Revealing the Mechanism of Agonist-Mediated Cannabinoid Receptor 1 (CB1) Activation and Phospholipid-Mediated Allosteric Modulation

Abstract

Cannabinoid receptor 1 (CB1) mediates the functional responses of Δ9-tetrahydrocannabinol. Although progress has been made in understanding cannabinoid binding and receptor activation, detailed knowledge of the dynamics involved in the activation mechanism of CB1 is lacking. Here, we use recently determined CB1 crystal structures to analyze its transition from inactive to active state by performing unbiased microsecond-length molecular dynamics (MD) simulations, totaling 32 μs, with and without bound potent cannabinoid agonist CP-55940. CB1 activation is characterized by an upward axial movement of transmembrane (TM) helix 3, inward movement of TM7, and outward movement of TM6. These conformational changes collectively allow Gi protein docking, although fully active states of the receptor occur only transiently during MD simulations. Additionally, positive allosteric modulation of CB1 by anionic phospholipids is found to increase action of the bound agonist. Specifically, this involves protein–lipid interactions at intracellular loop 3, TM6, and ionic lock residue Arg2143.50.

Insights into adenosine A2A receptor activation through cooperative modulation of agonist and allosteric lipid interactions

“…Moreover, these two different intermediate receptor conformations are notable because NMR experiments have identified two distinct intermediates of A2aR, which were resolved in apo and agonist-stabilized states, respectively [41]. Taken together, this demonstrates that like other homologous GPCRs [50,53], A2aR can be positively modulated by PG lipids even without bound agonist. On the other hand, neutral PC lipids induce little or no conformational change over a microsecond time-window and instead depend on agonist binding for reaching an intermediate receptor state.…”

“…Therefore, without bound ligand acting on TM7, TM6 conformational changes need to be triggered first by protein-lipid intracellular interactions, such as those provided by anionic DOPG phospholipids. As has been previously shown with homologous β2-adrenergic and CB1 receptors, protein-DOPG Hbonds and electrostatic interactions are able to exert an outward pull on TM6 [50,53]. These initial intracellular conformational changes in TM6 leave a space next to TM3, which TM7 can occupy by spontaneously moving inwards, creating intracellular and extracellular conformational changes that affect other receptor regions such as the G protein binding-site or ECL2.…”

“…Finally, in addition to allosteric protein-lipid interactions formed at the receptor surface, the outward orientation of TM6 in a DOPG membrane can be further assisted by formation of a specific protein-lipid salt-bridge at the intracellular side of the receptor interior, which has previously been observed to occur in homologous β2-adrenergic and CB1 receptors [50,53,105]. This particular allosteric interaction is possible because of the internalization of a single DOPG lipid from the lower leaflet into the G protein binding-site between TM6 and TM7 ( Fig 7A).…”

“…In similar fashion to β2-adrenergic and CB1 [50,53], DOPG is able to make specific protein-lipid interactions with A2aR through its negatively charged phosphate or polar head groups with positively charged or polar residues on the intracellular side of the receptor ( Fig 6). As a result, time-averaged membrane thickness profiles across DOPG-containing MD simulations reveal that DOPG phospholipids cluster more strongly around A2aR than DOPC ( Fig 6A-6C).…”

“…Moreover, data gathered in a β2AR ligand binding study suggests that protein and membrane interplay improves interaction between protein and ligand and could be important for drug development [49,52]. Likewise, our recent computational study on homologous cannabinoid receptor 1 (CB1) found compelling evidence for positive allosteric communication between bound CB1 agonists and DOPG membrane phospholipids, which appears to enhance the speed of receptor activation [53]. Still, an interesting question is how much different membranes might affect the activation process of other class A GPCRs and the functional effect of their agonists.…”