Using 7T resting-state fMRI in a double-blind, placebo-controlled crossover study, the authors show that acute 2C-B and psilocybin both reduce intra-network static connectivity while increasing between-network and subcortical–cortical coupling and brain complexity, but 2C-B causes less between-network desynchronisation and relative elevations in transmodal connectivity compared with psilocybin. These spatially divergent effects align with monoaminergic transporter and serotonergic receptor distributions and link transmodal-axis desynchronisation to behavioural measures, highlighting 2C-B as a distinct tool for psychedelic neuroscience and potential pharmacotherapies.
As psychedelic-assisted psychotherapy gains momentum, clinical investigation of next-generation psychedelics may lead to novel compounds tailored for specific populations. 2,5-dimethoxy-4-bromophenethylamine (2C-B) is a psychedelic phenethylamine reported to produce less dysphoria and subjective impairment than the psychedelic tryptamine psilocybin. Despite its popularity among recreational users and distinct pharmacodynamics, the neural correlates of 2C-B remain unexplored. Using 7 T resting-state functional MRI in 22 healthy volunteers, we mapped out the acute effects of matched doses of 20 mg 2C-B, 15 mg psilocybin and placebo across spatiotemporal benchmarks of functional brain organisation. In a within-subjects, double-blind, placebo-controlled crossover design, we evaluated the neuropharmacological and neurobehavioural correlates of an array of connectivity measures - including static (sFC) and global connectivity (gFC), dynamic connectivity variability (dFC), and spontaneous brain complexity. Compared to placebo, 2C-B and psilocybin selectively reduced intranetwork sFC, while broadly increasing between-network and subcortical-cortical connectivity. Compared to psilocybin, 2C-B exhibited less pronounced reductions in between-network dFC but elicited elevations in transmodal sFC. Both compounds yielded spatially divergent increases in gFC yet produced similar increases in brain complexity. Using PET density modelling, the spatial distribution of neural effects aligned with documented differences in monoaminergic transporter and serotonergic receptor binding affinity beyond 5-HT2A, highlighting the role of pharmacology in shaping functional dynamics. Lastly, we show behavioural markers of psychedelic effects are reflected by the decoupling of the transmodal axis of functional brain organisation. Together, our findings highlight 2C-B as a useful new addition to the study of psychedelic neuroscience and may motivate new pharmacotherapy strategies.
Papers cited by this study that are also in Blossom
Daws, R. E., Timmermann, C., Giribaldi, B. et al. · Nature Medicine (2022)
As clinical interest in psychedelic-assisted psychotherapy grows, there is increasing motivation to characterise next-generation compounds that may offer improved tolerability profiles for specific patient populations. 2C-B (2,5-dimethoxy-4-bromophenethylamine) is a phenethylamine psychedelic reported by recreational users to produce less dysphoria and cognitive impairment than classical tryptamine psychedelics such as psilocybin, despite broadly comparable subjective effects at matched doses. Its neural correlates had not previously been characterised in humans. This study mapped the acute effects of 2C-B and psilocybin on functional brain organisation using 7T resting-state fMRI in healthy volunteers, benchmarking findings against established neuroimaging signatures of psychedelic action and examining the relationship between brain changes, pharmacological receptor maps, and subjective experience.
Twenty-two healthy volunteers participated in a within-subject crossover design receiving 20 mg 2C-B, 15 mg psilocybin, and placebo in separate sessions. Doses were selected to produce psychotropic equivalence verified by visual analogue scales administered immediately before and after resting-state fMRI acquisition at 7T. Spatiotemporal benchmarks of functional brain organisation were assessed using three primary outcome measures: static functional connectivity (sFC), dynamic functional connectivity variance (dFC), and sample entropy (sampEn). Drug effects were analysed using network-based statistics and linear mixed effects models. The pharmacological specificity of brain changes was assessed through dominance analysis relating regional drug effects to 5-HT2A receptor density and monoamine transporter maps from publicly available receptor atlases.
