A Receptor on Acid

Abstract

Wacker et al. report the crystal structure of LSD in complex with one of its major targets in the brain, the 5-HT_2B receptor, the first such structure for any psychedelic drug. The results shed light on the molecular mechanisms underlying its ability to induce hallucinations with greater duration and potency than closely related compounds.

Lysergic acid diethylamide (LSD), also known as ‘acid’, is one of the most potent psychedelic drugs. At low microgram dosages, LSD can cause hallucinations (alteration of perception, emotion and cognition), or “trips”, that last for up to 15 hours. Albert Hofmann first synthesized LSD in 1938, but its psychedelic effects were only discovered five years later when he accidentally exposed himself to trace amounts, resulting in a “not-unpleasant” condition characterized by an “extremely stimulated imagination” along with an “uninterrupted stream of fantastic pictures”. He later confirmed that LSD was the culprit by self-administering 250 μg, now considered a heavy dose, and experienced the world’s first intentional acid trip in part while bicycling home from lab(Hofmann, 1979). LSD gained popularity in early 60’s for recreational use but later became regulated as a type I substance. Use of LSD resurged in the middle 90’s in United States when, according to the Monitoring the Future study, ~8.8% of 12^th grade students reported that they had used LSD at least once. The percentage has declined to ~2% since 2002 presumably due to enhanced control over LSD precursors (Johnston, 2016). LSD has come to attention more recently because it shows promise for treating alcohol abuse and depression, and in alleviating the anxiety of patients suffering terminally ill diseases. Although the drug targets a broad spectrum of receptors in vivo, it is believed to exert its psychedelic effects primarily by binding to the 5-HT_2A serotonin receptor (Titeler et al., 1988), a G protein-coupled receptor. However, the molecular mechanisms underlying its potent psychedelic effect have remained elusive. On page X, Wacker et al. report the crystal structure of LSD in complex with the closely related 5-HT_2B receptor. The study establishes that LSD induces unique conformational changes in the receptor that likely underlie its psychedelic activity and sheds light on the structural basis for the high potency of LSD and the long duration of its effects (Fig. 1).

Figure 1. LSD interacts with a unique lid element of the 5-HT receptor to send the brain on an extended bicycle trip.

LSD, schematized as a bicycle, exerts its potent psychedelic effects by directly binding to the extracellular surfaces of transmembrane receptors in the brain such as the 5-HT2A and 5-HT2B serotonin receptors. By stabilizing them in conformations that preferentially favor the binding of arrestin (green lobes) at its intracellular surface, LSD induces a mental ‘trip’ from reality. A key feature of LSD required for this effect is its diethylamine substituent (handlebars of the bicycle). One extracellular loop of the receptor serves as a ‘lid’, blocking the exit for LSD, thus making the trip long lasting. This figure is created with the help from Stephanie King (University of Michigan Life Sciences Institute). The LSD bicycle graphic is adapted from the website http://bicycleday.la.

An intriguing question in the field is why LSD leads to hallucination while chemically similar 5-hydroxytryptamine (5-HT) receptor agonists do not. The phenomenon of different ligands targeting the same receptor but leading to different functional outcomes is known as biased signaling (Violin and Lefkowitz, 2007), or functional selectivity(Urban et al., 2007). Biased signaling can occur if each ligand stabilizes the receptor in a unique conformational state that differentially engages its downstream signaling partners. In the case of 5-HT receptors, the best understood downstream partners are heterotrimeric G proteins and scaffolding proteins known as arrestins. The psychedelic effect of LSD relative to related less-hallucinogenic analogs like lysergamide could be a consequence of bias towards arrestin recruitment, as suggested by the fact that LSD-bound 5-HT_2A and 5-HT_2B is ~100 fold more efficient in recruiting arrestin than in activating G proteins (Wacker et al., 2013). In Wacker et al., the authors suggest that the structure of the 5-HT_2B receptor in complex with LSD exhibits hallmarks of an arrestin-biased conformation because of its overall similarity to a prior structure of the 5-HT_2B receptor in complex with ergotamine (Wacker et al., 2013), which is a nearly fully arrestin biased ligand (but non-hallucinogenic because it cannot pass the blood-brain barrier). The authors however observe significant conformational differences in extracellular regions of the receptor that bind to LSD, which are likely coupled allosterically to conformational changes in the intracellular domain responsible for engaging G proteins and arrestins. The extent of these allosteric changes and how the ligand binding site couples to the intracellular domain remains to be fully understood, in part because the experimental approach used by Wacker et al. required extensive engineering of the receptor (e.g. introduction of a thermal stabilizing mutation and replacement of an intracellular loop with a rigid domain) and because of possible crystal packing effects. Even so, conformational changes unique to the LSD binding site may underlie some of its most interesting pharmacological properties.

So what does LSD look like when bound to the receptor? Like the endogenous agonist 5-HT, LSD contains a tryptamine moiety but also features a diethylamide substituent important for its psychedelic properties. The tryptamine moiety sits in the endogenous agonist binding site (the orthosteric pocket) and is anchored via a salt bridge between its nitrogen and a conserved aspartate side chain. The ergoline ring system of LSD is further stabilized by hydrophobic interactions with the receptor core. Interestingly, the tryptamine core of LSD is rotated upwards in the pocket relative to that of ergotamide, which has a much larger amide substituent. The diethylamide group of LSD binds with its ethyl moieties in a stereospecific manner reminiscent of the legs of a bicyclist. The authors go on to show using chemical analogs that if the ethyl “legs” are constrained in a configuration opposite to that observed in the complex, then there is a dramatic loss of arrestin recruitment, but not of G protein coupling. These findings are consistent with a previous study in which the spatial configuration of the diethylamide legs was shown to be important for their potency in rats(Nichols et al., 2002). Together, the data strongly indicate that stereoselective binding of LSD to the receptor is critical for its ability to preferentially recruit arrestin and, ultimately, for its psychedelic properties in animals.

An LSD-triggered acid trip lasts 6–15 hours even though its clearance from the body is much faster (t_1/2 of around 3.6 hours) (Dolder et al., 2015). So what underlies the prolonged effect of LSD? Wacker et al. provide a potential molecular explanation: an extracellular loop of the HT_2B receptor (and likely of HT_2A) forms a lid covering the ligand-binding site, thereby dramatically slowing the dissociation of LSD from the receptor. Residues in the lid directly interact with LSD, and disrupting this interaction by site-directed mutagenesis results in ~10× faster off-rate and hence ~10× shorter residency time of LSD. Remarkably, a mutant that reduces interactions between the lid and LSD also attenuates arrestin recruitment while leaving the activation of G protein signaling intact, supporting the idea that specific contacts between the lid and LSD are critical for stabilizing the receptor in a conformational state that promotes the psychedelic effects of the drug.

An important caveat is that although LSD preferentially promotes the recruitment of arrestin relative to compounds like lysergamide, not all hallucinogenic drugs acting at 5HT receptors are necessarily dependent on arrestin (Schmid et al., 2008). In fact, 5-HT differs from its psychoactive analogs by its ability to recruit β-arrestin2 (Schmid and Bohn, 2010). Thus, the molecular bases underlying a complex process such as hallucination are likely to be strongly ligand and context dependent, and the role of arrestin itself in the LSD-mediated psychedelic process merits deeper investigation. The work reported by Wacker et al. is nonetheless important not only because of its historical interest, but also because it provides a roadmap for the rational design of drugs that are more biased towards desirable outcomes. In the case of LSD, this would entail the synthesis of variants that retain therapeutic benefit without psychedelic effects, but it remains to be seen if these can truly be separated.