IAF) following manu- facturer instructions (Thermo Fisher Scientific, Waltham, MA, USA). 5-IAF dye was solubilised in dimethylsulfoxide (DMSO) and the labeling reaction was performed in 0.1 M NH4HCO3, 1% SDS, and 20 mM dithiothreitol (DTT), pH 9.0, with a 3-fold molar excess of dye versus CB1R-CTD peptide for 1 hour at 25 °C protected from light. Subsequently, a 1.00 Da cut-off dialysis membrane was used to eliminate non-reacted 5-IAF. The concentration of the labeled peptide was calculated by using the value of 68,000 cm–1 M–1 as the molar extinction coefficient of the dye at pH 8.0 and wavelength 494 nm. Saturation binding experiments were performed for measuring binding affin- ity (Kd) between IAF-labeled CB1R-CTD and GAP43 by applying an increasing amount of GAP43 (between 0–300 µM) to a fixed and low concentration of the probe (100 nM). Incubation time was 20–30 min RT, and the assay was performed in 20 mM Tris, pH 7.0, and 50 mM NaCl, in a final volume of 0.4 mL Fluores- cence polarization (FP) was performed in a PerkinElmer Life Sciences MPF 44-E spectrofluorometer. Polarization of the IAF-labeled CB1R-CTD was measured at excitation/emission values of 488/530 nm (bandwidth, 10 nm). The fluorescence anisotropy (r) values were obtained by using the FP values with the equation r = 2FP/(3-FP). The initial anisotropy (ri) in the absence of protein was measured. The FP values were fitted to the equation (FP-FP0) = (FPmax-FP0)[GAP43]/(Kd + [GAP43]), where FP is the measured fluorescence polarization, FPmax is the maximal fluorescence polarization value, FP0 is the fluorescence polarization in the absence of added GAP43, and Kd is the dissociation constant. Cell culture and transfection The HEK293T cell line was from the American Type Culture Collection (Manassas, VA, USA). HEK293T cells stably expressing FLAG-CB1R (HEK-CB1R) were kindly provided by Dr. Maria Waldhoer (InterAx Biotech, PARK innovAARE, Villigen, Switzerland). Cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) 54 supplemented with 10% v/v fetal bovine serum (FBS), 1 mM sodium pyruvate, 2 mM L-glutamine, 1% streptomycin/penicillin and, exclusively for HEK-CB1R cell line and to ensure stable expression of FLAG-CB1R, 0.25 mg/mL zeocin was in- cluded as selection marker. Cells were maintained at 37ºC in an atmosphere with 5% CO2. Cells growing in 100 mm-diameter dishes were transiently transfected with the corresponding protein-encoding cDNA by the polyethylenimine method (Sigma, Steinheim, Germany) in a 4:1 mass ratio to DNA for HEK293T cells or by using Lipofectamine 2000 (Invitrogen, MA, USA, 11668-019) protocol according to manufacturer‘s instructions for HEK-CB1R cells. For Western blot analysis, 48 hours after transfection cells were transferred to 6-well dishes, FBS-starved overnight, and treated for 5 min with vehicle (DMSO, 0.1% v/v final concentration) or WIN (100 nM) prediluted in DMSO. For analyzing PKA phosphorylated substrates, forskolin (0.5 µM)/vehicle was added right after WIN/vehicle addition for 10 min. To block Gαq/11, YM-254890 (1 µM)/vehicle was added to cells 30 min before WIN/vehicle application for 5 min. Cells were washed in ice-cold PBS and directly frozen for further analysis. For immunocytochemistry, 12 µm-diameter coverslips were coated with 0.1 mg/mL poly-D-Lysine for 1 hour at 37ºC. Cells were seeded on the coverslips at a density of 30.000 cells/cm2 and trans- fected with the corresponding protein cDNA. Forty-eight hours after transfection, cells were starved overnight, treated and fixed in paraformaldehyde solution (PFA) (4% in PBS) for 15 min at RT. Primary hippocampal and cortical neurons were obtained from 0-2-day-old C57BL /6N mice. Dissection was followed by mechanical disaggregation and then a papain dissociation system (Worthington, OH, USA) was used. Cells were nucleofected with pCDNA3.1 empty plasmid or pCDNA3.1-CAG-hGAP43-WT / GAP43-S41A / GAP43-S41D (described above in Plasmids section) using an Amaxa mouse neuron Nucleofector kit (Lonza, Basel, Switzerland). Cells were seeded on plates pre- coated with 0.1 mg/mL poly-D-Lysine and 3 µg/mL laminin at a density of 200,000 cells/cm2 in Neurobasal medium supplemented with 10% v/v FBS, 1 mM sodium pyruvate, 0.5 mM L-glutamine, and 1% streptomycin/penicillin. Three hours later, FBS is substituted for B27 supplement (Blázquez et al., 2011). This media was renewed every 2 days. At 7 days in vitro (DIV), neurons were treated if required with WIN or vehicle for 5 min. Cultured neurons were fixed for inmunoflourescence or frozen for Western blot/co-IP assays. Western blot HEK293T cells, primary cultured neurons or mouse brain tissue samples were ho- mogenized and lysed for protein extraction in a ice-cold lysis buffer containing 50 mM Tris, 0.1 % Triton X-100, 1 mM ethylenedi¬aminetetraacetic acid (EDTA), 1 mM ethylene glycol tetraacetic acid (EGTA), 50 mM NaF, 10 mM sodium β- glicerophosphate, 5 mM sodium pyrophosphate, and 1 mM sodium orthovanadate, pH 7.5, supplemented with a protease inhibitor cocktail (Roche, Basel, Switzer- land), 0.1 mM phenylmethane-sulphonylfluoride, 0.1 % β-mercaptoethanol and 1 µM microcystin. Samples were cleared by centrifugation at 12,000 g for 15 min at 4ºC, and the supernatants collected. Protein concentration was determined by the Brad- ford method. Then, extracts were mixed with 5× SDS sample buffer and boiled 55 at 95ºC for 5 min (except for CB1R detection, in which samples are heated at 55ºC for 10 min). Equal amounts of protein samples were electrophoretically re- solved on 12% SDS–PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Whatman Schleicher Schuell, Keene, NH, USA). After incubation for at least 1 hour in blocking buffer containing 10% non-fat powdered milk in Tris buffered saline-Tween (TBS-T) or 5% bovine serum albumin (BSA), membranes were blotted overnight at 4°C with the indicated primary antibodies (Table 1.4). PVDF membranes were then rinsed 3 times with TBS-T and incubated with the corresponding secondary antibodies coupled to horseradish peroxidase for 1.5 hours at RT. After washing 3 times for 10 min with TBS-T, membranes were developed using an enhanced chemiluminescence (ECL) kit (Santa Cruz Biotechnology, Dal- las, TX, USA). Optical density of the specific immunoreactive bands was quantified with FIJI ImageJ open source software (NIH, Bethesda, MD) and normalized to the intensity of the β-actin or α-tubulin band (loading control) in the same membranes within a linear range of detection for the ECL reagent. Co-immunoprecipitation HEK293T cells, primary cultured neurons or mouse brain tissue samples were sol- ubilized and lysed in TAP buffer (10 mM Tris, pH 8.0, 150 mM NaCl, and 10% glycerol) containing 0.5% NP-40 plus a cocktail of protease and phosphatase inhib- itors (protease inhibitor cocktail 1x: 0.5 mM PMSF, 1 mM NaF, 1 mM Na2MoO4, and 0.5 mM NaVO4), referred from now on as co-IP Lysis Buffer. This buffer offers sufficiently gentle conditions to preserve protein complexes. Samples were centri- fuged at 10,000 rpm for 15 min at 4ºC. Solubilised-protein concentrations were then quantified by using the abovementioned Bradford method. For interaction experi- ments, 1 mg of total protein per condition was used. To pre-clarify, each sample was incubated with 20 µl of Protein G-Sepahrose beads (GE Healthcare Bio-Sciences, MA, USA, 17-0618-01), properly washed 3 times in PBS/co-IP Lysis Buffer, for 45 min at 4ºC on a rotating wheel. Beads were discarded by centrifugation (2000 rpm, 4ºC, 3 min), followed by incubation with 2 µg of anti-GAP43 antibody (200 µg/ml) per 1 mg of total protein (Table 1.4) and the same amount of mouse IgG isotype control (Invitrogen, Carlsbad, CA, USA, 10400C) overnight under rotation at 4ºC. On the next day, 25 µl of Protein G-Sepahrose beads was added to each sample and incubated for 2 hours. Beads with bound anti-GAP43 antibody and GAP43-associated complexes were recovered by centrifugation (2000 rpm, 4ºC, 3 min), and washed 3 times with co-IP Lysis Buffer. Immunoprecipitated proteins and its associated complexes were eluted in 30 µl of 2× SDS sample buffer. Alternatively, protein lysates were incubated with Protein G-Sepharose beads coupled to anti-HA antibody (Thermo Scientific, Waltham, MA, USA, 26181) within Pierce HA-Tag IP/Co-IP Kit, from 2 hours to overnight, at 4°C. The anti-HA af- finity matrix was then sedimented at 2000 rpm for 5 min and washed 3 times with co-IP Lysis Buffer. Immunoprecipitated proteins and their associated complexes were eluted in 25 µl of 2× Non-Reducing Sample Buffer. The resulting proteins were resolved by SDS-PAGE, and transferred to PVDF membranes. Samples were heated for 10 min at 55 ºC to detect GAP43 and co-precipitated CB1R by West- ern blotting, and 40 µg of input was loaded on the gel. The corresponding PDFV membrane was incubated with guinea pig anti-CB1R antibody (1:1000; Table 1.4). 56 Alternatively, samples were boiled at 95ºC for 5 min to detect co-precipitated HA and GAP43, and 20 µg of input protein were loaded on the gel. The corresponding PDVF membrane was incubated with rabbit anti-HA antibody (1:1000; Table 1.4). Finally, proteins were detected by using the aforementioned ECL kit. Immunofluorescence Mice were perfused transcardially with PBS followed by PFA solution. Brains were dissected and post-fixed overnight in the same solution, cryoprotected with sucrose and mounted on standard cryomold with OCT compound. Serial coronal cryostat sections (30 µm-thick) through the whole brain were collected in PBS as free-floating sections and stored at -20ºC. These sections, or alternatively cultured cells fixed in 4% PFA solution (see Cell culture and transfection section), were rinsed twice in PBS, and permeabilized and blocked in PBS containing 0.25% Triton X-100 and 10% or 5% respectively goat serum (Pierce Biotechnology, Rockford, IL, USA) for 1 hour at RT. Primary antibodies were diluted directly into the blocking buffer, and floating slices or coverslips were incubated overnight at 4°C with the indicated primary antibodies (Table 1.4). After 3 washes with PBS for 10 min, samples were subsequently incubated for 2 hours at RT with the appropriate highly cross-adsorbed anti-mouse, rat, guinea pig, and rabbit AlexaFluor 488, AlexaFluor 546, Alexa Fluor 594, and AlexaFluor 647 secondary antibodies (1:500 or 1:1000) (all from Invitrogen, Carlsbad, CA, USA) together with the nuclear dye 4’,6-diamidino-2-phenylindole (DAPI) (diluted 1:10,000 in blocking buffer) (Roche, Basel, Switzerland) to visualize nuclei . After washing 3 times in PBS, sections or cover glasses cell-side down were mounted onto microscope slides using Mowiol mounting media. Confocal fluorescence images were acquired by using either Leica TCS-SP2 or LAS-X software with an SP2 or SP8 confocal microscope, respectively (Leica Mi- crosystems, Mannheim, Germany), and a 1024 x 1024 or a 2048 x 2048 collection box, respectively. All quantifications were obtained from a minimum of 4 sections per condition. Images were taken using apochromatic oil-immersion 20X, 40X or 63X objective and standard (1 Airy disc) pinhole and 3X digital zoom in colocalization experiments in CB1R conditional knockout slices as well as in cells. Immunoreactive area was measured using FIJI ImageJ open source software, establishing a threshold to measure only specific signal. For colocalization assays, the resulting binary mask was then used along the built-in measure function to acquire the total immunoreact- ive area among all the pixels inside the binary mask overlaid on top of the original image. The obtained value was then referred to the number of DAPI+ cell nuc- lei or to the number of GFP+ transfected cells present in the optic field. Data were then expressed as percentage of control. None of the secondary antibodies produced any signal in preparations incubated in the absence of the corresponding primary antibodies. Representative images for each condition were prepared for fig- ure presentation by applying brightness, contrast and other adjustments uniformly. X-gal staining For X-gal staining, the protocol was described previously (Kokubu and Lim, 2014). Briefly, fixed brain cyro-sections (30 µm) are directly mounted on the slides and washed with the staining buffer (1 M MgCl2, 10% sodium deocycholate and 20% 57 Target Host Species Application and Concentration Company Reference pCofilin Rabbit WB (1:1000), IF (1:500) Cell Signalling 3313 Cofilin Mouse WB (1:1000) Santa Cruz Biotech. sc-376476 pROCK2 Rabbit WB (1:1000), IF (1:500) Gene Tex GTX122651 ROCK2 Mouse WB (1:1000) Santa Cruz Biotech. sc-398519 p-p44/42 MAPK Rabbit WB (1:1000) Cell Signaling 9102 p44/42 MAP Kinase Mouse WB (1:1000) Cell Signaling 4696 p-PKA substrates Rabbit WB (1:1000) Cell Signaling 9621 GAP43 Mouse WB (1:1000), IF (1:1000), IP (2µg), PLA (1:400; 1:100) Santa Cruz Biotech. sc-33705 GFP Rabbit WB (1:1000), PLA (1:200) Abcam ab290 c-myc Mouse PLA (1:200) Roche/Sigma 11667149001,00 α-tubulin Mouse WB (1:5000) SIGMA T 9026 β-tubulin III Mouse IF (1:500) SIGMA T 8660 β-tubulin III Rabbit IF (1:500) BioLegend 802001 actin Mouse WB (1:5000) SIGMA A5441 CB1R Rabbit WB (1:1000), IF (1:500), PLA (1:200) Frontier Science Af380-1 CB1R Guinea Pig WB (1:1000), IF (1:500) Frontier Science Af530-1 HA Rabbit WB (1:1000), IP (1µg) Cell Signaling 3724S Flag M2 Mouse IF (1:750), WB (1:1000) SIGMA F3165 EEA1 Rabbit IF (1:100) Cell Signalling 2411 LAMP1 Rabbit IF (1:750) Abcam ab24170 calretinin Rabbit IF (1:500) Swant 7699/3H Table 1.4: Primary antibodies used in this Thesis. NP-40; all from Sigma) for 10 min at room temperature. They were subsequently incubated with 1 mg/ml X-gal in the staining buffer supplemented with 5 mM Po- tassium Ferricyanide (Sigma) and 5 mM Potassium Ferrocyanide (Sigma) at 37 °C until color develops. This step usually took 6 hours to overnight. Stained samples were washed with PBS three times (5 min each) and dehydrated with series of eth- anol (50%, 75%, 90%, 100%, 2 min each). Finally, slides are mounted using Entellan resin and Images were adquired by using an Zeiss Axioplan 2 microscope (QImaging, Roper Technologies, Sarasota, FL, USA). Receptor internalization The protocol for quantification of receptor internalization in HEK-CB1R cells has been extensively described (Zhuang and Matsunami, 2008). Briefly, for live-cell im- munofluorescence, cells seeded on cover glasses were transferred from their culture dishes to a new 150 mm-diameter dish cell-side up and placed on ice. Mouse anti- FLAG M2 antibody (FLAG tag was cloned at the N -terminal, extracellular domain of CB1R) was diluted 1:1000 in a staining solution containing Minimum Essential Media (MEM) with 10 mM HEPES, pH 7.0 - 7.4, and 15 mM NaN3. Then, 100 µL of the solution was applied per cover glass and incubated on ice for 1 hour. At the end of the incubation, the cover glasses were carefully transferred back to its original 150-mm dishes, and washed 3 times by using cold washing solution for live-cell im- munofluorescence composed of Hank‘s Balanced Salt Solution (HBSS) with 10 mM HEPES, pH 7.0 – 7.4 and 15 mM NaN3. The corresponding secondary antibody (anti-mouse AlexaFluor 647, 1:1000) was diluted as well in staining solution and incubated for 30 min on ice. Afterwards, cover glasses were washed again 3 times. Finally the washing solution was replaced with 4% PFA for 15 min to fix the cells. A subsequent immunostaining step was performed as described previously. Cells were then permeabilized with PBS containing 0.25% Triton X-100 and 5% goat serum for 1 hour. Rabbit anti-CB1R primary antibody (1:500) directed against CB1R-CTD and anti-rabbit AlexaFluor 546 (1:1000) as secondary antibody were used. In situ proximity ligation assay (PLA) 58 PLA assays for CB1R and GAP43 were performed in HEK293T cells transfected with pcDNA3.1-CB1R-myc and pEGFP-GAP43 or, alternatively, untagged pcDNA3.1- GAP43. Cells were grown on glass coverslips and fixed in 4% PFA for 15 min. For conducting PLA in mouse hippocampal brain slices, mice were deeply anesthetized and immediately perfused transcardially with PBS followed by 4% PFA and post- fixed as described above. Serial coronal cryostat sections (30 µm-thick) through the whole brain were collected in cryoprotective solution and stored at -20ºC until PLA experiments were performed. Immediately before the assay, mouse brain sections were mounted on glass slides, and washed in PBS. In all cases, complexes were detected using the Duolink II in situ PLA detection Kit (OLink; Bioscience, Uppsala, Sweden) following supplier’s instructions. First, samples were permeabilized in PBS supplemented with 20 mM glycine and 0.05% Triton-X100 for 5 min (cell cultures) or 10 min (mounted slices) at RT. Slides were next incubated with Blocking Solution (one drop per 1 cm2) in a pre-heated humidity chamber for 1 hour at 37ºC. Primary antibodies were diluted in Antibody Diluent Reagent from the Kit: mouse anti-c-myc (1:200), rabbit anti-GFP (1:200), mouse anti-GAP43 (1:400), and rabbit anti-CB1R (1:200) antibodies for cell culture; and mouse anti-GAP43 (1:100) and rabbit anti-CB1R (1:100) antibodies for slices, as shown in Table 1.4. Samples were incubated overnight at 4ºC. They were used to visualize GAP43-CB1R complexes together with PLA probes detecting mouse or rabbit antibodies (PLA probe MINUS and PLA probe PLUS). Then, samples were washed and processed for ligation (ligase 1:40) and amplification (polymerase 1:80) with a Detection Reagent Red (546 nm), stained for DAPI, and mounted. Negative controls were performed with just primary antibody. Samples were analyzed with a Leica SP2 or SP8 confocal microscope and pro- cessed with FIJI ImageJ software. For cultured cells, 22 random fields along the whole preparation were used for analysis. For each field of view, a stack of two channels (one per staining) and 30 Z-stacks with a step size of 0.5 µm were ac- quired. For brain tissue, 3 serial coronal sections (30-µm thick) per animal spaced 0.24 mm apart containing the DG were used. For each field of view, a stack of two channels (one per staining) and 9 to 13 Z-stacks with a step size of 1 µm were acquired. Cells containing one or more red dots versus total GFP+ cells or altern- atively total cells (blue nuclei) were determined for quantification in experiments with cultured cells. The total number of red dots per area was determined for quantification of PLA signal in tissue slices. Nuclei and red dots were counted on the maximum projections of each image stack. After getting the projection, each channel was processed individually. The blue nuclei and red dots were segmented by subtracting the background and applying a threshold to obtain the binary im- age. PLA dots were counted upon restricting size and circularity. Additionally, for slices, the contrast was enhanced with the Contrast Limited Adaptive Histogram Equalization (CLAHE) plug-in, and regions of interest (ROIs) are established for quantification. Representative images for each condition were prepared for figure presentation by applying brightness, contrast and other adjustments uniformly. Bioluminiscence resonance energy transfer (BRET) 59 HEK293T cells grown in 6-well plates were transiently co-transfected with a constant amount (0.025 µg) of a cDNA encoding CB1R fused to Rluc protein (CB1R-Rluc) as BRET donor, and with increasing amounts (0.05 to 1.5 µg) of a cDNA of the different GAP43 forms fused to GFP (GAP43-WT-GFP, GAP43-S41A-GFP, GAP43-S41D- GFP) as BRET acceptor. Cells were harvested, washed, resuspended in PBS and distributed (approximately 20 µg protein per well) in 96-well microplates (black plates with a transparent bottom). For quantification of GAP43-GFP expression, fluorescence at 488 nm was ana- lyzed in a FLUOstar Optima fluorimeter (BMG Labtech, Offenburg, Germany). Fluorescence of cells expressing only the BRET donor was subtracted from these measurements. BRET signal was analyzed 1 minute after addition of the substrate DeepBlueC (Molecular Probes, Eugene, OR, USA) to each well with a Mithras LB 940 (Labnet Biotecnica, Bad Wildbad, Germany) that allows the integration of the signals detected in the short-wavelength filter at 400 nm and the long-wavelength filter at 510 nm. To quantify receptor-Rluc expression luminescence, readings were also performed after 10 min of adding 5 µM DeepBlueC. The net BRET signal was defined as [(long-wavelength emission)/(short-wavelength emission)] - Cf, where Cf corresponds to [(long-wavelength emission)/(short-wavelength emission)] for the Rluc construct expressed alone in the same experiment. BRET signal was expressed as milli-BRET units (mBU; net BRET × 1,000). In BRET curves, BRET was ex- pressed as a function of the ratio between fluorescence and luminescence × 100 (GFP/Rluc). To calculate maximum BRET (BRETmax) from saturation curves, data were fitted using a nonlinear regression equation and assuming a single phase with GraphPad Prism software (San Diego, CA, USA). Dynamic mass redistribution (DMR) The overall cell signaling signature was determined by using an EnSpire® Mul- timode Plate Reader (PerkinElmer, Waltham, MA, USA) with a label-free tech- nology. Refractive waveguide grating optical biosensors, integrated in 384-well mi- croplates, allow extremely sensitive measurements of changes in local optical dens- ity in a detecting zone up to 150 nm above the surface of the sensor. Cellular mass movements induced upon receptor activation were detected by illuminating the underside of the biosensor with polychromatic light and measured as changes in wavelength of the reflected monochromatic light, that is a sensitive function of the index of refraction. The magnitude of this wavelength shift (in pm) is directly proportional to the amount of DMR. Briefly, 24 hours before the assay, HEK-CB1R cells or HEK293T cells stably expressing CB2R (HEK-CB2R), already developed in our laboratory (Moreno et al., 2014), were seeded at a density of 10,000 cells per well in 384-well sensor microplates with 30 µL growth medium, and cultured for 24 hours (37°C, 5% CO2) to obtain 70-80% confluent monolayers. Prior to the assays, cells were preincubated for 30min with vehicle, YM-254890 (1µM) or Y-27326 (10 µM). Then, cells were washed twice with assay buffer (HBSS with 20 mM Hepes, pH 7.15) and incubated for 2 hours in 30 µL per well of assay buffer with 0.1% DMSO in the reader at 24°C. Hereafter, the sensor plate was scanned and a baseline optical signature was recorded before adding 10 µL of the CB1R agonist WIN (100 nM final concentration) or alternatively HU-308 (100 nM final concentration) dis- solved in assay buffer containing 0.1% DMSO. Then, the resulting shifts of reflected 60 light wavelength in picometers (pm) or DMR responses were monitored for at least 5,000 s. Kinetic results were analyzed by using EnSpire Workstation Software v 4.10. Determination of cAMP concentration Homogeneous time-resolved fluorescence energy transfer (HTRF) assays were per- formed using the Lance Ultra cAMP kit (PerkinElmer), based on competitive dis- placement of a europium chelate-labelled cAMP tracer bound to a specific antibody conjugated to acceptor beads. The optimal cell density for an appropriate fluorescent signal (to cover the dynamic range of cAMP standard curve) has been established before (Moreno et al., 2014). The forskolin dose-response curves were related to the cAMP standard curve in order to establish which cell density provides a response that covers most of the dynamic range of the cAMP standard curve. Cells (103 per well) were seeded in medium containing 50 µM zardeverine in white ProxiPlate 384-well microplates (PerkinElmer) at 25°C for the indicated time, and challenged with 100 nM WIN for 15 min before adding 0.5 µM forskolin or vehicle, and in- cubating for an additional 15-min period. Fluorescence at 665 nm was analyzed on a PHERAstar Flagship microplate reader equipped with an HTRF optical module (BMG Lab technologies, Offenburg, Germany). cAMP values produced in each con- dition minus basal stimulation in the absence of forskolin or agonists were expressed as the percentage of the forskolin-treated cells in each condition. Synaptosomal preparations Hippocampal synaptosomes were isolated from 8 week-old C57BL/6N mice at Dr. Jose Sánchez-Prieto lab (Complutense University, Madrid, Spain) as described pre- viously (Mart́ın et al., 2010). Briefly, the hippocampus was isolated from adult mice (2–3-month-old) and homogenized in medium containing 0.32 M sucrose (pH 7.4). The homogenate was centrifuged for 2 min at 2,000 g and 4 °C, and the supernatant spun again at 9,500 g for 12 min. From the pellets formed, the white loosely com- pacted layer containing the majority of the synaptosomes was gently resuspended in 8 mL of 0.32 M sucrose (pH 7.4). An aliquot of this synaptosomal suspension (2 mL) was placed onto a 3-mL Percoll discontinuous gradient containing 0.32 M sucrose, 1 mM EDTA, 0.25 mM DL-DTT, and 3, 10 or 23% Percoll (pH 7.4). After centrifu- gation at 25,000 g for 10 min at 4 °C, the synaptosomes were recovered from the 10 and 23% Percoll bands, and they were diluted in a final volume of 30 mL of HEPES buffer medium (HBM) composed of 140 mM NaCl, 5 mM KCl, 5 mM NaHCO3, 1.2 mM NaH2PO4, 1 mM MgCl2, 10 mM glucose, and 10 mM HEPES (pH7.4). Fol- lowing a further centrifugation step at 22,000 g for 10 min, the synaptosome pellet was resuspended in 6 mL of HBM and the protein content was determined by the Biuret method. Finally, 1 mg of the synaptosomal suspension was diluted in 2 mL of HBM and spun at 3,000 g for 10 min. The supernatant was discarded and the pellets containing the synaptosomes were seeded in polylysine-coated coverslips for 1 hour, and then fixed for 5 min in 4% PFA in 0.1 M phosphate buffer (pH 7.4) for conducting subsequent immunofluorescence procedures as described above. Antibody-capture [35S]GTPS scintillation proximity assay (SPA) 61 Extracts from HEK293T cells (approximately 500 mg total protein) were thawed at 4ºC and homogenized with a glass/Teflon grinder (IKA labortechnik, Satufen, Ger- many) (10 strokes at maximum speed) in 30 volumes of homogenization buffer (50 mM Tris-HCl, 1 mM EGTA, 3 mM MgCl2, and 1 mM DTT; pH 7.4; supplemented with 250 mM sucrose). The homogenates were centrifuged at 1,100 g for 10 min at 4ºC. The pellets were discarded, and the supernatants were then recentrifuged at 40,000 g for 10 min (4ºC). The resultant pellets were resuspended in 20 volumes of fresh cold centrifugation buffer (50 mM Tris-HCl, 1 mM EGTA, 3 mM MgCl2, and 1 mM DTT; pH 7.4) and recentrifuged. The resulting pellets were then resuspended in 5 volumes of centrifugation buffer. Protein content was determined by the Brad- ford method. Specific activation of different subtypes of Gα proteins (Gαi1, Gαi2, Gαi3, Gαo, Gαz, Gαs and Gαq/11) was determined using a homogeneous protocol of [35S]GTPS scintillation proximity assay coupled with the use of specific antibodies as previously described (Diez-Alarcia et al., 2016). [35S]GTPS binding was performed in 96-well isoplates (PerkinElmer Life Sciences, Maanstraat, Germany) and in a final volume of 200 µL containing 1 mM EGTA, 3 mM MgCl2, 100 mM NaCl, 0.2 mM DTT, 50 mM Tris–HCl at pH 7.4, 0.4 nM [35S] GTPS, 10 µg of protein per well, and different concentrations of GDP depending on the Gα subunit subtype tested. At the end of the 2-hour incubation period at 30 ºC, 20 µL of Igepal 1% plus 0.1% SDS was added to each well, and plates were incubated at 22 ºC for 30 min with gentle agitation. Specific antibody for the Gα subunit of interest was then added to each well before an additional 90-min RT incubation period. Polyvinyltoluene (PVT) SPA beads coated with protein A (PerkinElmer, S.L., Tres Cantos, Madrid, Spain) were then added (0.75 mg of beads per well), and plates were incubated for 3 hours at RT with gentle agitation. Finally, plates were centrifuged (5 min at 1,000 g), and bound radioactivity was detected on a MicroBeta TriLux scintillation counter (PerkinElmer S.L., Tres Cantos, Madrid, Spain). In order to test their ef- fect on the [35S]GTPS binding to the different Gα subunit subtypes in the different experimental conditions, a single submaximal concentration (10 µM) of WIN was tested. Nonspecific binding was defined as the remaining [35S]GTPS binding in the presence of 10 µM unlabelled GTPS. Specific [35S]GTPS binding values were trans- formed to percentage of basal [35S]GTPS binding (binding values in the absence of any exogenous drug) obtained for each Gα protein. rAAV production and injection Chimeric AAV serotype 1/2 vectors (virions) containing a 1:1 ratio of AAV1 and AAV2 capsid proteins with AAV2 ITRs were used. They were produced as previ- ously described (Monory et al., 2006; Ruiz-Calvo et al., 2018). We standardized this protocol in our laboratory. Briefly, HEK293T cells were transfected with polyethyl- eneimine in a 4:1 mass ratio to DNA with the AAV cis plasmid, the AAV1 and AAV2 helper plasmids, and either pAM-CBA-GAP43-S41D-CFP or pAM-CBA-GAP43- S41A-CFP for each of the viral batch required. Sixty hours after transfection, cells were harvested and the virions were purified by iodixanol (OptiPrep)-based density gradients, consisting of decreasing concentrations of OptiPrep - 54%, 40%, 25%, 15%) and ultracentrifugation at 160,000 g. Virion solutions were concentrated by centrifugation in Centricon® tubes (Millipore, Burlington, MA, USA). Purity ana- lysis was performed through PAGE-SDS with 20 µL of concentrated virion prepar- 62 ations per lane. The genomic titers determined using an Applied Biosystems ABI 7500 real time PCR cycler. The viral vectors rAAV1/2-CAG-GAP43-S41D-CFP and rAAV1/2-CAG-GAP43- S41 A-CFP (in 1 µL PBS) were injected stereotactically into the hilus of 3 week-old WT mice. Each animal received one ipsilateral injection at coordinates (mm to bregma): antero-posterior +2.18, lateral ±1.5, dorso-ventral -2.2, for a bregma to lambda distance of 4.2 mm. We conducted the electrophysiological experiments 3-4 weeks after surgery to allow viral expression. We describe the placement and ex- pression of the rAAV vectors in the mouse DG under these conditions in Figure 12A. Electrophysiology Both C57Bl/6J males and females were used. Animals were anesthetized with iso- flurane before sacrifice. Experiments on PKC signaling cascade were performed us- ing acute coronal slices from young mice (P13-P15). Experiments using the different form of GAP43 (S41A and S41D) were performed using acute transverse hippocam- pal slices from AAV-injected mice (3-4 weeks post-injection, ∼6-7 weeks old). Brains were removed and rapidly transferred into ice-cold dissection buffer maintained in 5% CO2/95% O2 and containing (in mM): 215 sucrose, 20 D-glucose, 26 NaHCO3, 4 MgCl2, 4 MgSO4, 1.6 NaH2PO4, 2.5 KCl, and 1 CaCl2. For injected mice, a choline-based solution was used, containing (in mM): 25 NaHCO3, 1.25 NaH2PO4, 2.5 KCl, 0.5 CaCl2, 7 MgCl2, 25 glucose, 110 choline chloride, 11.6 ascorbic acid, 3.1 pyruvic acid. Slices (300 µm-thick) were prepared using a vibratome (LeicaVT 1200 or DTK-2000 Dosaka EM Co., Ltd.) and then transferred in artificial cerebral spinal fluid (ACSF) and containing (in mM): 124 NaCl, 26 NaHCO3, 10 D-glucose, 2.5 KCl, 1 NaH2PO4, 2.5 CaCl2, and 1.3 MgSO4. Slices from young mice recovered at RT for 1 hour in ACSF whereas slices from injected mice were incubated at 37ºC for 30 min in ACSF. All solutions were maintained at 95% O2/5% CO2. Recordings were performed at 28 ± 1°C in a submersion-type recording chamber perfused at ∼2 mL/min with ACSF. For extracellular field recordings, a single borosilicate glass stimulating pipette filled with ACSF and a glass recording pipette filled with 1 M NaCl were placed ap- proximately 100 µm apart in the IML (<40 µm from the cell body layer and 150–200 µm slice depth). To elicit synaptic responses, paired, monopolar square-wave voltage pulses (100–200 µs pulse width) separated by 100 ms and triggered every 20 s were delivered through a stimulus isolator (Isoflex, AMPI) connected to the broken tip (∼10–20 µm, 2.5-3.0 µm) stimulating pipette. Stimulus intensity was adjusted in order to get comparable magnitude synaptic responses across experiments (∼ 0.6 mV). Excitatory synaptic transmission was monitored in the continuous presence of 100 µM picrotoxin (GABAA receptor antagonist). The CB1R agonist WIN (5 µM) was bath-applied for 25 min and chased with the CB1R inverse agonist/antagonist AM251 (5 µM) to halt CB1R signaling. The amplitude of the fiber volley and field EPSPs (fEPSP) are measured and bined per 1 min. The magnitude of fEPSPs depression was calculated as the percentage change in EPSP amplitude between baseline (averaged excitatory responses for 10 min before drug application) and 25 min of stable responses post WIN/AM251 application, and plotted as a function of time. Representative traces for fEPSPs were obtained by averaging 15 individual traces. 