Both 2C-B and psilocybin produced comparable subjective effects during fMRI acquisition, confirming psychotropic equivalence. Widespread alterations in spatiotemporal brain organisation were observed under both compounds, broadly consistent with prior reports of diminished unimodal-transmodal network segregation under classical psychedelics. Psilocybin selectively reduced within-network connectivity in visual networks A and B and default mode network C, whilst 2C-B showed a distinct spatial profile with differences in the degree and distribution of network changes. Dominance analysis revealed that changes in dFC under both compounds were best explained by an inverse relationship with 5-HT2A receptor density — regions of higher 5-HT2A expression showed greater reductions in dynamic FC variance. However, 2C-B's effects were somewhat less spatially constrained to 5-HT2A expression than psilocybin's, consistent with 2C-B's additional agonism at other receptor subtypes beyond 5-HT2A.
The finding that 2C-B and psilocybin produce partially overlapping but spatially distinct patterns of functional reorganisation — despite subjective equivalence — supports the hypothesis that pharmacological differences beyond 5-HT2A agonism shape the neural signature of psychedelic action. The tight coupling between regional 5-HT2A receptor density and dynamic FC changes for both compounds confirms 5-HT2A as the primary mediator of broad-network psychedelic effects, whilst compound-specific residuals implicate contributions from other binding targets. These results provide the first characterisation of 2C-B's neural correlates in humans, addressing a notable gap given the compound's recreational prevalence. Psychotropic equivalence does not imply neuroimaging equivalence — a distinction with practical implications for the use of novel compounds in clinical and research contexts where specific neural signatures may carry therapeutic relevance.
7T resting-state fMRI demonstrates that 2C-B and psilocybin produce widespread yet spatially distinct alterations of functional brain organisation despite comparable subjective effects at matched doses. Both compounds' effects are primarily governed by 5-HT2A receptor density, consistent with this receptor's role as the principal mediator of classical psychedelic brain changes. The distinct spatial profile of 2C-B relative to psilocybin — attributable in part to its additional receptor interactions — warrants further investigation in both basic and clinical psychedelic research settings.
There is currently intense clinical interest in the use of psychedelics for a range of neuropsychiatric disorders. Current research suggests that "classical" candidate compounds such as psilocybin partly exert their therapeutic effects by temporarily disrupting maladaptive functional brain organisation via 5-HT 2A agonism (1, 2). Several human resting- state functional magnetic resonance imaging (rsfMRI) studies have reported classical psychedelics acutely decrease blood-oxygen-level-dependent (BOLD) signal variance and broadly increase functional connectivity (FC) across the neocortex. These findings underlie several theories implicating reduced functional network segregation and increased neural complexity (often referred to as entropy) as the basis of psychedelic action. As the field advances, there is now growing attention towards next-generation psychedelics which may elicit similar cortical effects. Developing compounds that induce milder, shorter, or more tolerable altered states of consciousness may be more feasible to scale in clinical practice while offering new mechanistic insights due to diverging pharmacodynamics. One such compound is the novel phenethylamine 4-bromo-2,5-dimethoxyphenethylamine (2C-B). While 2C-B is the most frequently used novel psychedelic among recreational drug usersand has a history of therapeutic use, scarce clinical research data exists on its effects. Compared to the prototypical psychedelic tryptamine psilocybin, 2C-B appears to induce less subjective impairment and dysphoria at comparable doses, while producing similar perceptual and cognitive effects, making it potentially easier to tolerate. A partial agonist, 2C-B engages multiple serotonin receptors beyond 5-HT 2A , including 5-HT 1A , 5-HT 1B and 5-HT 2C while sharing a similar primary binding profile to psilocin, the active biometabolite of psilocybin. Experiential differences between these two compounds may be attributed to their functional selectivity for 5-HT 2A . receptors; with 2C-B showing approximately tenfold greater selectivity for 5-HT 2A over 5-HT 1A compared to psilocybin, which may bias downstream signalling cascades. In parallel, it is plausible that any of 2C-B's distinct entactogenic and cortical effects are due to its reversal of monoamine transporters otherwise trivial to classical psychedelics. Polypharmacology is thus likely to contribute to distinct neural circuits being modulated by distinct psychedelics, given the differing role of monoaminergic systems on cortical signalling. Neuromodulatory systems tune the neural 'gain' of receptive neuronal populationsthereby modulating inter-areal communication, with pronounced effects at a network level (22). Under psychedelics, changes in neural dynamics and phenomenology have been attributed to distinct serotonergic receptors. Some studies have retrospectively investigated such dependencies for brain organisation and subjective experience across psychedelics. However, within-subject comparative studies, which minimise interstudy and intersubject variability, are still lacking. Moreover, heterogeneous rsfMRI approaches lacking independent replication have been assessed as experiential markers with varying degrees of success. Understanding that no single measure of psychedelic brain effects is yet superior, a perhaps more cogent aim is to define focal neural regions across psychedelics that are differentially sensitive to drug-induced perturbations -regardless of the assessment method. Here, we examined 2C-B's effects on functional brain organisation using a double-blind, crossover design against placebo and matched doses of psilocybin as an active drug control condition in healthy volunteers using 7 Tesla rsfMRI. A first direct comparison of two structurally distinct psychedelics, we hypothesised that 2C-B would elicit alterations to spatiotemporal benchmarks of brain organisation consistent with a primary 5-HT 2A mechanism of action, that is -the presence of a small-world network organisation and increased neural complexity. Additionally, we hypothesised that monoamine reuptake inhibition would significantly explain topographical differences between 2C-B and psilocybin. To do so, we employed a pharmacology-guided approach leveraging positron emission tomography (PET) mapping. Finally, we investigated potential relationships between multivariate markers of brain organisation and subjective experience.
