GPR37 is an orphan G protein-coupled receptor (GPCR) implicated in several neurological diseases and important physiological pathways in the brain. We previously reported that its long N-terminal ectodomain undergoes constitutive metalloprotease-mediated cleavage and shedding, which have been rarely described for class A GPCRs. Here, we demonstrate that the protease that cleaves GPR37 at Glu167↓Gln168 is a disintegrin and metalloprotease 10 (ADAM10). This was achieved by employing selective inhibition, RNAi-mediated downregulation, and genetic depletion of ADAM10 in cultured cells as well as in vitro cleavage of the purified receptor with recombinant ADAM10. In addition, the cleavage was restored in ADAM10 knockout cells by overexpression of the wild type but not the inactive mutant ADAM10. Finally, postnatal conditional depletion of ADAM10 in mouse neuronal cells was found to reduce cleavage of the endogenous receptor in the brain cortex and hippocampus, confirming the physiological relevance of ADAM10 as a GPR37 sheddase. Additionally, we discovered that the receptor is subject to another cleavage step in cultured cells. Using site-directed mutagenesis, the site (Arg54↓Asp55) was localized to a highly conserved region at the distal end of the ectodomain that contains a recognition site for the proprotein convertase furin. The cleavage by furin was confirmed by using furin-deficient human colon carcinoma LoVo cells and proprotein convertase inhibitors. GPR37 is thus the first multispanning membrane protein that has been validated as an ADAM10 substrate and the first GPCR that is processed by both furin and ADAM10. The unconventional N-terminal processing may represent an important regulatory element for GPR37.

Increasing evidence shows that astrocytes, by releasing and uptaking neuroactive molecules, regulate synaptic plasticity, considered the neurophysiological basis of memory. This study investigated the impact of l-α-aminoadipate (l-AA) on astrocytes which sense and respond to stimuli at the synaptic level and modulate hippocampal long-term potentiation (LTP) and memory. l-AA selectivity toward astrocytes was proposed in the early 70's and further tested in different systems. Although it has been used for impairing the astrocytic function, its effects appear to be variable in different brain regions. To test the effects of l-AA in the hippocampus of male C57Bl/6 mice we performed two different treatments (ex vivo and in vivo) and took advantage of other compounds that were reported to affect astrocytes. l-AA superfusion did not affect the basal synaptic transmission but decreased LTP magnitude. Likewise, trifluoroacetate and dihydrokainate decreased LTP magnitude and occluded the effect of l-AA on synaptic plasticity, confirming l-AA selectivity. l-AA superfusion altered astrocyte morphology, increasing the length and complexity of their processes. In vivo, l-AA intracerebroventricular injection not only reduced the astrocytic markers but also LTP magnitude and impaired hippocampal-dependent memory in mice. Interestingly, d-serine administration recovered hippocampal LTP reduction triggered by l-AA (2 h exposure in hippocampal slices), whereas in mice injected with l-AA, the superfusion of d-serine did not fully rescue LTP magnitude. Overall, these data show that both l-AA treatments affect astrocytes differently, astrocytic activation or loss, with similar negative outcomes on hippocampal LTP, implying that opposite astrocytic adaptive alterations are equally detrimental for synaptic plasticity.