63 Whole-cell patch clamp experiments were performed in GCs from the dorsal blade of the GCL. Typically, stimulating pipettes were filled with ACSF and placed in the IML <40 µm from the cell body layer as well. Cells were continuously voltage- clamped at -60 mV by using patch-type pipettes filled with intracellular solution containing (in mM): 131 Cs-gluconate, 8 NaCl, 1 CaCl2, 10 EGTA, 10 glucose, 10 HEPES, pH 7.2, (285-290 mOsm). Excitatory synaptic transmission was monitored in the continuous presence of 100 µM picrotoxin. Series resistance (typically 15-30 MΩ) was monitored throughout the experiment with a -5 mV, 80 ms voltage step, and recordings with a greater than 20 % change in series resistance were excluded from analysis. Stimulation was achieved by delivering paired pulses 100 ms apart. In drug-delivery experiments, stimulation at MC axons was triggered every 20 s, and in DSE experiments, every 6 or 10 s. In DSE experiments, we used a 3-5-s depolarizing voltage step to trigger eCB release from the postsynaptic cell after baseline recording of at least 5 min. When required, drugs were perfused in the recording chamber or to the recording pipette as stated in the figure legends. The amplitude of the EPSCs in whole-cell patch clamp experiments are measured and DSE magnitude was calculated as the percentage change between the mean amplitude of 10 consecutive EPSCs preceding depolarization and the mean amplitude of 4 consecutive EPSCs following depolarization. Representative traces designated as basal were generated by averaging the 10 individual traces just prior to depolarization, and as post-dep were generated by averaging the first 3 individual traces post depolarization. The pre- or postsynaptic origin was determined by the paired-pulse ratio (PPR), which is defined as the ratio of the amplitude of the second EPSC (R2) to the amplitude of the first EPSC (R1; R2/R1) of the paired pulses 100 ms apart. Statistical analysis was performed by using OriginPro 7.0 software (OriginLab Corporation, Northampton, MA) or GraphPad Prism 6.07 (GraphPad Software, La Jolla, CA, USA). In Universidad Autónoma de Madrid, recordings were performed in voltage- clamp mode using a Cornerstone PC-ONE amplifier (DAGAN). Data were low-pass filtered at 3.0 kHz and sampled at 10.0 kHz, through a Digidata 1200 (Molecu- lar Devices). The pClamp program (Molecular Devices) was used to acquire and analyze data. In Albert Einstein College of Medicine, extracellular and whole-cell patch clamp recordings were performed using a Multiclamp 700B amplifier (Axon Instruments, Union City, CA, USA). Stimulation and acquisition were controlled by custom written software in Igor Pro (Wavemetrics, Inc., Lake Oswego, OR, USA). Picrotoxin, DCG-IV, WIN and AM251 were purchased from Tocris-Cookson Inc. (Ellisville, MO, USA). WIN and AM251 were dissolved in DMSO and added to the bath. Total DMSO in the bath solution was maintained at 0.1% in all experiments. Behavioral tests All the behavioral tests were video-recorded for subsequent blind analysis by a differ- ent trained observer using Smart3.0 Software (Panlab, Barcelona, Spain), except for epilepsy experiments. Adult (3-month old) Glu-GAP43−/− or GABA-GAP43−/−) mice and their WT (GAP43fl/fl) littermates were used for behavioral analysis. Body weight and temperature was measured, the latter with a thermo-coupled flexible probe (Panlab, Madrid, Spain) located in the rectum for 10 s. Analgesia was eval- uated using the hot-plate paradigm. The test consists of placing a mouse on an enclosed hot plate (Columbus Instruments, Columbus, OH, USA) and measuring 64 the latency to lick a hindpaw or jump out of the enclosure. Walking patterns were monitored as well with painted feet on paper sheets (blue, fore; red, hind) of 70-cm length. Locomotor responses were assessed in the open field test, conducted in an arena on a grey methacrylate open field of 70 x 70 x 40 cm under uniform lightning conditions of 25 lux. Total distance, global activity and resting time were analyzed for 10 min. Motor coordination (RotaRod) analysis was conducted as previously described (Blázquez et al., 2011) with acceleration from 4 to 40 rpm over a period of 570 s in an LE8200 device (Harvard Apparatus, Barcelona, Spain). Briefly, mice were tested on three consecutive days, for three trials per day with a rest period of 30 min between trials. Data from the three trials per day were averaged for each animal, and the mean value of each day averaged for each animal. Data from the first day were not used in statistical analyses. To study anxiety-like behaviors in the elevated plus-maze test, a 5-min observation session was performed, in which each mouse was placed in the central neutral zone facing one of the open arms. The cumulative time spent in open and closed arms was then recorded. Likewise, one entry was considered when the animal had placed at least both forelimbs in the arm. Data are represented as percentage of time spent in the open arms and the number of visits in the open arms. The Y-maze test was performed to assess short-term memory in the mice. The Y-maze apparatus consists of three identical arms (35 cm-long, 10 cm-wide, 25 cm- high). The subject mouse was placed in the center, allowing it to explore all three arms of the maze for 5 min driven by the innate curiosity of rodents to explore previously unvisited areas. A mouse with intact working memory will remember the arms previously visited and show a tendency to enter a less recently visited arm. Spontaneous alternation is measured as a memory index. For the food location task, mice were habituated for 5 min in a V-maze with two arms, similarly as previously described (Azevedo et al., 2019). Mice were allowed to freely explore the V-maze for 5 min the day before to let them become familiar with the maze and minimize distractions during the subsequent test. After exploring the maze, mice were returned to their home cage and fasted for 16-23 hours. The next day, mice were returned to the same V-maze containing two clean, empty cups at the end of each arm and allowed to explore the context for 5 min. During the training session, one of the cups was filled with food pellets and mice were allowed to explore for 5 min. They were then returned after training to their home cage, fed, and before the dark phase started, mice were fasted for an additional 16-23 hours. For assessing memory encoding of the location during the test session on the next day, mice were allowed to explore for 5 min the maze containing two empty food cups again. Cups were cleaned with 70% ethanol in between trials to remove any food crumbles or mouse smell. For both the context, the training and the test condition, the ratio between the time spent at the end of a given arm and the sum of time spent at the end of the two arms containing cups was calculated. A preference index (PI) was calculated to measure recognition memory: (tcontainingfood) / (tcontainingfood + tempty). For induction of acute excitotoxic seizures, KA (Sigma) was dissolved in isotonic saline, pH 7.4, and administered intraperitoneally (i.p.) to adult the age- and weight- matched mice. Seizures were induced by injecting 30 mg/kg of KA in a volume of 10 mL/kg body weight (Monory et al., 2006; Marsicano et al., 2003). Where indicated, 65 i.p. injection with vehicle (1% v/v DMSO in 1:18 v/v Tween-80/saline solution) or 10 mg/kg THC (THC Pharm) was performed 15 min prior to KA injection. After KA injections, mice were placed in clear plastic cages. Mice were constantly recor- ded and monitored for 2 hours for motor activity and manifestations of seizures. They were scored every 5 min according to a modified Racine scale (Racine, 1972): immobility (stage 1); forelimb and/or tail extension, rigid posture (stage 2); repetit- ive movements and head bobbing (stage 3); rearing and falling (stage 4); continuous rearing and falling, jumping, and/or wild running (stage 5); generalized tonic–clonic seizures (stage 6); and death (stage 7) (Monory et al., 2006). Quantified paramet- ers included time course of Racine scale, onset of seizure activity and mortality during SE. Seizure severity was determined as previously described (Armas-Capote et al., 2020): Seizure severity = Σ(all scores of a given mouse) / time of experiment . Statistics Data are presented as mean ± S.E.M., and the number of experiments is indic- ated in every case. Statistical analysis was performed with GraphPad Prism v8.0.1 (GraphPad Software, La Jolla, CA, USA). All variables were first tested for normal- ity (Kolmogorov-Smirnov test) and homocedasticity (Levene’s test). When variables satisfied these conditions, one-way ANOVA with Tukey, Sidak or Dunnett’s post hoc test, or by paired or unpaired Student’s t test were used as appropriate. p-values of <0.05 were considered as statistically significant. 66 Results 4.1 Results of Aim 1 Identification of GAP43 as a potential new interacting part- ner of CB1R CTD We aimed at defining the neural interactome of CB1R in order to identify new po- tentially relevant receptor-associated proteins. For this purpose, we conducted a high-throughput screening by affinity purification and subsequent tandem MS/MS (Figure 4.12 A). The cytoplasmic regions of the receptor (mainly its CTD and IL3) are involved in G protein and β-arrestin binding, as well as in receptor desensitization and trafficking (Howlett et al., 2010). As most of the cytoplasmic CB1R-interacting proteins known to date bind to specific residues of the receptor CTD (Stadel et al., 2011), we decided to use this region as bait. Purified hCB1R-CTD (amino acids 408- 472) was bound to lectin upon cloning into the expression vector pKLSLt. Lectin allows specific Sepharose binding and subsequent lactose elution for affinity purific- ation chromatography, and increases the solubility of fused peptides (jing Li et al., 2016). This procedure substantially reduces the amount of nonspecific background compared to bait protein immunoprecipitation, in which centrifugation steps can lead to significant dragging of contaminants. A sheep whole-brain homogenate was used as starting biological material in or- der to obtain large amounts of neural proteins (Figure 4.12 A). As a clarifying step, we loaded the homogenate first onto a Sepharose 4B chromatography column to avoid nonspecific interactions with the resin. The flow-through was subsequently challenged to a lectin-hCB1R-CTD-Sepharose 4B column, searching for specific in- teractions of brain proteins with the receptor CTD, or alternatively to an additional lectin-Sepharose column. After washing and elution with lactose, the resulting pro- teins were digested with trypsin and subjected to tandem MS/MS (Figure 4.12 A). The resulting spectra were subsequently challenged to databases of mammalian spe- cies by using Proteome Discoverer software. Peptide identification was validated by the Percolator algorithm using q ≤ 0,01, which defines the probability that the ob- served match between the experimental data and the database sequence is a random event. We conducted a rigorous statistical analysis of MS data for correct protein identification by peptide matching within the databases. Common contaminants in these experiments, such as cytokeratins, chaperones, histones, DNA polymerase subunits and mitochondrial components, were removed from the list. Next, we sub- stracted from the list the peptides that bound specifically to lectin in the additional column. In addition, to exclude nonspecific data, we compared our list of proteins with those from other MS/MS studies by using the CRAPome database with a co- incidence cut-off value of <10%. After all these steps, we obtained a final list of ∼50 potential CB1R-interacting proteins (Table 4.5). Some of the hits we identified, such 67 Figure 4.12: The CTD of CB1R interacts with GAP43. A. Schematic workflow of the affinity purification and tandem MS/MS conducted. A sheep whole-brain homogenate was clarified in a Sepharose 4B chromatography column and the flow-throw was loaded in a Lectin-CB1R-CTD (amino acids 408-472)-bound Sepharose 4B column. After washing, elution with lactose, eluted-fraction separation by PAGE-SDS and digestion with trypsin, peptides were subjected to nLC/MS-MS proteomic analysis. From the resulting spectra, a list of peptides was identified, and GAP43 was selected. B. Representative saturation binding curve obtained by plotting increasing concentrations of purified GAP43 in the presence of 5IAF-labelled CB1R-CTD against increasing values of calculated fluorescence polarization (FP) for each concentration value expressed as mili FP units (mFP). A representative experiment is shown (n=3). A non-linear regression fitting the curve to a hyperbola was performed and a Kd=59.1 µM was obtained. as plasma membrane Ca2+ ATPases, G-protein α subunits (specifically, Gαi1), Na+ and Cl−-dependent GABA transporters, Hsp70 and MAPK family members, coin- cided with those found in previous CB1R proteome high-throughput studies (Njoo et al., 2015; Mattheus et al., 2016). Distinctively, our list of potential interactors with high statistical significance included the pleiotropic growth-associated protein 43 ((GAP43/neuromodulin)) (Table 4.5; Box1). A Gene Ontology (GO) analysis of the main clusters obtained by the network analysis of the 50 identified proteins, by using the STRING Database and MCL Clustering, identified two enriched functional GO terms, both of them including CB1R and GAP43: GO.0008037-Cell recognition (with matching proteins CNR1, CRTAC1, GAP43 and MFGE8), and GO.0008038- Neuronal recognition (with matching proteins CNR1, CRTAC1 and GAP43) (p value = 3.84 x 10−2 for both terms). Based on its space-temporal location in the CNS and its possible anatomical and functional relations with CB1R (see below), we selected GAP43 out from the other hits we obtained as a potential new CB1R interactor. To evaluate a direct protein-protein interaction between CB1R and GAP43, we first expressed and purified GAP43 and 5IAF-labelled CB1R-CTD. In vitro protein- protein interaction was measured by performing fluorescence polarization/anisotropy binding experiments. Fluorescence polarization is a measure of the space-orientation changes of a molecule with respect to the time between the absorption and emission events. If the fluorophore population is excited with vertically-polarized light, the emitted light will retain some of that polarization based on how fast it is rotating in solution. When proteins interact, they form bigger complexes, which are slower 68 Protein name Score Unique peptides Coverage Serum albumin 135,42 9 22,41 LanC-like protein 2 44,24 7 17,23 DNA damage-binding protein 1 36,54 4 5,41 Aldehyde dehydrogenase 9 family, member A1 16,85 4 7,69 Heat shock cognate 71 kDa protein 41,91 3 25,42 Protein cereblon 14,97 3 9,01 Lumican 12,90 3 8,5 Desmoplakin 9,61 3 1,36 Cannabinoid receptor 1 71,16 2 9,49 Desmoglein-1 17,97 2 2,77 F-box only protein 3 15,70 2 11,52 Desmocollin-1 10,03 2 3,57 Filaggrin-2 8,30 2 1,74 Glycodelin 21,88 1 6,18 Creatine kinase B-type 14,95 1 12,88 Immunoglobulin heavy variable 3-23 9,53 1 16,49 Neuromodulin/GAP43 7,94 1 8,29 Dihydrolipoyl dehydrogenase, mitochondria 7,41 1 3,51 Complement factor B 5,34 1 7,77 Synaptic vesicle glycoprotein 2A 4,52 1 1,75 Serpin B12 4,47 1 8,74 Sodium- and chloride-dependent GABA transporter 3 4,39 1 12,88 Submaxillary gland androgen-regulated protein 3B 4,22 1 37,5 Kynurenine–oxoglutarate transaminase 3 4,14 1 2,79 Protein farnesyltransferase/geranylgeranyltransferase type-1 subunit alpha 4,10 1 22,41 Cartilage acidic protein 1 4,00 1 7,25 Glial fibrillary acidic protein 3,93 1 8,47 Corneodesmosin 3,93 1 16,36 Secretory phospholipase A2 receptor 3,78 1 1,08 Protein S100-A8 3,61 1 11,83 Glyceraldehyde-3-phosphate dehydrogenase 3,61 1 6,01 Kelch repeat and BTB domain-containing protein 11 3,58 1 2,4 Lactadherin 3,54 1 3,4 Plasma membrane calcium-transporting ATPase 1 3,44 1 8,66 Dermcidin 3,25 1 10 ATP-dependent 6-phosphofructokinase, muscle type 3,28 1 2,28 Mitogen-activated protein kinase 1 3,17 1 6 EEF1A1 protein 3,17 1 16,98 Tubulin alpha-8 chain 3,08 1 16,22 Neutrophil defensin 1 2,97 1 30 ATP synthase subunit beta 2,96 1 10,1 Lysozyme C 2,74 1 40 Myelin proteolipid protein 2,69 1 16,42 Catalase 2,68 1 1,71 Solute carrier family 7 member 13 0,00 1 5,2 Guanine nucleotide-binding protein G(i) subunit alpha-1 3,46 1 3,26 Sushi repeat-containing protein SRPX2 0,00 1 40,59 Pyridoxal kinase 0,00 1 11,58 Neurexin-2 0,00 1 2,8 Table 4.5: List of potential CB1R interactors identified in proteomic-MS analyses with statistical significance. The list was obtained after subtraction of common contaminants lectin-bound proteins and Crapome cut-off application. The score is the relative peptide query score. First, an ion score is given as a function of the p-value [-10Log(P)] as the probability that the peptide identification is a false positive; the protein score is the sum of the highest ion score for each distinct sequence identified. Coverage is the percentage of the protein sequence that has been identify. Unique peptides are identified peptides that are unique for that protein. 69 in motion than single protomers, so the emitted light retains higher polarization. The faster the orientation motion, the higher depolarization of the emitted light will be. In our setting, increasing concentrations of GAP43 (0-325 µM) were used in the presence of 10 µM 5IAF-labelled CB1R-CTD (Figure 4.12 B). This approach showed a saturating polarization curve, thus supporting a direct interaction between GAP43 and CB1R-CTD, with relatively high affinity (Kd = 59.1 µM). Box 1: GAP43/NEUROMODULIN The growth-associated protein-43 (GAP43) has been known under various designations over time such as B-50, F1, P-57, pp46 or neuromodulin, in the course of separate yet overlapping investigations into a neural-specific axonal- associated membrane-bound phosphoprotein that contributes to plasticity and growth of the presynaptic terminal. It is considered a key component of a larger presynaptic proteome, playing a role in the initial development of neural outgrowth and subsequent modulation, as would occur during regeneration and LTP (Karns et al., 1987; Holahan et al., 2007). The human GAP43 gene is localized in chromosome 3, while the mouse and rat GAP43 gene are localized in chromosomes 16 and 11 respectively, and is translated into the corresponding neural-specific GAP43 proteins in similar ways. The GAP43 gene includes three exons. The GAP43 promoter region contains seven E-boxes (E1 to E7) that are organized in two clusters, a distal cluster (P1-E3 to 7) and a proximal cluster (P2-E1 and E2). The regulation of GAP43 expression occurs mainly via the conserved E1 E-box positioned in active P2 promoter, which is neuronally restricted. Additionally, the repress- ive element named “SNOG” contributes to inhibit transcription in a variety of non-neuronal cells and tissues (Chiaramello et al., 1996). GAP43 mRNA expression is regulated by HuD expression. HuD is a neuronal RNA-binding protein, targetted by PKC, that is known as the main stabilizing agent for GAP43 mRNA. Therefore, PKC appears to firstly regulate GAP43 mRNA through a translation-independent mechanism. Two protein isoforms by al- ternative splicing of GAP43 mRNA are described. The predominant form of GAP43 protein has 238 aminoacids in humans and 227 or 226 in mouse or rat, respectively (A). It is mainly a hydrophilic protein lacking a membrane- spanning domain and containing no sites for glycosylation. However, a short hydrophobic N -terminal region (from residue 1 to 10) highly conserved in all vertebrates serves as a membrane-targeting domain. The presence of residues C3 and C4, that undergo reversible palmitoylations (Gauthier-Campbell et al., 2004; Tomatis et al., 2010), functions to anchor GAP43 protein to the cyto- plasmic side of the presynaptic plasma membrane. The adjacent polybasic cluster, containing residues R6, R7, K9 and K13, stabilizes membrane bind- ing by establishing electrostatic interactions. This mimics the cytoplasmic tail of GPCRs (Denny, 2006). Following the N -terminal region, a positively charged effector domain (from residues 32 to 52) contains a calmodulin (CaM)- binding region (B; known as the IQ domain) and a phosphatidilinositol-4,5- bisphosphate (PIP2)-binding motif through electrostatic interactions. A serine residue in position 41 (S41) within this domain is crucial because it is the only 70 site that can be phosphorylated by PKC. The CaM binding region forms an α- helix to bind CaM and contains five positively-charged amino acids following S41, which presumably facilitate binding to negatively-charged calmodulin (or PIP2). The PKC-mediated phosphorylation at S41 conceivably generates re- pulsion of electrical charges and steric hindrances in the helix, which precludes CaM binding (Chapman et al., 1991; Leu et al., 2010; Holahan, 2017). Finally, we can find a mostly unknown C -terminal domain, which is specific to GAP43. Thus, GAP43 functional attributes can be ascribed to the transcriptional and post-transcriptional regulation of the mRNA, and the post-translational modi- fication (phosphorylation) and protein localization at the biochemical level, most of them PKC-dependent (Perrone-Bizzozero et al., 1993; Sanna et al., 2014). GAP43 is initially synthesized as a soluble protein on free ribosomes in the cell body and palmitoylation takes place at the endoplasmic reticulum-Golgi intermediate compartment, where it binds to membranes of the early secret- ory pathway. Once palmitoylated, GAP43 becomes part of lipid rafts. Then, it is sorted onto vesicles and travels through the secretory pathway antero- gradelly at fast axonal transport rate (Hooff et al., 1989). It is mostly found at the cytosolic face of the plasma membrane in extending neurites and adult axon terminals. Ultrastructural analyses have revealed that a minor portion of GAP43 is also associated with vesicles, where it colocalizes with the synaptic vesicle marker synaptophysin. Residual GAP43 was found in the cytosol of axon terminals, but was absent at the plasma membranes of dendrites (Gorgels et al., 1989; Verkade et al., 1996; Campagne et al., 1990). A reduction in the palmitoylation of GAP43 is sufficient to inhibit axon fasciculation (Gauthier- Campbell et al., 2004). Additionally, the activation state of GAP43 appears to correlate with its localization (Kristjansson et al., 1982). Thus, two posttrans- lational mechanisms have been shown to determine polarized membrane traf- ficking to sort and deliver GAP43 to the distal axonal plasma membrane and vesicles: palmitoylation tags GAP43 for global sorting by piggybacking on exocytic vesicles, whereas phosphorylation locally regulates protein mobility and plasma membrane targeting of palmitoylated GAP43 (Gauthier-Kemper et al., 2014). 71 GAP43 interacts with CB1R in neural tissue Both GAP43 and CB1R are mostly neuron-specific proteins and share a high abund- ance in the hippocampus (Benowitz et al., 1988, 1989; Kano et al., 2009; Herkenham et al., 1990). Hence, primary hippocampal-neuron cultures were prepared from P0- 2 newborn mouse. In 7-DIV cultures, a marked overlapping was observed between CB1R and GAP43 immunoreactivity at the tips of neural processes labeled with βIII-tubulin (Figure 4.13 A). In order to validate a possible protein-protein interac- tion, we conducted co-immunoprecipitation assays in primary hippocampal neurons (Figure 4.13 B, left panel) as well as in dissected mouse hippocampal tissue (Figure 4.13 B, right panel). Lysates were immunoprecipitated with an anti-GAP43 anti- body, immunocomplexed proteins were resolved on SDS-PAGE, and the coimmuno- precipitate was developed with an anti-CB1R C -terminal antibody. This procedure revealed a band at ∼50 kDa that was absent in a CB1R-KO sample and in the co-IP negative control, thus indicating that endogenous CB1R and GAP43 interact in hippocampal-neuron cultures. GAP43 and CB1R are predominantly located on presynaptic terminals (Schlicker and Kathmann, 2001; Verkade et al., 1996). Hence, we asked whether they inter- act at these sites by using synaptosomal preparations isolated from hippocampal mouse tissue. Synaptosomes provide a very versatile experimental system in which epitopes for antibody recognition are accessible while maintaining much of the nat- ural synaptic architecture (Mart́ın et al., 2010). Immunostaining of synaptosomes, labeled with the marker protein synaptophysin (Syn), for GAP43 and CB1R re- vealed GAP43/CB1R colocalization (Figure 4.13 C, lower panel). These proteins were found at 19.7 ± 3.1 % (GAP43) or 20.1 ± 4.3 % (CB1R) of Syn-positive but- tons, and 8.2 ± 1.8 % of Syn-positive buttons were double-positive for both GAP43 and CB1R (Figure 4.13 C, upper panel). To test a potential protein-protein inter- action, we performed proximity ligation assays (PLA) in synaptosomes (Figure 4.13 D, upper left panel). This technique allows detecting the close proximity (usually less than ∼40 nm) between proteins in situ (Taura et al., 2015). Hippocampal syn- aptosomes from WT mice showed GAP43/CB1R PLA-positive dots, and this signal was drastically reduced in synaptosomes from CB1R−/− mice (Figure 4.13 D, lower panel). Specifically, quantification of PLA-positive signal relative to synaptosomes per field reveal a decrease of 62 % of signal in CB1R−/− animals compared to WT values (Figure 4.13 D, upper right panel). This observation supports an interaction between GAP43 and CB1R in hippocampal terminals. GAP43 phosphorylation favors its interaction with CB1R in HEK293T cells To gain further insight into the GAP43-CB1R interaction, we next used the HEK293T cell line to express different forms of GAP43. It is widely accepted that GAP43 activ- ation depends on PKC-mediated phosphorylation of the S41 residue (Box 2). Thus, when GAP43 is phosphorylated at S41, it can be released from CaM-inhibitory binding and interact with a set of proteins to mediate its functions (Holahan, 2017). In line with previous studies (Leu et al., 2010), to modulate the activation state of GAP43, we designed two mutant versions of the protein (Figure 4.14 A). One version bore an S41A point mutation, so the resulting protein is refractory to phosphoryla- 72 Figure 4.13: CB1R interacts with GAP43 in neural tissue. A. Confocal images of hippocampal primary neur- ons at DIV7 labeled for CB1R in red, GAP43 in green and β-tubulin III in grey. Doted boxes indicate the position of the inset below, pointing an axonal growth cone (white arrow). B. Representative Western blots showing CB1R- GAP43 immunocomplex in (left) hippocampal primary neurons or (right) hippocampal tissue. Immunoprecipitation (IP) was conducted with anti-GAP43 antibody. Cell extracts from WT and CB1R−/− mice are shown, together with the IgG control. C. Up, Representative confocal images of synaptosomes immunostained for GAP43 in green, CB1R in red and Syn-1 in grey, in hippocampal synaptosomes from WT mice. Down, Quantification of the percentage of Syn+/GAP43+, Syn+/CB1R+ or Syn+/GAP43+/CB1R+ synaptosomes (n=5 mice, 3 fields per animal). D. PLA for CB1R and GAP43 was performed in hippocampal synaptosomes from WT mice. Up left, Schematic of PLA, showing CB1R in purple and GAP43 in pink. Rolling circle amplification of fluorescently tagged oligonucleotides (represented in red) occurs after primary and secondary probes bind. Up right, Quantification of the number of PLA-positive signal per synaptosome per field (n=3 animals per genotype; p<0.05 by unpaired Student’s t-test). (Down), Representative confocal images show CB1R-GAP43 complexes appearing as red dots. PLA signal was absent from hippocampal synaptosomes in CB1R−/− mice. 73 tion and activation. The other version was achieved by an S41D point mutation, which thus allows its constitutive phosphorylated-like conformation and activation. To test whether the phosphorylation status of S41 affects GAP43-CB1R in- teraction, we conducted PLA and co-IP assays. For PLA, HEK293T cells were co-transfected with CB1R-myc plus the different forms of GAP43 fused to GFP (GAP43-WT-GFP, GAP43-S41D-GFP, GAP43-S41A-GFP). Primary antibodies against c-myc and GFP were used. Control experiments were conducted in the absence of one of the two primary antibodies, as well as in cells transfected with CB1R-myc and empty GFP (Figure 4.14 B) or, alternatively, in cells transfected only with GAP43-WT-GFP in absence of CB1R (Figure 4.14 C). GAP43-S41D-CB1R and GAP43-WT-CB1R complexes were readily detected and quantified as PLA-positive red dots in GFP-positive cells normalized to control condition (3.0 ± 0.8 and 2.6 ± 0.5 dots per cell, respectively), while remarkably lower complex levels were found in cells transfected with GAP43-S41A-GFP (1.6 ± 0.3 dots per cell) (Figure 4.14 B). Similar data were obtained by using primary antibodies against CB1R and GAP43 in HEK293T cells transfected with untagged plasmids as a control (Figure 4.14 D). For co-IP assays, HEK293T cells were co-transfected with HA-tagged CB1R plus the different versions of GAP43 (Figure 4.15 left panel). Upon HA-CB1R precipitation, GAP43-S41D was the predominant co-inmunoprecipitated form of GAP43, whereas GAP43-WT and GAP43-S41A appeared mostly in the unbound fraction. Similar data were obtained when we precipitated GAP43 and blotted the co-inmunoprecipitated HA-CB1R (Figure 4.15 A right panel). It is worth noting that the anti-GAP43 mAb7B10 antibody used in this study recognizes all post- translationally modified forms of GAP43 (He et al., 1997; Meiri et al., 1991). Finally, we used BRET as a third technique to assess protein-protein interactions in cultured cells (Figure 4.15 B). Thus, a constant amount of a donor RLuc-tagged version of CB1R and increasing amounts of mutant forms of GAP43 fused to GFP as acceptors were used. If the two fusion proteins interact and this positions the energy donor and acceptor within a distance of less than ∼10 nm, resonance energy transfer occurs upon donor substrate addition and an additional light signal corresponding to the acceptor reemission can be detected. We found a positive and saturating BRET signal for CB1R–RLuc plus GAP43-WT, and for CB1R–RLuc plus GAP43- S41D (Bmax = 75.0 ± 8.3 and 63.0 ± 5,7 mBRET units; BRET50 Kd = 15.2 ± 4.5 and 7.1 ± 2,5 mBRET units µM for GAP43-WT and S41D, respectively). The pair CB1R–RLuc plus GAP43-S41A gave a basically linear, non-specific BRET signal, thus providing further support to the specificity of the interaction between CB1R and phosphorylated/active GAP43 in HEK293T cells. Additionally, CB1R 1-458, lacking the last stretch of the receptor CTD, exerted a partial blockade of binding (Figure 4.15 B). Moreover, two fragments of GAP43 were cloned in an attempt to approach the region of interaction within the GAP43 protein (Figure 4.15 C). One fragment encompassed amino acids 5 to 68 of GAP43, which includes the N -terminal region and effector domain. The other fragment encompassed amino acids 60 to 238, comprising the whole C -terminal domain of GAP43. Interestingly, neither of these fragments yielded a BRET saturable curve. This observation may indicate that the entire protein is needed for the interaction. In any event, taken together, all these data show that GAP43 and CB1R interact specifically in HEK293T cells, and that phosphorylated GAP43 favors this interaction. Taken together, these results show that GAP43 can interact with CB1R in 74 Figure 4.14: GAP43 phosphorylation favors its interaction with CB1R in HEK293T cells by in situ PLA detection. A. Scheme of mutant constructs aimed to modify GAP43 activation state. It shows the inactive form and the phosphomimetic form of GAP43 obtained by a change mutation at S41 in the original protein by A (S41A) or D (S41D), respectively. B. PLA for CB1R and GAP43 was performed in HEK293T cells transfected with CB1R-myc plus empty GFP, GAP43-WT-GFP, GAP43-S41D-GFP or GAP43-S41A-GFP, with anti-c-myc and anti-GFP antibodies. Up, Quantification of CB1R-GAP43 complex expression. Results are expressed as PLA ratio (number of red dots per cell). Data are given as mean ± SEM (n = 5 experiments, 22 fields per n; p<0.05 by ordinary one-way ANOVA with Dunnett’s multiple comparisons test). Down, Representative confocal microscopy images show CB1R-GAP43 complexes appearing as red spots. Cell nuclei were stained with DAPI (blue). C. Control PLA experiments were performed in WT HEK293T cells not expressing CB1R. D. Control PLA experiments in HEK-293T cells transfected with untagged GAP43 and CB1R and using antibodies against the full GAP43 protein and CB1R-CTD. 75 HEK293T cells, and that the GAP43 S41-phosphorylated state favors this inter- action. Figure 4.15: GAP43 phosphorylation favors its interaction with CB1R in HEK293T cells by co- immunoprecipitation and BRET. A. Representative Western blots of coimmunoprecipitation protein com- plexes from HEK293T lysates co-transfected with HA-tagged CB1R and GAP43-WT, GAP43-S41D, GAP43-S41A (Left) Immunoprecipitation (IP) was conducted with anti-HA antibody. (Right) Immunoprecipitation (IP) was conducted with anti-GAP43 antibody. Note co-IP is evident when GAP43-S41D is co-transfected. B. BRET sat- uration experiments in HEK293T cells expressing CB1R-Rluc cDNA and increasing amounts of GAP43-WT-GFP (blue), GAP43-S41D-GFP (red) or GAP43-S41A-GFP (green). BRET is expressed as milli BRET units (mBU) (mean ± standard deviation (SD); n=3; fitted to a nonlinear regression curve). Kd=7.1±2.5 for GAP43-WT and Kd=15.2±4.5 for GAP43-S41D. C. BRET experiments in HEK-293T cells expressing Rluc-CB1R and increasing amounts of GAP43-WT-GFP together or not with two fragments of GAP43 as competitors. Maximum BRET signal of each curve is showed (n=3). 