All aspects of this study (trial register NL8813) were approved (NL73539.068. (approximately 90 and 120 minutes post-drug intake). These assessments were used to reassess dose equivalency.For more details, see the Supplementary Materials.
Participants underwent one imaging session per acute dosing visit separated by a two-week washout, comprising structural imaging, magnetic resonance spectroscopy, task and rsfMRI. Across visits, imaging was performed within subjects at a fixed time of day to minimise diurnal effects. Images were acquired on a Siemens MAGNETOM 7T MRI scanner. Imaging parameters and procedures are detailed in the Supplementary Materials. Neuroimaging data was preprocessed using fMRIPrep 1.1.147
We derived spatiotemporal measures of resting-state network organization at both edge and regional levels to benchmark 2C-B's effects. These measures have been extensively studied in the psychedelic literature. All connectomic measures were contextualized using the Yeo et al. 17 rsfMRI network classes. Static functional connectivity (sFC) was computed by generating Pearson's r between the timeseries of all pairwise combinations of 232 regions, with all r-values Fisher ztransformed. Global functional connectivity (gFC) serving as a proxy for node centrality, was calculated as the average Fisher z-transformed Pearson correlation from each region to all others.We focused on significant edges flagged by sFC statistics. To assess dynamic FC variability (dFC), we computed the variance of dynamic conditional correlations (DCC). DCC is a non-parametric model-based approach devised for assessing time-varying correlations. A framewise approach, it offers superior test-retest reliability compared to other dFC methods, which often face window artefacts and require arbitrary hyperparameters. Variance was calculated for all possible pairwise regional DCC combinations. We selected and ranked spontaneous signal complexity measures according to previous replicability and association with psychedelic effects based on McCulloch et al.'s work, using the COPBET toolbox (31). Regional signal complexity was characterised using sample entropy (sampEn), defined as the negative logarithm of the conditional probability that two vectors of length m (set to 2) are dissimilar within a threshold distance r (set at 0.3 times the standard deviation of the regional timeseries) and that this dissimilarity persists when the vector length increases to m + 1. From the same selection criteria, we also derived edgewise DCC Shannon entropy (dccEn), whole-brain Lempel-Ziv entropy (ZivEn) and pathlength degree-distribution entropy (degreeEn). See Supplementary Materials for information on each measure.
Molecular densities were estimated from PET tracer studies for relevant serotonin receptors (5-HT 1A , 5-HT 1B , 5-HT 2A ) and monoamine transporters (serotonin reuptake [SERT], dopamine reuptake [DAT] and norepinephrine reuptake transporter [NET]). Targets were selected from literature demonstrating functional affinity of either 2C-B or psilocybin for these sites. Volumetric PET images were registered to standard space, averaged across participants within each study, and then parcellated into ROIs. Receptors having more than one mean image of the same tracer (5-HT 1B ) were combined using a weighted average. Regional values for cortical and subcortical areas were z-scored independently before their integration for analysis, considering established differences in radioligand uptake between these structures. For details on each PET dataset, see. Dominance analysis was used to determine the relative importance of each PET density map in predicting regional contrast t-statistics for our primary imaging benchmarks (nodal sum sFC & dFC, sampEn). Dominance analysis aims to establish the relative significance (or "dominance") of each independent variable in relation to the overall fit (adjusted R 2 ) of the multiple linear regression. We also independently investigated the associative validity of the top topographical contributors for each metric and by performing spatial autocorrelationpreserving Pearson correlations. We generated spatially constrained nulls (10 000 iterations) using Moran Spectral Randomisation as implemented in the BrainSpace toolbox, a process first proposed in the ecology literature. This approach is best suited to assess the concordance between two spatial maps in which a geodesic projection cannot be ensured (e.g atlases comprising subcortical structures). See Supplementary Materials for additional details.
