Conformational changes are observed in the LBDs after ligand

Conformational changes are observed in the LBDs after ligand binding, including partial cleft closure for GluN2A and complete cleft closure for GluN1. For GluN2A, additional conformational rearrangements in the protein and ligand may be required before the ligand can be accommodated within lobe 2. For example, it is likely that the glutamate ligand must rotate such that its α-carboxylate interacts with lobe 1 in order to allow full cleft closure (Figure 7). The predicted glutamate-binding modes can be experimentally tested using nuclear magnetic resonance (NMR). The glutamate carboxylates sample vastly different electronic environments in the two binding modes, and signatures for these modes can be probed using various NMR methods (Palmer, 2014). For GluN1, our simulations reveal that the cross-lobe interactions that stabilize the closed state are actually quite dynamic. The formation and breakage of these interactions can give rise to a range of conformational dynamics in the LBD (Yao et al., 2013). Why do glutamate and glycine bind to the MHY1485 in such different ways? Given the overall structural similarity between the GluN2A and GluN1 LBDs, one might conclude that the LBDs also bind ligands via similar processes. NMDA receptors with engineered disulfide linkages that lock the GluN1 lobes shut conduct current with kinetic profiles similar to that of wild-type NMDA receptors, suggesting that glutamate is the primary neurotransmitter, whereas D-serine, the endogenous agonist for GluN1 at synapses, or glycine, play more modulatory roles (Mothet et al., 2000, Kussius and Popescu, 2010). We speculate that the distinct binding mechanisms of glutamate and glycine may reflect the differing physiological roles the ligands play during NMDA receptor activation. Differences in these binding mechanisms may inform strategies for the design of therapeutic agents.
STAR★Methods
Acknowledgments We thank Dominique Frueh for helpful discussion. Anton computer time was provided by the Pittsburgh Supercomputing Center (PSC) through NIH grant R01GM116961. The Anton and Anton2 machines at PSC were generously made available by D.E. Shaw Research. This study also used resources provided by the Maryland Advanced Research Computing Center (MARCC) at Johns Hopkins University. This work was supported by NIH grant R01GM094495 (to A.Y.L.).
Introduction Cannabis has been used to relieve chronic pain in human for centuries (Johnson et al., 2010, Martin-Sanchez et al., 2009, Murray et al., 2007). Both Δ9-tetrahydrocannabinol (THC) and cannabidiol (CBD), the major psychoactive and nonpsychoactive components of cannabis (Wang et al., 2008), are found to produce analgesic effects in chronic pain (Costa et al., 2007, Johnson et al., 2013). While the primary action site of THC in brain is cannabinoid receptor type 1 (CB1) (Kawamura et al., 2006), emerging evidence suggests that some of THC-induced analgesic effects are independent of CB1 receptors (LedentOV et al., 1999, Costa et al., 2005). For instance, the analgesic effect of THC remained intact in the tail-flick reflex (TFR) test in CB1 receptor knock-out mice (Wilson and Nicoll, 2002), and this THC-induced effect was not affected by intrathecal injection of a selective CB1 receptor antagonist either (Hohmann et al., 2005). In addition, CBD could also induce analgesic effects without activating CB1 receptors (Costa et al., 2007, Izzo et al., 2009, Long et al., 2006). These observations suggest that there are additional targets other than CB1 receptors contributing to cannabis analgesia. Emerging evidence has suggested that some cannabinoids can directly or indirectly act on the glycine receptor (GlyR), an inhibitory ion channel, through a CB receptor-independent mechanism in the central nervous system (Anderson et al., 2009, Hejazi et al., 2006, Lozovaya et al., 2005, Xiong et al., 2011). Molecular cloning has identified four isoforms of the GlyR α subunit (α1-4) and a single isoform of the β subunit (Betz and Laube, 2006). In general, these GlyRs are functional as pentameric assemblies containing homomeric α subunits or heteromeric α/β subunits (Lynch, 2009). In adults, α1 and α3 are the most abundant GlyR subunits in the spinal cord (Shiang et al., 1995, Robert Harvey et al., 2004). The α1 GlyR is considered to be critical for neuromotor activity since several naturally occurring point-mutations that cause functional deficiency of α1 GlyRs are closely associated with hyperekplexia, an exaggerated startle disease, in human (Zhou and ChillagNigro, 2002). Meanwhile, the α3 GlyR, which is abundantly expressed in the superficial laminae of spinal cord dorsal horn, plays an important role in the suppression of inflammatory-induced hypersensitivity to pain (Robert Harvey et al., 2004).