76 Box 2: GAP43 PHOSPHORYLATION GAP43 is subjected to phosphorylation by several kinases, including clas- sical Ca2+-dependent type PKC isoforms (mostly βII), casein kinase II (Apel et al., 1990, 1991; Taniguchi et al., 1994), as well as proline-directed kinases such as JNK, which phosphorylates GAP43 at S96 in developing and regen- erating axons of mouse brain (Kawasaki et al., 2018; Igarashi et al., 2020). However, regarding biochemical and physiological roles of the protein, phos- phorylation of GAP43 by PKC at a unique site (S41) appears to be the most relevant modification. In vivo, PKCβ shows a pattern of activation and ex- pression consistent with phosphorylated GAP43 (Meiri et al., 1991; Gauthier- Kemper et al., 2014). In the presynaptic terminal, CaM shows a high affinity for GAP43. If PKC is not sufficiently activated by transient changes in second messengers as Ca2+ or DAG, S41 is not phosphorylated and CaM remains as- sociated to GAP43. When Ca2+ levels rise, as during high plasticity events, Ca2+-dependent PKC activation phosphorylates GAP43, and CaM is released. In a feedback loop, the slower Ca2+/CaM-dependent activation of calcineurin results in GAP43 dephosphorylation and CaM re-association (Caprara et al., 2016; Baumgärtel and Mansuy, 2012; Lautermilch and Spitzer, 2000). GAP43 undergoes regulated proteolysis as well. It is a substrate for calcium-activated cysteine protease m-calpain, that cleaves GAP43 near S41 (Zakharov and Mosevitsky, 2007), and it is degraded by the ubiquitin/ proteasome system (Denny, 2006). Phosphorylated GAP43 is known to interact with other proteins to fulfill its functions in the cell. GAP43 is associated with membrane actin-cytoskeleton fraction to facilitate axonal elongation/outgrowth and adhesion to substrate. It regulates actin filament length by establishing lateral interaction and sta- bilization of filamentous (F)-actin (He et al., 1997; Moss et al., 1990) or with cytoskeleton-associated proteins as brain spectrin (Riederer and Routtenberg, 1999). GAP43 tightly associates to lipid rafts, where it sequesters PI(4,5)P2 and modulates its local availability for binding to actin-controlling proteins to increase assembly. This retention can be blocked by S41 phosphorylation (Denny, 2006; Laux et al., 2000; Leu et al., 2010; Larsson, 2006). GAP43 interacts as well with SNAP25, syntaxin and VAMP, all of them components of the SNARE complex involved in vesicle fusion (Haruta et al., 1997), and also with the vesicle-associated protein Rabaptin-5, an effector of the small GTPase Rab5 that mediates membrane fusion in endocytosis of incoming ves- icles to expand early endosomes. The overexpression of GAP43 in neurons causes a decrease in the size of Rab5-containing endosomes, accompanied by an acceleration of uptake of material into endosomes and of synaptic vesicle recycling. These data implicate GAP43 in an earlier step of endocytosis than the fusion and expansion of endosomes, which is mediated by the recruit- ment of rabaptin-5 by activated Rab5 and is rather involved in the endocytic stage of neurotransmitter release, when vesicular membrane is retrieved for reuse (Neve et al., 1998). These interactions occur in axonal terminals in a phosphorylation-dependent manner and allow GAP43 to modulate the release of neurotransmitters such as acetylcholine, dopamine, noradrenaline and vari- ous neuropeptides (Kumagai-Tohda et al., 2006; Ivins et al., 1993; Hens et al., 1995; Dekker et al., 1989). 77 GAP43 inhibits CB1R-mediated signaling in HEK293T cells We next asked about the functional consequences of GAP43 binding to CB1R. We first performed dynamic mass redistribution (DMR) assays (a technique that allows quantifying changes in overall signaling triggered by a receptor agonist) in HEK293T cells stably expressing FLAG-CB1R (herein referred to as HEK-CB1R cells. Changes are detected in light diffraction at the bottom 150 nm of a cell monolayer. When cells were treated with the CB1R agonist WIN at 100 nM we registered a response that lasted at least 2,000 s. GAP43-WT or GAP43-S41D co-expression blunted WIN-evoked action (Figure 4.16 A). This inhibitory effect was not evident when GAP43-S41A was co-transfected, pointing to the involvement of the interaction between active GAP43 and CB1R. This effect seemed specific for CB1R as it did not occur with the HU-308-induced activation of CB2R, the GPCR that shows, by large, the highest homology with CB1R. As the aforementioned DMR assays revealed a similar inhibitory effect of both GAP43-WT and GAP43-S41D on CB1R signaling, we decided to use GAP43-WT hereafter. This provides a more physiological approach as the GAP43 phosphoryla- tion–dephosphorylation cycle is preserved, therefore allowing a physiological fine- tuning of its activity. As active GAP43 may conceivably decrease G protein-coupled CB1R signaling, we sought to evaluate the coupling of the receptor to different G proteins in the absence or presence of GAP43. For this purpose we conduc- ted [35S]GTPγS scintillation proximity assays coupled to immunoprecipitation with specific antibodies raised against different Gα subunits (Figure 4.16 B). An over- all decrease in CB1R-mediated G protein recruitment pattern was observed in the presence of GAP43 as Gαi2, Gαi3 and Gαq/11 coupling was abolished. Nevertheless, Gαi1 binding was preserved. To determine the functional impact of this reduced G-protein coupling we per- formed a series of Western blot assays in HEK-CB1R cells co-transfected with GAP43-WT-GFP or control vector. We observed that the two ectopically-expressed proteins colocalized in the plasma membrane as well as in the cytoplasm (Figure 4.17 A), and we tested transfection efficiency prior to performing the experiments (Figure 4.17 B). PKA phosphorylated substrates after WIN (100 nM) and/or for- skolin (FSK; 0.5 µM) treatment were not substantially altered (Figure 4.16 D), and neither were cAMP levels (Figure 4.16 C). Likewise, WIN-evoked ERK phosphoryla- tion was unaffected by GAP43 (Figure 4.16 D). These data suggest that the amount of Gαi1 coupling remaining upon GAP43 co-expression was sufficient to trigger full Gαi/o-mediated CB1R signaling. Given the prominent role of GAP43 in cytoskeletal remodeling, the Rho/ROCK/ cofilin/actin cytoskeleton pathway was subsequently tested. In this case, 5-min WIN (100 nM) treatment increased ROCK2 and cofilin phosphorylation, and GAP43 coexpression prevented this effect (Figure 4.17 C). As the coupling of CB1R to Gαq/11 was reduced as well by GAP43 (see above), we evaluated whether this route may affect signaling events as triggered by CB1R upon GAP43 ectopic expression. We tested an inhibitor of Gαq/11 (YM-2514890, at 1 µM) by Western blot. This drug decreased similarly the phosphorylation status of ROCK2 and cofilin (Figure 4.17 D), and DMR experiments showed a blockade of the WIN-mediated response by YM-2514890 and the ROCK inhibitor Y-27632 at 10 µM (Figure 4.17 E). Taken together, these data show that GAP43 reduces CB1R signaling in a spe- cific manner: it does not affect Gαi/o-mediated routes, such as the decrease in 78 Figure 4.16: GAP43 reduces CB1R-mediated signaling independently of Gαi/o in HEK293T cells. A. DMR assays were conducted in (left) HEK-CB1R cells transfected with GAP43-WT, GAP43-S41D, GAP43-S41A and exposed to 100nM WIN, or in (right) HEK-CB2R cells transfected with GAP43-WT exposed to 100 nM HU- 308. A representative experiment is shown (n=3). B. Coupling of CB1R to Gα proteins in membrane extracts from HEK293T cells expressing CB1R, together or not with GAP43-WT. Data are shown as percentage of [35S]GTPγS basal binding values obtained for each specific subunit. Bars represent mean ± SEM *p<0.05 from basal (dashed line) by one-sample Student’s t-test (n=4). C. cAMP concentration was determined in cells exposed to 100 nM WIN or vehicle and 15 min later with 0.5µM forskolin (F, or F + WIN). Values are graphed as mean ± SEM (n=3; **p<0.01 from vehicle or p<0.01 from F alone, by one-way ANOVA with Tukey’s multiple comparisons test). D. HEK-CB1R cells transfected with empty vector or GAP43-WT were incubated with vehicle or 100nM WIN for 5 min and subsequently treated with FSK for 10 min when indicated. Representative Western blots are shown. Quantification of optical density (OD) values of PKA phospho-substrates and phosphoERK1/2 relative to those of the loading control are shown (n=6; *p<0.05 by one-way ANOVA with Tukey’s multiple comparisons test). 79 Figure 4.17: GAP43 reduces CB1R-activated ROCK pathway signaling via Gαq/11 in HEK293T cells. A. Confocal images of HEK-CB1R cells transfected with GAP43-GFP. CB1R-FLAG is labelled in red.B. Western blotting of GFP and GAP43 showed as controls of transfection. C. HEK-CB1R cells transfected with empty vector or GAP43-WT were treated with vehicle or 100nM WIN for 5 min. Representative experiments showing phosphoROCK2 and phosphocofilin and levels of total ROCK2 and cofilin are depicted. Quantification of optical density (OD) values of phosphorylated proteins relative to those of the loading control are shown. Data are the mean ± SEM of counts (n=6–8; *p<0.05 by one-way ANOVA with Sidak’s multiple comparisons test or unpaired Student’s t-test from control). D. HEK-CB1R cells were treated with vehicle or 1 µM YM-254890 for 30 min and with vehicle or 100 nM WIN the last 5 min. Representative Western blots are shown. Quantification of OD values of phosphoROCK2 and phosphocofilin relative to those of the loading control are depicted (right). Data are the mean ± SEM of counts (n=6–8; *p<0.05 by one-way ANOVA with Sidak’s multiple comparisons test or unpaired Student’s t-test from control). E. Representative experiment of DMR assays in HEK-CB1R cells that were pretreated for 30 min with (left) vehicle or 1 µM YM-254890 or (right) vehicle or 10 µM Y-27632, and further treated with 100nM WIN. 80 cAMP/AC/PKA or the increase in ERK1/2 phosphorylation, but occludes ROCK pathway activation possibly by a reduction of Gαq/11 coupling. We next conducted ROCK2 phosphorylation analyses in 7-DIV cultures of primary neurons nucleofected with GAP43-WT or control vector (Figure 4.18). The WIN- mediated phosphorylation of ROCK2 (which shows a labeling pattern predominantly somatic) was abrogated by GAP43, as shown by both Western blot and immuno- fluorescence assays. Figure 4.18: GAP43 reduces CB1R-mediated signaling in 7-DIV primary cultured neurons. A. Western blotting of GAP43 as a control of nucleofection. Note an increased expression of GAP43 compared to endogenous levels. B. DIV7 primary neurons nucleofected with GAP43WT were treated with vehicle or 100 nM WIN for 5 min. Left, Representative experiments showing phosphoROCK2 are depicted. Right, Quantification of optical density values of phosphoROCK2 relative to those of the loading control are shown. Data are the mean ± SEM of counts in 10 different experiments (*p<0.05 by Student‘s t-test to GAP43 vehicle; p<0,05 to control WIN) C. Left, Representative confocal images of phoshoROCK2. Right, Quantification of phosphoROCK2 immunoreactivity relative to vehicle in these same neuronal samples are shown. Data are the mean ± SEM of counts (n=3; *p<0.05 by one-way ANOVA with Dunnett’s multiple comparisons test). GAP43 does not affect CB1R internalization in HEK293T cells As GAP43 affects CB1R signalling and has been shown to induce AMPA receptor internalization (Han et al., 2013), we sought to evaluate the effect of GAP43-WT overexpression on agonist-evoked CB1R internalization. We used HEK293T cells stably expressing an N-terminally FLAG-tagged CB1R, which allowed labelling sur- face receptors and relate them to total receptors, as stained with an anti-CB1R-Ct antibody. Incubation with WIN (100 nM) for 40 min decreased cell-surface CB1R compared to 10-minute drug treatment (Figure 4.19 A). Internalized CB1R turned to be localized in early endosomes, as identified by early endosome antigen 1 (EEA1) at the same time points of treatment (Figure 4.19 B), with negligible trafficking to lysosomes, as identified by lysosomal-associated membrane protein 1 (LAMP1) 81 Figure 4.19: GAP43 does not affect CB1R internalization in HEK293T cells. A. HEK-CB1R cells express- ing GAP43-WT plasmid or empty vector were incubated with 100 nM WIN for 10 and 40 min. Left, Representative confocal images are shown. Right, Quantification of internalization is expressed as the ratio of surface CB1R to total CB1R, normalized to vehicle and expressed as a percentage of control condition. CRIP1a transfection was used as negative control (n=3-6; **p<0,01 by one-way ANOVA with Tukey’s multiple comparisons test.). B. Left, Representative confocal images showing colocalization of CB1R with EEA1 at 40 min of WIN treatment. Right, Quantification of colocalization at 10 and 40 min is expressed relative to control condition at 10 min. C. Quantifica- tion of colocalization with LAMP1 expressed relative to control condition at 10 min (n=3-6). **p<0,01 by one-way ANOVA with Tukey’s multiple comparisons test. 82 (Figure 4.19 C). GAP43 did not substantially modulate CB1R internalization and intracellular location after WIN treatment (Figure 4.19 A-C). CRIP1a-mediated oc- clusion of CB1R internalization was used as a positive control (Blume et al., 2016) (Figure 4.19 A). 4.2 Results of Aim 2 GAP43-CB1R interaction occurs in MC terminals of DG GAP43 is widely expressed in the peripheral nervous system and the CNS during the perinatal period. The levels of GAP43 decline in most neurons when mature synapses are formed, but continues to be expressed in selected brain regions, as the hippocampus, that retain high activity of synaptic plasticity (Box 3). To map the GAP43-CB1R interaction in the brain we first performed immunofluorescence co-localization assays in representative regions of the WT mouse brain. CB1R and GAP43 showed a pronounced immunoreactivity in the striatum, cortex and hippo- campus (Figure 4.20). Figure 4.20: CB1R and GAP43 are present in the mouse striatum, cortex and hippocampus. GAP43 and CB1R immunoreactivity in brain sections of striatum (STR), cortex (CX), and hippocampal CA1 area and dentate gyrus (DG) of WT mice. Representative images are shown with GAP43 in green and CB1R in red. Nuclei are colored in blue by DAPI staining. CC: corpus callosum; I-IV: I-IV cortical layers, Pyr: Pyramidal cell layer; Mol: Molecular Layer; Hil: Hilus. The DG called our attention as a high expression of both GAP43 (Benowitz et al., 1988) and glutamatergic-neuron CB1R (Monory et al., 2006) had been reported in a restricted area, namely the IML. We next used conditional knockout mice for CB1R in telencephalic GABAergic neurons (herein referred to as GABA-CB1R−/−) and glutamatergic neurons (herein referred to as Glu-CB1R−/−). We found abundant double-positive puncta for GAP43 and (glutamatergic-neuron) CB1R in GABA- CB1R−/− mice, while a very scant colocalization between GAP43 and (GABAergic- neuron) CB1R was evident in Glu-CB1R−/− animals (Figure 4.21 A). Colocalization 83 Figure 4.21: CB1R and GAP43 colocalize exclusively in glutamatergic MC axons of the DG. A. GAP43 and CB1R immunoreactivity in the DG of GABAergic-neuron (up) or glutamatergic-neuron (down) CB1R KO mice. Representative images are shown. Nuclei are colored in blue by DAPI staining. The dotted line depicts the high-magnification inset of the IML just below. Arrows depict colocalizing boutons, which are absent in Glu-CB1R/ mice. The quantification of the percentage of colocalization is graphed (n=3 mice per group, *p < 0,05 by unpaired Student’s t-test). B. Triple labelling of GAP43, CB1R and calretinin immunoreactivity in the DG of GABAergic- neuron CB1R KO mice. Representative images are shown. Nuclei are colored in blue by DAPI staining. Arrows depict colocalizing boutons. 84 analyses showed a 8.3 ± 1.1 % of colocalization area between GAP43 and CB1R in GABA-CB1R−/− mice and 4.2 ± 0.8 % (CB1R) in Glu-CB1R−/− animals. Calretinin is, among others, a cellular marker for MC (Scharfman, 2016). In order to confirm the identity of the glutamatergic terminals in the IML, an anti-calretinin antibody was used in triple colocalization assays in GABA-CB1R−/− mice. As shown in (Figure 4.21 B), calretinin (stained in grey) clearly labeled the IML and MC somas within the hilus. Inconveniently, calretinin is also expressed in newborn adult GCs, so somas located in the bottom of GC layer and their dendrites arising onto IML were labelled as well. Nevertheless, triple-positive punta for glutamatergic-neuron CB1R, GAP43 and calretinin were noticeably in high magnitude micrographs. Box 3: GAP43 EXPRESSION GAP43 protein (A, adapted from Benowitz et al, 1988) is highly ex- pressed in the adult brain in associative regions of the neocortex, particularly within layers 1 and 6, entorhinal cortex, retinal cells, olfactory bulb and adult- born olfactory sensory neurons, as well as in cerebellum -only in outer portions of the molecular layer and granule cells, but not Purkinje cells. Within the hip- pocampus, GCs in the DG show dense GAP43 labeling. The highest GAP43 immunoreactivity is found in the IML (B, C, adapted from Naffah-Mazzacorati et al., 1999, and Benowitz et al, 1988, respectively); GAP43 is evident in CA1 field but absent in the pyramidal cell layer itself. Pronounced labeling of GAP43 has also been found in a variety of adult subcortical structures such as the caudate-putamen, amigdala, olfactory tubercle, nucleus accumbens and medial preoptic area of the hypothalamus. There is also GAP43 labeling over- lapping with specific neurotransmitter systems such as those in the substantia nigra pars compacta (dopamine), the locus coeruleus (norepinephrine), and the dorsal raphe (serotonin). Within the spinal cord and brainstem, unmyelin- ated or moderately myelinated areas, such as the nucleus of the solitary tract, express high levels of GAP43. GAP43 staining is detected in small unmyelin- ated axons and small axon terminals. Within motor neurons of the brainstem and spinal cord, GAP43 is present at all vertebral levels, with higher concen- trations in cervical and thoracic regions. In contrast, GAP43 is present at low levels in primary sensory and motor regions of the neocortex and portions of dorsal thalamus (Benowitz et al., 1988, 1989; McNamara and Lenox, 1997; Holahan et al., 2007). The pronounced variations in GAP43 immunostaining among various areas of the human brain may reflect different potentials for functional and/or structural remodeling. GAP43 mRNA (D, Image credit: Allen Institute) is detected at pro- nounced levels in the hippocampus, particularly in hilar neurons, followed by CA3 region, the granular layer of the cerebellar cortex and inferior olivary nucleus (originating climbing fibers), mitral olfactory cells, locus coeruleus, raphe nuclei, certain thalamic nuclei, dopaminergic nigral and ventral teg- mental nuclei, and several nuclei of the hypothalamus and basal forebrain. Many of the neurons showing substantial levels of GAP43 mRNA have both long axonal paths and extensive arborization of terminals (Meberg and Rout- tenberg, 1991; Kruger et al., 1993). 85 Regarding cell types, it is highly expressed in excitatory neurons (Nemes et al., 2017), but there is also some expression in inhibitory neurons (Takano and Matsui, 2015). It is also expressed in the serotonergic and catecholamin- ergic (monoaminergic) neurons (Bendotti et al., 1991). In contrast, choliner- gic neurons generally express little or no GAP43 (Meberg and Routtenberg, 1991). In neuroglial cells, GAP43 shows plasma membrane localization in astrocytes (Vitkovic et al., 1988) as well as in oligodendrocytes (Deloulme et al., 1990), but it is absent from Schwann cells (Meiri et al., 1988). At a subcellular level, the highest GAP43 density appears in presynaptic ter- minals and it is largely absent from dendrites, somata and myelinated axons. Exceptionally, GAP43 was reported postsynaptically in subcellular fractions prepared from mouse brains and cultured neurons (Han et al., 2013), however, the GAP43-homologous PKC-substrate located preferentially in postsynapses is neurogranin (Ramakers et al., 1999). We subsequently conducted a PLA analysis in the IML of GABA-CB1R−/− and Glu-CB1R-/ mice to seek for CB1R-GAP43 complexes. Consistent with our immun- ostaining data, an overt PLA signal, shown as positive dots, was present in GABA- CB1R−/− and CB1Rfl/fl mice (23.9 ± 2.1 and 25.6 ± 4.4 dots/area for GABA- CB1R−/− and CB1Rfl/fl, respectively), but it notably decreased in Glu-CB1R−/− and full CB1R−/− animals (15.3 ± 0.6 and 10.3 ± 1.6 dots/area for Glu-CB1R−/− and full CB1R−/−, respectively) (Figure 4.22 A). To unequivocally ascribe CB1R-GAP43 complexes to glutamatergic terminals we made use of a Cre-mediated, lineage-specific CB1R re-expression/rescue strategy in a CB1R-null background (herein referred to as Stop-CB1R mice) (Ruehle et al., 2013; de Salas-Quiroga et al., 2015) (Figure 4.22 B). The rescue CB1R expression select- ively in forebrain glutamatergic neurons (herein referred to as Glu-CB1R-RS mice) was achieved by expressing Cre under the regulatory elements of the Nex1 gene. In parallel, we rescued CB1R expression selectively in dorsal telencephalic GABAergic neurons (herein referred to as GABA-CB1R-RS mice) by using a Dlx5/6-Cre mouse line. As a control, an EIIa-Cre-mediated, global CB1R expression-rescue in a CB1R- null background was conducted (herein referred to as CB1R-RS mice). PLA signal for CB1R-GAP43 complexes was notably restored in Glu-CB1R-RS mice (30.8 ± 1.2 and 32.5 ± 1.5 dots/area for CB1R-RS and Glu-CB1R-RS mice, respectively). No significant rescue of the complex expression was observed in GABA-CB1R-RS anim- als (15.4 ± 1.8 for GABA-CB1R-RS mice and 15.2 ± 1.8 dots/area for Stop-CB1R mice). 86 Figure 4.22: CB1R and GAP43 interact in glutamatergic MC axons of the DG. PLA assays were performed in hippocampal sections from 3–4-month-old mice of different genotypes. GAP43-CB1R complexes are shown as red dots. Nuclei are colored in blue by DAPI staining. A. Representative images of DG sections from CB1R-floxed, GABA-CB1R−/−, Glu-CB1R−/− and full CB1R−/−. Arrows depict complexes, which are absent in WT and Glu- CB1R−/− mice. Quantification of PLA-positive dots per field is shown. B. Representative images of DG sections from Stop-CB1R, CB1R-RS, GABA-CB1R-RS and Glu-CB1R-RS mice. Data are the mean ± SEM (n=6-7 mice per group; **p < 0.01 from the corresponding CB1R-floxed group or the corresponding CB1R-RS group by one-way ANOVA with Tukey’s multiple comparisons test). 87 Taken together, these findings support that the GAP43-CB1R interaction occurs in glutamatergic terminals of the IML, presumably MCs. CB1R function is modulated by PKC at MC-GC synapses The glutamatergic terminals of the IML mostly correspond to MC axons impinging on proximal dendrites of GCs, the main neuronal population of the DG. At MC-GC synapses, CB1R mediates short-term but not long-term plasticity (Chiu and Castillo, 2008). Furthermore, GAP43 is involved in the control of synaptic plasticity (Box 4). Specifically, in MCs, GAP43 expression mediates LTP events (Namgung et al., 1997). Taken together, previous findings are compatible with the notion that GAP43 could affect the signalling and function of the pool of CB1R molecules specifically located on MC terminals of the DG. Thus, we aimed to determine whether CB1R- mediated regulation of transmission and plasticity at the MC-GC synapse is affected by the interaction with GAP43. Box 4: GAP43 AND HIPPOCAMPAL SYNAPTIC PLASTICITY GAP 43 located in presynaptic terminals is involved in short and long- term synaptic plasticity. It has been extensively studied in the hippocampus, where GAP43 phosphorylation and expression are dynamically regulated by synaptic activity. In CA1, GAP43 phosphorylation is selectively increased immediately after LTP induction by HFS in vitro and in vivo, and requires NMDA activation. The increase persists for 60 min if induced in hippocampal slices (Ramakers et al., 1999) and for several days if induced in vivo (Rout- tenberg et al., 1985; Gianotti et al., 1992). Moreover, the extent of GAP43 phosphorylation correlates with the level of LTP (Lovinger et al., 1986). In vivo tetanic stimulation of the mossy fiber pathway was associated with in- creased PKC-mediated GAP43 phosphorylation 1 and 5 but not 60 min after stimulation, indicative of a role of GAP43 phosphorylation in the induction but not the maintenance of LTP (Son et al., 1997). In the DG, LTP induction in PP showed augmented GAP43 phosphorylation for three days in dorsal hippocampus (Lovinger et al., 1985). Moreover, not only phosphorylation but GAP43 expression levels can be modified by LTP. Induction of LTP in PP showed augmented GAP43 mRNA in MCs of DG in a NMDA-dependent manner (Namgung et al., 1997) and in CA3 pyramidal cells (Meberg et al., 1995). Arachidonic acid (AA) has been recently found to act as a retrograde mes- senger potentiating excitatory transmission in a process called depolarization- induced potentiation of excitation (DPE). Interestingly, AA increases sensit- ivity of GAP43 phosphorylation and bump-up its maximal phosphorylation level. It has been shown that AA at LTP-inducing concentrations signific- antly increases translocation of PKC to the presynaptic membrane and GAP43 phosphorylation in hippocampal slices (Luo and Vallano, 1995). AA can in- duce a presynaptic LTP in the DG linked to an increased glutamate release in stimulated PP in vitro and in vivo. However, it is not known if GAP43 is specifically involved in this form of AA-mediated LTP in the DG (Willi- ams et al., 1989). The proposed mechanism of action of GAP43 in synaptic 88 transmission would be as follows: During Hebbian LTP induction, postsyn- aptic NMDAR activation would lead to the release of a retrograde messenger, suggested to be AA. Released AA would act presinaptically to activate VD- CCs, which ultimately activate PKC and increase GAP43 phosphorylation. In turn, active GAP43 influences vesicle release and cytoskeleton remodelling to increase LTP (Routtenberg et al., 2000). Consequently, GAP43 is implicated in the control of learning and memory. GAP43 phosphorylation increases after training on a memory task in the hippocampus (Cammarota et al., 1997). Heterozygous GAP43+/− mice have deficits in spatial learning in a contextual fear conditioning test, but also have multiple sensorimotor deficits and decreased sociability (Rekart et al., 2005; Zaccaria et al., 2010). Conversely, GAP43 overexpression has a boosting effect on memory-associated tasks. Dysregulation of GAP43 in the hippocam- pus has been associated with some neurodegenerative diseases as Alzheimer’s disease or other neurological alterations as epilepsy. In Alzheimer’s disease, reduced levels and activity of GAP43 are observed (de la Monte et al., 1995), together with aberrant sprouting (Rekart et al., 2004). In epileptic models of cortical dysplasia or temporal lobe epilepsy, GAP43 levels are increased, and it has been proposed as a promoting factor of epileptogenesis (Nemes et al., 2017), showing as well aberrant sprouting of mossy fibers (McNamara and Routtenberg, 1995). For this purpose, we conducted in vitro electrophyisiology experiments in hip- pocampal slices. Whole-cell patch clamp recordings of GCs were achieved upon selective stimulation of MC fibers in the IML (Figure 4.23 A) as previously de- scribed (Chiu and Castillo, 2008). Consistent with previous reports (Macek et al., 1996), MPP, but not MCF, are sensitive to group II metabotropic glutamate re- ceptors (mGluR-II) agonists. We corroborated that bath-perfusion of the agonist DCG-IV at 1 µM had no effect on excitatory postsynaptic currents (EPSCs) from MCs showing unaltered amplitudes (102 ± 9 % of baseline; Figure 4.23 B) (Macek et al., 1996). Typical CB1R-mediated DSE in MC-GC synapses (Chiu and Castillo, 2008) also occurred in our hands when using postsynaptic depolarizing steps of 3 or 5 seconds in order to trigger eCB release (73 ± 4 % of baseline) that was CB1R dependent since the CB1R antagonist SR141716 blocked DSE (Figure 4.23 C). This CB1R-mediated reduction in neurotransmission was accompanied by a significant in- crease in PPR (Figure 4.23 D), consistent with the typical CB1R-mediated inhibition of presynaptic release probability observed associated with eCBs-STD throughout the brain (Kano et al., 2009). In this experimental setting we used the PKC inhibitor BIM-I, which blocks the ATP-binding site of PKC, as a conceivable manipulation to inhibit GAP43 phosphorylation and thereby to de-inhibit CB1R. Firstly, a basal DSE was achieved and then BIM-I was applied in the recording chamber at 1 µM for 20 min. BIM-I facilitated DSE (control 75.3 ± 2.4 %; BIM-I 69.0 ± 1.2 % from baseline), which turned to last longer (at 2 min post-depolarization EPSC amplitudes were 93.4 ± 5.6 % of baseline in ACSF and 77.4 ± 5.3 % of baseline in BIM-I presence) (Figure 4.23 F). This observation suggests that PKC, maybe in part by phosphorylating GAP43, which is highly expressed in the IML, hampers DSE. Of note, BIM-I did not affect baseline EPSC amplitude (Figure 4.23 E), and the effect of BIM-I was not 89 Figure 4.23: eCB-mediated CB1R function at MC-GC synapses is enhanced by PKC pharmacological inhibition. A. Schematic diagram illustrating the DG recurrent circuit and the recording configuration. A stim- ulating electrode was placed in the IML while performing whole-cell patch-clamp recordings from GCs (∼100 µm apart from the stimulation pipette). Adapted from Hashimotodani et al., 2017. B. Representative traces (left) and time-course summary plot (right). MC EPSCs were recorded from GCs. After a stable (∼5-min) baseline, DCG-IV (1 µM) was bath-applied as indicated by the black bar (n=4 cells/3 mice). DCG-IV did not affect MC-GC synaptic transmission. C. Representative traces (left) and time-course summary plot (right) showing that 3-s postsynaptic depolarization of the GC (indicated by the arrow) induces DSE (n=11 cells/8 mice). After 10-min bath application of SR 141716A (4 µM), DSE was abolished (n=6 cells/4 mice). *p<0.05 from baseline by paired Student’s t-test. D. DSE was accompanied by a significant increase in PPR (n=11 cells/8 mice), consistent with a CB1R presynaptic action. *p<0.05 from baseline by paired Student’s t-test. E. Bath application of BIM-I (1 µM for 15 min) does not affect basal MC-GC synaptic transmission (n=3 cells/3 mice). Representative traces (up) and time-course summary plot (down) are shown. F. BIM-I (1 µM for 10 min) increases DSE as compared to baseline DSE (n=9 cells/9 mice). Representative traces (up) and time-course summary plot (down) are shown. G. BIM-I (5 µM) was added in the recording pipette to inhibit PKC selectively in the GC. BIM-I increases DSE (n=4 cells/2 mice), as compared to controls (n=6 cells/4 mice), suggesting a presynaptic effect of the drug. *p<0.05 from baseline by paired Student’s t-test or p<0.05 by unpaired Student’s t-test in BIM-I vs ACSF. 90 evident when the inhibitor at 5 µM was confined to the postsynaptic cell (i.e., when it was delivered through the recording pipette; Figure 4.23 G), therefore supporting a presynaptic site of PKC action. Active GAP43 inhibits CB1R at MC-GC synapses We next generated a tool to target presynaptic GAP43 more specifically than with BIM-I. Thus, we produced AAVs encoding mutant forms of GAP43 fused to the fluorescent reporter CFP under the control of a constitutive promoter (AAV1/2- CBA-GAP43-S41A-CFP and AAV1/2-CBA-GAP43-S41D-CFP). We injected these viruses ipsilaterally into the hilus (i.e., where MC somata are located) of 3 week- old WT mice. Each animal received one injection at coordinates (mm to bregma): antero-posterior +2.18, lateral ±1.5, dorso-ventral +2.2, for a distance lambda- bregma 4.2. The coordinates were adjusted to the lambda-bregma distance for each animal. We performed the recordings in the contralateral DG, thus taking advantage of the commissural nature of MC fibers. In this way we restrict transgene expression to presynatic commissural MC axons. The contralateral DG of the injected mice showed CFP-positive fibers specifically in the IML (Figure 4.24 A). Viral expression levels were consistent among the different conditions. At MC-GC synapses, CB1R mediates DSE and cannabinoid-evoked suppression of EPSP amplitude as shown by WIN application (Chiu and Castillo, 2008). The latter allows a direct and selective assessment of postsynaptic sensitivity by short- cutting the eCB-release step. Thus, we evaluated whether WIN-mediated depression of neurotransmitter release and DSE are affected by GAP43 expression at MC ter- minals. We conducted field EPSP (fEPSP) recordings in 6 week-old mice (i.e., 3 weeks after viral injection). A patch-type stimulation pipette was placed in the IML to activate commissural MC axons. The recording pipette was placed in the IML to record MC-GC fEPSPs. Given the possibility of crosstalk between the MC and MPP inputs, stimulation of specific inputs was confirmed distinguishing electrophysiolo- gical and pharmacological properties of these synapses: when two identical stimuli are delivered in a short period of time (here, 100 ms), two types of paired responses can occur. In synapses with low probability of release, during the first pulse, the action potential arriving to the presynaptic terminal triggers a Ca2+ influx that adds to the Ca2+ influx triggered by the entrance of the second action potential 100 ms apart. This increased Ca2+ concentration in the terminal induces a higher neur- otransmitter release and a higher amplitude response to the second stimuli. This phenomenon is named paired-pulse facilitation (PPF) (Debanne et al., 1996). MCF input exhibits typically PPF. An opposite phenomenon can happen in synapses with high release probability where the response amplitude to the first stimuli is higher than the second one, named paired-pulse depression (PPD). MPP exhibit PPD. We monitored fEPSPs as evoked by MC fibers before and after bath application of WIN (5 µM). Baseline and post-induction synaptic responses were monitored at 0.05 Hz. The typical waveform in an extracellular recording consists of a fiber volley, which is an indication of the presynaptic action potential arriving at the recording site and should be constant along the experiment, and the fEPSP itself. We found that, 25 min after bath application of WIN, the fEPSP amplitude decreased in mice injected with AAV1/2-CBA-GAP43-S41A-CFP to 77.9 ± 2.4 % from baseline (Figure 4.24 B). In contrast, when AAV1/2-CBA-GAP43-S41D-CFP was expressed, 91 Figure 4.24: Active GAP43 inhibits CB1R function at MC-GC synapses. A. Schematic diagram illustrating ipsilateral injection of AAV1/2-CBA-GAP43-S41A-CFP or S41D-CFP in the hilus of 3 week-old WT mice. Elec- trophysiological recordings were performed in contralateral DG. Infrared differential interference contrast (IR/DIC, left) and fluorescence images (right) showing the expression in the commissural MC axon terminals imaged in the contralateral DG of mice injected with AAV1/2-CBA-GAP43-S41A-CFP (up) or AAV1/2-CBA-GAP43-S41D-CFP (down). Note the presence of CFP-positive fibers in the IML of the dorsal blade and the absence of CFP expression in the GC layer. B. Up, Schematic diagram illustrating MC-GC fEPSP recording configuration. fEPSPs were performed from mice injected with AAV1/2-CBA-GAP43-S41A-CFP (GAP43-S41A; grey dots; n= 6 slices/ 3 mice) or AAV1/2-CBA-GAP43S41D-CFP (GAP43-S41D; white dots; n=5 slices/ 4 mice). WIN (5 µM, 25 min) was bath- applied, followed by a wash with AM251 (4 µM). (Top right) Time-course summary plot and (down right) Fiber Volley (FV) amplitude are shown. (Left) Representative fEPSP traces, before and after WIN application, are shown. In the presence of GAP43-S41D, WIN reduction of fEPSPs amplitude is impeded compared to GAP43-S41A. Data are presented as mean ± SEM. *p<0,05 by paired Student’s t-test from baseline; p<0,05 by unpaired Student’s t-test between conditions. C. Whole-cell patch-clamp recordings were performed on GCs from mice injected with GAP43-S41A (grey dots; n=9 cells/ 5 mice) or GAP43-S41D (white dots; n=10 cells/ 4 mice). Representative traces (up) and Time-course summary plot (down) are shown. *p<0.05 from baseline by paired Student’s t-test or p<0.05 by unpaired Student’s t-test between conditions. D. GAP43-S41D-mediated blockade of DSE was accompanied by a significant reduction in PPR (GAP43-S41A: 0.99 ± 0.15, n=9; GAP43-S41D: 0.85 ± 0.09%, n = 13). Representative traces (left) and quantification bar graph (right) are shown. *p<0.05 from baseline by paired Student’s t-test.). 