Spectral signatures of serotonergic psychedelics and glutamatergic dissociatives
Carla Pallavicini, Martina G. Vilas, Mirta Villarreal, Federico Zamberlan, Suresh Muthukumaraswamy, David Nutt, Robin Carhart-Harris, Enzo Tagliazucchi
Classic serotonergic psychedelics are remarkable for their capacity to induce reversible alterations in consciousness of the self and the surroundings, mediated by agonism at serotonin 5-HT2A receptors. The subjective effects elicited by dissociative drugs acting as N-methyl-D-aspartate (NMDA) antagonists (e.g. ketamine and phencyclidine) overlap in certain domains with those of serotonergic psychedelics, suggesting some potential similarities in the brain activity patterns induced by both classes of drugs, despite different pharmacological mechanisms of action. We investigated source-localized magneto-encephalography recordings to determine the frequency-specific changes in oscillatory activity and long-range functional coupling that are common to two serotonergic compounds (lysergic acid diethylamide [LSD] and psilocybin) and the NMDA-antagonist ketamine. Administration of the three drugs resulted in widespread and broadband spectral power reductions. We established their similarity by using different pairs of compounds to train and subsequently evaluate multivariate machine learning classifiers. After applying the same methodology to functional connectivity values, we observed a pattern of occipital, parietal and frontal decreases in the low alpha and theta bands that were specific to LSD and psilocybin, as well as decreases in the low beta band common to the three drugs. Our results represent a first effort in the direction of quantifying the similarity of large-scale brain activity patterns induced by drugs of different mechanism of action, confirming the link between changes in theta and alpha oscillations and 5-HT2A agonism, while also revealing the decoupling of activity in the beta band as an effect shared between NMDA antagonists and 5-HT2A agonists. We discuss how these frequency-specific convergences and divergences in the power and functional connectivity of brain oscillations might relate to the overlapping subjective effects of serotonergic psychedelics and glutamatergic dissociative compounds.
Keywords : serotonergic psychedelics; ketamine; dissociatives; consciousness; machine learning; magnetoencepalography
Historically, the term ‘hallucinogen’ has been used to refer to several categories of compounds capable of eliciting marked altered states of consciousness e.g. characterized by altered perception of self, time, space and one’s surroundings, plus eyes-closed complex and vivid visual imagery (Nichols 2004). Within this generic category exist the more specific categories of serotonergic psychedelics (SP) and N-methyl-D-aspartate (NMDA) receptor antagonist glutamatergic dissociatives (GD1). These drugs have been
investigated in humans at the molecular and systems level, as well as in terms of behavioral changes and subjective effects. Both SP and GD are valuable tools to dissect the neural mechanisms associated with conscious experience, and are currently being explored as potential treatments for a variety of neuropsychiatric disorders (Vollenweider and Kometer, 2010; Krystal, Sanacora, & Duman, 2013; Carhart-Harris et al., 2014; Nichols, 2016). The subjective effects elicited by SP and GD, together with human neuroimaging studies of individual compounds, suggest some overlap in their effects on global brain function (Pomarol-Clotet, Honey et al. 2006, Schartner, Carhart-Harris et al. 2017, Preller and Vollenweider 2018). To date, a quantitative comparison of the neurophysiological changes associated with the acute effects of SP and GD remains to be conducted.
