A mechanistic model of the neural entropy increase elicited by psychedelic drugs

Abstract

Psychedelic drugs, including lysergic acid diethylamide and other agonists of the serotonin 2A receptor (5HT2A-R), induce drastic changes in subjective experience, and provide a unique opportunity to study the neurobiological basis of consciousness. One of the most notable neurophysiological signatures of psychedelics, increased entropy in spontaneous neural activity, is thought to be of relevance to the psychedelic experience, mediating both acute alterations in consciousness and long-term effects. However, no clear mechanistic explanation for this entropy increase has been put forward so far. We sought to do this here by building upon a recent whole-brain model of serotonergic neuromodulation, to study the entropic effects of 5HT2A-R activation. Our results reproduce the overall entropy increase observed in previous experiments in vivo, providing the first model-based explanation for this phenomenon. We also found that entropy changes were not uniform across the brain: entropy increased in some regions and decreased in others, suggesting a topographical reconfiguration mediated by 5HT2A-R activation. Interestingly, at the whole-brain level, this reconfiguration was not well explained by 5HT2A-R density, but related closely to the topological properties of the brain's anatomical connectivity. These results help us understand the mechanisms underlying the psychedelic state and, more generally, the pharmacological modulation of whole-brain activity.

Figure 1

Modelling the effect of 5HT2A-R activation on the whole-brain topographical distribution of entropy. (A) Resting state activity is simulated using the Dynamic Mean-Field (DMF) model, in which each region’s activity is represented by a time series of excitatory firing rates (constrained to 0–15 Hz for visualisation). The probability density function (PDF) and differential entropy (h(X)) of each region is then estimated, obtaining a topographical distribution of entropy values. (B) 5HT2A-R agonism is modelled as a receptor-density-dependent response gain modulation. Black line is the frequency–current (F–I) curve of a population without 5HT2A-R agonism, and coloured curves show the resulting F–I curves of regions with increasing 5HT2A-R agonism. (C) 5HT2A-R activation changes the topographical distribution of entropy with respect to resting state activity, which constitutes the main subject of analysis in this study.

Figure 2

Non-linear heterogeneous increase of entropy following 5HT2A-R activation. (A) Effect of 5HT2A-R agonism on the local entropy each of region in the AAL atlas. See Supplementary Table 1 for abbreviations. Bars indicate the (bilateral) average relative change in local entropy, $\mathrm{\Delta }{h}_n$, and error bars indicate 1 standard deviation across 1000 simulations. (B) Histograms of local entropy values for the condition with (red) and without (blue) 5HT2A-R activation. 5HT2A-R activation increased both the average and the spread of the local entropy distribution. (C) Topographical map of entropy changes. Brain regions are coloured according to their $\mathrm{\Delta }{h}_n$ values. (D) 5HT2A-R agonism changed the topographical distribution of entropy in a heavily non-linear manner. Each circle indicates the averages of each region across 1000 simulations.

Figure 3

Changes in local entropy are explained best by connectivity strength, then receptor density. (A) Changes in entropy were overall independent from receptor density, although (B) they were well predicted by the connectivity strength of each region. We split into groups of low (blue), intermediate (grey) and higher (red) connectivity strength, only those regions with low and high connectivity strength were well predicted by their receptor density. Regions with intermediate strength show no significant relationship with receptor density, but are even more accurately predicted by their strength. (C) Topographical localisation of the three groups, following the same colour code. Low-strength regions are mainly located in the parietal area, while high-strength ones are in occipital and cingulate areas.

Figure 4

Relative changes in entropy are reproduced by a strength-preserving null model of the connectome. (A)–(D) Connectivity matrices used to control the role of local properties of the connectome on $\mathrm{\Delta }{h}_n$. See main text for the description of the matrices and randomisation algorithm. (E)–(G). Scatter plots of $\mathrm{\Delta }{h}_n$ for the human connectome against the three null models. DSPR yielded almost the same results than the human connectome, showing that only local network properties of human connectome are sufficient to capture the effect of 5HT2A-R activation.