92 fEPSPs decreased only to 88.5 ± 2.5 % from baseline. Thus, WIN-mediated depres- sion was significatively blunted by the S41D mutant compared to the S41A mutant. Of note, there was no significant difference in the recovery of basal fEPSPs during WIN washout when using the CB1R antagonist AM251, indicating a lack of LTD on this synapse upon GAP43 mutants expression. DSE experiments at excitatory synapses were subsequently conducted in 6 week- old mice (Figure 4.24 C). Whole-cell patch-clamp recordings were performed from GCs in the GC layer. A patch-type stimulation pipette was placed in the IML to activate commissural MC axons. Consistent with the field recording data, a reduction in DSE magnitude was evident in the animals injected with AAV1/2- CBA-GAP43-S41D-CFP (90.4 ± 2.7 % from baseline) compared to those injected with AAV1/2-CBA-GAP43-S41A-CFP (77.8 ± 4.9 % from baseline). This enhance- ment was accompanied by a decrease in paired-pulse ratio (PPR) (Figure 4.24 D), suggesting an increase in glutamate release probability. Taken together, these last observations strongly suggest that GAP43 negatively modulates CB1R-mediated regulation of MC-GC synapse. Generation of conditional GAP43 KO mice Many growth-permissive events in the CNS occur in concert with elevations in GAP43 levels. GAP43 is involved in input-dependent neuroplastic processes, such as LTP and memory formation, as mentioned previously, but it has also a key role in axonal regeneration, sprouting and structural plasticity during development, as well as following axonal injury (Box 5). As a next step in our study, and since we have delimitated the CB1R-GAP43 interaction to a particular glutamatergic synapse, we aimed at delineating the function of GAP43 in conditional knockout mouse models with specific deletion of GAP43 in excitatory or inhibitory neurons. To date, this approach to study GAP43 functionality has never been conducted before. Only full knockout models, with a very limited survival rate of the animals, were attempted (Strittmatter et al., 1995). For this purpose, we purchased GAP43 reporter knockout-first allele heterozy- gote embryos generated on a C57BL/6N background (Figure 4.25 A). This model was designed with a LacZ cassette and a Neo cassette with a STOP codon flanked by Frt sites followed by the Gap43 exon 2 flanked by LoxP sites. The location of primers used for detection of LacZ, Frt insertion and LoxP -flanked exon are shown (Figure 4.25 A and B, left panel). The LacZ cassette is only expressed in tissues where the gene of interest has been knocked-out. However, for many genes, skipping over the LacZ cassette restores gene expression to some extent, to constitute in fact a knockdown model. In order to generate conditionally-ready GAP43-floxed mice, it was necessary to first assess a F0 heterozygous-mouse generation by crossing the Gap43 knock- down heterozygous reporters with mice carrying the Actb-FLP transgene that en- codes a constitutively expressed Flp recombinase to remove LacZ and Neo cassettes. The resultant heterozygotes were positive for Flp recombinase and the Frt-flanked sequence deletion as tested by tail-snip PCR with the indicated plasmids. They were backcrossed to achieve homozygotes (Figure 4.25 B, mid panel). Homozygotes (herein referred as GAP43fl/fl mice, with exon 2 floxed) were bred with mice carrying Cre recombinase under the control of Nex1 or Dlx5/6 genes to generate telencephalic 93 Figure 4.25: Conditional GAP43 KO mouse generation. A. Gene targeting strategy for the production of conditional GAP43 KO mice. Knockout-first allele (A) required a first cross with Flp-mice (B) to achieve a conditional-ready allele. Subsequent crossing with Cre-expressing mice (C) yielded the conditional KO mice in the specific locations of Cre recombinase expression. B. PCR genotyping of (A) heterozygous or WT allele before Flp-mouse crossing (the heterozygous allele would harbor LacZ, Frt and LoxP sequences), (B) heterozygous or homozygous alleles after Flp-mouse crossing, and (C) GAP43fl/fl and Dlx5/6 or Nex1-Cre alleles. C. Distribution pattern of X-gal blue staining in MC somas matches with a lack of GAP43 staining in IML in Flp-heterozygous mice. D. Reduced levels of GAP43 protein in hippocampal lysates of conditional GABA-GAP43−/− mice (lane 2) or Glu-GAP43−/− mice (lane 3) compared to control GAP43fl/fl mice (lane 1), as revealed by Western blot analyses. A representative blot is shown. Quantification of normalized optical densities (OD) values of GAP43 relative to those of the loading control is shown (n=2-4; *p <0,05, **p < 0,01 by unpaired Student’s t-test from GAP43fl/fl mice). E. Reduced GAP43 immunoreactivity in the IML of Glu-GAP43−/− mice compared to GAP43fl/fl or GABA-GAP43−/− mice. Representative images are shown. Nuclei are colored in blue by DAPI staining. 94 glutamatergic-neuron (herein referred to as Glu-GAP43−/− mice) and GABAergic- neuron (herein referred to as GABA-GAP43−/− mice) conditional knockouts, re- spectively, by using Cre-LoxP recombination technology. The mouse-tail DNA was genotyped using the primers for LoxP sites and Cre recombinase (Figure 4.25 B, right panel). Box 5: GAP43 AND AXONAL GROWTH In the developing brain, GAP43 has a prominent role in neurite out- growth and axonal pathfinding. GAP43 orchestrates the amplification of pathfinding signals from numerous adhesion molecules or growth factors in the growth cone. During the early phases of development, GAP43 increases from birth to early postnatal days depending on the temporal developmental pattern of each brain region, and then declines during the ensuing sculpt- ing and maturation processes. Newly formed synapses show reduced levels of GAP43, which may constitute a signal to prevent the advancing of axons (Patterson and Skene, 1999). Interference of GAP43 function inhibits neurite outgrowth (Benowitz and Routtenberg, 1997). GAP43 was reported for the first time by the Fernández-Ruiz lab to colocalize with CB1R on axonal cones in extending myelin fibres in the developing rat brain (Gómez et al., 2008). Is relevance during development is proven by the observation that complete de- letion of the GAP43 gene in mice leads to the animals’ death very early in the postnatal period (Strittmatter et al., 1995). These mice present highly disrup- ted connectivity, topography and complex spatial patterning in the neocortex (Maier et al., 1999). Consequently, GAP43 has been related with neurodevel- opmental psychiatric disorders as autism (Zaccaria et al., 2010), bipolar dis- orders (Zarate and Manji, 2009) and schizophrenia (Weickert, 2001). In the adult brain, GAP43 triggers axonal regeneration, thus mimicking an early developmental phase. In regenerating neurons, GAP43 sustains regrowth after injury, promoting functional recovery. Upregulation of GAP43 is observed in the newly-formed sprouts following axonal damage in different models of axo- tomy such as crush of sciatic nerve, deafferentization of olfactory epithelium by bulbectomy, unilateral coclear ablation, and vibrisectomy (Mascaro et al., 2013). Regarding central axons, GAP43 is increased in PP fibers undergoing contralateral sprouting after physical lesions of EC (Lin et al., 1992). Like- wise, excitotoxic damage after kainite injection increases GAP43-dependent sprouting in mossy fibers (into the IML) (Cantallops and Routtenberg, 1996). The re-expression of GAP43 in neurons that ode not naturally express GAP43 gene can recapitulate a development axonal growth process, as found in GCs that promote spontaneous massive sprouting of mossy fibers (Aigner et al., 1995) or in Purkinje cell axons in the cerebellum (Zhang et al., 2005). Behavioural characterization of conditional GAP43 KO mice In order to assess any gross neurodevelopmental phenotype in this new mouse model we observed that Glu-GAP43−/− and GABA-GAP43−/− mice are viable did not exhibit any noticeable dysmorphology or reduction in fertility or survival, and they reach adulthood with normal postnatal growth. Normal gross morphology of the 95 brain of Glu-GAP43−/− and GABA-GAP43−/− mice at 8 weeks of age compared to their littermates is shown by staining with the nuclear marker DAPI (Figure 4.26 A). Body temperature (Figure 4.26 B) and weight (Figure 4.26 C) was similar among all genotypes and genders. Both conditional knockout genotypes showed similar latency to pain symptoms (Figure 4.26 D). There were no differences in anxiety-like behaviors between geno- types as measured by the number of entries in the open arms in an elevated plus maze (Figure 4.26 F). Normal stride patterns were also found as revealed by monit- oring the walking of mice with painted feet (blue, fore; red, hind) on paper (Figure 4.26 E). Locomotion and motor coordination were assessed in an open-field plat- form and in a RotaRod device, respectively. We found that Glu-GAP43−/− had a higher global activity and less resting time, and ambulated longer distances than their WT littermates (Figure 4.26 G-I). Consequently, they remain longer times in the rotating bar of the RotaRod (Figure 4.26 J). These differences were not found between GABA-GAP43−/− mice and their WT littermates. No differences between males and females among genotypes were noticed in all the tests conducted. Over- all, these data suggest that excitatory or inhibitory neuron-restricted deletion of GAP43 does not affect neurodevelopmental phenotypes despite the striking hyper- locomotory phenotype found in Glu-GAP43−/− mice. GAP43 overexpression has a boosting effect on memory-associated tasks. For example, mice with loss of GAP43 function show deficits in spatial learning and memory (Rekart et al., 2005; Zaccaria et al., 2010; Holahan and Routtenberg, 2008). It is known that the hippocampus plays a key role in short-term memory and helps to construct a representation of an experience or a memory (Josselyn et al., 2015). As we have narrowed the functional consequences of the CB1R-GAP43 complex to a specific synapse in the hippocampus, we conducted behavioral assays in the con- ditional GAP43 knockout mice to evaluate deficits that could be ascribed to this discrete brain area. MCs have been reported to control specifically spatial memory and spontaneous convulsive seizures (Bui et al., 2018). Regarding spatial memory, MCs encode multiple place fields and undergo robust remapping in response to contextual manipulation (Senzai and Buzsáki, 2017; Danielson et al., 2017), and spontaneous alternation in 3-branched mazes have been previously used to assess dysfunction of the DG (Senzai and Buzsáki, 2017). We used a Y-maze disposition, a test that is based on the willing of rodents to explore new environments choosing the arm not visited before, reflecting short-term spatial memory of the first choice and responsiveness to novelty (Figure 4.27 A). The total number of arm entries and the percentage of triads of arm entries in which the mouse sequentially visited each possible arm without repeating, termed % of alternance, were recorded. Noticeably, Glu-GAP43−/− mice were impaired in this task by showing a % of alternance of around 50%, which is considered a “chance” level, compared to their WT littermates (Figure 4.27 B). Differences in the total number of arm entries were evident, suggest- ing the aforementioned effects on motor behavior. Meanwhile, GABA-GAP43−/− showed a normal alternance, with values significantly above chance level (> 50%) (Figure 4.27 B). Recently, neurons in the hippocampus that express the D2R, which closely re- semble hilar MCs (Gangarossa et al., 2012; Senzai and Buzsáki, 2017; Botterill et al., 2019), have been shown to be specifically activated by food and to encode a spatial memory linking food to a specific location (Azevedo et al., 2019). Given that mod- 96 Figure 4.26: Conditional GAP43 KO mouse general characterization. Glu-GAP43−/− and GABA- GAP43−/− mice vs. their WT littermates at 8 week-old were evaluated. A. DAPI staining of coronal slices show normal gross morphology of the brain for both genotypes. Representative stitched images are depicted. B-E. No differences were found in mean temperature (B), body weight (C), nociceptive response (latency to show pain symptoms, s) (D) and walking patterns (mean stride length, cm) (E.). F Anxious behavior (normalized entries in open arms) in the EPM was similar in all genotypes. G-I. Ambulation (total distance travelled, cm) in the open-field test was increased in Glu-GAP43−/−(G), similar to global activity (H), showing a decreased resting time (I). J. RotaRod performance (time to fall, s) is increased in Glu-GAP43−/−. (n = 6-14 mice per group; *p < 0.05 by unpaired Student’s t-test from their respective littermates). 97 Figure 4.27: Conditional GAP43 KO mouse learning and memory performance. A. Left, Diagrams for the correct (up) and incorrect (down) alternation in the Y-maze test. B. Up, Summary of spontaneous alternation. The percent of alternation in the Y-maze was significantly above chance level (50%) in (left) WT littermates but not Glu-GAP43−/− mice; (right) both WT littermates and GABA-GAP43−/− were above chance level (n=6-14 per group, *p < 0,05 by one-sample t-test). Down, Summary of the total arm entries. The number of arm entries was increased in Glu-GAP43−/− mice compared to their littermates (p < 0,05 by unpaired Student’s t-test). C. Left, Schematic representation of the food location test. Mice were habituated to an empty arena with two empty cups on two different quadrants and fasted overnight (Context). On the next day, mice were exposed to the same context with one food cup containing mouse chow pellets (Training). An additional day (Test) exposing the mice to two empty cups again enabled to test memory acquisition and retrieval for food location. Right, Preference index after 5-minute exploration time of empty, food-containing cups (Food), and cups that previously contained food (No-food) for WT littermates (white bars) or Glu-GAP43−/− mice (grey bars) (n=8-12 mice per group; * p < 0,05 by paired Student’s t test). ulation of food-place associations is a behavioral outcome that has been specifically related to MCs in the DG, we aimed to test this paradigm in our mice models. In this protocol, fasted animals are placed in a chamber with food present in one of the quadrants. When animals receiving food during this interval are later placed in the same chamber but without food, they spend significantly more time in the quad- rant where the food was previously located, suggesting that the mice had learned to associate a specific location with food and that a memory of this association had been established (Figure 4.27 C, left panel). Discrimination during the training 98 session was not affected in Glu-GAP43−/− mice comparing to their WT littermates. However, Glu-GAP43−/− animals showed decreased encoding of a spatial memory for food during the test session (Figure 4.27 C, right panel). Pharmacological activation of CB1R in Glu-GAP43−/− mice reduces epileptic seizure susceptibility To address in further detail the functional consequences of the CB1R-GAP43 inter- action specifically at the MC-GC synapse, we used a combined genetic and pharma- cological approach. Specifically, we analyzed the effect of THC in Glu-GAP43−/− mice (in which GAP43 would be absent from excitatory cells, including MCs) after intraperitoneal injection of KA to induce epileptic seizures. CB1R, GAP43 and MCs have been previously related to the epileptic phenotype in the hippocampus. By targeting the ipsi- and contralateral DG along the dorsoventral axis of the hip- pocampus, MCs form an extensive recurrent excitatory circuit (GC-MC-GC) whose dysregulation can promote epilepsy (Botterill et al., 2019). KA-induced models have been used previously to test MC contribution to convulsions (Bui et al., 2018). On the other hand, selective deletion of CB1R in glutamatergic cells of the forebrain, most of it being receptors located on MC terminals, worsens KA-induced seizures (Monory et al., 2006). Regarding GAP43, it has been proposed as a promoting factor of epileptogenesis (Nemes et al., 2017), showing as well aberrant expression upon sprouting of mossy fibers in epileptic models of cortical dysplasia or TLE (McNamara and Routtenberg, 1995). Evaluation of the time course of KA-induced seizure behavior by scoring Ra- cine stage (averages of 5-minute intervals) throughout the experiment revealed an acute progression towards higher Racine scale stages during the first 30-45 min after KA administration (30 mg/kg) in both genotypes (Figure 4.28 A). After this ini- tial period, Glu-GAP43−/−mice progressively ensued a trend towards more severe seizure behaviour stages, which was significantly reduced by i.p. injection of 10 mg/kg THC at 15 min prior to KA (Figure 4.28 A). Meanwhile, WT littermates were on average stabilized and subsequently showed a reversion pattern over the 120-minute experimental period. In this group of mice, THC treatment did not sig- nificantly decrease seizure behaviour (Figure 4.28 A). We subsequently integrated the individual scores per mouse for the total duration of the experiment to better account for seizure severity, as described previously (Armas-Capote et al., 2020). As shown in Figure 4.28 B, seizure severity was significantly reduced in Glu-GAP43−/− mice treated with THC, from 135% to 65% severity, while in WT littermates THC caused a non-significant reduction of 20% in seizure severity. Additionally, WT mice showed a latency of 44 min to reach tonic–clonic seizures (stage 6; Figure 4.28 C) upon vehicle or 47 min upon THC treatment. In contrast, Glu-GAP43−/− mice showed a reduced latency (29 min) when treated with vehicle, which tended to be reverted (to 42 min) with THC treatment. This latter trend will indeed be the focus of further analyses with a larger cohort of animals. Evaluation of seizure-induced mortality showed survival rates of 83 % in Glu-GAP43−/− mice (4/6 surviving mice treated with vehicle and 6/6 surviving mice treated with THC). In contrast, WT mice did not display lethality events (6/6 surviving mice treated with vehicle and 6/6 surviving mice treated with THC). Altogether, these data suggest that Glu-GAP43−/− mice display a worse epi- 99 Figure 4.28: THC protects from seizure severity and progression in Glu-GAP43−/− vs WT mice after systemic KA administration. A. Raw Racine stage score (mean ± SEM) in 5-min intervals over the monitored period of time for (left) WT littermates or (right) Glu-GAP43−/− mice treated i.p with vehicle (black circles) or THC at 10 mg/kg (white circles); n=6 mice per group, *p < 0.05 by unpaired Student t-test. B. Integrated seizure severity corresponding to WT littermates or Glu-GAP43−/− mice treated with vehicle or THC and expressed as normalized percentage to WT vehicle (mean ± SEM; n=6 mice per group; * p < 0,05 by unpaired Student’s t test). C. Latency to SE, expressed as time in min to reach score 6 in the Racine scale, for WT littermates or Glu-GAP43−/− mice treated with vehicle or THC (mean ± SEM; n=3-4 mice per group reaching SE). leptic pattern than WT mice, and that THC might revert this phenotype in Glu- GAP43−/− mice. However, these pilot experiments comprise a still reduced number of animals, and should therefore be taken with caution. Active research is cur- rently being conducted in our lab to assess the precise functional outcomes of the CB1R-GAP43 interaction in the mouse brain in vivo. 100 Discussion CB1R-mediated signaling in the brain controls neuronal activity and a wide spec- trum of brain functions. Several studies have aimed at discovering CB1R-interacting partners that modulate CB1R-evoked action. The findings included in Aim 1 of this Thesis unveil the GAP43 protein, classically known for decades for its functions in axonal plasticity, as a new CB1R interactor. We demonstrated the interaction and found that its strength depends on posttranslational modifications. Additionally, we showed, as previously reported on other CB1R-CTD interactors (such as CRIP1a or SGIP), that it modifies some CB1R-mediated signaling events in an inhibitory manner, while other pathways remain unchanged. This may imply a signaling bias towards some selected pathways. On the other hand, the evidence shown in Aim 2 of this Thesis goes a step beyond and links this interaction to a physiological output, supporting its relevance for brain functioning. To the best of our knowledge, this is the first study that confers this specificity to a CB1R-interactor complex at the synapse level. Validation of GAP43 as a new CB1R interator The CB1R interactome has been characterized in diverse studies along the last years. We, and others previously, have used MS proteomic-based approaches to define the CB1R interactome. MS proteomic methodologies have enabled large-scale studies of protein interactors. The brain presents a particular challenge for proteo- mics as neurons are believed to contain more different proteins than any other type of cell (Deisseroth et al., 2003). Moreover, the vast diversity of cell types and pop- ulations in the brain introduces an additional complexity to this kind of studies. Using proteomics on mouse cortex in vivo, Njoo et al. (2015) pulled down a list of proteins that interacted with CB1R, highlighting multiple members of the WAVE1 complex and the Rho GTPase Rac1 family. A year later, neuron-population dif- ferences were taken into account to compare and analyze the composition of CB1R protein complexes in glutamatergic neurons and GABAergic INs of the mouse hip- pocampus (Mattheus et al., 2016). In our study, we were not able to detect any of the previously reported cytoplasmic proteins that interact with CB1R-CTD, such as CRIP1a, GASP1, FAN or WAVE1 complex, maybe because they were minor com- ponents compared to other cellular proteins in our starting brain sample. Both of the aforementioned studies provided higher specificity than our study regarding brain region and cell population. Nevertheless, obtaining enough quantities of material for these types of experiments is difficult, and MS–based protein identification is ex- tremely sensitive and highly dependent on the experimental conditions. Therefore, “interactosomes” likely represent a heterogeneous population of protein complexes present in the initial tissue used, which basically provides a launching pad that leads to reductionist-based methods focused on a target protein to validate the in- teraction. Of note, a previous high-throughput screening proteomics study, BioPlex 101 2.0 (Biophysical Interactions of ORFeome-derived complexes), which used robust affinity purification-MS methodology to elucidate protein interaction networks and complexes including more than 25% of protein-coding genes from the human gen- ome, predicted GAP43 to interact with sphingosine-1-phosphate receptor 2 (Huttlin et al., 2017), which is highly homologous to CB1R (Toman and Spiegel, 2002). Compared to other potential CB1R interactors studies to date (Hájková et al., 2016; Sánchez et al., 2001; Mart́ın et al., 2010; Njoo et al., 2015), and mainly to those extensive ones on CRIP1a (Niehaus et al., 2007), here we verified and confirmed the interaction with orthologously-expressed protein as well as with the endogenous pro- tein by several molecular techniques. We added experiments to prove the interaction of the two purified proteins in solution and techniques to detect the interaction on brain tissue in situ. The interaction described is dependent on the phosphorylation status of GAP43. When occurring in vivo, this interaction would conceivably not stand as a constant interaction but would only occur under specific biological trig- gers. Thus, the detection of the interaction would likely reflect a transient functional state. The precise definition of the region of interaction for both proteins is largely missing in this study. A truncated form of CB1R lacking the last 14 amino acids of the receptor (CB1R 1-458) show a drastically reduced interaction with GAP43 as detected by BRET. From this observation one could infer that GAP43 interacts, at least in part, with the distal region of CB1R-CTD. On the other hand, GAP43 inhibited the coupling of CB1R to Gαi2, Gαi3 and Gαq/11 in HEK293T cells. Within CB1R-CTD, Hx9, which encompasses residues A440-M461, has been suggested to bind Gαq/11 (Fletcher-Jones et al., 2019). Therefore, it is conceivable that GAP43 binds to distal CTD and biases the coupling to Gαq/11 according to our data. Re- garding Gαi/o, R400-E416 region including H8 is involved in Gαi3 activation. Recent studies have shown that the region comprising amino acids 401–417, including H8 in the CTD, can directly couple to Gαo and Gαi3 but not to Gαi1 or Gαi2 pro- teins (Mukhopadhyay et al., 2000). Additional studies showed that residues located carboxyl-terminally to H8 (i.e., distal to G417) regulate CB1R constitutive (agonist- independent) activity, and Gαi3 coupling was pointed as responsible for CB1R tonic activity of the receptor (Anavi-Goffer et al., 2007; Nie and Lewis, 2001). On the contrary, the association of CB1R with Gαi1/2 and Gαs occurs in IL2 and IL3 (Ulfers et al., 2002; Howlett et al., 2010; Kumar et al., 2019), although a role of C415 in func- tional coupling CB1R to Gαi2 protein has also been described (Oddi et al., 2018). Therefore, it is plausible that GAP43, by displacing Gαi3 (and maybe also Gαi2) binding site on central regions of the CTD, could affect signaling and/or constitutive activity of the receptor. In our assays to analyze Gα protein-subtype binding, Gα0 is surprisingly not coupled to CB1R in control conditions. The absence of endogenous Gαo in HEK293T cells was shown in some reports (Anavi-Goffer et al., 2007) but not in others (Burford et al., 1998). It could be a plausible explanation for that find- ing. However, it is important to have in mind that an agonist (WIN)-mediated bias can occur. In HEK293T cells stably expressing CB1R, WIN was reported to stabil- ize CB1R in a conformation that enables Gαq/11 signaling (Lauckner et al., 2005). GAP43 displays widespread protein-protein interactions, including those with Gαo. The first 74 amino acids of GAP43 are key for the interaction with Gαo, which en- hances Gαo sensitivity by increasing the rate of GDP/GTP interchange. Its action in stimulating GDP release is quite similar to that of GPCRs, but, strikingly, pertussis 102 toxin does not alter the activation of Gαo protein by GAP43 (Strittmatter et al., 1990, 1991; Strittmatter, 1992; Nakamura et al., 2002; Igarashi et al., 1995). Given that GAP43 can bind Gαo, one may envisage the formation of a bigger complex, including CB1R, GAP43 and Gαo that reduces the availability of Gαo on the re- ceptor, thus reducing CB1R-mediated signaling and short-term plasticity processes. Future research would be necessary on this issue. Regarding GAP43 structure, except for the CaM-binding domain, which forms an α-helix to bind CaM, GAP43 is primarily an unfolded protein following its syn- thesis in the cytoplasm. It is mainly considered as an intrinsically disordered protein (IDP), with no defined structural patterns, which makes difficult to predict its in- teracting regions (Flamm et al., 2016). In a GAP43-rabaptin-5 interaction study, co-IP of both proteins could occur when CaM was released and GAP43 was phos- phorylated (Neve et al., 1998), as for CB1R in the present study. Specifically, GAP43 binds to rabaptin-5 within its C -terminal 311 amino acids, the region that also con- tains the site of interaction of rabaptin-5 with Rab5, thus the two interaction sites are mutually exclusive (Neve et al., 1998). We generated two constructs of GAP43 that were tested to define the interaction region on GAP43, one encompassing amino acids 5 to 68 and the other encompassing amino acids 60 to 238. Neither of them could bind CB1R, as shown by BRET, in competition studies with the whole GAP43 protein. In sum, as GAP43 is a small IDP protein, a specific folding comprising the whole protein seems necessary to establish an interaction with CB1R. GAP43 biases CB1R-mediated signaling The impact of GAP43 on the coupling of CB1R to several Gα protein subtypes leads to the inhibition of receptor’s signaling cascades, specifically the ROCK/cofilin pathway. Previous evidence had suggested a functional link between CB1R signal- ing and RhoA activity. Indeed, RhoA/ROCK activation by CB1R was found in both neuronal (Berghuis et al., 2007; Dı́az-Alonso et al., 2016) and non-neural (Nithip- atikom et al., 2012; Mai et al., 2015) cells. CB1R modulates ROCK and the cyto- skeletal motor non-muscle myosin II (NMII) via Gα12/13 coupling in growth cones for neurite retraction and reshaping of dendritic morphology (Roland et al., 2014). However Gα12/13 was reported to bind preferably to IL3 regions of CB1R rather to the receptor’s CTD (Inoue et al., 2019). In fact, it does not necessarily imply the recruitment of Gα12/13, as RhoA activation by CB1R was found to be pertussis toxin-sensitive (and thus Gαi/o-dependent) in macrophages (Mai et al., 2015). A bias to Gαq/11 coupling, which can signal via RhoA/ROCK (Chikumi et al., 2002), is also a very feasible option. Here we show that GAP43 reduces both ROCK2 and cofilin phosphorylation, acting in the same direction as a Gαq/11 pharmacological inhibitor. Furthermore, in platelets, exposure to 2-AG was found to activate ROCK through PI3K and Akt (Signorello and Leoncini, 2014), a preferential pathway of the G-protein βγ subunits, an unexplored option too. Cofilin is an actin-binding protein which binds to both monomeric globular (G)- actin and filamentous (F)-actin, and plays an essential role in actin-filament dy- namics and reorganization. It promotes actin turnover by stimulating severance and depolymerization of actin filaments. It can contribute likewise to actin-filament assembly by increasing the actin-monomer concentration for polymerization and creating new barbed ends on actin filaments for polymerization. Cofilin becomes 103 inactive upon phosphorylation. LIM kinases specifically phosphorylate cofilin at S3, thereby inhibiting actin binding, severing and depolymerizing activities, and therefore the actin turnover effect. The main upstream kinase responsible for this phosphorylation is ROCK, identified mainly as effector of the RhoA small GTPase. It plays vital roles in facilitating actomyosin cytoskeleton contractility as well as actin polymerization. Different protein phosphatases dephosphorylate and react- ivate cofilin (Mizuno, 2013; Julian and Olson, 2014). GAP43 can be considered another actin-binding protein due to its involvement in lateral interaction and sta- bilization of long F-actin filaments (He et al., 1997; Moss et al., 1990), as well as interaction with other cytoskeleton-associated proteins to promote a polymeriza- tion state (Riederer and Routtenberg, 1999). WIN-mediated CB1R activation in the presence of GAP43 in HEK293T cells, as shown here, would abrogate cofilin inactivation and therefore F-actin stabilization, which goes in line with described functions of GAP43 in the cell. Suppression of CB1R-mediated ROCK phosphorylation upon GAP43 expression was recapitulated in primary neuron cultures from mouse hippocampus at DIV7. GAP43 was reported to colocalize for the first time with CB1R on axonal cones in vivo in extending myelin fibres in the developing rat brain at E21 (Gómez et al., 2008). Here we have shown colocalization of both proteins in growth cones of primary neurons. Phosphorylated GAP43 is a classical marker for expanding growth cones. Conversely, levels