We assessed the behavioural relevance of "net" drug effects of both 2C-B and psilocybin on functional organisation by investigating the predictive value of regional multivariate coherence. Multivariate coherence, as presented herein, is defined as the Pearson correlation across measurements for a particular node. For each participant, primary nodal psychedelic effect measures (nodal sum sFC & dFC, sampEn) were z-scored across regions to control for inter-feature variability. Pearson correlations were then performed for each pair of regional feature vectors, forming an ROI x ROI matrix per participant, ultimately summed to generate a measure of nodal coherence strength per region. By performing a multivariate integration of measures of interest, this approach provides a simple regional summary of "net" drug effects while diminishing the number of potential multiple comparisons. This approach is comparable to generating annotation similarity networks. We employed a multilevel partial-least-squares (PLS, accounting for inter-individual variability) analysis to assess the relationship between drug-induced effects on regional coherence and corresponding subjective effects (). PLS is an unsupervised multivariate statistical approach that relates two feature matrices (X, Y) by estimating ther weighted linear combination to maximise their covariance. Here, X corresponds to multivariate coherence, with n being subjects and r being coherence change scores per condition (X nxr ) and Y reflecting experiential measures, with n being subjects and d corresponding to change scores for each of our 6 retrospective dimensions (5D-ASC, EDI) per condition (Y nxd ). This decomposition yields orthogonal latent variables (LVs) associated with a pattern of neural activity (i.e., multivariate coherence) and subjective effects. We assessed overall LV significance using permutation testing (10 000 permutations), and loading stability by generating bootstrap ratios (10 000 bootstraps). Overall model findings were also supported using 5-fold cross-validation, to derive out-of-sample correlation estimates. See Supplementary Materials.
For all analyses, the alpha criterion was set at p <G0.05 following appropriate multiple comparison correction. A network-based statistic (NBS) approach (57) derived from the publicly available NBS toolbox was used to assess alterations to connectomic measures in a repeated measures design (see Supplementary Materials). Significant NBS (10 000 permutations) multiplecomparisons corrected edges were used as masks for post-hoc t-tests assessments. Changes to regional rsfMRI measures and behavioural outcomes were assessed by linear mixed effect models (LMEMs) with drug condition as a main fixed effect and participant as a random intercept. Follow-up assessments were performed with paired t-tests after Benjamini-Hochberg false discovery rate (FDR) correction across significant ROIs. For behavioural outcomes, Tukey's method was applied.
We sought to define 2C-B's effects on resting-state functional brain organisation and explore relative differences to psilocybin across a range of established psychedelic neuroimaging benchmarks. Administration of 2C-B and psilocybin yield comparable psychoactive effects. We first aimed to ensure psychotropic equivalence during resting-state fMRI acquisition (Figure, Table). Analyses revealed a significant drug effect on all VAS items immediately before (approximately 90 min post-administration) and after (approximatley 120 min postadministration) resting-state acquisition indicating comparable subjective effect intensity at the time of acquisition. The maximal drug effects (e max ) for our real-time measures were equivalent for both 2C-B and psilocybin, consistent with previous findings. Retrospective dysphoric experiences related to changes in to selfhood ('ego dissolution'), specifically anxious ego dissolution (AED), but not the Ego Dissolution Index (EDI), as well as the overall 5D-ASC (ASC) score, were significantly higher under psilocybin compared to both placebo and 2C-B. Following scanner exit, participants reported the phenomenological content of their restingstate mentation using the ARSQ. Analyses demonstrated a significant drug effect across 6/10 dimensions of resting-state cognition, with numerically large increases for both drugs relative to placebo in discontinuity of mind and self-referential thinking. However, neither compound exhibited significant differences in any ARSQ dimension during acquisition. Particularly, measures of in-scanner arousal (sleepiness) and comfort did not show compelling differences.
