GAP43 Located on Corticostriatal Terminals Restrains Novelty-Induced Hyperactivity in Mice.

Growth-associated protein of 43 kDa (GAP43) is a key cytoskeleton-associated component of the presynaptic terminal that facilitates neuroplasticity. Downregulation of GAP43 expression has been associated to various psychiatric conditions in humans and evokes hippocampus-dependent memory impairments in mice. Despite the extensive studies conducted on hippocampal GAP43 in past decades, however, very little is known about its roles in modulating the excitatory versus inhibitory balance in other brain regions. We recently generated conditional knock-out mice in which the Gap43 gene was selectively inactivated in either telencephalic glutamatergic neurons (Gap43fl/fl ;Nex1Cre mice, hereafter Glu-GAP43-/- mice) or forebrain GABAergic neurons (Gap43fl/fl ;Dlx5/6Cre mice, hereafter GABA-GAP43-/- mice). Here, we show that Glu-GAP43-/- but not GABA-GAP43-/- mice of either sex show a striking hyperactive phenotype when exposed to a novel environment. This behavioral alteration of Glu-GAP43-/- mice was linked to a selective activation of dorsal-striatum neurons, as well as to an enhanced corticostriatal glutamatergic transmission and an abrogation of corticostriatal endocannabinoid-mediated long-term depression. In line with these observations, GAP43 was abundantly expressed in corticostriatal glutamatergic terminals of wild-type mice. The novelty-induced hyperactive phenotype of Glu-GAP43-/- mice was abrogated by chemogenetically inhibiting corticostriatal afferences with a Gi-coupled "designer receptor exclusively activated by designer drugs" (DREADDs), thus further supporting that novelty-induced activity is controlled by GAP43 at corticostriatal excitatory projections. Taken together, these findings show an unprecedented regulatory role of GAP43 in the corticostriatal circuitry and provide a new mouse model with a delimited neuronal-circuit alteration for studying novelty-induced hyperactivity, a phenotypic shortfall that occurs in diverse psychiatric diseases.

The CB1 receptor interacts with cereblon and drives cereblon deficiency-associated memory shortfalls.

Cereblon/CRBN is a substrate-recognition component of the Cullin4A-DDB1-Roc1 E3 ubiquitin ligase complex. Destabilizing mutations in the human CRBN gene cause a form of autosomal recessive non-syndromic intellectual disability (ARNSID) that is modelled by knocking-out the mouse Crbn gene. A reduction in excitatory neurotransmission has been proposed as an underlying mechanism of the disease. However, the precise factors eliciting this impairment remain mostly unknown. Here we report that CRBN molecules selectively located on glutamatergic neurons are necessary for proper memory function. Combining various in vivo approaches, we show that the cannabinoid CB1 receptor (CB1R), a key suppressor of synaptic transmission, is overactivated in CRBN deficiency-linked ARNSID mouse models, and that the memory deficits observed in these animals can be rescued by acute CB1R-selective pharmacological antagonism. Molecular studies demonstrated that CRBN interacts physically with CB1R and impairs the CB1R-Gi/o-cAMP-PKA pathway in a ubiquitin ligase-independent manner. Taken together, these findings unveil that CB1R overactivation is a driving mechanism of CRBN deficiency-linked ARNSID and anticipate that the antagonism of CB1R could constitute a new therapy for this orphan disease.

Selective inhibition of cannabinoid CB1 receptor-evoked signalling by the interacting protein GAP43.

Cannabinoids exert pleiotropic effects on the brain by engaging the cannabinoid CB1 receptor (CB1R), a presynaptic metabotropic receptor that regulates key neuronal functions in a highly context-dependent manner. We have previously shown that CB1R interacts with growth-associated protein of 43 kDa (GAP43) and that this interaction inhibits CB1R function on hippocampal excitatory synaptic transmission, thereby impairing the therapeutic effect of cannabinoids on epileptic seizures in vivo. However, the underlying molecular features of this interaction remain unexplored. Here, we conducted mechanistic experiments on HEK293T cells co-expressing CB1R and GAP43 and show that GAP43 modulates CB1R signalling in a strikingly selective manner. Specifically, GAP43 did not affect the archetypical agonist-evoked (i) CB1R/Gi/o protein-coupled signalling pathways, such as cAMP/PKA and ERK, or (ii) CB1R internalization and intracellular trafficking. In contrast, GAP43 blocked an alternative agonist-evoked CB1R-mediated activation of the cytoskeleton-associated ROCK signalling pathway, which relied on the GAP43-mediated impairment of CB1R/Gq/11 protein coupling. GAP43 also abrogated CB1R-mediated ROCK activation in mouse hippocampal neurons, and this process led in turn to a blockade of cannabinoid-evoked neurite collapse. An NMR-based characterization of the CB1R-GAP43 interaction supported that GAP43 binds directly and specifically through multiple amino acid stretches to the C-terminal domain of the receptor. Taken together, our findings unveil a CB1R-Gq/11-ROCK signalling axis that is selectively impaired by GAP43 and may ultimately control neurite outgrowth.