of phosphorylated GAP43 were low in actively retracting branches (Dent and Meiri, 1998). PKC phosphorylates GAP43 in axonal growth cones in re- sponse to several extracellular attractive signals –nerve growth factor (NGF), basic fibroblast growth factor (bFGF) and the neural cell adhesion molecules (NCAMs)- to direct the functional behavior of the growth cone. Interference of GAP43 func- tion inhibits neurite outgrowth (Benowitz and Routtenberg, 1997). On the other hand, CB1R activation induces repulsive growth cone turning and eventual neur- ite collapse, at least in vitro. A number of studies would specifically find CB1R- activation to have a repulsive effect on axonal growth cone pathfinding (Berghuis et al., 2007; Argaw et al., 2011; Zhou and Song, 2001). In neural progenitor cells and PC12 pheochromocytoma cells transfected with CB1R, NGF-induced neurite extension was inhibited (Rueda et al., 2002). Although some studies have contra- dicted these results, showing a positive effect of CB1R agonism on neurite outgrowth (Williams et al., 2003), the generally negative effects of CB1R activation could re- flect a dual, seemingly contradictory effect on neurites, where the negative effect on neurite branching would induce a positive effect on general neurite length (Oudin et al., 2011). CB1R-mediated repulsive growth cone turning involves neurite ex- tension, while growth cone collapse may be as well followed by neurite retraction, growth cone reinstatement and renewed extension along alternative paths (Maccar- rone et al., 2014). The molecular mechanism of CB1R-mediated cytoskeletal instability in growth cones is believed to involve RhoA/ROCK (Berghuis et al., 2007), Rap1/Src/STAT3 (He et al., 2005) and PI3K/Akt signaling (Bromberg et al., 2008). CB1R couples to neurite growth via Rho GTPases by modulating upstream Ca2+ signaling in axonal growth cones as triggered by guidance factors, thereby representing a dir- ect link to downstream actomyosin cytoskeleton (Jin, 2005). In human SH-SY5Y neuroblastoma cells stably expressing CB1R, the Gαi/o inhibitor pertussis toxin, the Gβγ inhibitor gallein and the β-arrestin inhibitor barbadin reduced CB1R-mediated 104 short projections, while the ROCK inhibitor Y27632 increased long projections (Ly- ons et al., 2020). Thus, several pathways triggered by CB1R may control neurite growth. As CB1R and GAP43 show opposed functions in regulating growth cone ad- vance, an inhibitory effect of GAP43 on CB1R upon interaction fits in this scenario. Studying the existence and the putative role of CB1R-GAP43 complexes in growth cone dynamics and integrating signalling to extrinsic cues in developing neurons constitutes an exciting perspective for future research. Overall, our results support that GAP43, when phosphorylated at S41, interacts directly with CB1R and reduces CB1R-mediated ROCK/cofilin signaling cascade, while other canonical routes (cAMP/PKA and ERK) remain unaffected (Figure 5.29). We suggest that GAP43 reduces the upstream coupling of Gαq/11 activation to ROCK, but we cannot exclude potential alterations on some other upstream triggers of this cascade such as Gαi or βγ subunits depending on the cell type or the signaling time. Figure 5.29: Proposed mechanism of action of GAP43 interaction on CB1R signaling. Left, CB1R- coupled signaling pathways in the absence of GAP43. The canonical Gαi-triggered inhibition of AC/PKA pathway is shown. Gαq protein, and also possibly Gαi or βγ subunits (dashed arrow), activate RhoA/ROCK/LIMK/cofilin phosphorylation to modulate cytoskeletal dynamics. Right, CB1R-GAP43 interacting scenario, in which Gαq- mediated ROCK cascade activation is decreased while AC/PKA inhibition remains unaltered. Active GAP43 has been related to the process of synaptic vesicle trafficking. A weak, phosphorylation-dependent interaction of GAP43 with SNAP25, syntaxin and VAMP components of the SNARE complex in differentiated neuronal terminals and rat brain tissue has been described. Thus, GAP43 can interact with the synaptic core complex in the active zone of the presynaptic membrane, suggesting that it is involved in the Ca2+-dependent docking and/or fusion of synaptic vesicles for neuro- transmitter release (Haruta et al., 1997). In addition, a Ca2+-dependent interaction occurs between GAP43 and the vesicle-associated protein Rabaptin-5, an effector of the small GTPase Rab5 that mediates membrane fusion in the endocytosis of incoming vesicles to expand early endosomes. Moreover, GAP43 mediates AMPA receptor endocytosis in postsynaptic membranes, and therefore LTD, as a substrate of caspase-3 action (Han et al., 2013). Thus, it was conceivable that GAP43 could modulate CB1R recycling and its availability at the plasma membrane upon activa- tion. We ruled out this notion in our studies conducted in HEK293T cells, in which 105 GAP43 did not substantially affect CB1R internalization upon prolonged challenge to WIN. Conversely, we also hypothesized that the CB1R-mediated modulation of neuro- transmitter release could be affected by the interaction of the receptor with GAP43. Several studies have shown that CB1R activation may induce a depletion of syn- aptic vesicles from the docked presynaptic pool in the active zone (Garćıa-Morales et al., 2015; Ramı́rez-Franco et al., 2014). CB1R activation induces a decrease in synaptic vesicle exocytosis but does not affect total pool size through ROCK and actomyosin modulation (McFadden et al., 2018). In this same study, high-definition imaging revealed an actomyosin-induced synaptic vesicle redistribution under CB1R activation, with fewer synaptic vesicles proximal to the active zone and clustered within the axonal bouton. Both effects were abolished by ROCK inhibition. To explore this possibility, we studied the synaptic plasticity events mediated by the receptor in Aim 2 of this Doctoral Thesis. Presynaptic plastic changes are mostly transient and primarily shape the probability and the amount of neurotransmitter release triggered by a given action potential. This involves synaptic vesicle- and act- ive zone-associated proteins that regulate vesicle availability, docking and priming (Gogolla et al., 2007). CB1R-mediated plasticity is controlled by GAP43 at the MC-GC synapse By mapping GAP43 and CB1R in the mouse brain we found a similar distribution of both proteins in the DG, specifically showing a high expression at MC terminals in the IML and a low expression in GCs. This is the first time that a highly re- stricted expression of CB1R together with an interacting protein is defined. In order to modulate GAP43 activity at the MC-GC synapse we used an inhibitor of PKC, the main enzyme responsible for GAP43 activation. Pharmacological modulations of PKC activity have been used in electrophysiology previously. Presynaptically, the DAG/PKC axis is a central event in short-term synaptic plasticity by indu- cing potentiation of synaptic transmission in hippocampal neurons. PKC activat- ors such as phorbol esters enhance the amplitude of evoked synaptic responses in many structures, including the hippocampus (Malenka et al., 1986), by increasing the probability of neurotransmitter release and the ready releasable pool (RRP) of neurotransmitter vesicles (Lonart and Südhof, 2000; Berglund et al., 2002; Gillis et al., 1996). These compounds also increase vesicle recovery following depletion, and the amplitude of EPSCs (Stevens and Sullivan, 1998). Additionally, the dila- tion of the fusion pore for transient neurotransmitter release is promoted by PKC (Fulop and Smith, 2006). In line with these data, our results show an additional reduction of neurotransmitter release when DSE of MC-GC synapse is accompanied by PKC inhibition in hippocampal slices. PKC activation targets a wide spectrum of synaptic proteins and is crucial for synaptic adaptations. However, the precise targets of PKC action are not clear yet. PKC-evoked phosphorylation could act at many levels in this scenario to modulate transmission. For example, PKC is an important activator of presynaptic Ca2+ channels (Kamatchi et al., 2000; Hamid et al., 1999; Stea et al., 1995). Synaptotagmin-1, Munc-18 and SNAP-25 are down- stream vesicle-associated substrates of PKC (Brown and Sihra, 2008), as well as myristoylated alanine-rich C-kinase substrate (MARCKS) and GAP43, so it is con- ceivable that these proteins could mediate the multiple mechanisms of PKC-induced 106 activation of exocytosis. Nevertheless, it has also been shown that bath applic- ation of phorbol 12,13-diacetate elicits a comparable increase in the frequency of miniature EPSCs in WT and GAP43 KO neuron cultures (Capogna et al., 1999). PKC might regulate the activity of CB1R itself as well. Thus phosphorylation of CB1R by PKC at S317 in the IL3 disrupts the activation of inward-rectifying K+ channels and the depression of Ca2+ channels (Garcia et al., 1998). Additionally, PKC selectively regulates behavioral sensitivity, binding and tolerance of CB1R to WIN (Wallace et al., 2009). Lastly, PKC could modulate eCB release from the postsynapse (Valentinova and Mameli, 2016). Postsynaptic PKCα may be critical to control dendritic spine structural plasticity, synaptic potentiation, and learning and memory by phosphorylation of postsynaptic receptor subunits (Dell’Acqua and Woolfrey, 2018). However, we can rule out a contribution of postsynaptic PKCε in our setting as loading specifically the postsynaptic cell (GC) with BIM-I through the patch pipette did not exert any effect on DSE. Here, the low specificity of PKC pharmacological modulation was overcome by the use of the phosphomimetic version of its substrate GAP43 by an AAV-based delivery strategy. Several reports have used these phosphomimetic models previ- ously. Mutant mice with a generalized expression of GAP43-S41D exhibited en- hanced levels of hebbian LTP as induced in the PP of the DG in vivo (Routtenberg et al., 2000) and at SC-CA1 synapses (Hulo et al., 2002) compared to WT mice. An enhanced PPF (a short-term form of synaptic plasticity, explained in the Res- ults section) and an increased synaptic response summation during HFS at SC-CA1 synapses compared to WT mice have been described as well (Hulo et al., 2002). In contrast, GAP43-WT over-expression did not affect the level of LTP, similarly to the GAP43-S41A mutant or a GAP43 form with a deletion of the entire effector domain. Additionally, there was no difference in the amount of LTP as induced by HFS in GAP43 KO when compared to WT animals. Thus, elimination of GAP43 did not prevent LTP, indicating that GAP43 is not necessary for LTP, although its phosphorylation may affect synaptic mechanisms to increase levels of LTP (Rout- tenberg et al., 2000; Hulo et al., 2002). On the contrary, LTD is not affected by overexpression of the phosphomimetic mutant (Hulo et al., 2002), and the induc- tion of LTD in WT animals led to a reduction in GAP43 phosphorylation status (Ramakers et al., 1999). Distinctively, instead of a generalized knock-in model, our approach of AAV delivery assures a specific transgene expression in presynaptic terminals corresponding to MC axons. In line with previous reports with phospho- mimetic models, our results show an inhibitory effect on WIN-mediated depression of neurotransmission and in DSE only upon GAP43-S41D expression, and no effect on fEPSP amplitude was detected when expressing S41A. We were able to detect such robust effects despite the coexistence of mutant forms of GAP43 with endo- genous levels of the protein, together with the contribution of ipsilateral MC axons that are being stimulated and recorded -though not infected- as part of the field response. The first time that CB1R modulation at the MC-GC synapse was shown (Chiu and Castillo, 2008), WIN was bath-applied for 15 min and washed out in the presence of a CB1R antagonist. The authors observed that this activation of CB1R was not sufficient to trigger eCB-LTD of MC inputs. Long term-plasticity involves structural changes by which the size or presence of pre-existing synapses is altered, and eCB- LTD has never been achieved at the MC-GC synapse with any other established 107 paradigm. Alternatively, when they perfused WIN continuously, a plateau was reached at 40 min, with a fEPSP reduction of 76%. For evaluating CB1R-mediated depression of transmission in our experimental setting, we bath-applied WIN for 25 min (an intermediate time between 15 and 40 min), pursuing this way to detect an LTD. Unfortunately, when WIN was washed out, fEPSPs returned to control values, thus excluding the existence of long-term changes triggered by GAP43 presence in MC terminals. However, the application of other LTD protocols in this context remains untested. Several studies have found that eCB-LTD is Gαi/o protein and cAMP/PKA- dependent (Chevaleyre et al., 2007). In our studies with HEK293T cells, GAP43 does not affect CB1R-modulated cAMP concentration and PKA phosphorylated substrates, but it boosts ROCK-mediated pathways. Assuming that these observa- tions could be extrapolated to the MC-GC neural circuit, one might not expect any modulation of PKA-mediated plasticity as triggered by GAP43-S41D. An earlier study reported that eCB-LTD but not eCB-STD depends on the contractile prop- erties of the presynaptic actomyosin cytoskeleton, specifically on NMII and ROCK signalling, at both inhibitory CA1 and excitatory corticostriatal synapses (McFad- den et al., 2018). An additional report showed that, in LPP fibers, WIN does not decrease baseline fEPSPs. Instead, activation of CB1R preferentially engages sig- naling mechanisms leading to a potentiated glutamatergic transmission due to a bias towards the FAK/RhoA/ROCK cascade. This also requires activation of a subclass of β1 integrins by retrograde co-release of unspecified factors that activ- ate them (Wang et al., 2018). Both studies clearly show the high specificity and unique identity of every synapse and every CB1R subpopulation in the brain. There- fore, our ongoing research is directed towards evaluating the ROCK dependence of CB1R short-term plasticity in MCs (which, to the best of our knowledge, has not been tested before), and, furthermore, of CB1R short-term plasticity in MCs upon expression of GAP43 mutants. Our findings show an effect induced by a GAP43 gain of function (GAP43-S41D expression), which implies a loss of CB1R activation or a lower level of CB1R basal activity. In fact, a high constitutive activity of CB1R in MC axons has been recently described (Jensen et al., 2021). Tonically active CB1Rs inhibit glutamate release at MC-GC synapses as revealed by and increase in MC-GC synaptic transmission upon bath-application of the inverse agonists AM251. This enhancement was accompan- ied by a decrease in PPR, suggesting an increase in glutamate release probability. This constitutively-active receptor subpopulation selectively inhibited MC inputs onto GCs, presumably via βγ subunits. Interestingly, our results show that PPR in mice expressing GAP43-S41D is decreased compared to GAP43-S41D in basal con- ditions, thus suggesting an increase in neurotransmitter release. So, on mechanistic grounds, one could speculate as well that a blockade of CB1R constitutive activity in MC terminals by expression of GAP43-S41D occurs by impeding the βγ limb of G protein-evoked signaling, with a connection with downstream ROCK. Nevertheless, further research is needed to shed light on this matter. Overall, we propose that CB1R in MCs can signal preferentially by a ROCK-mediated cascade, as it happens at the LPP-GC synapse in the DG, maybe by a shift in the preferred Gα protein sub- type (Figure 5.30). Thus, GAP43 would be phosphorylated in an activity-dependent manner by depolarization of MC afferents. This way, it presumably interacts with CB1R in the MC terminal. The interaction would displace Gαq/11-protein coupling 108 and this, probably in concert with a reduction of Gβγ signaling, leads to a reduction of CB1R-mediated suppression of glutamate release. Future experiments will there- fore have to uncover the specific mechanisms underlying the observed effects, as well as the possible involvement of the downstream actin cytoskeleton on the structural changes of the terminal. Figure 5.30: Proposed mechanism of action of CB1R-GAP43 complexes on eCB-STD at the MC- GC synapse. 1. Activation of the MC releases glutamate to the synaptic cleft. 2. Glutamate depolarizes the postsynaptic cell (GC) through glutamatergic ionotropic (NMDAR) and metabotropic (mGluRI) receptors. This activation triggers eCB production and eCB-STD. 3. PKC-phosphorylated GAP43 interacts with CB1R in the presynaptic membrane and decreases CB1R-evoked signaling through the RhoA/ROCK/actin pathway, thus reducing the suppression of glutamate release. 4. As a result, eCB-STD at the MC-GC synapse is compromised. Classically, LTP has not been observed at the MC-GC synapse. A potential explanation for the lack of LTP was provided by experiments showing that the de- polarization of GCs suppresses the MC input to GC, an effect that was mediated by the retrograde signalling of eCBs through CB1R. Thus, MC input can be suppressed by HFS that sufficiently depolarizes the GC (Scharfman, 2016). However, recently, a robust presynaptic LTP was shown for the first time at MC-GC synapses medi- ated by BDNF release and presynatic PKA (Hashimotodani et al., 2017), which can be modulated by CB1R tonic activity (Jensen et al., 2021). Also, remarkably to our study, MCs influence the LTP of the PP-GC synapse by an increase of GAP43 expression. Thus, in response to HFS of the PP, GAP43 mRNA was increased in MCs, which could support the persistence of LTP. It has been suggested that the MC and PP inputs to GCs are cooperative (Namgung et al., 1997). However, the possibility of a direct connection of PP coming from EC impinging on MC dendrites has also been suggested (Azevedo et al., 2019). Thus, PP activity patterns could be the physiological trigger to upregulate GAP43 levels in MCs when needed. Like- wise, Ca2+ would increase in MCs in an activity-dependent manner, thus enhancing PKC-mediated GAP43 activation and, consequently, protein-protein interactions. GAP43 impacts on motor and learning behavior The battery of behavioral tests that we have conducted in conditional GAP43 KO mice shows some remarkable phenotypes. Thus, basal motor activity of Glu- GAP43−/− mice is altered. We observed a hyperlocomotor phenotype that was 109 specific of this mutant line. Based on previous evidence, we propose that this hy- perlocomotor phenotype most likely does not rely on CB1R (as an opposite, i.e. hypolocomotor phenotype would be expected) and the hippocampus (as, for ex- ample, the modified Rotarod performance may align closer with a corticostriatal- dependent behavior). No matter how, we believe that this is a promising finding as there are no previous reports relating GAP43 to motor phenotype and therefore interesting mechanisms may be behind it. We will indeed study this issue in the future. Several studies have shown that the phosphorylation status of GAP43 and the amount of GAP43 protein regulate information storage in different learning experi- ences. Regarding phosphorylation, mutant mice overexpressing GAP43-WT possess superior learning and memory abilities as well as enhanced behavioral flexibility, whereas mice overexpressing GAP43-S41D reveal a persistence of enhanced behavi- oral responsiveness that leads to inflexibility and apparent difficulty in unlearning. In addition, mice overexpressing GAP43-S41A have retention deficits in some tests of spatial memory ability (Holahan and Routtenberg, 2008). GAP43 expression levels have been linked to multiple types of learning and memory behaviors in different paradigms upon using GAP43 heterozygous mice (Rekart et al., 2005; Zaccaria et al., 2010). Our data support that glutamatergic-neuron GAP43 is involved in Y-maze proper alternance and food location memory. It has been suggested that the DG is involved in the ability to define what is familiar and what is novel in the environ- ment and to develop fine spatial discrimination. However, proper alternation in Y maze performance has also an important cortical component (Dias and Aggleton, 2000) that impedes us to directly relate this behavior with the DG. Multiple studies reveal memory impairment produced by cannabinoids in paradigms involving spatial tasks, including the alternation in a T-shaped or a Y-shaped maze (Puighermanal et al., 2012; Basavarajappa and Subbanna, 2014). However, in most of the studies, cannabinoid agonists are administered systemically, and so a selective contribution of the hippocampus is not determined. Recently, a specific behavioural outcome for D2R-expressing hippocampal neur- ons, preferentially MCs, identified as a food-sensitive hippocampal subpopulation, has been proposed (Azevedo et al., 2019). The activity of this neuronal population is modulated by feeding behaviour, since they are silenced by fasting. When activating these D2R-expressing neurons, food intake is reduced, and also was the time explor- ing a quadrant that previously contained food -so the encoding for food-related memories-. MC axons uniquely display high expression levels of CB1R (Monory et al., 2006). Therefore, it is conceivable that an involvement of the receptor in this particular behavior occurs. The more active MCs are, the higher release of eCBs from the postsynaptic neuron and the higher retrograde activation of CB1R are con- ceivably achieved. According to our proposed mechanism, when deleting GAP43, CB1R would show an increased functionality. Hence, one of our next steps will be using a CB1R antagonist to try to revert this phenotype. GAP43 impacts on excitotoxic seizures In the mature brain, CB1R acts as a synaptic circuit breaker that is crucial for the control of brain excitability (Katona and Freund, 2008; Soltesz et al., 2015). Accordingly, acute activation of CB1R exerts anticonvulsant effects in various an- 110 imal models (Karler et al., 1974; Turkanis et al., 1979; Monory et al., 2006). The analysis of conditional mouse mutants lacking CB1R on different neuron popula- tions subjected to KA-induced seizures revealed that CB1R located on hippocampal glutamatergic but not GABAergic neurons is required for protection against ex- citotoxic seizures (Monory et al., 2006). Specifically, glutamatergic CB1R in the DG was reported as crucial for KA-induced epileptic phenotype modulation. By large, the higher proportion of glutamatergic CB1R found in the DG was the pool present in MC axons (Monory et al., 2006). A specific down-regulation of CB1R protein and mRNA on glutamatergic -but not on GABAergic- axon terminals has been reported in epileptic human hippocampal tissue (Ludanyi et al., 2008). Im- portantly, although the capacity of the eCB control system may be substantially downgraded in epilepsy, it is not completely lost. Thus, how to boost the residual circuit-breaking function without compromising temporal and spatial specificity is a crucial issue (Soltesz et al., 2015). To complement conditional loss-of-function studies with the corresponding gain-of-function approach, a somatic transfer of the CB1R gene to glutamatergic hippocampal neurons was proven as sufficient to pro- tect against acute seizures and neuronal damage (Guggenhuber et al., 2010). The results presented in this Doctoral Thesis, in line with the latter study, suggest that a GAP43 deletion in glutamatergic neurons –presumably MCs- would boost the pre-existing anticonvulsant functionality of CB1R located on those cells. MCs can regulate GC activity directly through innervation of GCs or indirectly through modulation of GABAergic INs. MC excitation of GCs is normally weak, but under pathological conditions such as pro-convulsant insults, it is dramatically strengthened. Inhibiting MCs selectively during severe seizures reduced phenotyp- ical manifestations of those seizures. In contrast, activating MCs selectively was pro-convulsant (Botterill et al., 2019). CB1R activation could play an important role in stabilizing the recurrent GC-MC-GC circuit. If, according to our hypothesis, CB1R was de-inhibited by eliminating its inhibitory interaction with GAP43, its functioning would be boosted at this synapse. In our Racine Scale measurements, we mostly observed differences in the receptor sensitivity to THC rather than in the maximal effect of the drug, so as a next step it would be interesting to perform a dose-response study of the effects of THC in Glu-GAP43 KO mice compared to WT littermates, and to include the GABA-GAP43 KO line as a control. Noticeably, we observed a tendency to higher susceptibility to KA-induced epi- lepsy in Glu-GAP43−/− compared to WT littermates, contrary to the notion of GAP43 as a pro-epileptic marker as described in the literature. GAP43 is increased in patients with cortical dysplasia, a condition that causes refractory epilepsy. Si- lencing GAP43 in the cortex, where it is mainly expressed in glutamatergic neur- ons, reverts the pro-epileptic phenotype in cortical dysplasia mouse models (Nemes et al., 2017). Hippocampal alterations following pro-convulsant drug administra- tion include increases in glutamate release, prostaglandin levels, changes in ATPase activity, as well as neuronal loss and mossy fiber sprouting (Naffah-Mazzacoratti et al., 1999). In the KA model, GAP43 expression increases in concert with an ab- errant sprouting of mossy fibers from GCs to the IML. Mossy fiber sprouting in TLE might underlie the appearance of an aberrant circuitry which promotes the gener- ation or spread of spontaneous seizure activity. GAP43 mRNA expression in GCs and mossy fiber sprouting increase 24 hours after KA injection, and no changes were observed in the hippocampus of the animals 6 or 12 hours after the excitatory insult 111 (McNamara and Routtenberg, 1995; Meberg et al., 1993). However, another report showed rapid increases of the GAP43 labelling content in IML after 5 hours of seizure activity onset as evoked by the injection of pilocarpine, another pro-convulsant drug. This may be related to increased protein transport or increased GAP43 gene expres- sion in pre-existing synaptic terminals of MCs (Naffah-Mazzacoratti et al., 1999). However, this issue is not well understood yet. Of note, the pro-convulsant tendency of our Glu-GAP43−/− mice would be in line with an activation of MCs (Botterill et al., 2019) and a lack of food memory encoding, conceivably upon activation of MCs too (Azevedo et al., 2019). Thus, a deletion of glutamatergic-neuron GAP43 somehow may cause MC activation. Nev- ertheless, to address all these questions, we should achieve a precise assessment of synapse specificity and to delete GAP43 selectively in hippocampal MCs in addition to conduct a cannabinoid-based treatment. Otherwise, GAP43 deletion in all telen- cephalic glutamatergic cells might have multiple interpretations that are difficult to ascribe solely to the behavior of MCs. Understanding of the cell type/population-specificity of cannabinoid signalling is crucial to advance in cannabinoid-based treatments for epilepsy and other diseases. Without mechanistic knowledge of the structural and functional organization of the ECS under normal conditions, it will be impossible to design personalized inter- ventions to restore or depress the eCB control system in selective spatiotemporal pathological contexts. General treatments with exogenous cannabinoids as THC may cause unwanted cognitive effects. Thus, an ideal approach would be to induce the selective targeting of a CB1R pool in excitatory cells in a spatially, temporally and neuronal population-specific manner. In other words, to develop strategies that enhance the ability to release eCBs when and where needed, or to prolong the activ- ity of eCBs only when and where naturally released, by procedures that selectively target CB1R molecules on the bases of their tissue and time of expression (Soltesz et al., 2015; Hofmann and Frazier, 2013). 112 Conclusions The data obtained in this Doctoral Thesis allow us to delineate the main following conclusions: 1. GAP43, mainly in its S41-phosphorylated form, interacts with CB1R-CTD in HEK293T cells heterologously expressing the two proteins. 2. GAP43, presumably by its interaction with CB1R, decreases cannabinoid- evoked signaling in HEK293T cells, especially the Gαq/11 protein-driven ROCK /cofilin cascade, while the canonical Gαi/o protein-driven cAMP/PKA and ERK pathways remain unaltered. 3. The GAP43-CB1R interaction occurs selectively in glutamatergic terminals of MCs, being restricted to the IML of the DG. The interaction is not observed in GABAergic INs within that hippocampal region. 4. The expression of the GAP43-S41D phosphomimetic mutant, presumably by its interaction with CB1R, reduces short-term cannabinoid-mediated plasticity in MC terminals, i.e. DSE and WIN-mediated suppression of neurotransmis- sion. Conversely, the blockade of PKC-dependent GAP43-S41 phosphorylation at MC-GC synapses facilitates short-term CB1R-mediated plasticity. 5. GAP43 expression in telencephalic glutamatergic terminals is relevant for mo- tor performance and memory encoding in mice, but it does not seem to affect other behavioral traits. 6. GAP43 expression in telencephalic glutamatergic terminals seems to modulate CB1R-mediated anti-seizure activity in an acute KA-injection mouse model of epilepsy. 113 References Aguado, T., Monory, K., Palazuelos, J., Stella‡, N., Cravatt, B., Lutz, B., Marsicano, G., Kokaia, Z., Guzmán, M., and Galve-Roperh, I. (2005). The endocannabinoid system drives neural progenitor proliferation. The FASEB Journal, 19(12):1704– 1706. Aigner, L., Arber, S., Kapfhammer, J. P., Laux, T., Schneider, C., Botteri, F., Brenner, H. R., and Caroni, P. (1995). Overexpression of the neural growth- associated protein gap-43 induces nerve sprouting in the adult nervous system of transgenic mice. Cell, 83(2):269–278. Amaral, D. and Witter, M. (1989). The three-dimensional organization of the hip- pocampal formation: A review of anatomical data. Neuroscience, 31(3):571–591. Amaral, D. G., Scharfman, H. E., and Lavenex, P. (2007). The dentate gyrus: fundamental neuroanatomical organization (dentate gyrus for dummies). Progress in Brain Research, 163:3–22. Anavi-Goffer, S., Fleischer, D., Hurst, D. P., Lynch, D. L., Barnett-Norris, J., Shi, S., Lewis, D. L., Mukhopadhyay, S., Howlett, A. C., Reggio, P. H., and Abood, M. E. (2007). Helix 8 leu in the cb1 cannabinoid receptor contributes to selective signal transduction mechanisms. Journal of Biological Chemistry, 282(34):25100– 25113. Apel, E. D., Byford, M. F., Au, D., Walsh, K. A., and Storm, D. R. (1990). Identific- ation of the protein kinase c phosphorylation site in neuromodulin. Biochemistry, 29(9):2330–2335. Apel, E. D., Litchfield, D. W., Clark, R. H., Krebs, E. G., and Storm, D. R. (1991). Phosphorylation of neuromodulin (gap-43) by casein kinase ii. identification of phosphorylation sites and regulation by calmodulin. The Journal of biological chemistry, 266(16):10544–51. Araque, A., Castillo, P. E., Manzoni, O. J., and Tonini, R. (2017). Synaptic functions of endocannabinoid signaling in health and disease. Neuropharmacology, 124:13– 24. Araque, A. and Navarrete, M. (2010). Glial cells in neuronal network function. Philo- sophical Transactions of the Royal Society B: Biological Sciences, 365(1551):2375– 2381. Argaw, A., Duff, G., Zabouri, N., Cecyre, B., Chaine, N., Cherif, H., Tea, N., Lutz, B., Ptito, M., and Bouchard, J.-F. (2011). Concerted action of cb1 cannabinoid receptor and deleted in colorectal cancer in axon guidance. Journal of Neuros- cience, 31(4):1489–1499. 114 Armas-Capote, N., Maglio, L. E., Pérez-Atencio, L., Martin-Batista, E., Reboreda, A., Barios, J. A., Hernandez, G., de la Rosa, D. A., Lamas, J. A., Barrio, L. C., and Giraldez, T. (2020). Sgk1.1 reduces kainic acid-induced seizure severity and leads to rapid termination of seizures. Cerebral cortex (New York, N.Y. : 1991), 30(5):3184–3197. Augustin, S. M. and Lovinger, D. M. (2018). Functional relevance of endocannabinoid-dependent synaptic plasticity in the central nervous system. ACS Chemical Neuroscience, 9(9):2146–2161. Azad, S. C. (2004). Circuitry for associative plasticity in the amygdala involves endocannabinoid signaling. Journal of Neuroscience, 24(44):9953–9961. Azevedo, E. P., Pomeranz, L., Cheng, J., Schneeberger, M., Vaughan, R., Stern, S. A., Tan, B., Doerig, K., Greengard, P., and Friedman, J. M. (2019). A role of drd2 hippocampal neurons in context-dependent food intake. Neuron, 102(4):873– 886.e5. Bacci, A., Huguenard, J. R., and Prince, D. A. (2004). Long-lasting self-inhibition of neocortical interneurons mediated by endocannabinoids. Nature, 431(7006):312– 316. Bagher, A. M., Laprairie, R. B., Kelly, M. E., and Denovan-Wright, E. M. (2013). Co-expression of the human cannabinoid receptor coding region splice variants (hcb1) affects the function of hcb1 receptor complexes. European Journal of Phar- macology, 721(1-3):341–354. Basavarajappa, B. S. and Subbanna, S. (2014). Cb1 receptor-mediated signaling un- derlies the hippocampal synaptic, learning, and memory deficits following treat- ment with jwh-081, a new component of spice/k2 preparations. Hippocampus, 24(2):178–188. Baumgärtel, K. and Mansuy, I. M. (2012). Neural functions of calcineurin in synaptic plasticity and memory. Learning and Memory, 19(9):375–384. Bellocchio, L., Lafenêtre, P., Cannich, A., Cota, D., Puente, N., Grandes, P., Chaouloff, F., Piazza, P. V., and Marsicano, G. (2010). Bimodal control of stimulated food intake by the endocannabinoid system. Nature Neuroscience, 13:281–283. Bendotti, C., Servadio, A., and Samanin, R. (1991). Distribution of gap-43 mrna in the brain stem of adult rats as evidenced by in situ hybridization: localization within monoaminergic neurons. The Journal of Neuroscience, 11(3):600–607. Benowitz, L. I., Apostolides, P. J., Perrone-Bizzozero, N., Finklestein, S. P., and Zwiers, H. (1988). Anatomical distribution of the growth-associated protein gap- 43/b-50 in the adult rat brain. Journal of Neuroscience, 8(1):339–352. Benowitz, L. I., Perrone-Bizzozero, N. I., Finkelstein, S. P., and Bird, E. D. (1989). Localization of the growth-associated phosphoprotein gap-43 (b-50, f1) in the human cerebral cortex. Journal of Neuroscience, 9(3):990–995. 