Widespread changes in spatiotemporal brain organisation were observed under both drugs, consistent with prior reports of diminished unimodal-transmodal organization under psychedelics (see Figure). All outcomes were largely consistent after controlling for FD, across different parcellation schemes and global signal regression (see Supplementary). For sFC, NBS identified widespread differences in sFC as indicated by a main effect of drug ( p NBS = 0.0421, 229 nodes, 1922 edges). Consistent with previous studies, psilocybin significantly reduced within-network connectivity for visual network A & B (VIS A/B ) and the default mode network C (DMN C ), compared to placebo. Network segregation, particularly across the frontoparietal network A & B (FPN A/B ), salience network A & B (SAN A/B ), and DMN A/B/C , was also significantly reduced. In addition, cortical-subcortical connectivity was broadly increased. 2C-B showed similar effects, with reductions in within-network sFC for VIS B and DMN C , but showed converse increases in FPN C . Reductions in network segregation were focal, notably between FPN A/B and VIS A/B , and SAL B with transmodal networks. Subcortical-cortical sFC, especially involving DMN B and VIS B , was significantly increased. Between-drug comparisons revealed that psilocybin had a generally greater reducing effect on between-network sFC relative to 2C-B, particularly for connectivity spanning the DMN B and somatomotor network B (SMN B ), whereas 2C-B showed greater increases in DMN A -FPN C connectivity. These results align with a predominantly cortical effect, highlighting the significance of transmodal cortices. For dFC, NBS analyses also revealed a widespread effect of drug (p NBS = 0.0135, 231 nodes, 1755 edges). Both psilocybin and 2C-B significantly induced strong, nonspecific whole-brain reductions in between-network dFC variance compared to placebo, with the greatest decreases observed in subcortical-cortical (DMN B , FPN A/B ) and VIS A -FPN A connectivity variance. No clear within-network effects were visible under either compound. Post-hoc comparisons across compounds revealed overall spatially distributed greater reductions in dFC under psilocybin relative to 2C-B for between-network dFC. Edgewise complexity (dccEn) did not exhibit significant changes (p NBS > 0.1). In parallel, we calculated regional measures of gFC and several complexity measures (Figure, Supplementary). gFC analyses showed significant increases in mean FC across most regions for both substances compared to placebo. Under psilocybin, notable increases were seen in the left dorsolateral and bilateral ventral prefrontal cortex, inferior frontal gyrus, and left ventral posterior thalamus, with decreases in the right temporal pole, medial amygdala, dorsal posterior thalamus, and occipital lobe. For 2C-B, the strongest gFC increases occurred in the right precuneus, lateral prefrontal cortex, left posterior cingulate cortex, and right insula, with reductions in the bilateral temporal poles, right dorsal posterior thalamus, and left nucleus accumbens. Between drugs, psilocybin generally led to greater gFC increases compared to 2C-B, particularly in the left postcentral gyrus, medial cingulate cortex, ventral posterior thalamus, and frontal eye fields. In contrast, 2C-B showed greater increases in the right extrastriate visual cortex, bilateral posterior cingulate, medial prefrontal cortex, and left dorsal anterior thalamus. Controlling for motion, GSR exhibited drug-dependent effects (see Supplementary Materials). Analyses of regional complexity using sampEn revealed comparable effects for 2C-B and psilocybin, indicating an increase in time series complexity. Specifically, both compounds significantly increased sampEn in the bilateral extrastriate/primary visual cortex and the anterior ventral thalamus, with no significant differences observed between the compounds. Whole-brain signal complexity (zivEn but not degreeEn) was also significantly increased relative to placebo, with no significant differences between compounds (see Supplementary Materials).
