Cortical structural differences following repeated ayahuasca use hold molecular signatures

Abstract

Introduction: Serotonergic psychedelics such as ayahuasca are reported to promote both structural and functional neural plasticity via partial 5-HT_2A agonism. However, little is known about how these molecular mechanisms may extend to repeated psychedelic administration in humans, let alone neuroanatomy. While early evidence suggests localised changes to cortical thickness in long-term ayahuasca users, it is unknown how such findings may be reflected by large-scale anatomical brain networks comprising cytoarchitecturally complex regions.

Methods: Here, we examined the relationship between cortical gene expression markers of psychedelic action and brain morphometric change following repeated ayahuasca usage, using high-field 7 Tesla neuroimaging data derived from 24 members of an ayahuasca-using church (Santo Daime) and case-matched controls.

Results: Using a morphometric similarity network (MSN) analysis, repeated ayahuasca use was associated with a spatially distributed cortical patterning of both structural differentiation in sensorimotor areas and de-differentiation in transmodal areas. Cortical MSN remodelling was found to be spatially correlated with dysregulation of 5-HT_2A gene expression as well as a broader set of genes encoding target receptors pertinent to ayahuasca's effects. Furthermore, these associations were similarly interrelated with altered gene expression of specific transcriptional factors and immediate early genes previously identified in preclinical assays as relevant to psychedelic-induced neuroplasticity.

Conclusion: Taken together, these findings provide preliminary evidence that the molecular mechanisms of psychedelic action may scale up to a macroscale level of brain organisation in vivo. Closer attention to the role of cortical transcriptomics in structural-functional coupling may help account for the behavioural differences observed in experienced psychedelic users.

Keywords: 5-HT2A; ayahuasca; morphometry; psychedelics; transcriptomics; ultra-high field MRI.

Figure 1

Morphometric similarity analyses of repeat ayahuasca usage. (A) Regional distribution of morphometric similarity (MS) in Santo Daime members and matched controls. (B) Case–control distributions of residual morphometric similarity, following regression of sex and age. (C) t-statistic and FDR flagged (p < 0.05) regions for differences in MS between groups (ayahuasca – controls). (D) Top – kernel density scatterplot of the mean regional MS scores of controls (x-axis) and the ayahuasca-control t-statistic (y-axis), bottom – schematic of functional implication of MS scatter plot value distribution. Lighter hues reflect higher value densities. (E) Case–control MS differences relative to Yeo functional and von Economo cytoarchitectural communities. Absolute t-statistics are displayed. Yeo abbreviations correspond to the following: VIS, visual network; DAN, dorsal attention network; SMN, somato-motor network; DA, dorsal attentional network; VA, ventral attention network; L, limbic network; FPN, fronto-parietal network; DMN, default mode network. Von Economo labels reflect the following: Prim motor, granular primary motor cortex; Asso1, granular association isocortex type I; Asso2, granular association isocortex type 2; Sec sens, secondary sensory cortex; Prim sens, primary sensory cortex; Limbic, limbic regions (allocortex including entorhinal, retrosplenial, presubicular and cingulate); Insula, insular cortex (containing granular, agranular and dysgranular regions). For all renders, local maximum values are displayed.

Figure 2

Cortical thickness and ayahuasca use frequency correlations. (A) Spearman correlations of ceremony attendance rates with MS scores. MS scores in FDR flagged regions are aggregated per contrast (positive negative and global, indicated by arrows) and averaged per participant. Scatter plots depict mean regional MS scores of Santo Daime members (x-axis) and corresponding ceremony attendance rates (y-axis). (B) t-statistic and FDR flagged (p < 0.05) regions for differences in CT between groups (ayahuasca – controls). For all renders, local maximum values are displayed.

Figure 3

Transcriptional profiles associated with Santo Daime differences in morphometric similarity. (A) Cortical map of left hemispheric t-values used for PLS. (B) Regional loadings of PLS1 weights. (C) Kernel density scatterplot of the regional PLS1 scores of controls (x-axis) and regional ayahuasca-control left-hemispheric t-statistic (y-axis). Lighter hues reflect higher densities. (D) Significant PLS1 loadings following FDR correction. Gene targets reflect selected markers encoded by gene expression maps. Lighter hues representing positive loadings and vice versa. (E) Scatterplots of top gene target normalised gene expression values derived from the AHBA atlas in relation to regional differences in MS, paired with corresponding renders of their spatial distribution. For all renders, local maximum values are displayed.