115 Benowitz, L. I. and Routtenberg, A. (1997). Gap-43: An intrinsic determinant of neuronal development and plasticity. Trends in Neurosciences, 20(2):84–91. Berghuis, P., Rajnicek, A. M., Morozov, Y. M., Ross, R. A., Mulder, J., Urbán, G. M., Monory, K., Marsicano, G., Matteoli, M., Canty, A., Irving, A. J., Katona, I., Yanagawa, Y., Rakic, P., Lutz, B., Mackie, K., and Harkany, T. (2007). Hardwiring the bain: Endocannabinoids shape neuronal connectivity. Science, 316(5828):1212–1216. Berglund, K., Midorikawa, M., and Tachibana, M. (2002). Increase in the pool size of releasable synaptic vesicles by the activation of protein kinase c in goldfish retinal bipolar cells. The Journal of Neuroscience, 22(12):4776–4785. Bisogno, T., Hanuš, L., Petrocellis, L. D., Tchilibon, S., Ponde, D. E., Brandi, I., Moriello, A. S., Davis, J. B., Mechoulam, R., and Marzo, V. D. (2001). Molecular targets for cannabidiol and its synthetic analogues: effect on vanilloid vr1 recept- ors and on the cellular uptake and enzymatic hydrolysis of anandamide. British Journal of Pharmacology, 134(4):845–852. Bisogno, T., Melck, D., Petrocellis, L., and Marzo, V. (1999). Phosphatidic acid as the biosynthetic precursor of the endocannabinoid 2-arachidonoylglycerol in intact mouse neuroblastoma cells stimulated with ionomycin. Journal of Neurochemistry, 72(5):2113–2119. Blume, L. C., Eldeeb, K., Bass, C. E., Selley, D. E., and Howlett, A. C. (2015). Cannabinoid receptor interacting protein (crip1a) attenuates cb1r signaling in neuronal cells. Cellular Signalling, 27(3):716–726. Blume, L. C., Leone-Kabler, S., Luessen, D. J., Marrs, G. S., Lyons, E., Bass, C. E., Chen, R., Selley, D. E., and Howlett, A. C. (2016). Cannabinoid receptor interact- ing protein suppresses agonist-driven cb1 receptor internalization and regulates receptor replenishment in an agonist-biased manner. Journal of Neurochemistry, 139(3):396–407. Blume, L. C., Patten, T., Eldeeb, K., Leone-Kabler, S., Ilyasov, A. A., Keegan, B. M., O’Neal, J. E., Bass, C. E., Hantgan, R. R., Lowther, W. T., Selley, D. E., and Howlett, A. C. (2017). Cannabinoid receptor interacting protein 1a com- petition with -arrestin for cb 1 receptor binding sites. Molecular Pharmacology, 91(2):75–86. Blázquez, C., Chiarlone, A., Sagredo, O., Aguado, T., Pazos, M. R., Resel, E., Palazuelos, J., Julien, B., Salazar, M., Börner, C., Benito, C., Carrasco, C., Diez- Zaera, M., Paoletti, P., Dı́az-Hernández, M., Ruiz, C., Sendtner, M., Lucas, J. J., de Yébenes, J. G., Marsicano, G., Monory, K., Lutz, B., Romero, J., Alberch, J., Ginés, S., Kraus, J., Fernández-Ruiz, J., Galve-Roperh, I., and Guzmán, M. (2011). Loss of striatal type 1 cannabinoid receptors is a key pathogenic factor in huntington’s disease. Brain, 134(1):119–136. Blázquez, C., Ruiz-Calvo, A., Bajo-Grañeras, R., Baufreton, J. M., Resel, E., Var- ilh, M., Zottola, A. C. P., Mariani, Y., Cannich, A., Rodŕıguez-Navarro, J. A., 116 Marsicano, G., Galve-Roperh, I., Bellocchio, L., and Guzmán, M. (2020). Inhib- ition of striatonigral autophagy as a link between cannabinoid intoxication and impairment of motor coordination. eLife, 9. Bodor, A. L. (2005). Endocannabinoid signaling in rat somatosensory cortex: Lam- inar differences and involvement of specific interneuron types. Journal of Neur- oscience, 25(29):6845–6856. Booth, W. T., Walker, N. B., Lowther, W. T., and Howlett, A. C. (2019). Cannabin- oid receptor interacting protein 1a (crip1a): Function and structure. Molecules, 24(20):3672. Bosier, B., Muccioli, G. G., Hermans, E., and Lambert, D. M. (2010). Functionally selective cannabinoid receptor signalling: Therapeutic implications and opportun- ities. Biochemical Pharmacology, 80(1):1–12. Botterill, J. J., Lu, Y.-L., LaFrancois, J. J., Bernstein, H. L., Alcantara-Gonzalez, D., Jain, S., Leary, P., and Scharfman, H. E. (2019). An excitatory and epilep- togenic effect of dentate gyrus mossy cells in a mouse model of epilepsy. Cell Reports, 29(9):2875–2889.e6. Bouaboula, M., Hilairet, S., Marchand, J., Fajas, L., Fur, G. L., and Casellas, P. (2005). Anandamide induced ppar transcriptional activation and 3t3-l1 preadipo- cyte differentiation. European Journal of Pharmacology, 517(3):174–181. Bouaboula, M., Poinot-Chazel, C., Bourrié, B., Canat, X., Calandra, B., Rinaldi- Carmona, M., Fur, G. L., and Casellas, P. (1995). Activation of mitogen-activated protein kinases by stimulation of the central cannabinoid receptor cb1. Biochem- ical Journal, 312(2):637–641. Breivogel, C. S., Selley, D. E., and Childers, S. R. (1998). Cannabinoid receptor agonist efficacy for stimulating [35s]gtps binding to rat cerebellar membranes cor- relates with agonist-induced decreases in gdp affinity. Journal of Biological Chem- istry, 273(27):16865–16873. Breivogel, C. S., Sim, L. J., and Childers, S. R. (1997). Regional differences in can- nabinoid receptor/g-protein coupling in rat brain. The Journal of pharmacology and experimental therapeutics, 282(3):1632–42. Breivogel, C. S. and Sim-Selley, L. J. (2009). Basic neuroanatomy and neurophar- macology of cannabinoids. International Review of Psychiatry, 21(2):113–121. Breivogel, C. S. and Vaghela, M. S. (2015). The effects of beta-arrestin1 deletion on acute cannabinoid activity, brain cannabinoid receptors and tolerance to can- nabinoids in mice. Journal of Receptors and Signal Transduction, 35(1):98–106. Bromberg, K. D., Ma’ayan, A., Neves, S. R., and Iyengar, R. (2008). Design logic of a cannabinoid receptor signaling network that triggers neurite outgrowth. Science, 320(5878):903–909. Brown, D. A. and Sihra, T. S. (2008). Presynaptic signaling by heterotrimeric g- proteins. Handbook of experimental pharmacology, (184):207–60. 117 Buckmaster, P. S. and Schwartzkroin, P. A. (1994). Hippocampal mossy cell func- tion: A speculative view. Hippocampus, 4(4):393–402. Buckmaster, P. S., Wenzel, H. J., Kunkel, D. D., and Schwartzkroin, P. A. (1996). Axon arbors and synaptic connections of hippocampal mossy cells in the rat in vivo. The Journal of Comparative Neurology, 366(2):270–292. Bui, A. D., Nguyen, T. M., Limouse, C., Kim, H. K., Szabo, G. G., Felong, S., Maroso, M., and Soltesz, I. (2018). Dentate gyrus mossy cells control spon- taneous convulsive seizures and spatial memory. Science (New York, N.Y.), 359(6377):787–790. Burford, N. T., Tolbert, L. M., and Sadee, W. (1998). Specific g protein activation and mu-opioid receptor internalization caused by morphine, damgo and endo- morphin i. European journal of pharmacology, 342(1):123–6. Busquets-Garcia, A., Desprez, T., Metna-Laurent, M., Bellocchio, L., Marsicano, G., and Soria-Gomez, E. (2015). Dissecting the cannabinergic control of behavior: The where matters. BioEssays, 37(11):1215–1225. Bénard, G., Massa, F., Puente, N., Lourenço, J., Bellocchio, L., Soria-Gómez, E., Matias, I., Delamarre, A., Metna-Laurent, M., Cannich, A., Hebert-Chatelain, E., Mulle, C., Ortega-Gutiérrez, S., Mart́ın-Fontecha, M., Klugmann, M., Guggen- huber, S., Lutz, B., Gertsch, J., Chaouloff, F., López-Rodŕıguez, M. L., Grandes, P., Rossignol, R., and Marsicano, G. (2012). Mitochondrial cb1 receptors regulate neuronal energy metabolism. Nature Neuroscience, 15(4):558–564. Caballero-Florán, R. N., Conde-Rojas, I., Chávez, A. O., Cortes-Calleja, H., Lopez- Santiago, L. F., Isom, L. L., Aceves, J., Erlij, D., and Florán, B. (2016). Cannabinoid-induced depression of synaptic transmission is switched to stimu- lation when dopaminergic tone is increased in the globus pallidus of the rodent. Neuropharmacology, 110:407–418. Callén, L., Moreno, E., Barroso-Chinea, P., Moreno-Delgado, D., Cortés, A., Mal- lol, J., Casadó, V., Lanciego, J. L., Franco, R., Lluis, C., Canela, E. I., and McCormick, P. J. (2012). Cannabinoid receptors cb1 and cb2 form functional heteromers in brain. Journal of Biological Chemistry, 287(25):20851–20865. Cammarota, M., Paratcha, G., Stein, M. L. D., Bernabeu, R., Izquierdo, I., and Medina, J. H. (1997). B-50/gap-43 phosphorylation and pkc activity are increased in rat hippocampal synaptosomal membranes after an inhibitory avoidance train- ing. Neurochemical Research, 22(4):499–505. Campagne, M. V. L., Oestreicher, A. B., Henegouwen, P. M. P. V. B. E., and Gispen, W. H. (1990). Ultrastructural double localization of b-50/gap43 and synaptophysin (p38) in the neonatal and adult rat hippocampus. Journal of Neurocytology, 19(6):948–961. Cantallops, I. and Routtenberg, A. (1996). Rapid induction by kainic acid of both axonal growth and f1/gap-43 protein in the adult rat hippocampal granule cells. Journal of Comparative Neurology, 366(2):303–319. 118 Capogna, M., Fankhauser, C., Gagliardini, V., Gähwiler, B. H., and Thompson, S. M. (1999). Excitatory synaptic transmission and its modulation by pkc is unchanged in the hippocampus of gap-43- deficient mice. European Journal of Neuroscience, 11(2):433–440. Caprara, G. A., Morabito, C., Perni, S., Navarra, R., Guarnieri, S., and Mariggiò, M. A. (2016). Evidence for altered ca2+ handling in growth associated protein 43-knockout skeletal muscle. Frontiers in Physiology, 7:493. Carlson, G., Wang, Y., and Alger, B. E. (2002). Endocannabinoids facilitate the induction of ltp in the hippocampus. Nature Neuroscience, 5(8):723–724. Castillo, P. E., Younts, T. J., Chávez, A. E., and Hashimotodani, Y. (2012). En- docannabinoid signaling and synaptic function. Neuron, 76(1):70–81. Chakravarthy, B., Rashid, A., Brown, L., Tessier, L., Kelly, J., and Ménard, M. (2008). Association of gap-43 (neuromodulin) with microtubule-associated protein map-2 in neuronal cells. Biochemical and Biophysical Research Communications, 371(4):679–683. Chapman, E. R., Au, D., Alexander, K. A., Nicolson, T. A., and Storm, D. R. (1991). Characterization of the calmodulin binding domain of neuromodulin. functional significance of serine 41 and phenylalanine 42. Journal of Biological Chemistry, 266(1):207–13. Chen, K., Neu, A., Howard, A. L., Foldy, C., Echegoyen, J., Hilgenberg, L., Smith, M., Mackie, K., and Soltesz, I. (2007). Prevention of plasticity of endocannabin- oid signaling inhibits persistent limbic hyperexcitability caused by developmental seizures. Journal of Neuroscience, 27(1):46–58. Chen, K., Ratzliff, A., Hilgenberg, L., Gulyás, A., Freund, T. F., Smith, M., Dinh, T. P., Piomelli, D., Mackie, K., and Soltesz, I. (2003). Long-term plasticity of endocannabinoid signaling induced by developmental febrile seizures. Neuron, 39(4):599–611. Chevaleyre, V. and Castillo, P. E. (2003). Heterosynaptic ltd of hippocampal gabaer- gic synapses: a novel role of endocannabinoids in regulating excitability. Neuron, 38(3):461–72. Chevaleyre, V. and Castillo, P. E. (2004). Endocannabinoid-mediated metaplasticity in the hippocampus. Neuron, 43(6):871–881. Chevaleyre, V., Heifets, B. D., Kaeser, P. S., Südhof, T. C., and Castillo, P. E. (2007). Endocannabinoid-mediated long-term plasticity requires camp/pka sig- naling and rim1. Neuron, 54(5):801–812. Chevaleyre, V., Takahashi, K. A., and Castillo, P. E. (2006). Endocannabinoid- mediated synaptic plasticity in the cns. Annual Review of Neuroscience, 29(1):37– 76. Chiaramello, A., Neuman, T., Peavy, D. R., and Zuber, M. X. (1996). The gap- 43 gene is a direct downstream target of the basic helix-loop-helix transcription factors. Journal of Biological Chemistry, 271(36):22035–22043. 119 Chiarlone, A., Bellocchio, L., Blazquez, C., Resel, E., Soria-Gomez, E., Cannich, A., Ferrero, J. J., Sagredo, O., Benito, C., Romero, J., Sanchez-Prieto, J., Lutz, B., Fernandez-Ruiz, J., Galve-Roperh, I., and Guzman, M. (2014). A restricted population of cb1 cannabinoid receptors with neuroprotective activity. Proceedings of the National Academy of Sciences, 111(22):8257–8262. Chikumi, H., Vázquez-Prado, J., Servitja, J.-M., Miyazaki, H., and Gutkind, J. S. (2002). Potent activation of rhoa by gq and gq-coupled receptors. Journal of Biological Chemistry, 277(30):27130–27134. Childers, S. R. and Deadwyler, S. A. (1996). Role of cyclic amp in the actions of cannabinoid receptors. Biochemical Pharmacology, 52(6):819–827. Chiu, C. Q. and Castillo, P. E. (2008). Input-specific plasticity at excitatory syn- apses mediated by endocannabinoids in the dentate gyrus. Neuropharmacology, 54(1):68–78. Chávez, A. E., Chiu, C. Q., and Castillo, P. E. (2010). Trpv1 activation by endo- genous anandamide triggers postsynaptic long-term depression in dentate gyrus. Nature Neuroscience, 13(12):1511–1518. Citri, A. and Malenka, R. C. (2008). Synaptic plasticity: Multiple forms, functions, and mechanisms. Neuropsychopharmacology, 33(1):18–41. Cravatt, B. F. and Lichtman, A. H. (2004). The endogenous cannabinoid system and its role in nociceptive behavior. Journal of Neurobiology, 61(1):149–160. Cui, Y., Prokin, I., Xu, H., Delord, B., Genet, S., Venance, L., and Berry, H. (2016). Endocannabinoid dynamics gate spike-timing dependent depression and potentiation. eLife, 5:e13185. d. H. Ratzliff, A. (2004). Rapid deletion of mossy cells does not result in a hyperex- citable dentate gyrus: Implications for epileptogenesis. Journal of Neuroscience, 24(9):2259–2269. Dalton, G. D. and Howlett, A. C. (2012). Cannabinoid cb1 receptors transactivate multiple receptor tyrosine kinases and regulate serine/threonine kinases to activate erk in neuronal cells. British Journal of Pharmacology, 165(8):2497–2511. Daniel, H. and Crepel, F. (2001). Control of ca2+ influx by cannabinoid and metabotropic glutamate receptors in rat cerebellar cortex requires k+ channels. The Journal of Physiology, 537(3):793–800. Daniel, H., Rancillac, A., and Crepel, F. (2004). Mechanisms underlying cannabin- oid inhibition of presynaptic ca 2+ influx at parallel fibre synapses of the rat cerebellum. The Journal of Physiology, 557(1):159–174. Danielson, N. B., Turi, G. F., Ladow, M., Chavlis, S., Petrantonakis, P. C., Poirazi, P., and Losonczy, A. (2017). In vivo imaging of dentate gyrus mossy cells in behaving mice. Neuron, 93(3):552–559.e4. 120 Davis, M. I., Ronesi, J., and Lovinger, D. M. (2003). A predominant role for inhibition of the adenylate cyclase/protein kinase a pathway in erk activation by cannabinoid receptor 1 in n1e-115 neuroblastoma cells. Journal of Biological Chemistry, 278(49):48973–48980. de la Monte, S. M., Ng, S. C., and Hsu, D. W. (1995). Aberrant gap-43 gene expression in alzheimer’s disease. The American journal of pathology, 147(4):934– 46. de Salas-Quiroga, A., Dı́az-Alonso, J., Garćıa-Rincón, D., Remmers, F., Vega, D., Gómez-Cañas, M., Lutz, B., Guzmán, M., and Galve-Roperh, I. (2015). Prenatal exposure to cannabinoids evokes long-lasting functional alterations by targeting cb1 receptors on developing cortical neurons. Proceedings of the National Academy of Sciences of the United States of America, 112(44):13693–8. Deadwyler, S. A., Hampson, R. E., Bennett, B. A., Edwards, T. A., Mu, J., Pacheco, M. A., Ward, S. J., and Childers, S. R. (1993). Cannabinoids modulate potassium current in cultured hippocampal neurons. Receptors channels, 1(2):121–34. Deadwyler, S. A., Hampson, R. E., Mu, J., Whyte, A., and Childers, S. (1995). Can- nabinoids modulate voltage sensitive potassium a-current in hippocampal neurons via a camp-dependent process. The Journal of pharmacology and experimental therapeutics, 273(2):734–43. Debanne, D., Guérineau, N. C., Gähwiler, B. H., and Thompson, S. M. (1996). Paired-pulse facilitation and depression at unitary synapses in rat hippocam- pus: quantal fluctuation affects subsequent release. The Journal of Physiology, 491(1):163–176. Degroot, A., Köfalvi, A., Wade, M. R., Davis, R. J., Rodrigues, R. J., Rebola, N., Cunha, R. A., and Nomikos, G. G. (2006). Cb 1 receptor antagonism increases hippocampal acetylcholine release: Site and mechanism of action. Molecular Phar- macology, 70(4):1236–1245. Deisseroth, K., Mermelstein, P. G., Xia, H., and Tsien, R. W. (2003). Signaling from synapse to nucleus: the logic behind the mechanisms. Current Opinion in Neurobiology, 13(3):354–365. Dekker, L. V., Graan, P. N. E., Versteeg, D. H. G., Oestreicher, A. B., and Gispen, W. H. (1989). Phosphorylation of b-50 (gap43) is correlated with neurotransmitter release in rat hippocampal slices. Journal of Neurochemistry, 52(1):24–30. del Pulgar, T. G., Velasco, G., and Guzmán, M. (2000). The cb1 cannabinoid receptor is coupled to the activation of protein kinase b/akt. The Biochemical journal, 347(Pt 2):369–73. Delgado-Peraza, F., Ahn, K. H., Nogueras-Ortiz, C., Mungrue, I. N., Mackie, K., Kendall, D. A., and Yudowski, G. A. (2016). Mechanisms of biased -arrestin- mediated signaling downstream from the cannabinoid 1 receptor. Molecular Phar- macology, 89(6):618–629. 121 Dell’Acqua, M. L. and Woolfrey, K. M. (2018). Fretting over postsynaptic pkc signaling. Nature Neuroscience, 21(8):1021–1022. Deloulme, J. C., Janet, T., Au, D., Storm, D. R., Sensenbrenner, M., and Baudier, J. (1990). Neuromodulin (gap43): a neuronal protein kinase c substrate is also present in 0-2a glial cell lineage. characterization of neuromodulin in secondary cultures of oligodendrocytes and comparison with the neuronal antigen. Journal of Cell Biology, 111(4):1559–1569. Demuth, D. G. and Molleman, A. (2006). Cannabinoid signalling. Life Sciences, 78(6):549–563. Denny, J. (2006). Molecular mechanisms, biological actions, and neuropharmacology of the growth-associated protein gap-43. Current Neuropharmacology, 4(4):293– 304. Dent, E. W. and Meiri, K. F. (1998). Distribution of phosphorylated gap-43 (neur- omodulin) in growth cones directly reflects growth cone behavior. Journal of Neurobiology, 35(3):287–299. Derkinderen, P., Ledent, C., Parmentier, M., and Girault, J.-A. (2001). Cannabin- oids activate p38 mitogen-activated protein kinases through cb1 receptors in hip- pocampus. Journal of Neurochemistry, 77(3):957–960. Deshmukh, S. S. and Knierim, J. J. (2011). Representation of non-spatial and spatial information in the lateral entorhinal cortex. Frontiers in Behavioral Neuroscience, 5:69. Devane, W., Hanus, L., Breuer, A., Pertwee, R., Stevenson, L., Griffin, G., Gibson, D., Mandelbaum, A., Etinger, A., and Mechoulam, R. (1992). Isolation and structure of a brain constituent that binds to the cannabinoid receptor. Science, 258(5090):1946–1949. Devinsky, O., Cilio, M. R., Cross, H., Fernandez-Ruiz, J., French, J., Hill, C., Katz, R., Marzo, V. D., Jutras-Aswad, D., Notcutt, W. G., Martinez-Orgado, J., Robson, P. J., Rohrback, B. G., Thiele, E., Whalley, B., and Friedman, D. (2014). Cannabidiol: Pharmacology and potential therapeutic role in epilepsy and other neuropsychiatric disorders. Epilepsia, 55(6):791–802. Dias, R. and Aggleton, J. P. (2000). Effects of selective excitotoxic prefrontal le- sions on acquisition of nonmatching- and matching-to-place in the t-maze in the rat: differential involvement of the prelimbic-infralimbic and anterior cingulate cortices in providing behavioural flexibility. European Journal of Neuroscience, 12(12):4457–4466. Diez-Alarcia, R., Ibarra-Lecue, I., Ángela P. Lopez-Cardona, Meana, J., Gutierrez- Adán, A., Callado, L. F., Agirregoitia, E., and Urigüen, L. (2016). Biased agonism of three different cannabinoid receptor agonists in mouse brain cortex. Frontiers in Pharmacology, 7. Doller, H. J. and Weight, F. F. (1982). Perforant pathway activation of hippocam- pal ca1 stratum pyramidale neurons: Electrophysiological evidence for a direct pathway. Brain Research, 237(1):1–13. 122 Dı́az-Alonso, J., de Salas-Quiroga, A., Paráıso-Luna, J., Garćıa-Rincón, D., Garcez, P. P., Parsons, M., Andradas, C., Sánchez, C., Guillemot, F., Guzmán, M., and Galve-Roperh, I. (2016). Loss of cannabinoid cb 1 receptors induces cortical migration malformations and increases seizure susceptibility. Cerebral Cortex. Echegoyen, J., Armstrong, C., Morgan, R. J., and Soltesz, I. (2009). Single applica- tion of a cb1 receptor antagonist rapidly following head injury prevents long-term hyperexcitability in a rat model. Epilepsy Research, 85(1):123–127. Egertová, M., Simon, G. M., Cravatt, B. F., and Elphick, M. R. (2008). Localiz- ation ofn-acyl phosphatidylethanolamine phospholipase d (nape-pld) expression in mouse brain: A new perspective onn-acylethanolamines as neural signaling molecules. The Journal of Comparative Neurology, 506(4):604–615. Falenski, K., Blair, R., Sim-Selley, L., Martin, B., and DeLorenzo, R. (2007). Status epilepticus causes a long-lasting redistribution of hippocampal cannabinoid type 1 receptor expression and function in the rat pilocarpine model of acquired epilepsy. Neuroscience, 146(3):1232–1244. Falenski, K. W., Carter, D. S., Harrison, A. J., Martin, B. R., Blair, R. E., and DeLorenzo, R. J. (2009). Temporal characterization of changes in hippocampal cannabinoid cb1 receptor expression following pilocarpine-induced status epilepti- cus. Brain Research, 1262:64–72. Fanselow, M. S. and Dong, H.-W. (2010). Are the dorsal and ventral hippocampus functionally distinct structures? Neuron, 65(1):7–19. Flamm, A. G., Żerko, S., Zawadzka-Kazimierczuk, A., Koźmiński, W., Konrat, R., and Coudevylle, N. (2016). 1h, 15n, 13c resonance assignment of human gap-43. Biomolecular NMR assignments, 10(1):171–4. Fletcher-Jones, A., Hildick, K. L., Evans, A. J., Nakamura, Y., Wilkinson, K. A., and Henley, J. M. (2019). The c-terminal helix 9 motif in rat cannabinoid receptor type 1 regulates axonal trafficking and surface expression. eLife, 8. Flores, A., Maldonado, R., and Berrendero, F. (2013). Cannabinoid-hypocretin cross-talk in the central nervous system: what we know so far. Frontiers in Neuroscience, 7:256. Fortin, D. A. and Levine, E. S. (2006). Differential effects of endocannabinoids on glutamatergic and gabaergic inputs to layer 5 pyramidal neurons. Cerebral Cortex, 17(1):163–174. Frazier, C. J. (2007). Endocannabinoids in the dentate gyrus. Progress in Brain Research, 163:319–37. Freund, T. F. and Buzsáki, G. (1996). Interneurons of the hippocampus. Hippo- campus, 6(4):247–470. Freund, T. F., Katona, I., and Piomelli, D. (2003). Role of endogenous cannabinoids in synaptic signaling. Physiological Reviews, 83(3):1017–1066. 123 Fukudome, Y., Ohno-Shosaku, T., Matsui, M., Omori, Y., Fukaya, M., Tsubokawa, H., Taketo, M. M., Watanabe, M., Manabe, T., and Kano, M. (2004). Two distinct classes of muscarinic action on hippocampal inhibitory synapses: M2- mediated direct suppression and m1/m3-mediated indirect suppression through endocannabinoid signalling. European Journal of Neuroscience, 19(10):2682–2692. Fulop, T. and Smith, C. (2006). Physiological stimulation regulates the exocytic mode through calcium activation of protein kinase c in mouse chromaffin cells. Biochemical Journal, 399(1):111–119. Galve-Roperh, I., Aguado, T., Palazuelos, J., and Guzmán, M. (2007). The en- docannabinoid system and neurogenesis in health and disease. The Neuroscientist, 13(2):109–114. Galve-Roperh, I., Rueda, D., del Pulgar, T. G., Velasco, G., and Guzmán, M. (2002). Mechanism of extracellular signal-regulated kinase activation by the cb 1 cannabinoid receptor. Molecular Pharmacology, 62(6):1385–1392. Galve-Roperh, I., Sánchez, C., Cortés, M. L., del Pulgar, T. G., Izquierdo, M., and Guzmán, M. (2000). Anti-tumoral action of cannabinoids: Involvement of sus- tained ceramide accumulation and extracellular signal-regulated kinase activation. Nature Medicine, 6(3):313–319. Gangarossa, G., Longueville, S., Bundel, D. D., Perroy, J., Hervé, D., Girault, J.-A., and Valjent, E. (2012). Characterization of dopamine d1 and d2 receptor- expressing neurons in the mouse hippocampus. Hippocampus, 22(12):2199–2207. Garcia, D. E., Brown, S., Hille, B., and Mackie, K. (1998). Protein kinase c disrupts cannabinoid actions by phosphorylation of the cb1 cannabinoid receptor. The Journal of Neuroscience, 18(8):2834–2841. Garćıa-Morales, V., Montero, F., González-Forero, D., Rodŕıguez-Bey, G., Gómez- Pérez, L., Medialdea-Wandossell, M. J., Domı́nguez-Vı́as, G., Garćıa-Verdugo, J. M., and Moreno-López, B. (2015). Membrane-derived phospholipids control synaptic neurotransmission and plasticity. PLoS biology, 13(5):e1002153. Gauthier-Campbell, C., Bredt, D. S., Murphy, T. H., and El-Husseini, A. E.-D. (2004). Regulation of dendritic branching and filopodia formation in hippocam- pal neurons by specific acylated protein motifs. Molecular Biology of the Cell, 15(5):2205–2217. Gauthier-Kemper, A., Igaev, M., Sündermann, F., Janning, D., Brühmann, J., Moschner, K., Reyher, H. J., Junge, W., Glebov, K., Walter, J., Bakota, L., and Brandt, R. (2014). Interplay between phosphorylation and palmitoylation mediates plasma membrane targeting and sorting of gap43. Molecular Biology of the Cell, 25(21):3284–3299. Gerdeman, G. L., Ronesi, J., and Lovinger, D. M. (2002). Postsynaptic endocan- nabinoid release is critical to long-term depression in the striatum. Nature Neur- oscience, 5(5):446–451. 124 Gianotti, C., Nunzi, M., Gispen, W., and Corradetti, R. (1992). Phosphorylation of the presynaptic protein b-50 (gap-43) is increased during electrically induced long-term potentiation. Neuron, 8(5):843–848. Gillis, K. D., Mößner, R., and Neher, E. (1996). Protein kinase c enhances exocyt- osis from chromaffin cells by increasing the size of the readily releasable pool of secretory granules. Neuron, 16(6):1209–1220. Glass, M. and Felder, C. C. (1997). Concurrent stimulation of cannabinoid cb1 and dopamine d2 receptors augments camp accumulation in striatal neurons: Evidence for a g s linkage to the cb1 receptor. The Journal of Neuroscience, 17(14):5327– 5333. Gogolla, N., Galimberti, I., and Caroni, P. (2007). Structural plasticity of axon terminals in the adult. Current Opinion in Neurobiology, 17(5):516–524. Gorgels, T. G., Campagne, M. V. L., Oestreicher, A. B., Gribnau, A. A., and Gispen, W. H. (1989). B-50/gap43 is localized at the cytoplasmic side of the plasma membrane in developing and adult rat pyramidal tract. The Journal of neuroscience : the official journal of the Society for Neuroscience, 9(11):3861–9. Groen, T. V. and Wyss, J. M. (1990). Extrinsic projections from area ca1 of the rat hippocampus: Olfactory, cortical, subcortical, and bilateral hippocampal forma- tion projections. The Journal of Comparative Neurology, 302(3):515–528. Grotenhermen, F. (2003). Pharmacokinetics and pharmacodynamics of cannabin- oids. Clinical Pharmacokinetics, 42(4):327–360. 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., and Klugmann, M. (2015). Cannabin- oid receptor-interacting protein crip1a modulates cb1 receptor signaling in mouse hippocampus. Brain Structure and Function, pages 2061–2074. Guggenhuber, S., Monory, K., Lutz, B., and Klugmann, M. (2010). Aav vector- mediated overexpression of cb1 cannabinoid receptor in pyramidal neurons of the hippocampus protects against seizure-induced excitoxicity. PLoS ONE, 5:e15707. Gulyas, A. I., Cravatt, B. F., Bracey, M. H., Dinh, T. P., Piomelli, D., Boscia, F., and Freund, T. F. (2004). Segregation of two endocannabinoid-hydrolyzing enzymes into pre- and postsynaptic compartments in the rat hippocampus, cerebellum and amygdala. European Journal of Neuroscience, 20(2):441–458. Gérard, C. M., Mollereau, C., Vassart, G., and Parmentier, M. (1991). Molecu- lar cloning of a human cannabinoid receptor which is also expressed in testis. Biochemical Journal, 279(1):129–134. Gómez, M., Hernández, M. L., Pazos, M. R., Tolón, R. M., Romero, J., and Fernández-Ruiz, J. (2008). Colocalization of cb1 receptors with l1 and gap-43 in forebrain white matter regions during fetal rat brain development: Evidence for a role of these receptors in axonal growth and guidance. Neuroscience, 153(3):687– 699. 125 Gómez-Gonzalo, M., Navarrete, M., Perea, G., Covelo, A., Mart́ın-Fernández, M., Shigemoto, R., Luján, R., and Araque, A. (2015). Endocannabinoids induce lateral long-term potentiation of transmitter release by stimulation of gliotrans- mission. Cerebral Cortex, 25(10):3699–3712. Haines, D. E. and Mihailoff, G. A. (2017). Fundamental Neuroscience for Basic and Clinical Applications: Fifth Edition. Hamid, J., Nelson, D., Spaetgens, R., Dubel, S. J., Snutch, T. P., and Zamponi, G. W. (1999). Identification of an integration center for cross-talk between pro- tein kinase c and g protein modulation of n-type calcium channels. Journal of Biological Chemistry, 274(10):6195–6202. Han, J., Kesner, P., Metna-Laurent, M., Duan, T., Xu, L., Georges, F., Koehl, M., Abrous, D. N., Mendizabal-Zubiaga, J., Grandes, P., Liu, Q., Bai, G., Wang, W., Xiong, L., Ren, W., Marsicano, G., and Zhang, X. (2012). Acute cannabinoids im- pair working memory through astroglial cb 1 receptor modulation of hippocampal ltd. Cell, 148(5):1039–1050. Han, M. H., Jiao, S., Jia, J. M., Chen, Y., Chen, C. Y., Gucek, M., Markey, S. P., and Li, Z. (2013). The novel caspase-3 substrate gap43 is involved in ampa re- ceptor endocytosis and long-term depression. Molecular and Cellular Proteomics, 12(12):3719–3731. Harkany, T., Dobszay, M., Cayetanot, F., Härtig, W., Siegemund, T., Aujard, F., and Mackie, K. (2005). Redistribution of cb1 cannabinoid receptors during evol- ution of cholinergic basal forebrain territories and their cortical projection areas: A comparison between the gray mouse lemur (microcebus murinus, primates) and rat. Neuroscience, 135(2):595–609. Haruta, T., Takami, N., Ohmura, M., Misumi, Y., and Ikehara, Y. (1997). Ca2+- dependent interaction of the growth-associated protein gap-43 with the synaptic core complex. Biochemical Journal, 325(2):455–463. Hashimotodani, Y., Nasrallah, K., Jensen, K. R., Chávez, A. E., Carrera, D., and Castillo, P. E. (2017). Ltp at hilar mossy cell-dentate granule cell synapses mod- ulates dentate gyrus output by increasing excitation/inhibition balance. Neuron, 95(4):928–943.e3. Hashimotodani, Y., Ohno-Shosaku, T., and Kano, M. (2007). Presynaptic monoacyl- glycerol lipase activity determines basal endocannabinoid tone and terminates ret- rograde endocannabinoid signaling in the hippocampus. Journal of Neuroscience, 27(5):1211–1219. Hashimotodani, Y., Ohno-Shosaku, T., Tsubokawa, H., Ogata, H., Emoto, K., Maejima, T., Araishi, K., Shin, H.-S., and Kano, M. (2005). Phospholipase c serves as a coincidence detector through its ca2+ dependency for triggering ret- rograde endocannabinoid signal. Neuron, 45(2):257–268. He, J. C., Gomes, I., Nguyen, T., Jayaram, G., Ram, P. T., Devi, L. A., and Iyengar, R. (2005). The go/i-coupled cannabinoid receptor-mediated neurite out- growth involves rap regulation of src and stat3. Journal of Biological Chemistry, 280(39):33426–33434. 126 He, Q., Dent, E. W., and Meiri, K. F. (1997). Modulation of actin filament behavior by gap-43 (neuromodulin) is dependent on the phosphorylation status of serine 41, the protein kinase c site. The Journal of Neuroscience, 17(10):3515–3524. Hebert-Chatelain, E., Desprez, T., Serrat, R., Bellocchio, L., Soria-Gomez, E., Busquets-Garcia, A., Zottola, A. C. P., Delamarre, A., Cannich, A., Vincent, P., Varilh, M., Robin, L. M., Terral, G., Garćıa-Fernández, M. D., Colavita, M., Mazier, W., Drago, F., Puente, N., Reguero, L., Elezgarai, I., Dupuy, J.-W., Cota, D., Lopez-Rodriguez, M.-L., Barreda-Gómez, G., Massa, F., Grandes, P., Bénard, G., and Marsicano, G. (2016). A cannabinoid link between mitochondria and memory. Nature, 539(7630):555–559. Heifets, B. D. and Castillo, P. E. (2009). Endocannabinoid signaling and long-term synaptic plasticity. Annual Review of Physiology, 71(1):283–306. Henderson-Redmond, A. N., Nealon, C. M., Davis, B. J., Yuill, M. B., Sepulveda, D. E., Blanton, H. L., Piscura, M. K., Zee, M. L., Haskins, C. P., Marcus, D. J., Mackie, K., Guindon, J., and Morgan, D. J. (2020). c-jun n terminal kinase signaling pathways mediate cannabinoid tolerance in an agonist-specific manner. Neuropharmacology, 164:107847. Hens, J. J., Wit, M. D., Boomsma, F., Mercken, M., Oestreicher, A. B., Gispen, W. H., and Graan, P. N. D. (1995). N-terminal-specific anti-b-50 (gap-43) anti- bodies inhibit ca2+-induced noradrenaline release, b-50 phosphorylation and de- phosphorylation, and calmodulin binding. Journal of Neurochemistry, 64:1127–36. Hentges, S. T. (2005). Differential regulation of synaptic inputs by constitutively released endocannabinoids and exogenous cannabinoids. Journal of Neuroscience, 25(42):9746–9751. Herkenham, M., Lynn, A., Johnson, M., Melvin, L., de Costa, B., and Rice, K. (1991). Characterization and localization of cannabinoid receptors in rat brain: a quantitative in vitro autoradiographic study. The Journal of Neuroscience, 11(2):563–583. Herkenham, M., Lynn, A. B., Little, M. D., Johnson, M. R., Melvin, L. S., de Costa, B. R., and Rice, K. C. (1990). Cannabinoid receptor localization in brain. Pro- ceedings of the National Academy of Sciences, 87(5):1932–1936. Hofmann, M. E. and Frazier, C. J. (2013). Marijuana, endocannabinoids, and epi- lepsy: potential and challenges for improved therapeutic intervention. Experi- mental neurology, 244:43–50. Hofmann, M. E., Nahir, B., and Frazier, C. J. (2006). Endocannabinoid-mediated depolarization-induced suppression of inhibition in hilar mossy cells of the rat dentate gyrus. Journal of Neurophysiology, 96(5):2501–2512. Holahan, M. and Routtenberg, A. (2008). The protein kinase c phosphorylation site on gap-43 differentially regulates information storage. Hippocampus, 18(11):1099– 1102. 