Through dominance analysis, we compared the strength of primary regional metrics' associations with pharmacological receptor and transporter maps (see Figure). Separate models were generated for each primary outcome measure (sFC, dFC and sampEn nodal sum t-stats) per contrast and were supported by independent correlations (Figure). Across 54 possible models, we found that for both 2C-B and psilocybin, dFC changes from placebo were best explained (Psilocybin: 56.89%, 2C-B: 64.38 %) by an inverse relationship with 5-HT 2A receptor density indicating that dFC variance decreases relative to placebo, 5- HT 2A spatial density increases. Similarly, sampEn was best explained for each drug (Psilocybin: 48.26%, 2C-B: 45.40 %) by an inverse relationship with 5-HT 1A spatial density, suggesting increased signal complexity relative to placebo correlates with lower 5-HT 1A levels. For 5-HT 1B , there was a significant positive relationship with sFC under psilocybin (41.89%), suggesting greater increases in sFC for regions of high 5-HT 1B density. Importantly, when assessing relative differences across drugs, DAT density was a strong determinant of topological differences in sFC between 2C-B and psilocybin (65.75%). Specifically, areas with greater sFC under 2C-B relative to psilocybin correlated with higher DAT density and vice versa. Whereas SERT was found to be a strong contributor to sampEn between-drug differences (49.25%) independent correlations revealed a non-significant, positive relationship. Lastly, 5-HT 1A was found to also significantly contribute to betweendrug differences (34.36%): in regions with lower 5-HT 1A density, psilocybin tended to exert a stronger effect on reducing dFC variance compared to 2C-B ( and vice versa).] Multivariate neural effects show distinct associations with subjective experience We lastly sought to understand how net effects on functional brain organisation may relate to a participant's subjective experience. To do so, we generated multivariate coherence change scores per region and drug, summarising the nonlinear effects of each compound across largely uncorrelated measures (see Supplementary Materials). As depicted in Figure, multivariate coherence followed a unimodal-transmodal topographical organisation. Primary networks (e.g., VIS, SM) closely coupled to sensory input, showed high multivariate coherence and responded similarly across measures, while more flexible heteromodal networks exhibited lower coherence, responding asynchronously. Levene's testing indicated the variance of regional multivariate coherence scores differed across conditions (W = 22 p < 0.0001). We next applied a multilevel PLS analysis to identify a multivariate mapping of maximum covariance between drug-induced effects on brain organisation and subjective effects. Permutation testing revealed a statistically significant latent variable (PLS1), explaining 33.13 % of total cross-drug experiential scores (p perm = 0.0291). Across contrasts, regional coherence PLS1 scores were positively associated with experiential PLS1 scores (overall: r in- sampl e = 0.40, r out--sample = 0.32, p FDR = 0.0151; Psilocybin: r in-sample = 0.79, p FDR = 0.0001, 2C-B: r in-sample = 0.53 , p FDR = 0.0151) , suggesting PLS1 captured a pattern in multivariate coherence that is positively associated to subjective effects. To evaluate spatial patterns of covariance relevant to each drug's effects, we extracted PLS1 regional loadings per condition after bootstrapping (Figure). Regions with positive loadings in PLS1 indicate that greater coherence following drug intake is linked to stronger subjective effects, while regions with negative loadings suggest that greater coherence in those areas is linked to weaker subjective effects and vice versa. Under psilocybin, the strongest negative loadings were in the prefrontal cortex (PFC), including SAL B , DMN A , and FPN B (e.g., bilateral ventral, dorsal, and lateral PFC), while positive loadings appeared in primary sensory regions (e.g., striate/extrastriate cortex, precentral gyrus). 2C-B showed a similar spatial pattern, with subtle regional differences. Negative loadings were strongest in regions associated with FPN B/C and DMN C (e.g., bilateral inferior parietal lobe, dorsal and lateral PFC), and positive loadings appeared in areas linked to SM A/B and DAN B (e.g., primary/secondary motor cortex, postcentral gyrus). Overall, PLS1 suggests that a transmodal cortical gradient best characterises psychedelic effects. Cortical PLS1 bootstrap ratios correlated with the primary gradient of cortical FCand the sensorimotor-associative (S-A) axis of brain organisation (59) following spin permutation tests (FC gradient: r = 0.50 , p FDRspin < 0.0001; S-A axis: r = -0.51 , p FD Rspin < 0.0001, see Supplementary). Across psychedelics, regional neurobehavioral markers are thus marked by reduced multivariate coherence in transmodal networks (including regions such as the PFC and temporoparietal junction) and increased coherence in sensorimotor cortices.
This study aimed to characterise the neural correlates of 2C-B using an integrative approach that combines spatiotemporal measures of functional organisation, molecular pharmacology, and subjective experience. Consistent with our hypothesis, we found that 2C-B exhibits empirical properties similar to classical psychedelics, including network desegregation and increased spontaneous BOLD complexity, irrespective of variations in rsfMRI thought content or effect intensity. By examining functional (sub)networks, we identified coupling changes to subcortical and prefrontal regions as well as transmodal network desegregation across compounds. Furthermore, we confirm that regional differences between 2C-B and psilocybin effects are associated with reported pharmacodynamic differences in 5-HT 1A and DAT affinity, with serotonergic receptors sharing a common role in defining functional dynamics. Characterising nonlinear psychedelic effects, we isolated neurobehavioral markers for each compound, contributing to a net destabilisation of prefrontal cortices. Our findings suggest that 2C-B is a valuable tool for psychedelic neuroscience, while enhancing our understanding of the differing neurobiological mechanisms of psychedelics. One of the strongest findings in psychedelic research is the reduction of intranetwork sFC for the VIS and DMN, often termed network 'disintegration'. This effect is observed across various psychedelics and entactogens, such as lysergic acid diethylamide (LSD), ayahuasca, dimethyltryptamine (DMT), psilocybin, and MDMA. Given 2C-B's compatible phenomenology, its similarity here is expected. Previous studieshave often used broad network definitions (e.g.or methods like independent components analysis (ICA), which may overestimate intranetwork effects by examining coherent neural activity across extended parts of the brain. Here for example, while within-network sFC was broadly reduced for the VIS under 2C-B and psilocybin, only the DMN C (comprising the retrosplenial cortex, parahippocampus, and anterior hippocampus) showed reduced sFC. This supports the need for finer (sub)network definition, as highlighted by a recent precision hypersampling approach demonstrating that psilocybin's enduring effects are localised to altered hippocampal-cortical connectivity. It is theorised that the degree of whole-brain functional integration underlies the depth of the psychedelic experience. Psychedelics with rich phenomenological profiles, such as DMT show a substantial compression of functional cortical organisation, whereas minimally "disruptive" compounds such as MDMA or dexamphetamine, result in less extensive changes. In our study, compared to psilocybin, 2C-B demonstrated less extensive unimodal-transmodal internetwork sFC (e.g., DMN B , SM A ) and subcortical-cortical connectivity, but exhibited enhanced transmodal connectivity (e.g. DMN C -FPN A ). The use of gFC revealed marked increases in connectivity within thalamic nuclei involved in sensorimotor gatingand flexible network reconfigurationand substantial portions of the neocortex for both compounds. Interestingly, 2C-B was associated with marked increases in gFC and sFC compared to psilocybin for areas such as the mPFC (subdivisions of corresponding to the FPN and DMN), which impacted a broader extent of the neocortex. Sparse increases in FPN sFC after MDMAand reductions in centrality after SSRIs have been observed. Whereas our findings follow those of groups showing increased transmodal and reduced unimodal gFCdiverging findings have been identified. Further work is needed to reconcile these differences given that incorporating GSR seemingly taps into latent pharmacological properties (see Supplementary). Temporal assessments of brain organisation across substances revealed significant reductions in dFC variance across the brain and increased spontaneous signal complexity. Notably, 2C-B yielded less widespread reductions in dFC variance compared to psilocybin while increases in regional entropy were consistent between the two compounds, indicating that the between-drug differences in the minimisation of drastic connectivity changes are independent of an increased prevalence of unique signalling motifs. This interpretation may be mutually equivalent to results from cluster-based dynamical analyses, showing greater occurrence rates of hyperconnected FC states under psychedelics. Acute reductions to dFC variance parallel to numerically increased entropy, have also been observed after inhalation of the dissociative hallucinogen Salvia divinorum. Doss and colleagues however report subacute increases in dFC variance following psilocybin-assisted therapy for depression, which they refer to as a marker of increased 'neural flexibility' following treatment -otherwise absent in healthy volunteers. Further replication is needed to validate the concept of neural flexibility in the acute psychedelic state and its relationship to longer-term effects. Whole-brain complexity, which refers to the irreducibility of brain activity from functional integration and differentiation, is considered a hallmark of the phenomenological richness of psychedelics. Eliciting similar behavioural effects, 2C-B also produced comparable increases in complexity to psilocybin. It remains to be studied however whether this is a feature inherent to 5-HT 2A agonism, since MDMA or amphetamines have yet to be assessed. We particularly observed increased complexity in thalamic