Numerous assays have been developed to investigate the interactions between G-protein–coupled receptors (GPCRs) and their ligands since GPCRs are key therapeutic targets. Reporter-based assays using the cAMP response element (CRE) coupled with bioluminescence from a luciferase reporter have been used extensively in vitro with high-throughput screens (HTS) of large chemical compound libraries. We have generated a transgenic mouse model (CRE luc) with a luciferase reporter under the control of a synthetic promoter that contains several CREs, which supports real-time bioimaging of GPCR ligand activity in whole animals, tissues, or primary cells. In the CRE luc model, GPCR signaling through the cAMP pathway can be detected from the target GPCR that is in a native cellular environment with a full complement of associated receptors and membrane constituents. Multiple independent lines have been produced by random integration of the transgene, resulting in tissue expression profiles covering the major organs. The goal of the CRE luc model is to accelerate the transition from HTS to profiling of GPCR small-molecule leads in preclinical animal disease models, as well as define the mechanism of action of GPCR drugs in three experimental formats: primary cells, tissue homogenates, and whole animals.

Cannabinoid receptor CB1 is expressed abundantly in the brain and presumably in the peripheral tissues responsible for energy metabolism. It is unclear if the antiobesity effects of rimonabant, a CB1 antagonist, are mediated through the central or the peripheral CB1 receptors. To address this question, we generated transgenic mice with central nervous system (CNS)-specific knockdown (KD) of CB1, by expressing an artificial microRNA (AMIR) under the control of the neuronal Thy1.2 promoter. In the mutant mice, CB1 expression was reduced in the brain and spinal cord, whereas no change was observed in the superior cervical ganglia (SCG), sympathetic trunk, enteric nervous system, and pancreatic ganglia. In contrast to the neuronal tissues, CB1 was undetectable in the brown adipose tissue (BAT) or the liver. Consistent with the selective loss of central CB1, agonist-induced hypothermia was attenuated in the mutant mice, but the agonist-induced delay of gastrointestinal transit (GIT), a primarily peripheral nervous system-mediated effect, was not. Compared to wild-type (WT) littermates, the mutant mice displayed reduced body weight (BW), adiposity, and feeding efficiency, and when fed a high-fat diet (HFD), showed decreased plasma insulin, leptin, cholesterol, and triglyceride levels, and elevated adiponectin levels. Furthermore, the therapeutic effects of rimonabant on food intake (FI), BW, and serum parameters were markedly reduced and correlated with the degree of CB1 KD. Thus, KD of CB1 in the CNS recapitulates the metabolic phenotype of CB1 knockout (KO) mice and diminishes rimonabant's efficacy, indicating that blockade of central CB1 is required for rimonabant's antiobesity actions.