127 Holahan, M. R. (2017). A shift from a pivotal to supporting role for the growth- associated protein (gap-43) in the coordination of axonal structural and functional plasticity. Frontiers in Cellular Neuroscience, 11(August):1–19. Holahan, M. R., Honegger, K. S., Tabatadze, N., and Routtenberg, A. (2007). Gap-43 gene expression regulates information storage. Learning and Memory, 14(6):407–415. Hooff, C. O. V., Holthuis, J. C., Oestreicher, A. B., Boonstra, J., Graan, P. N. D., and Gispen, W. H. (1989). Nerve growth factor-induced changes in the intracellu- lar localization of the protein kinase c substrate b-50 in pheochromocytoma pc12 cells. Journal of Cell Biology, 108(3):1115–1125. Houser, C. R. (2007). Interneurons of the dentate gyrus: an overview of cell types, terminal fields and neurochemical identity. Progress in Brain Research, 163:217– 32. Howlett, A., Blume, L., and Dalton, G. (2010). Cb1 cannabinoid receptors and their associated proteins. Current Medicinal Chemistry, 17(14):1382–1393. Howlett, A. C. (1985). Cannabinoid inhibition of adenylate cyclase. biochem- istry of the response in neuroblastoma cell membranes. Molecular pharmacology, 27(4):429–36. Howlett, A. C., Breivogel, C. S., Childers, S. R., Deadwyler, S. A., Hampson, R. E., and Porrino, L. J. (2004). Cannabinoid physiology and pharmacology: 30 years of progress. Neuropharmacology, 47:345–358. Howlett, A. C. and Fleming, R. M. (1984). Cannabinoid inhibition of adenylate cy- clase. pharmacology of the response in neuroblastoma cell membranes. Molecular pharmacology, 26(3):532–8. Howlett, A. C., Qualy, J. M., and Khachatrian, L. L. (1986). Involvement of gi in the inhibition of adenylate cyclase by cannabimimetic drugs. Molecular pharmacology, 29(3):307–13. Hu, S. S.-J. and Mackie, K. (2015). Distribution of the endocannabinoid system in the central nervous system. Handbook of Experimental Pharmacology, 231:59–93. Hua, T., Li, X., Wu, L., Iliopoulos-Tsoutsouvas, C., Wang, Y., Wu, M., Shen, L., Johnston, C. A., Nikas, S. P., Song, F., Song, X., Yuan, S., Sun, Q., Wu, Y., Jiang, S., Grim, T. W., Benchama, O., Stahl, E. L., Zvonok, N., Zhao, S., Bohn, L. M., Makriyannis, A., and Liu, Z.-J. (2020). Activation and signaling mechanism revealed by cannabinoid receptor-gi complex structures. Cell, 180(4):655–665.e18. Hua, T., Vemuri, K., Nikas, S. P., Laprairie, R. B., Wu, Y., Qu, L., Pu, M., Korde, A., Jiang, S., Ho, J.-H., Han, G. W., Ding, K., Li, X., Liu, H., Hanson, M. A., Zhao, S., Bohn, L. M., Makriyannis, A., Stevens, R. C., and Liu, Z.-J. (2017). Crystal structures of agonist-bound human cannabinoid receptor cb1. Nature, 547(7664):468–471. 128 Hua, T., Vemuri, K., Pu, M., Qu, L., Han, G. W., Wu, Y., Zhao, S., Shui, W., Li, S., Korde, A., Laprairie, R. B., Stahl, E. L., Ho, J.-H., Zvonok, N., Zhou, H., Kufareva, I., Wu, B., Zhao, Q., Hanson, M. A., Bohn, L. M., Makriyannis, A., Stevens, R. C., and Liu, Z.-J. (2016). Crystal structure of the human cannabinoid receptor cb1. Cell, 167(3):750–762.e14. Hulo, S., Alberi, S., Laux, T., Muller, D., and Caroni, P. (2002). A point mutant of gap-43 induces enhanced short-term and long-term hippocampal plasticity. European Journal of Neuroscience, 15(12):1976–1982. Hunter, S. A., Burstein, S., and Renzulli, L. (1986). Effects of cannabinoids on the activities of mouse brain lipases. Neurochemical Research, 11(9):1273–1288. Huttlin, E. L., Bruckner, R. J., Paulo, J. A., Cannon, J. R., Ting, L., Baltier, K., Colby, G., Gebreab, F., Gygi, M. P., Parzen, H., Szpyt, J., Tam, S., Zarraga, G., Pontano-Vaites, L., Swarup, S., White, A. E., Schweppe, D. K., Rad, R., Erickson, B. K., Obar, R. A., Guruharsha, K. G., Li, K., Artavanis-Tsakonas, S., Gygi, S. P., and Harper, J. W. (2017). Architecture of the human interactome defines protein communities and disease networks. Nature, 545(7655):505–509. Hájková, A., Šárka Techlovská, Dvořáková, M., Chambers, J. N., Kumpošt, J., Hubálková, P., Prezeau, L., and Blahos, J. (2016). Sgip1 alters internalization and modulates signaling of activated cannabinoid receptor 1 in a biased manner. Neuropharmacology, 107:201–214. Häring, M., Enk, V., Rey, A. A., Loch, S., de Azua, I. R., Weber, T., Bartsch, D., Monory, K., and Lutz, B. (2015). Cannabinoid type-1 receptor signaling in central serotonergic neurons regulates anxiety-like behavior and sociability. Frontiers in Behavioral Neuroscience, 9. Igarashi, M., Kawasaki, A., Ishikawa, Y., Honda, A., Okada, M., and Okuda, S. (2020). Phosphoproteomic and bioinformatic methods for analyzing signaling in vertebrate axon growth and regeneration. Journal of Neuroscience Methods, 339:108723. Igarashi, M., Li, W., Sudo, Y., and Fishman, M. (1995). Ligand-induced growth cone collapse: amplification and blockade by variant gap-43 peptides. The Journal of Neuroscience, 15(8):5660–5667. Inoue, A., Raimondi, F., Kadji, F. M. N., Singh, G., Kishi, T., Uwamizu, A., Ono, Y., Shinjo, Y., Ishida, S., Arang, N., Kawakami, K., Gutkind, J. S., Aoki, J., and Russell, R. B. (2019). Illuminating g-protein-coupling selectivity of gpcrs. Cell, 177:1933–1947. Isokawa, M. and Alger, B. E. (2005). Retrograde endocannabinoid regulation of gabaergic inhibition in the rat dentate gyrus granule cell. The Journal of Physiology, 567(3):1001–1010. Ivins, K. J., Neve, K. A., Feller, D. J., Fidel, S. A., and Neve, R. L. (1993). Anti- sense gap-43 inhibits the evoked release of dopamine from pc12 cells. Journal of Neurochemistry, 60(2):626–633. 129 Jensen, K. R., Berthoux, C., Nasrallah, K., and Castillo, P. E. (2021). Multiple can- nabinoid signaling cascades powerfully suppress recurrent excitation in the hippo- campus. Proceedings of the National Academy of Sciences, 118(4):e2017590118. Jimenez-Blasco, D., Busquets-Garcia, A., Hebert-Chatelain, E., Serrat, R., Vicente- Gutierrez, C., Ioannidou, C., Gómez-Sotres, P., Lopez-Fabuel, I., Resch-Beusher, M., Resel, E., Arnouil, D., Saraswat, D., Varilh, M., Cannich, A., Julio-Kalajzic, F., Ŕıo, I. B.-D., Almeida, A., Puente, N., Achicallende, S., Lopez-Rodriguez, M.-L., Jollé, C., Déglon, N., Pellerin, L., Josephine, C., Bonvento, G., Panatier, A., Lutz, B., Piazza, P.-V., Guzmán, M., Bellocchio, L., Bouzier-Sore, A.-K., Grandes, P., Bolaños, J. P., and Marsicano, G. (2020). Glucose metabolism links astroglial mitochondria to cannabinoid effects. Nature, 583(7817):603–608. Jin, K., Xie, L., Kim, S. H., Parmentier-Batteur, S., Sun, Y., Mao, X. O., Childs, J., and Greenberg, D. A. (2004). Defective adult neurogenesis in cb1 cannabinoid receptor knockout mice. Molecular Pharmacology, 66(2):204–208. Jin, M. (2005). Ca2+-dependent regulation of rho gtpases triggers turning of nerve growth cones. Journal of Neuroscience, 25(9):2338–2347. jing Li, X., ling Liu, J., sheng Gao, D., yan Wan, W., Yang, X., tao Li, Y., tao Chang, H., Chen, L., qing Wang, C., and Zhao, J. (2016). Single-step affinity and cost-effective purification of recombinant proteins using the sepharose-binding lectin-tag from the mushroom laetiporus sulphureus as fusion partner. Protein Expression and Purification, 119:51–56. Jordan, C. J. and Xi, Z.-X. (2019). Progress in brain cannabinoid cb2 receptor research: From genes to behavior. Neuroscience Biobehavioral Reviews, 98:208– 220. Josselyn, S. A., Köhler, S., and Frankland, P. W. (2015). Finding the engram. Nature Reviews Neuroscience, 16(9):521–534. Julian, L. and Olson, M. F. (2014). Rho-associated coiled-coil containing kinases (rock): structure, regulation, and functions. Small GTPases, 5(2):e29846. Jung, K.-M., Astarita, G., Zhu, C., Wallace, M., Mackie, K., and Piomelli, D. (2007). A key role for diacylglycerol lipase- in metabotropic glutamate receptor-dependent endocannabinoid mobilization. Molecular Pharmacology, 72(3):612–621. Kamatchi, G. L., Tiwari, S. N., Durieux, M. E., and Lynch, C. (2000). Effects of volatile anesthetics on the direct and indirect protein kinase c-mediated en- hancement of alpha1e-type ca(2+) current in xenopus oocytes. The Journal of pharmacology and experimental therapeutics, 293(2):360–9. Kano, M., Ohno-Shosaku, T., Hashimotodani, Y., Uchigashima, M., and Watanabe, M. (2009). Endocannabinoid-mediated control of synaptic transmission. Physiolo- gical Reviews, 89(1):309–380. Karler, R., Cely, W., and Turkanis, S. A. (1974). Anticonvulsant properties of delta 9-tetrahydrocannabinol and other cannabinoids. Life sciences, 15(5):931–47. 130 Karlócai, M. R., Tóth, K., Watanabe, M., Ledent, C., Juhász, G., Freund, T. F., and Maglóczky, Z. (2011). Redistribution of cb1 cannabinoid receptors in the acute and chronic phases of pilocarpine-induced epilepsy. PLoS ONE, 6(11):e27196. Karns, L., Ng, S., Freeman, J., and Fishman, M. (1987). Cloning of complementary dna for gap-43, a neuronal growth-related protein. Science, 236(4801):597–600. Katona, I. (2006). Molecular composition of the endocannabinoid system at glutamatergic synapses. Journal of Neuroscience, 26(21):5628–5637. Katona, I. and Freund, T. F. (2008). Endocannabinoid signaling as a synaptic circuit breaker in neurological disease. Nature Medicine, 14(9):923–930. Katona, I. and Freund, T. F. (2012). Multiple functions of endocannabinoid signaling in the brain. Annual Review of Neuroscience, 35(1):529–558. Katona, I., Sperlágh, B., Maglóczky, Z., Sántha, E., Köfalvi, A., Czirják, S., Mackie, K., Vizi, E., and Freund, T. (2000). Gabaergic interneurons are the targets of cannabinoid actions in the human hippocampus. Neuroscience, 100(4):797–804. Katona, I., Sperlágh, B., Sık, A., Käfalvi, A., Vizi, E. S., Mackie, K., and Freund, T. F. (1999). Presynaptically located cb1 cannabinoid receptors regulate gaba release from axon terminals of specific hippocampal interneurons. The Journal of Neuroscience, 19(11):4544–4558. Kawamura, Y. (2006). The cb1 cannabinoid receptor is the major cannabinoid re- ceptor at excitatory presynaptic sites in the hippocampus and cerebellum. Journal of Neuroscience, 26(11):2991–3001. Kawasaki, A., Okada, M., Tamada, A., Okuda, S., Nozumi, M., Ito, Y., Kobayashi, D., Yamasaki, T., Yokoyama, R., Shibata, T., Nishina, H., Yoshida, Y., Fujii, Y., Takeuchi, K., and Igarashi, M. (2018). Growth cone phosphoproteomics reveals that gap-43 phosphorylated by jnk is a marker of axon growth and regeneration. iScience, 4:190–203. Kearn, C. S., Blake-Palmer, K., Daniel, E., Mackie, K., and Glass, M. (2005). Concurrent stimulation of cannabinoid cb1 and dopamine d2 receptors enhances heterodimer formation: A mechanism for receptor cross-talk? Molecular Phar- macology, 67(5):1697–1704. Keeler, M. H. and Reifler, C. B. (1967). Grand mal convulsions subsequent to marijuana use. case report. Diseases of the nervous system, 28(7 Pt 1):474–5. Kempermann, G., Song, H., and Gage, F. H. (2015). Neurogenesis in the adult hippocampus. Cold Spring Harbor Perspectives in Biology, 7(9):a018812. Kishimoto, Y. (2006). Endogenous cannabinoid signaling through the cb1 receptor is essential for cerebellum-dependent discrete motor learning. Journal of Neuros- cience, 26(34):8829–8837. Kokubu, H. and Lim, J. (2014). X-gal staining on adult mouse brain sections. Bio Protocol, 4(5):e1064. 131 Koppel, B. S., Fife, T., and Youssof, S. (2014). Systematic review : Efficacy and safety of medical marijuana in selected neurologic disorders report of the guideline development subcommittee of the american academy of neurology. Neurology, 82(17):1556–63. Kreitzer, A. C. (2005). Dopamine modulation of state-dependent endocannabin- oid release and long-term depression in the striatum. Journal of Neuroscience, 25(45):10537–10545. Kreitzer, A. C. and Regehr, W. G. (2001). Retrograde inhibition of presynaptic calcium influx by endogenous cannabinoids at excitatory synapses onto purkinje cells. Neuron, 29(3):717–727. Kristjansson, G. I., Zwiers, H., Oestreicher, A. B., and Gispen, W. H. (1982). Evid- ence that the synaptic phosphoprotein b-50 is localized exclusively in nerve tissue. Journal of Neurochemistry, 39(2):371–378. Kruger, L., Bendotti, C., Rivolta, R., and Samanin, R. (1993). Distribution of gap-43 mrna in the adult rat brain. Journal of Comparative Neurology, 333(3):417–434. Kumagai-Tohda, C., Tohda, M., and Nomura, Y. (2006). Increase in neurite form- ation and acetylene line release by transfection of growth-associated protein-43 cdna into ng108-15 cells. Journal of Neurochemistry, 61(2):526–532. Kumar, K. K., Shalev-Benami, M., Robertson, M. J., Hu, H., Banister, S. D., Hollingsworth, S. A., Latorraca, N. R., Kato, H. E., Hilger, D., Maeda, S., Weis, W. I., Farrens, D. L., Dror, R. O., Malhotra, S. V., Kobilka, B. K., and Skiniotis, G. (2019). Structure of a signaling cannabinoid receptor 1-g protein complex. Cell, 176(3):448–458.e12. Laprairie, R. B., Bagher, A. M., Kelly, M. E. M., Dupré, D. J., and Denovan-Wright, E. M. (2014). Type 1 cannabinoid receptor ligands display functional selectivity in a cell culture model of striatal medium spiny projection neurons. Journal of Biological Chemistry, 289(36):24845–24862. Larimer, P. and Strowbridge, B. W. (2008). Nonrandom local circuits in the dentate gyrus. Journal of Neuroscience, 28(47):12212–12223. Larsson, C. (2006). Protein kinase c and the regulation of the actin cytoskeleton. Cellular Signalling, 18(3):276–284. Lau, T. and Schloss, P. (2008). The cannabinoid cb1 receptor is expressed on sero- tonergic and dopaminergic neurons. European Journal of Pharmacology, 578(2- 3):137–141. Lauckner, J. E., Hille, B., and Mackie, K. (2005). The cannabinoid agonist win55,212-2 increases intracellular calcium via cb1 receptor coupling to gq/11 g proteins. Proceedings of the National Academy of Sciences, 102(52):19144–19149. Lautermilch, N. J. and Spitzer, N. C. (2000). Regulation of calcineurin by growth cone calcium waves controls neurite extension. The Journal of neuroscience : the official journal of the Society for Neuroscience, 20(1):315–25. 132 Laux, T., Fukami, K., Thelen, M., Golub, T., Frey, D., and Caroni, P. (2000). Gap43, marcks, and cap23 modulate pi(4,5)p2 at plasmalemmal rafts, and regulate cell cortex actin dynamics through a common mechanism. Journal of Cell Biology, 149(7):1455–1472. Lee, S.-E., Jeong, S., Lee, U., and Chang, S. (2019). Sgip1 functions as a selective en- docytic adaptor for the internalization of synaptotagmin 1 at synapses. Molecular Brain, 12(1):41. Lerner, T. N. and Kreitzer, A. C. (2012). Rgs4 is required for dopaminergic control of striatal ltd and susceptibility to parkinsonian motor deficits. Neuron, 73(2):347– 359. Leu, B., Koch, E., and Schmidt, J. T. (2010). Gap43 phosphorylation is critical for growth and branching of retinotectal arbors in zebrafish. Developmental Neuro- biology, 70(13):897–911. Lin, L.-H., Bock, S., Carpenter, K., Rose, M., and Norden, J. J. (1992). Synthesis and transport of gap-43 in entorhinal cortex neurons and perforant pathway during lesion-induced sprouting and reactive synaptogenesis. Molecular Brain Research, 14(1-2):147–153. Lisman, J. E., Talamini, L. M., and Raffone, A. (2005). Recall of memory sequences by interaction of the dentate and ca3: A revised model of the phase precession. Neural Networks, 18(9):1191–1201. Liu, J., Gao, B., Mirshashi, F., Sanyal, A. J., Khanolkar, A. D., Makriyannis, A., and Kunos, G. (2000). Functional cb1 cannabinoid receptors in human vascular endothelial cells. Biochemical Journal, 346(3):835. Liu, J., Wang, L., Harvey-White, J., Osei-Hyiaman, D., Razdan, R., Gong, Q., Chan, A. C., Zhou, Z., Huang, B. X., Kim, H.-Y., and Kunos, G. (2006). A biosynthetic pathway for anandamide. Proceedings of the National Academy of Sciences, 103(36):13345–13350. Llano, I., Leresche, N., and Marty, A. (1991). Calcium entry increases the sensitiv- ity of cerebellar purkinje cells to applied gaba and decreases inhibitory synaptic currents. Neuron, 6(4):565–574. Lonart, G. and Südhof, T. C. (2000). Assembly of snare core complexes prior to neurotransmitter release sets the readily releasable pool of synaptic vesicles. Journal of Biological Chemistry, 275(36):27703–27707. Lovinger, D. M. (2008). Presynaptic modulation by endocannabinoids. Handbook of Experimental Pharmacology, (184):435–77. Lovinger, D. M., Akers, R. F., Nelson, R. B., Barnes, C. A., McNaughton, B. L., and Routtenberg, A. (1985). A selective increase in phosphorylation of protein f1, a protein kinase c substrate, directly related to three day growth of long term synaptic enhancement. Brain Research, 343(1):137–143. 133 Lovinger, D. M., Colley, P. A., Akers, R. F., Nelson, R. B., and Routtenberg, A. (1986). Direct relation of long-term synaptic potentiation to phosphorylation of membrane protein f1, a substrate for membrane protein kinase c. Brain Research, 399(2):205–211. Lu, H.-C. and Mackie, K. (2016). An introduction to the endogenous cannabinoid system. Biological psychiatry, 79(7):516–25. Ludanyi, A., Eross, L., Czirjak, S., Vajda, J., Halasz, P., Watanabe, M., Palkovits, M., Magloczky, Z., Freund, T. F., and Katona, I. (2008). Downregulation of the cb1 cannabinoid receptor and related molecular elements of the endocannabinoid system in epileptic human hippocampus. Journal of Neuroscience, 28(12):2976– 2990. Luo, Y. and Vallano, M. L. (1995). Arachidonic acid, but not sodium nitroprusside, stimulates presynaptic protein kinase c and phosphorylation of gap-43 in rat hip- pocampal slices and synaptosomes. Journal of Neurochemistry, 64(4):1808–1818. Lutz, B. (2004). On-demand activation of the endocannabinoid system in the control of neuronal excitability and epileptiform seizures. Biochemical Pharmacology, 68(9):1691–1698. Lutz, B., Marsicano, G., Maldonado, R., and Hillard, C. J. (2015). The endocan- nabinoid system in guarding against fear, anxiety and stress. Nature Reviews Neuroscience, 16(12):705–718. Lyons, E. L., Leone-Kabler, S., Kovach, A. L., Thomas, B. F., and Howlett, A. C. (2020). Cannabinoid receptor subtype influence on neuritogenesis in human sh- sy5y cells. Molecular and Cellular Neuroscience, 109:103566. Maa, E. and Figi, P. (2014). The case for medical marijuana in epilepsy. Epilepsia, 55(6):783–786. Maccarrone, M., Battista, N., and Centonze, D. (2007). The endocannabinoid path- way in huntington’s disease: A comparison with other neurodegenerative diseases. Progress in Neurobiology, 81(5-6):349–379. Maccarrone, M., Guzmán, M., Mackie, K., Doherty, P., and Harkany, T. (2014). Programming of neural cells by (endo)cannabinoids: from physiological rules to emerging therapies. Nature reviews. Neuroscience, 15(12):786–801. Macek, T. A., Winder, D. G., Gereau, R. W., Ladd, C. O., and Conn, P. J. (1996). Differential involvement of group ii and group iii mglurs as autoreceptors at lateral and medial perforant path synapses. Journal of Neurophysiology, 76(6):3798–3806. Mackie, K. (2005). Cannabinoid receptor homo- and heterodimerization. Life Sci- ences, 77(14):1667–1673. Mackie, K. and Hille, B. (1992). Cannabinoids inhibit n-type calcium channels in neuroblastoma-glioma cells. Proceedings of the National Academy of Sciences, 89(9):3825–3829. 134 Maglio, L. E., Noriega-Prieto, J. A., Maraver, M. J., and de Sevilla, D. F. (2018). Endocannabinoid-dependent long-term potentiation of synaptic transmission at rat barrel cortex. Cerebral Cortex, 28(5):1568–1581. Mai, P., Tian, L., Yang, L., Wang, L., Yang, L., and Li, L. (2015). Cannabinoid receptor 1 but not 2 mediates macrophage phagocytosis by g ()i/o /rhoa/rock signaling pathway. Journal of Cellular Physiology, 230(7):1640–1650. Maier, D. L., Mani, S., Donovan, S. L., Soppet, D., Tessarollo, L., Mccasland, J. S., and Meiri, K. F. (1999). Disrupted cortical map and absence of cortical barrels in growth-associated protein (gap)-43 knockout mice. Proceedings of the National Academy of Sciences of the United States of America, 96(16):9397–9402. Malenka, R. C., Madison, D. V., and Nicoll, R. A. (1986). Potentiation of synaptic transmission in the hippocampus by phorbol esters. Nature, 321(6066):175–177. Marinelli, S., Pacioni, S., Cannich, A., Marsicano, G., and Bacci, A. (2009). Self- modulation of neocortical pyramidal neurons by endocannabinoids. Nature Neur- oscience, 12(12):1488–1490. Marsicano, G., Goodenough, S., Monory, K., Hermann, H., Eder, M., Cannich, A., Azad, S. C., Cascio, M. G., Gutierrez, S. O., van der Stelt, M., Lopez-Rodriguez, M. L., Casanova, E., Schutz, G., Zieglgansberger, W., Marzo, V. D., Behl, C., and Lutz, B. (2003). Cb1 cannabinoid receptors and on-demand defense against excitotoxicity.[comment]. Science, 302(5642):84–88. Marsicano, G. and Lutz, B. (1999). Expression of the cannabinoid receptor cb1 in distinct neuronal subpopulations in the adult mouse forebrain. European Journal of Neuroscience, 11(12):4213–4225. Marsicano, G., Wotjak, C. T., Azad, S. C., Bisogno, T., Rammes, G., Cascio, M. G., Hermann, H., Tang, J., Hofmann, C., Zieglgänsberger, W., Marzo, V. D., and Lutz, B. (2002). The endogenous cannabinoid system controls extinction of aversive memories. Nature, 418(6897):530–534. Martini, L., Thompson, D., Kharazia, V., and Whistler, J. L. (2010). Differential regulation of behavioral tolerance to win55,212-2 by gasp1. Neuropsychopharma- cology, 35(6):1363–1373. Mart́ın, R., Durroux, T., Ciruela, F., Torres, M., Pin, J.-P., and Sánchez-Prieto, J. (2010). The metabotropic glutamate receptor mglu7 activates phospholipase c, translocates munc-13-1 protein, and potentiates glutamate release at cerebrocor- tical nerve terminals. Journal of Biological Chemistry, 285(23):17907–17917. Marzo, V. D. and Piscitelli, F. (2015). The endocannabinoid system and its modu- lation by phytocannabinoids. Neurotherapeutics, 12(4):692–698. Marzo, V. D., Stella, N., and Zimmer, A. (2015). Endocannabinoid signalling and the deteriorating brain. Nature Reviews Neuroscience, 16(1):30–42. Mascaro, A. L. A., Cesare, P., Sacconi, L., Grasselli, G., Mandolesi, G., MacO, B., Knott, G. W., Huang, L., Paola, V. D., Strata, P., and Pavone, F. S. (2013). 135 In vivo single branch axotomy induces gap-43-dependent sprouting and synaptic remodeling in cerebellar cortex. Proceedings of the National Academy of Sciences of the United States of America, 110(26):10824–10829. Mascia, F., Klotz, L., Lerch, J., Ahmed, M. H., Zhang, Y., and Enz, R. (2017). Crip1a inhibits endocytosis of g-protein coupled receptors activated by endocan- nabinoids and glutamate by a common molecular mechanism. Journal of Neuro- chemistry, 141(4):577–591. Mato, S., Lafourcade, M., Robbe, D., Bakiri, Y., and Manzoni, O. J. (2008). Role of the cyclic-amp/pka cascade and of p/q-type ca++ channels in endocannabinoid- mediated long-term depression in the nucleus accumbens. Neuropharmacology, 54(1):87–94. Matsuda, L. A., Lolait, S. J., Brownstein, M. J., Young, A. C., and Bonner, T. I. (1990). Structure of a cannabinoid receptor and functional expression of the cloned cdna. Nature, 346(6284):561–564. Mattheus, T., Kukla, K., Zimmermann, T., Tenzer, S., and Lutz, B. (2016). Cell type-specific tandem affinity purification of the mouse hippocampal cb1 receptor- associated proteome. Journal of Proteome Research, 15(10):3585–3601. McFadden, M. H., Xu, H., Cui, Y., Piskorowski, R. A., Leterrier, C., Zala, D., Venance, L., Chevaleyre, V., and Lenkei, Z. (2018). Actomyosin-mediated nano- structural remodeling of the presynaptic vesicle pool by cannabinoids induces long-term depression. bioRxiv preprint, pages 1–18. McNamara, R. K. and Lenox, R. H. (1997). Comparative distribution of myris- toylated alanine-rich c kinase substrate (marcks) and f1/gap-43 gene expression in the adult rat brain. The Journal of Comparative Neurology, 379(1):48–71. McNamara, R. K. and Routtenberg, A. (1995). Nmda receptor blockade prevents kainate induction of protein f1/gap-43 mrna in hippocampal granule cells and subsequent mossy fiber sprouting in the rat. Molecular Brain Research, 33(1):22– 28. Meberg, P. and Routtenberg, A. (1991). Selective expression of protein f1/(gap- 43) mrna in pyramidal but not granule cells of the hippocampus. Neuroscience, 45(3):721–733. Meberg, P. J., Gall, C. M., and Routtenberg, A. (1993). Induction of f1/gap-43 gene expression in hippocampal granule cells after seizures [corrected]. Brain research. Molecular brain research, 17(3-4):295–9. Meberg, P. J., Valcourt, E. G., and Routtenberg, A. (1995). Protein f1/gap-43 and pkc gene expression patterns in hippocampus are altered 1–2 h after ltp. Molecular Brain Research, 34(2):343–346. Mechoulam, R., Ben-Shabat, S., Hanus, L., Ligumsky, M., Kaminski, N. E., Schatz, A. R., Gopher, A., Almog, S., Martin, B. R., Compton, D. R., Pertwee, R. G., Griffin, G., Bayewitch, M., Barg, J., and Vogel, Z. (1995). Identification of an endogenous 2-monoglyceride, present in canine gut, that binds to cannabinoid receptors. Biochemical Pharmacology, 50(1):83–90. 136 Mechoulam, R., Fride, E., Hanu, L., Sheskin, T., Bisogno, T., Marzo, V. D., Baye- witch, M., and Vogel, Z. (1997). Anandamide may mediate sleep induction. Nature, 389(6646):25–26. Mechoulam, R., Hanuš, L. O., Pertwee, R., and Howlett, A. C. (2014). Early phytocannabinoid chemistry to endocannabinoids and beyond. Nature Reviews Neuroscience, 15(11):757–764. Mechoulam, R. and Parker, L. A. (2013). The endocannabinoid system and the brain. Annual Review of Psychology, 64(1):21–47. Meiri, K., Willard, M., and Johnson, M. (1988). Distribution and phosphorylation of the growth-associated protein gap- 43 in regenerating sympathetic neurons in culture. The Journal of Neuroscience, 8(7):2571–2581. Meiri, K. F., Bickerstaff, L. E., and Schwob, J. E. (1991). Monoclonal antibodies show that kinase c phosphorylation of gap-43 during axonogenesis is both spatially and temporally restricted in vivo. The Journal of cell biology, 112(5):991–1005. Melis, M. (2004). Prefrontal cortex stimulation induces 2-arachidonoyl-glycerol- mediated suppression of excitation in dopamine neurons. Journal of Neuroscience, 24(47):10707–10715. Merino-Gracia, J., Costas-Insua, C., Ángeles Canales, M., and Rodŕıguez-Crespo, I. (2017). Insights into the c-terminal peptide binding specificity of the pdz domain of neuronal nitric-oxide synthase. Journal of Biological Chemistry, 291(22):11581– 11595. Mizuno, K. (2013). Signaling mechanisms and functional roles of cofilin phosphoryla- tion and dephosphorylation. Cellular Signalling, 25(2):457–469. Molina-Holgado, E., Vela, J. M., Arévalo-Martın, A., Almazán, G., Molina-Holgado, F., Borrell, J., and Guaza, C. (2002). Cannabinoids promote oligodendrocyte pro- genitor survival: Involvement of cannabinoid receptors and phosphatidylinositol-3 kinase/akt signaling. The Journal of Neuroscience, 22(22):9742–9753. Monday, H. R., Bourdenx, M., Jordan, B. A., and Castillo, P. E. (2020). Cb1- receptor-mediated inhibitory ltd triggers presynaptic remodeling via protein syn- thesis and ubiquitination. eLife, 9(e54812). Monory, K., Blaudzun, H., Massa, F., Kaiser, N., Lemberger, T., Schütz, G., Wotjak, C. T., Lutz, B., and Marsicano, G. (2007). Genetic dissection of behavioural and autonomic effects of delta(9)-tetrahydrocannabinol in mice. PLoS Biology, 5(10):e269. Monory, K., Massa, F., Egertová, 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., Wölfel, B., Dodt, H.- U., Zieglgänsberger, W., Wotjak, C. T., Mackie, K., Elphick, M. R., Marsicano, G., and Lutz, B. (2006). The endocannabinoid system controls key epileptogenic circuits in the hippocampus. Neuron, 51(4):455–466. 137 Morena, M., Patel, S., Bains, J. S., and Hill, M. N. (2016). Neurobiological interac- tions between stress and the endocannabinoid system. Neuropsychopharmacology, 41(1):80–102. Moreno, E., Andradas, C., Medrano, M., Caffarel, M. M., Pérez-Gómez, E., Blasco- Benito, S., Gómez-Cañas, M., Pazos, M. R., Irving, A. J., Llúıs, C., Canela, E. I., Fernández-Ruiz, J., Guzmán, M., McCormick, P. J., and Sánchez, C. (2014). Targeting cb2-gpr55 receptor heteromers modulates cancer cell signaling. Journal of Biological Chemistry, 289(32):21960–21972. Moreno, E., Chiarlone, A., Medrano, M., Puigdelĺıvol, M., Bibic, L., Howell, L. A., Resel, E., Puente, N., Casarejos, M. J., Perucho, J., Botta, J., Suelves, N., Ciruela, F., Ginés, S., Galve-Roperh, I., Casadó, V., Grandes, P., Lutz, B., Monory, K., Canela, E. I., Llúıs, C., McCormick, P. J., and Guzmán, M. (2018). Singular loc- ation and signaling profile of adenosine a2a-cannabinoid cb1 receptor heteromers in the dorsal striatum. Neuropsychopharmacology, 43(5):964–977. Moss, D. J., Fernyhough, P., Chapman, K., Baizer, L., Bray, D., and Allsopp, T. (1990). Chicken growth-associated protein gap-43 is tightly bound to the actin- rich neuronal membrane skeleton. Journal of Neurochemistry, 54(3):729–736. Mukhopadhyay, S., McIntosh, H. H., Houston, D. B., and Howlett, A. C. (2000). The cb(1) cannabinoid receptor juxtamembrane c-terminal peptide confers activation to specific g proteins in brain. Molecular pharmacology, 57(1):162–70. Munro, S., Thomas, K. L., and Abu-Shaar, M. (1993). Molecular characterization of a peripheral receptor for cannabinoids. Nature, 365(6441):61–65. Murataeva, N., Straiker, A., and Mackie, K. (2014). Parsing the players: 2- arachidonoylglycerol synthesis and degradation in the cns. British Journal of Pharmacology, 171(6):1379–1391. Naffah-Mazzacoratti, M. G., Funke, M. G., Sanabria, E. R., and Cavalheiro, E. A. (1999). Growth-associated phosphoprotein expression is increased in the supra- granular regions of the dentate gyrus following pilocarpine-induced seizures in rats. Neuroscience, 91(2):485–492. Nakamura, F., Strittmatter, P., and Strittmatter, S. M. (2002). Gap-43 augmenta- tion of g protein-mediated signal transduction is regulated by both phosphoryla- tion and palmitoylation. Journal of Neurochemistry, 70(3):983–992. Namgung, U., Matsuyama, S., and Routtenberg, A. (1997). Long-term potentiation activates the gap-43 promoter: Selective participation of hippocampal mossy cells. Proceedings of the National Academy of Sciences of the United States of America, 94(21):11675–11680. Narushima, M., Uchigashima, M., Hashimoto, K., Watanabe, M., and Kano, M. (2006). Depolarization-induced suppression of inhibition mediated by endocan- nabinoids at synapses from fast-spiking interneurons to medium spiny neurons in the striatum. European Journal of Neuroscience, 24(8):2246–2252. 138 Navarrete, M. and Araque, A. (2008). Endocannabinoids mediate neuron-astrocyte communication. Neuron, 57(6):883–893. Navarrete, M. and Araque, A. (2010). Endocannabinoids potentiate synaptic trans- mission through stimulation of astrocytes. Neuron, 68(1):113–126. Nemes, A. D., Ayasoufi, K., Ying, Z., Zhou, Q. G., Suh, H., and Najm, I. M. (2017). Growth associated protein 43 (gap-43) as a novel target for the diagnosis, treatment and prevention of epileptogenesis. Scientific Reports, 7(1):1–13. Neve, R. L., Coopersmith, R., McPhie, D. L., Santeufemio, C., Pratt, K. G., Murphy, C. J., and Lynn, S. D. (1998). The neuronal growth-associated protein gap-43 interacts with rabaptin-5 and participates in endocytosis. Journal of Neuroscience, 18(19):7757–7767. Nguyen, P. T., Schmid, C. L., Raehal, K. M., Selley, D. E., Bohn, L. M., and Sim-Selley, L. J. (2012). -arrestin2 regulates cannabinoid cb1 receptor signaling and adaptation in a central nervous system region–dependent manner. Biological Psychiatry, 71(8):714–724. Nie, J. and Lewis, D. L. (2001). Structural domains of the cb1 cannabinoid receptor that contribute to constitutive activity and g-protein sequestration. Journal of Neuroscience, 21:8758–64. Niehaus, J. L., Liu, Y., Wallis, K. T., Egertová, M., Bhartur, S. G., Mukhopadhyay, S., Shi, S., He, H., Selley, D. E., Howlett, A. C., Elphick, M. R., and Lewis, D. L. (2007). Cb 1 cannabinoid receptor activity is modulated by the cannabinoid receptor interacting protein crip 1a. Molecular Pharmacology, 72(6):1557–1566. Nithipatikom, K., Gomez-Granados, A. D., Tang, A. T., Pfeiffer, A. W., Williams, C. L., and Campbell, W. B. (2012). Cannabinoid receptor type 1 (cb1) activation inhibits small gtpase rhoa activity and regulates motility of prostate carcinoma cells. Endocrinology, 153(1):29–41. Njoo, C., Agarwal, N., Lutz, B., and Kuner, R. (2015). The cannabinoid receptor cb1 interacts with the wave1 complex and plays a role in actin dynamics and structural plasticity in neurons. PLoS Biology, 13(10):1–36. Nogueras-Ortiz, C. and Yudowski, G. A. (2016). The multiple waves of cannabinoid 1 receptor signaling. Molecular Pharmacology, 90(5):620–626. Nomura, D. K., Morrison, B. E., Blankman, J. L., Long, J. Z., Kinsey, S. G., Marcondes, M. C. G., Ward, A. M., Hahn, Y. K., Lichtman, A. H., Conti, B., and Cravatt, B. F. (2011). Endocannabinoid hydrolysis generates brain prostaglandins that promote neuroinflammation. Science, 334(6057):809–813. Nýıri, G., Szabadits, E., Cserép, C., Mackie, K., Shigemoto, R., and Freund, T. F. (2005). Gaba b and cb 1 cannabinoid receptor expression identifies two types of septal cholinergic neurons. European Journal of Neuroscience, 21(11):3034–3042. O’Connell, B. K., Gloss, D., and Devinsky, O. (2017). Cannabinoids in treatment- resistant epilepsy: A review. Epilepsy Behavior, 70:341–348. 139 Oddi, S., Stepniewski, T. M., Totaro, A., Selent, J., Scipioni, L., Dufrusine, B., Fezza, F., Dainese, E., and Maccarrone, M. (2017). Palmitoylation of cysteine 415 of cb 1 receptor affects ligand-stimulated internalization and selective interaction with membrane cholesterol and caveolin 1. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids, 1862(5):523–532. Oddi, S., Totaro, A., Scipioni, L., Dufrusine, B., Stepniewski, T. M., Selent, J., Maccarrone, M., and Dainese, E. (2018). Role of palmitoylation of cysteine 415 in functional coupling cb 1 receptor to g i2 protein. Biotechnology and Applied Biochemistry, 65(1):16–20. Ohno-Shosaku, T. and Kano, M. (2014). Endocannabinoid-mediated retrograde modulation of synaptic transmission. Current Opinion in Neurobiology, 29:1–8. Ohno-Shosaku, T., Shosaku, J., Tsubokawa, H., and Kano, M. (2002). Cooperative endocannabinoid production by neuronal depolarization and group i metabotropic glutamate receptor activation. European Journal of Neuroscience, 15(6):953–961. Onaivi, E. S., Chaudhuri, G., Abaci, A. S., Parker, M., Manier, D. H., Martin, P. R., and Hubbard, J. R. (1999). Expression of cannabinoid receptors and their gene transcripts in human blood cells. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 23(6):1063–1077. Ong, W. and Mackie, K. (1999). A light and electron microscopic study of the cb1 cannabinoid receptor in primate brain. Neuroscience, 92(4):1177–1191. Oropeza, V. C., Mackie, K., and Bockstaele, E. J. V. (2007). Cannabinoid receptors are localized to noradrenergic axon terminals in the rat frontal cortex. Brain Research, 1127:36–44. O’Sullivan, S. E. (2007). Cannabinoids go nuclear: evidence for activation of peroxisome proliferator-activated receptors. British Journal of Pharmacology, 152(5):576–582. Oudin, M. J., Hobbs, C., and Doherty, P. (2011). Dagl-dependent endocannabinoid signalling: roles in axonal pathfinding, synaptic plasticity and adult neurogenesis. European Journal of Neuroscience, 34(10):1634–1646. Ozaita, A., Puighermanal, E., and Maldonado, R. (2007). Regulation of pi3k/akt/gsk-3 pathway by cannabinoids in the brain. Journal of Neurochem- istry, 102:1105–14. Palazuelos, J., Aguado, T., Egia, A., Mechoulam, R., Guzmán, M., Galve-Roperh, I., Palazuelos, J., Aguado, T., Egia, A., Mechoulam, R., Guzmán, M., and Galve- Roperh, I. (2006). Non-psychoactive cb 2 cannabinoid agonists stimulate neural progenitor proliferation. The FASEB Journal, 20(13):2405–2407. Pamplona, F. A., Ferreira, J., de Lima, O. M., Duarte, F. S., Bento, A. F., Forner, S., Villarinho, J. G., Bellocchio, L., Wotjak, C. T., Lerner, R., Monory, K., Lutz, B., Canetti, C., Matias, I., Calixto, J. B., Marsicano, G., Guimaraes, M. Z. P., and Takahashi, R. N. (2012). Anti-inflammatory lipoxin a4 is an endogenous allosteric enhancer of cb1 cannabinoid receptor. Proceedings of the National Academy of Sciences, 109(51):21134–21139. 140 Pan, B., Hillard, C. J., and s. Liu, Q. (2008). Endocannabinoid signaling medi- ates cocaine-induced inhibitory synaptic plasticity in midbrain dopamine neurons. Journal of Neuroscience, 28(6):1385–1397. Pan, X., Ikeda, S. R., and Lewis, D. L. (1996). Rat brain cannabinoid receptor modulates n-type ca2+ channels in a neuronal expression system. Molecular phar- macology, 49(4):707–14. Pandey, P., Roy, K. K., and Doerksen, R. J. (2020). Negative allosteric modulators of cannabinoid receptor 2: protein modeling, binding site identification and mo- lecular dynamics simulations in the presence of an orthosteric agonist. Journal of Biomolecular Structure and Dynamics, 38(1):32–47. Panikashvili, D., Simeonidou, C., Ben-Shabat, S., Hanuš, L., a Breuer, Mechoulam, R., and Shohami, E. (2001). An endogenous cannabinoid (2-ag) is neuroprotective after brain injury. Nature, 413(6855):527–531. Panlilio, L., Goldberg, S., and Justinova, Z. (2015). Cannabinoid abuse and ad- diction: Clinical and preclinical findings. Clinical Pharmacology Therapeutics, 97(6):616–627. Patterson, S. I. and Skene, J. H. P. (1999). A shift in protein s-palmitoylation, with persistence of growth-associated substrates, marks a critical period for synaptic plasticity in developing brain. Journal of Neurobiology, 39(3):423–437. Perrone-Bizzozero, N. I., Cansino, V. V., and Kohn, D. T. (1993). Posttranscrip- tional regulation of gap-43 gene expression in pc12 cells through protein kinase c-dependent stabilization of the mrna. The Journal of Cell Biology, 120(5):1263– 1270. Pertwee, R. G. (2012). Targeting the endocannabinoid system with cannabinoid receptor agonists: pharmacological strategies and therapeutic possibilities. Philo- sophical Transactions of the Royal Society B: Biological Sciences, 367(1607):3353– 3363. Pertwee, R. G., Howlett, A. C., Abood, M. E., Alexander, S. P. H., Marzo, V. D., Elphick, M. R., Greasley, P. J., Hansen, H. S., Kunos, G., Mackie, K., Mechoulam, R., and Ross, R. A. (2010). International union of basic and clinical pharmacology. lxxix. cannabinoid receptors and their ligands: Beyond cb 1 and cb 2. Pharma- cological Reviews, 62(4):588–631. Petrucci, V., Chicca, A., Glasmacher, S., Paloczi, J., Cao, Z., Pacher, P., and Gertsch, J. (2017). Pepcan-12 (rvd-hemopressin) is a cb2 receptor positive al- losteric modulator constitutively secreted by adrenals and in liver upon tissue damage. Scientific Reports, 7(1):9560. Peñasco, S., Rico-Barrio, I., Puente, N., Gómez-Urquijo, S. M., Fontaine, C. J., Egaña-Huguet, J., Achicallende, S., Ramos, A., Reguero, L., Elezgarai, I., Nahir- ney, P. C., Christie, B. R., and Grandes, P. (2019). Endocannabinoid long-term depression revealed at medial perforant path excitatory synapses in the dentate gyrus. Neuropharmacology, 153:32–40. 141 Piazza, P. V., Cota, D., and Marsicano, G. (2017). The cb1 receptor as the corner- stone of exostasis. Neuron, 93(6):1252–1274. Pinar, C., Fontaine, C. J., Triviño-Paredes, J., Lottenberg, C. P., Gil-Mohapel, J., and Christie, B. R. (2017). Revisiting the flip side: Long-term depression of synaptic efficacy in the hippocampus. Neuroscience Biobehavioral Reviews, 80:394–413. Piomelli, D. (2003). The molecular logic of endocannabinoid signalling. Nature reviews. Neuroscience, 4(11):873–84. Pitler, T. and Alger, B. (1992). Postsynaptic spike firing reduces synaptic gabaa responses in hippocampal pyramidal cells. The Journal of Neuroscience, 12(10):4122–4132. Price, M. R., Baillie, G. L., Thomas, A., Stevenson, L. A., Easson, M., Goodwin, R., McLean, A., McIntosh, L., Goodwin, G., Walker, G., Westwood, P., Marrs, J., Thomson, F., Cowley, P., Christopoulos, A., Pertwee, R. G., and Ross, R. A. (2005). Allosteric modulation of the cannabinoid cb 1 receptor. Molecular Phar- macology, 68(5):1484–1495. Priestley, R., Glass, M., and Kendall, D. (2017). Functional selectivity at cannabin- oid receptors. Advances in Pharmacology, 80:207–221. Puente, N., Cui, Y., Lassalle, O., Lafourcade, M., Georges, F., Venance, L., Grandes, P., and Manzoni, O. J. (2011). Polymodal activation of the endocannabinoid system in the extended amygdala. Nature Neuroscience, 14(12):1542–1547. Puighermanal, E., Busquets-Garcia, A., Maldonado, R., and Ozaita, A. (2012). Cel- lular and intracellular mechanisms involved in the cognitive impairment of can- nabinoids. Philosophical Transactions of the Royal Society B: Biological Sciences, 367(1607):3254–3263. Puighermanal, E., Marsicano, G., Busquets-Garcia, A., Lutz, B., Maldonado, R., and Ozaita, A. (2009). Cannabinoid modulation of hippocampal long-term memory is mediated by mtor signaling. Nature neuroscience, 12(9):1152–8. Pérez-Gómez, E., Andradas, C., Flores, J. M., Quintanilla, M., Paramio, J. M., Guzmán, M., and Sánchez, C. (2013). The orphan receptor gpr55 drives skin carcinogenesis and is upregulated in human squamous cell carcinomas. Oncogene, 32(20):2534–2542. Qian, W.-J., Yin, N., Gao, F., Miao, Y., Li, Q., Li, F., Sun, X.-H., Yang, X.-L., and Wang, Z. (2017). Cannabinoid cb1 and cb2 receptors differentially modulate l- and t-type ca 2+ channels in rat retinal ganglion cells. Neuropharmacology, 124:143–156. Qin, N., Platano, D., Olcese, R., Stefani, E., and Birnbaumer, L. (1997). Direct interaction of gbetagamma with a c-terminal gbetagamma-binding domain of the ca2+ channel alpha1 subunit is responsible for channel inhibition by g protein- coupled receptors. Proceedings of the National Academy of Sciences of the United States of America, 94(16):8866–71. 142 Quarta, C., Bellocchio, L., Mancini, G., Mazza, R., Cervino, C., Braulke, L. J., Fekete, C., Latorre, R., Nanni, C., Bucci, M., Clemens, L. E., Heldmaier, G., Watanabe, M., Leste-Lassere, T., Maitre, M., Tedesco, L., Fanelli, F., Reuss, S., Klaus, S., Srivastava, R. K., Monory, K., Valerio, A., Grandis, A., Giorgio, R. D., Pasquali, R., Nisoli, E., Cota, D., Lutz, B., Marsicano, G., and Pagotto, U. (2010). Cb1 signaling in forebrain and sympathetic neurons is a key determinant of endocannabinoid actions on energy balance. Cell Metabolism, 11(4):273–285. Racine, R. J. (1972). Modification of seizure activity by electrical stimulation: Ii. motor seizure. Electroencephalography and Clinical Neurophysiology, 32(3):281– 294. Rajagopal, S. and Shenoy, S. K. (2018). Gpcr desensitization: Acute and prolonged phases. Cellular Signalling, 41:9–16. Ramakers, G. M., McNamara, R. K., Lenox, R. H., and Graan, P. N. D. (1999). Differential changes in the phosphorylation of the protein kinase c substrates myristoylated alanine-rich c kinase substrate and growth-associated protein-43/b- 50 following schaffer collateral long-term potentiation and long-term depression. Journal of neurochemistry, 73(5):2175–83. Ramı́rez-Franco, J., Bartolomé-Mart́ın, D., Alonso, B., Torres, M., and Sánchez- Prieto, J. (2014). Cannabinoid type 1 receptors transiently silence glutamatergic nerve terminals of cultured cerebellar granule cells. PLoS ONE, 9(2):e88594. Ratzliff, A. H., Santhakumar, V., Howard, A., and Soltesz, I. (2002). Mossy cells in epilepsy: rigor mortis or vigor mortis? Trends in Neurosciences, 25(3):140–144. Rekart, J., Quinn, B., Mesulam, M.-M., and Routtenberg, A. (2004). Subfield- specific increase in brain growth protein in postmortem hippocampus of alzheimer’s patients. Neuroscience, 126(3):579–584. Rekart, J. L., Meiri, K., and Routtenberg, A. (2005). Hippocampal-dependent memory is impaired in heterozygous gap-43 knockout mice. Hippocampus, 15(1):1– 7. Rey, A. A., Purrio, M., Viveros, M.-P., and Lutz, B. (2012). Biphasic effects of cannabinoids in anxiety responses: Cb1 and gabab receptors in the balance of gabaergic and glutamatergic neurotransmission. Neuropsychopharmacology, 37(12):2624–2634. Riederer, B. M. and Routtenberg, A. (1999). Can gap-43 interact with brain spec- trin? Molecular Brain Research, 71(2):345–348. Rios, C., Gomes, I., and Devi, L. A. (2006). opioid and cb1 cannabinoid receptor interactions: reciprocal inhibition of receptor signaling and neuritogenesis. British Journal of Pharmacology, 148(4):387–395. Ritter, S. L. and Hall, R. A. (2009). Fine-tuning of gpcr activity by receptor- interacting proteins. Nature Reviews Molecular Cell Biology, 10(12):819–830. 143 Robbe, D., Kopf, M., Remaury, A., Bockaert, J., and Manzoni, O. J. (2002). En- dogenous cannabinoids mediate long-term synaptic depression in the nucleus ac- cumbens. Proceedings of the National Academy of Sciences, 99(12):8384–8388. Roland, A. B., Ricobaraza, A., Carrel, D., Jordan, B. M., Rico, F., Simon, A., Humbert-Claude, M., Ferrier, J., McFadden, M. H., Scheuring, S., and Lenkei, Z. (2014). Cannabinoid-induced actomyosin contractility shapes neuronal morpho- logy and growth. eLife, 3(e03159). Ronesi, J. (2004). Disruption of endocannabinoid release and striatal long-term depression by postsynaptic blockade of endocannabinoid membrane transport. Journal of Neuroscience, 24(7):1673–1679. Ross, H. R., Napier, I., and Connor, M. (2008). Inhibition of recombinant human t-type calcium channels by 9 -tetrahydrocannabinol and cannabidiol. Journal of Biological Chemistry, 283(23):16124–16134. Ross, R. A. (2009). The enigmatic pharmacology of gpr55. Trends in Pharmacolo- gical Sciences, 30(3):156–163. Routtenberg, A., Cantallops, I., Zaffuto, S., Serrano, P., and Namgung, U. (2000). Enhanced learning after genetic overexpression of a brain growth protein. Pro- ceedings of the National Academy of Sciences of the United States of America, 97(13):7657–7662. Routtenberg, A., Lovinger, D. M., and Steward, O. (1985). Selective increase in phosphorylation of a 47-kda protein (f1) directly related to long-term potentiation. Behavioral and Neural Biology, 43(1):3–11. Rueda, D., Galve-Roperh, I., Haro, A., and Guzmán, M. (2000). The cb 1 cannabin- oid receptor is coupled to the activation of c-jun n-terminal kinase. Molecular Pharmacology, 58(4):814–820. Rueda, D., Navarro, B., Martinez-Serrano, A., Guzman, M., and Galve-Roperh, I. (2002). The endocannabinoid anandamide inhibits neuronal progenitor cell differentiation through attenuation of the rap1/b-raf/erk pathway. The Journal of biological chemistry, 277(48):46645–50. Ruehle, S., Remmers, F., Romo-Parra, H., Massa, F., Wickert, M., Wortge, S., Haring, M., Kaiser, N., Marsicano, G., Pape, H.-C., and Lutz, B. (2013). Can- nabinoid cb1 receptor in dorsal telencephalic glutamatergic neurons: Distinctive sufficiency for hippocampus-dependent and amygdala-dependent synaptic and be- havioral functions. Journal of Neuroscience, 33(25):10264–10277. Ruehle, S., Wager-Miller, J., Straiker, A., Farnsworth, J., Murphy, M. N., Loch, S., Monory, K., Mackie, K., and Lutz, B. (2017). Discovery and characterization of two novel cb1 receptor splice variants with modified n-termini in mouse. Journal of Neurochemistry, 142(4):521–533. Ruiz-Calvo, A., Maroto, I. B., Bajo-Grañeras, R., Chiarlone, A., Ángel Gaudioso, Ferrero, J. J., Resel, E., Sánchez-Prieto, J., Rodŕıguez-Navarro, J. A., Marsicano, G., Galve-Roperh, I., Bellocchio, L., and Guzmán, M. (2018). Pathway-specific 144 control of striatal neuron vulnerability by corticostriatal cannabinoid cb1 recept- ors. Cerebral cortex (New York, N.Y. : 1991), 28(1):307–322. Ryberg, E., Larsson, N., Sjögren, S., Hjorth, S., Hermansson, N.-O., Leonova, J., Elebring, T., Nilsson, K., Drmota, T., and Greasley, P. J. (2007). The orphan receptor gpr55 is a novel cannabinoid receptor. British Journal of Pharmacology, 152(7):1092–1101. Sagredo, O., González, S., Aroyo, I., Pazos, M. R., Benito, C., Lastres-Becker, I., Romero, J. P., Tolón, R. M., Mechoulam, R., Brouillet, E., Romero, J., and Fernández-Ruiz, J. (2009). Cannabinoid cb 2 receptor agonists protect the striatum against malonate toxicity: Relevance for huntington’s disease. Glia, 57(11):1154–1167. Sanna, M. D., Quattrone, A., Mello, T., Ghelardini, C., and Galeotti, N. (2014). The rna-binding protein hud promotes spinal gap43 overexpression in antiretroviral- induced neuropathy. Experimental neurology, 261:343–53. Scharfman, H. E. (1995). Electrophysiological evidence that dentate hilar mossy cells are excitatory and innervate both granule cells and interneurons. Journal of Neurophysiology, 74(1):179–194. Scharfman, H. E. (2016). The enigmatic mossy cell of the dentate gyrus. Nature Reviews Neuroscience, 17(9):562–575. Scharfman, H. E. and Myers, C. E. (2013). Hilar mossy cells of the dentate gyrus: A historical perspective. Frontiers in Neural Circuits, 6(JAN):1–17. Schlicker, E. and Kathmann, M. (2001). Modulation of transmitter release via pre- synaptic cannabinoid receptors. Trends in Pharmacological Sciences, 22(11):565– 572. Schmitz, S. K., King, C., Kortleven, C., Huson, V., Kroon, T., Josta, T., Schut, D., Saarloos, I., Hoetjes, J. P., Wit, H. D., Stiedl, O., Li, K. W., Mansvelder, H. D., Smit, A. B., Cornelisse, L. N., and Toonen, R. F. (2016). Presynaptic inhibition upon cb1 or mglu2/3 receptor activation requires erk/mapk phosphorylation of munc18-1. EMBO Journal, 35(11):1236–1250. Senzai, Y. and Buzsáki, G. (2017). Physiological properties and behavioral correlates of hippocampal granule cells and mossy cells. Neuron, 93(3):691–704.e5. Shao, Z., Yin, J., Chapman, K., Grzemska, M., Clark, L., Wang, J., and Rosenbaum, D. M. (2016). High-resolution crystal structure of the human cb1 cannabinoid receptor. Nature, 540(7634):602–606. Shire, D., Calandra, B., Delpech, M., Dumont, X., Kaghad, M., Fur, G. L., Caput, D., and Ferrara, P. (1996). Structural features of the central cannabinoid cb1 receptor involved in the binding of the specific cb1 antagonist sr 141716a. Journal of Biological Chemistry, 271(12):6941–6946. Signorello, M. G. and Leoncini, G. (2014). Effect of 2-arachidonoylglycerol on myosin light chain phosphorylation and platelet activation: The role of phosphatidylin- ositol 3 kinase/akt pathway. Biochimie, 105:182–91. 145 Simcocks, A. C., O’Keefe, L., Jenkin, K. A., Mathai, M. L., Hryciw, D. H., and McAinch, A. J. (2014). A potential role for gpr55 in the regulation of energy homeostasis. Drug Discovery Today, 19(8):1145–1151. Sinor, A. D., Irvin, S. M., and Greenberg, D. A. (2000). Endocannabinoids protect cerebral cortical neurons from in vitro ischemia in rats. Neuroscience Letters, 278(3):157–160. Sjöström, P. J., Turrigiano, G. G., and Nelson, S. B. (2003). Neocortical ltd via coincident activation of presynaptic nmda and cannabinoid receptors. Neuron, 39(4):641–654. Sloviter, R. S. (1991). Permanently altered hippocampal structure, excitability, and inhibition after experimental status epilepticus in the rat: The ?dormant basket cell? hypothesis and its possible relevance to temporal lobe epilepsy. Hippocampus, 1(1):41–66. Smith, T. H., Blume, L. C., Straiker, A., Cox, J. O., David, B. G., McVoy, J. R. S., Sayers, K. W., Poklis, J. L., Abdullah, R. A., Egertová, M., Chen, C.-K., Mackie, K., Elphick, M. R., Howlett, A. C., and Selley, D. E. (2015). Cannabinoid receptor–interacting protein 1a modulates cb 1 receptor signaling and regulation. Molecular Pharmacology, 87(4):747–765. Smith, T. H., Sim-Selley, L. J., and Selley, D. E. (2010). Cannabinoid cb 1 receptor- interacting proteins: Novel targets for central nervous system drug discovery? British Journal of Pharmacology, 160(3):454–466. Soler-Llavina, G. J. and Sabatini, B. L. (2006). Synapse-specific plasticity and compartmentalized signaling in cerebellar stellate cells. Nature Neuroscience, 9(6):798–806. Soltesz, I., Alger, B. E., Kano, M., Lee, S.-H., Lovinger, D. M., Ohno-Shosaku, T., and Watanabe, M. (2015). Weeding out bad waves: towards selective cannabinoid circuit control in epilepsy. Nature Reviews Neuroscience, 16(5):264–277. Son, H., Davis, P. J., and Carpenter, D. O. (1997). Time course and involvement of protein kinase c-mediated phosphorylation of f1/gap-43 in area ca3 after mossy fiber stimulation. Cellular and molecular neurobiology, 17(2):171–94. Soria-Gómez, E., Bellocchio, L., Reguero, L., Lepousez, G., Martin, C., Bendah- mane, M., Ruehle, S., Remmers, F., Desprez, T., Matias, I., Wiesner, T., Cannich, A., Nissant, A., Wadleigh, A., Pape, H.-C., Chiarlone, A. P., Quarta, C., Verrier, D., Vincent, P., Massa, F., Lutz, B., Guzmán, M., Gurden, H., Ferreira, G., Lledo, P.-M., Grandes, P., and Marsicano, G. (2014). The endocannabinoid system con- trols food intake via olfactory processes. Nature neuroscience, 17(3):407–15. Stadel, R., Ahn, K. H., and Kendall, D. A. (2011). The cannabinoid type-1 receptor carboxyl-terminus, more than just a tail. Journal of Neurochemistry, 117(1):1–18. Stea, A., Soong, T. W., and Snutch, T. P. (1995). Determinants of pkc-dependent modulation of a family of neuronal calcium channels. Neuron, 15(4):929–40. 146 Steindel, F., Lerner, R., Häring, M., Ruehle, S., Marsicano, G., Lutz, B., and Monory, K. (2013). Neuron-type specific cannabinoid-mediated g protein sig- nalling in mouse hippocampus. Journal of Neurochemistry, 124(6):795–807. Stella, N. (2010). Cannabinoid and cannabinoid-like receptors in microglia, astro- cytes, and astrocytomas. Glia, 58(9):1017–1030. Stempel, A. V., Stumpf, A., Zhang, H.-Y., Özdoğan, T., Pannasch, U., Theis, A.- K., Otte, D.-M., Wojtalla, A., Rácz, I., Ponomarenko, A., Xi, Z.-X., Zimmer, A., and Schmitz, D. (2016). Cannabinoid type 2 receptors mediate a cell type-specific plasticity in the hippocampus. Neuron, 90(4):795–809. Stevens, C. F. and Sullivan, J. M. (1998). Regulation of the readily releasable vesicle pool by protein kinase c. Neuron, 21(4):885–893. Straiker, A., Wager-Miller, J., Hutchens, J., and Mackie, K. (2012). Differential signalling in human cannabinoid cb1 receptors and their splice variants in autaptic hippocampal neurones. British Journal of Pharmacology, 165(8):2660–2671. Strange, B. A., Witter, M. P., Lein, E. S., and Moser, E. I. (2014). Functional organization of the hippocampal longitudinal axis. Nature Reviews Neuroscience, 15(10):655–669. Strittmatter, S. M. (1992). Gap-43 as a modulator of g protein transduction in the growth cone. Perspectives on developmental neurobiology, 1(1):13–9. Strittmatter, S. M., Fankhauser, C., Huang, P. L., Mashimo, H., and Fishman, M. C. (1995). Neuronal pathfinding is abnormal in mice lacking the neuronal growth cone protein gap-43. Cell, 80(3):445–452. Strittmatter, S. M., Valenzuela, D., Kennedy, T. E., Neer, E. J., and Fishman, M. C. (1990). G0 is a major growth cone protein subject to regulation by gap-43. Nature, 344(6269):836–841. Strittmatter, S. M., Valenzuela, D., Vartanian, T., Sudo, Y., Zuber, M. X., and Fishman, M. C. (1991). Growth cone transduction: Go and gap-43. Journal of Cell Science, 1991(Supplement 15):27–33. Sugiura, T., Kondo, S., Sukagawa, A., Nakane, S., Shinoda, A., Itoh, K., Yamashita, A., and Waku, K. (1995). 2-arachidonoylgylcerol: A possible endogenous can- nabinoid receptor ligand in brain. Biochemical and Biophysical Research Com- munications, 215(1):89–97. Sánchez, C., Galve-Roperh, I., Canova, C., Brachet, P., and Guzmán, M. (1998). 9 -tetrahydrocannabinol induces apoptosis in c6 glioma cells. FEBS Letters, 436(1):6–10. Sánchez, C., Rueda, D., Ségui, B., Galve-Roperh, I., Levade, T., and Guzmán, M. (2001). The cb 1 cannabinoid receptor of astrocytes is coupled to sphingomyelin hydrolysis through the adaptor protein fan. Molecular Pharmacology, 59(5):955– 959. 147 Takano, T. and Matsui, K. (2015). Increased expression of gap43 in interneurons in a rat model of experimental polymicrogyria. Journal of Child Neurology, 30(6):716– 728. Tanigami, H., Yoneda, M., Tabata, Y., Echigo, R., Kikuchi, Y., Yamazaki, M., Kishimoto, Y., Sakimura, K., Kano, M., and Ohno-Shosaku, T. (2019). En- docannabinoid signaling from 2-arachidonoylglycerol to cb1 cannabinoid receptor facilitates reward-based learning of motor sequence. Neuroscience, 421:1–16. Taniguchi, H., Suzuki, M., Manenti, S., and Titani, K. (1994). A mass spectrometric study on the in vivo posttranslational modification of gap-43. The Journal of biological chemistry, 269(36):22481–4. Tanimura, A., Yamazaki, M., Hashimotodani, Y., Uchigashima, M., Kawata, S., Abe, M., Kita, Y., Hashimoto, K., Shimizu, T., Watanabe, M., Sakimura, K., and Kano, M. (2010). The endocannabinoid 2-arachidonoylglycerol produced by diacylglycerol lipase mediates retrograde suppression of synaptic transmission. Neuron, 65(3):320–327. Taura, J., Fernández-Dueñas, V., and Ciruela, F. (2015). Visualizing g protein- coupled receptor-receptor interactions in brain using proximity ligation in situ assay. Current Protocols in Cell Biology, 67(1). Tehran, A., López-Hernández, and Maritzen (2019). Endocytic adaptor proteins in health and disease: Lessons from model organisms and human mutations. Cells, 8(11):1345. Tham, M., Yilmaz, O., Alaverdashvili, M., Kelly, M. E. M., Denovan-Wright, E. M., and Laprairie, R. B. (2019). Allosteric and orthosteric pharmacology of cannabid- iol and cannabidiol-dimethylheptyl at the type 1 and type 2 cannabinoid receptors. British Journal of Pharmacology, 176(10):1455–1469. Toman, R. E. and Spiegel, S. (2002). Lysophospholipid receptors in the nervous system. Neurochemical Research, 27:619–627. Tomatis, V. M., Trenchi, A., Gomez, G. A., and Daniotti, J. L. (2010). Acyl-protein thioesterase 2 catalizes the deacylation of peripheral membrane-associated gap-43. PLoS ONE, 5(11):e15045. Trettel, J. and Levine, E. S. (2003). Endocannabinoids mediate rapid retrograde signaling at interneuron → pyramidal neuron synapses of the neocortex. Journal of Neurophysiology, 89(4):2334–2338. Tsou, K., Nogueron, M., Muthian, S., Sañudo-Peña, M., Hillard, C. J., Deutsch, D. G., and Walker, J. (1998). Fatty acid amide hydrolase is located preferentially in large neurons in the rat central nervous system as revealed by immunohisto- chemistry. Neuroscience Letters, 254(3):137–140. Turkanis, S. A., Smiley, K. A., Borys, H. K., Olsen, D. M., and Karler, R. (1979). An electrophysiological analysis of the anticonvulsant action of cannabidiol on limbic seizures in conscious rats. Epilepsia, 20(4):351–63. 148 Twitchell, W., Brown, S., and Mackie, K. (1997). Cannabinoids inhibit n- and p/q- type calcium channels in cultured rat hippocampal neurons. Journal of Neuro- physiology, 78(1):43–50. Tzounopoulos, T., Rubio, M. E., Keen, J. E., and Trussell, L. O. (2007). Co- activation of pre- and postsynaptic signaling mechanisms determines cell-specific spike-timing-dependent plasticity. Neuron, 54(2):291–301. Ulfers, A. L., McMurry, J. L., Kendall, D. A., and Mierke, D. F. (2002). Structure of the third intracellular loop of the human cannabinoid 1 receptor. Biochemistry, 41(38):11344–11350. Valentinova, K. and Mameli, M. (2016). mglur-ltd at excitatory and inhibitory synapses in the lateral habenula tunes neuronal output. Cell Reports, 16(9):2298– 2307. Vallée, M., Vitiello, S., Bellocchio, L., Hébert-Chatelain, E., Monlezun, S., Martin- Garcia, E., Kasanetz, F., Baillie, G. L., Panin, F., Cathala, A., Roullot-Lacarrière, V., Fabre, S., Hurst, D. P., Lynch, D. L., Shore, D. M., Deroche-Gamonet, V., Spampinato, U., Revest, J.-M., Maldonado, R., Reggio, P. H., Ross, R. A., Marsicano, G., and Piazza, P. V. (2014). Pregnenolone can protect the brain from cannabis intoxication. Science, 343(6166):94–98. Varma, N., Brager, D. H., Morishita, W., Lenz, R. A., London, B., and Alger, B. (2002). Presynaptic factors in the regulation of dsi expression in hippocampus. Neuropharmacology, 43(4):550–562. Verkade, P., Verkleij, A., Annaert, W., Gispen, W., and Oestreicher, A. (1996). Ultrastructural localization of b-50/growth-associated protein-43 to anterogradely transported synaptophysin-positive and calcitonin gene-related peptide-negative vesicles in the regenerating rat sciatic nerve. Neuroscience, 71(2):489–505. Vitkovic, L., Steisslinger, H. W., Aloyo, V. J., and Mersel, M. (1988). The 43-kda neuronal growth-associated protein (gap-43) is present in plasma membranes of rat astrocytes. Proceedings of the National Academy of Sciences, 85(21):8296– 8300. Viñals, X., Moreno, E., Lanfumey, L., Cordomı́, A., Pastor, A., de La Torre, R., Gasperini, P., Navarro, G., Howell, L. A., Pardo, L., Llúıs, C., Canela, E. I., McCormick, P. J., Maldonado, R., and Robledo, P. (2015). Cognitive impairment induced by delta9-tetrahydrocannabinol occurs through heteromers between can- nabinoid cb1 and serotonin 5-ht2a receptors. PLOS Biology, 13(7):e1002194. Vásquez, C., Navarro-Polanco, R. A., Huerta, M., Trujillo, X., Andrade, F., Trujillo- Hernández, B., and Hernández, L. (2003). Effects of cannabinoids on endogenous k + and ca 2+ currents in hek293 cells. Canadian Journal of Physiology and Pharmacology, 81(5):436–442. Wallace, M. J., Blair, R. E., Falenski, K. W., Martin, B. R., and DeLorenzo, R. J. (2003). The endogenous cannabinoid system regulates seizure frequency and dur- ation in a model of temporal lobe epilepsy. Journal of Pharmacology and Experi- mental Therapeutics, 307(1):129–137. 149 Wallace, M. J., Martin, B. R., and DeLorenzo, R. J. (2002). Evidence for a physiolo- gical role of endocannabinoids in the modulation of seizure threshold and severity. European Journal of Pharmacology, 452(3):295–301. Wallace, M. J., Newton, P. M., McMahon, T., Connolly, J., Huibers, A., Whistler, J., and Messing, R. O. (2009). Pkc regulates behavioral sensitivity, binding and tolerance to the cb1 receptor agonist win55,212-2. Neuropsychopharmacology, 34(7):1733–1742. Wang, L., Hagemann, T. L., Messing, A., and Feany, M. B. (2016). An in vivo pharmacological screen identifies cholinergic signaling as a therapeutic target in glial-based nervous system disease. The Journal of neuroscience : the official journal of the Society for Neuroscience, 36(5):1445–55. Wang, W., Jia, Y., Pham, D. T., Palmer, L. C., Jung, K. M., Cox, C. D., Rumbaugh, G., Piomelli, D., Gall, C. M., and Lynch, G. (2018). Atypical endocannabinoid signaling initiates a new form of memory-related plasticity at a cortical input to hippocampus. Cerebral cortex (New York, N.Y. : 1991), 28(7):2253–2266. Weickert, C. S. (2001). Reduced gap-43 mrna in dorsolateral prefrontal cortex of patients with schizophrenia. Cerebral Cortex, 11(2):136–147. Wettschureck, N., van der Stelt, M., Tsubokawa, H., Krestel, H., Moers, A., Pet- rosino, S., Schutz, G., Marzo, V. D., and Offermanns, S. (2006). Forebrain-specific inactivation of gq/g11 family g proteins results in age-dependent epilepsy and im- paired endocannabinoid formation. Molecular and Cellular Biology, 26(15):5888– 5894. Whalley, B. J., Lin, H., Bell, L., Hill, T., Patel, A., Gray, R. A., Roberts, C. E., Devinsky, O., Bazelot, M., Williams, C. M., and Stephens, G. J. (2019). Species- specific susceptibility to cannabis-induced convulsions. British Journal of Phar- macology, 176(10):1506–1523. Williams, E. J., Walsh, F. S., and Doherty, P. (2003). The fgf receptor uses the endocannabinoid signaling system to couple to an axonal growth response. Journal of Cell Biology, 160(4):481–486. Williams, J. H., Errington, M. L., Lynch, M. A., and Bliss, T. V. P. (1989). Arachidonic acid induces a long-term activity-dependent enhancement of synaptic transmission in the hippocampus. Nature, 341(6244):739–742. Wilson, R. I. (2002). Endocannabinoid signaling in the brain. Science, 296(5568):678–682. Wilson, R. I. and Nicoll, R. A. (2001). Endogenous cannabinoids mediate retrograde signalling at hippocampal synapses. Nature, 410(6828):588–592. Wittmann, G., Deli, L., Kalló, I., Hrabovszky, E., Watanabe, M., Liposits, Z., and Fekete, C. (2007). Distribution of type 1 cannabinoid receptor (cb1)- immunoreactive axons in the mouse hypothalamus. The Journal of comparative neurology, 503(2):270–9. 150 Xing, C., Zhuang, Y., Xu, T.-H., Feng, Z., Zhou, X. E., Chen, M., Wang, L., Meng, X., Xue, Y., Wang, J., Liu, H., McGuire, T. F., Zhao, G., Melcher, K., Zhang, C., Xu, H. E., and Xie, X.-Q. (2020). Cryo-em structure of the human cannabinoid receptor cb2-gi signaling complex. Cell, 180(4):645–654.e13. Yanovsky, Y., Mades, S., and Misgeld, U. (2003). Retrograde signaling changes the venue of postsynaptic inhibition in rat substantia nigra. Neuroscience, 122(2):317– 328. Yeckel, M. F. and Berger, T. W. (1990). Feedforward excitation of the hippocampus by afferents from the entorhinal cortex: redefinition of the role of the trisynaptic pathway. Proceedings of the National Academy of Sciences, 87(15):5832–5836. Yoshida, T. (2006). Localization of diacylglycerol lipase- around postsynaptic spine suggests close proximity between production site of an endocannabinoid, 2-arachidonoyl-glycerol, and presynaptic cannabinoid cb1 receptor. Journal of Neuroscience, 26(18):4740–4751. Younts, T. J., Monday, H. R., Dudok, B., Klein, M. E., Jordan, B. A., Katona, I., and Castillo, P. E. (2016). Presynaptic protein synthesis is required for long-term plasticity of gaba release. Neuron, 92(2):479–492. Zaccaria, K. J., Lagace, D. C., Eisch, A. J., and McCasland, J. S. (2010). Resistance to change and vulnerability to stress: autistic-like features of gap43-deficient mice. Genes, Brain and Behavior, 9(8):985–996. Zakharov, V. V. and Mosevitsky, M. I. (2007). M-calpain-mediated cleavage of gap-43 near ser41 is negatively regulated by protein kinase c, calmodulin and calpain-inhibiting fragment gap-43-3. Journal of Neurochemistry, 101(6):1539– 1551. Zarate, C. A. and Manji, H. K. (2009). Protein kinase c inhibitors. CNS Drugs, 23(7):569–582. Zhang, Y., Bo, X., Schoepfer, R., Holtmaat, A. J., Verhaagen, J., Emson, P. C., Lieberman, A. R., and Anderson, P. N. (2005). Growth-associated protein gap-43 and l1 act synergistically to promote regenerative growth of purkinje cell axons in vivo. Proceedings of the National Academy of Sciences of the United States of America, 102(41):14883–14888. Zhou, D. and Song, Z. (2001). Cb1 cannabinoid receptor-mediated neurite remod- eling in mouse neuroblastoma n1e-115 cells. Journal of Neuroscience Research, 65(4):346–353. Zhou, X. E., He, Y., de Waal, P. W., Gao, X., Kang, Y., Eps, N. V., Yin, Y., Pal, K., Goswami, D., White, T. A., Barty, A., Latorraca, N. R., Chapman, H. N., Hubbell, W. L., Dror, R. O., Stevens, R. C., Cherezov, V., Gurevich, V. V., Griffin, P. R., Ernst, O. P., Melcher, K., and Xu, H. E. (2017). Identification of phosphorylation codes for arrestin recruitment by g protein-coupled receptors. Cell, 170(3):457–469.e13. 151 Zhu, P. J. (2005). Retrograde endocannabinoid signaling in a postsynaptic neuron/synaptic bouton preparation from basolateral amygdala. Journal of Neur- oscience, 25(26):6199–6207. Zhuang, H. and Matsunami, H. (2008). Evaluating cell-surface expression and meas- uring activation of mammalian odorant receptors in heterologous cells. Nature Protocols, 3(9):1402–1413. Zou, S. and Kumar, U. (2018). Cannabinoid receptors and the endocannabinoid sys- tem: Signaling and function in the central nervous system. International Journal of Molecular Sciences, 19(3):833. Zygmunt, P. M., Petersson, J., Andersson, D. A., hu Chuang, H., Sørg̊ard, M., Marzo, V. D., Julius, D., and Högestätt, E. D. (1999). Vanilloid receptors on sens- ory nerves mediate the vasodilator action of anandamide. Nature, 400(6743):452– 457. 152 Tesis Irene Berenice Maroto Martínez PORTADA ACKNOWLEDGEMENTS CONTENTS LIST OF FIGURES LIST OF TABLES ABBREVIATIONS RESUMEN ABSTRACT INTRODUCTION AIMS MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES