Resting State Brain Networks under Inverse Agonist versus Complete Knockout of the Cannabinoid Receptor 1.

Suppressing the cannabinoid receptor 1 (CB1) has been proposed as a treatment for conditions like obesity and Parkinson's disease, but a clinical trial produced unanticipated psychological side effects. Using in vivo imaging in mice, the authors show that complete genetic knockout of the CB1 gene (cnr1-/-) alters white matter structure and functional directional uniformity (in male mice) across the brain while largely preserving network activity. In contrast, the CB1 inverse agonist rimonabant alters network activity in cortical regions known to be affected by THC but does not affect directional uniformity. Chronic loss of cnr1 differs substantially from short-term pharmacological suppression, which may explain why pathological CB1 mutations do not predict side effects of CB1-suppressing drugs.

Study at a glance

Abstract

The cannabinoid receptor 1 (CB1) is famous as the target of Δ9-tetrahydrocannabinol (THC), which is the active ingredient of marijuana. Suppression of CB1 is frequently suggested as a drug target or gene therapy for many conditions (e.g., obesity, Parkinson's disease). However, brain networks affected by CB1 remain elusive, and unanticipated psychological effects in a clinical trial had dire consequences. To better understand the whole brain effects of CB1 suppression we performed in vivo imaging on mice under complete knockout of the gene for CB1 (cnr1-/-) and also under the CB1 inverse agonist rimonabant. We examined white matter structural changes and brain function (network activity and directional uniformity) in cnr1-/- mice. In cnr1-/- mice, white matter (in both sexes) and functional directional uniformity (in male mice) were altered across the brain but network activity was largely unaltered. Conversely, under rimonabant, functional directional uniformity was not altered but network activity was altered in cortical regions, primarily in networks known to be altered by THC (e.g., neocortex, hippocampal formation). However, rimonabant did not alter many brain regions found in both our cnr1-/- results and previous behavioral studies of cnr1-/- mice (e.g., thalamus, infralimbic area). This suggests that chronic loss of cnr1 is substantially different from short-term suppression, subtly rewiring the brain but largely maintaining the network activity. Our results help explain why pathological mutations in CB1 (e.g., chronic pain) do not always provide insight into the side effects of CB1 suppression (e.g., clinical depression), and thus urge more preclinical studies for any drugs that suppress CB1.