At the molecular level, it is known that both SP and GD elicit their pharmacological action by non-selectively binding to receptors associated with different endogenous neurotransmitters. SP (e.g. lysergic acid diethylamide [LSD], psilocybin, N,Ndimethyltryptamine [DMT], mescaline) are at least partial agonists of certain serotonin (5-hydroxytryptamine [5-HT]) receptor subtypes (González-Maeso et al., 2007; Halberstadt, 2015; Nichols, 2016), but have also been shown to bind to other receptors (Ray 2010, Rickli, Luethi et al. 2015, Rickli, Moning et al. 2016), which could influence their subjective effects (Zamberlan, Sanz et al. 2018). Several animal and human studies have established agonism at 5-HT2A receptors as a necessary condition for the characteristic subjective effects elicited by SP (Glennon, Titeler et al. 1984, Titeler, Lyon et al. 1988, Hanks and González-Maeso 2013, Kraehenmann, Pokorny et al. 2017, Kraehenmann, Pokorny et al. 2017, Preller, Herdener et al. 2017, Barrett, Preller et al. 2018, Preller, Burt et al. 2018). Activation of 5-HT2A receptors by SP in turn increases levels of the excitatory neurotransmitter glutamate (Aghajanian and Marek, 1999, 1997). GD (e.g. ketamine, phencyclidine (PCP), dextromethorphan) act as non-competitive antagonists at NMDA (N-methyl-D-aspartate) receptors (Anis et al., 1983; Homayoun and Moghaddam, 2007; Stahl, 2013), also resulting in the elevation of extracellular glutamate levels (Moghaddam et al., 1997). Like SP, GD also modulate the activity of diverse monoamine transporters and opioid receptors (Stahl, 2013); in particular, it has been shown that ketamine and PCP are likely agonists at 5-HT2A and dopamine D2 receptors, with binding affinities comparable with those at NMDA receptors (Kapur and Seeman 2002). Adding to the plausibility of the serotonergic action of ketamine, it has been reported that pre-treatment with 5-HT2A inverse agonist risperidone attenuates metabolic response to ketamine (Deakin, Lees et al. 2008), even though arterial spin labeling measurements revealed the opposite effect (Shcherbinin, Doyle et al. 2015). Animal studies with animals show that ketamine enhances 5-HT2A receptor-mediated vasoconstriction (Park, Noh et al. 2016).
Although SP and GD have some overlapping effects in terms of their effects on glutamatergic activity, their primary molecular sites of action are different. However, the complex series of neurochemical events triggered by these drugs could result in converging changes at the systems level. Administration of SP (e.g. LSD and psilocybin) (Carhart-Harris, Erritzoe et al. 2012, Roseman, Leech et al. 2014) Palhano-Fontes et al., 2015; Carhart-Harris et al., 2016a) and GD (ketamine) (Niesters et al., 2012; Scheidegger et al., 2012) have been shown to decrease the functional integrity of the default mode network (DMN) and increase its functional connectivity with other brain systems (Kometer et al., 2015; Tagliazucchi et al., 2016), and to reduce fronto-parietal connectivity (Muthukumaraswamy et al., 2015, 2013). Reduced broadband oscillatory power (Kometer et al., 2015; Muthukumaraswamy et al., 2013; Riba et al., 2004, 2002) has been observed following the administration of LSD and psilocybin. Such reductions are especially marked in the alpha (8 – 12 Hz) band. On the other hand, sub-anesthetic doses of ketamine are associated with increases in gamma and theta power in anteriorregions, decreases in theta power in the posterior cortex, and decreased amplitude of low frequencies in posterior areas (Muthukumaraswamy et al., 2015). Also convergent between SP and GD were increased global connectivity (Driesen et al., 2013; Tagliazucchi et al., 2016), and increased entropy (Carhart-Harris et al, 2014; Lebedev et al, 2016; Schartner et al, 2017; Tagliazucchi et al, 2014). While a review of the current literature suggests overlapping neural effects of SP and GD, the reliability of such similarities remains hard to assess lacking quantitative comparisons of the drugs.
We introduced a framework to quantitatively estimate the similarity of changes in brain oscillations induced by the administration of different psychoactive drugs. Brain activity under the acute effects of two SP (LSD and psilocybin) and one GD (ketamine) was recorded using magnetencephalography (MEG). We identified changes at the source level (spectral power and functional connectivity) associated with the administration of each substance relative to a placebo following a double-blind experimental design. Next, we modeled linear relationships between the spatial patterns of the changes induced by every pair of drugs. Finally, we trained multivariate machine learning models, i.e. random forests (Breiman 2001), to distinguish each drug from the corresponding placebo using frequency-specific features, and then evaluated the accuracy of generalizing each classifier to distinguish other drugs from the placebo conditions. Since this method relies in relative differences across regions of interest in source space, it circumvents the problem of directly comparing absolute values by means of mass univariate tests, a procedure prone to be confounded by non-comparable doses of the administered drugs.
We hypothesized that spectral features associated with common molecular and subjective effects of SP and GD would be informative of general signatures of the relevant subjective effects, while others would generalize only between both SP, reflecting changes in neural activity specific to agonism at 5-HT2A receptors.