nuclei under psychedelics, aligning with the 'thalamic gating' model of psychedelic effects that suggest psychedelics disinhibit thalamic connections, allowing more sensory information to reach the neocortex. It should be noted that sparse regional elevations in entropy were identified, with no significant changes in edge-based complexity. This may be reflective of our choice of eyesopen rsfMRI to minimise arousal differences as it provides a more 'grounded' psychological state, akin to task-based approaches where attention is directed towards external stimuli rather than internal psychological processes. Studies indicate that complexity scales with cognitive load (77) and external stimulation under psilocybin, with no significant changes observed during eyes-open resting EEG. A more definitive consensus on the applicability of different measures of complexity could thus be gleamed from block designs increasing in task demand. For the development of novel compounds with refined affinities in psychedelic-based pharmacotherapy, it is crucial to understand how molecular pharmacodynamics translate into cortical markers of therapeutic benefit. Extending a body of work relating 5-HT 2A agonism to altered regional dynamics (24-26) our ranked modelling approach identifies regional 5-HT 2A expression to most strongly align with greater reductions in dFC across psychedelics relative to placebo. Concurrently, regional 5-HT 1A density was found to correlate with changes in sampEn relative to placebo and dFC differences under psilocybin relative to 2C-B. Coexpressed, 5-HT 1A autoreceptors act in opponency to 5-HT 2A receptors to dampen the spiking output of cortical pyramidal neurons, attenuating changes to cortical excitability (80) and perhaps by extension, signal complexity. Recent theoretical modelling of cortical dynamics has proposed that between-compound differences in 5-HT 1A agonism may mediate psychological tolerability. Considered a target for antidepressant treatment (82), 5HT 1B emerged as a novel marker for psilocybin's effects on sFC. 5-HT 1B autoreceptors are expressed on cortical serotonergic neurons as well as the raphe and basal ganglia, reducing synaptic 5-HT release via negative feedback mechanisms. Furthermore, 5-HT 1B receptors act as heteroreceptors regulating the release of other neurotransmitters, including GABA and dopamine. Across compounds, we also found that higher DAT density was associated with increased sFC under 2C-B relative to psilocybin. This aligns with 2C-B's secondary monoaminergic binding affinity and preclinical work showing increased striatal dopamine and reduced EEG coherence post-administration. With 2C-B reported to yield less discontinuity of mind compared to other psychedelics, it is worth speculating whether its dopaminergic effects may contribute to this observation. The degree of cognitive lability under psychedelics has been described to be reliant on cognitive control mechanisms which mediate the balance between "stable" and "divergent" thinking via the optimisation of prefrontal-striatal dopaminergic transmission. Exploring ways to operationalise dopaminergic tone to either enhance cognitive flexibility or reduce impairment may provide new options for clinical populations characterised by cognitive rigidity. Our analyses also sought to distinguish behavioural correlates of "net" drug effects on brain function. Here, we found that asynchronicity in transmodal cortices (i.e. reduced coherence), which are involved in high-level cognitive processingserves as a useful neurobehavioral marker of psychedelic effects. Additionally, our results indicate that the behavioural impacts of each drug are particularly contingent on the destabilisation of specific subnetworks within the DMN, FPN and SAL -together forming the triple network of psychopathology posited to underlie the symptomatology of mood disorders. Given that interindividual variability in complex cognition under both normative (91) and psychedelic stateslies in these evolutionarily recent brain regions, clarifying compound-specific regional markers may provide future avenues for personalising psychedelic-assisted therapy. Theoretically, compound selection based on their ability to maximise neural therapeutic effects could be further steered by intervention-based modelling approaches to identify cortical perturbation points. It should be noted that psychedelics impair some aspects of memory recollection in humans (96) which may impact the fidelity of post-hoc subjective ratings and how they correlate with acute, objective brain measures. Analysing task-based fMRI, we will next seek to characterise how nonlinear dynamics may map onto relevant real-time cognitive processes such as emotional decision-making. The present findings are not without limitations. Firstly, a larger sample size would have enabled a more robust detection of between-drug effects and behavioural correlates. Efforts are ongoing to integrate our results into a multi-group consortium characterising psychedelics. Secondly, we acknowledge there is a broader range molecular targets, as well as relevant complexity measures beyond those validated by McCullough et al. This is pertinent given the current absence of in vivo 2C-B receptor occupancy data and the nonspecificity of 5-HT 2A PET tracers (CIMBI36) for 5-HT 2A/2C receptors due to high sequence homology. Thirdly, the timing of our subjective intensity queries makes it difficult to ascertain whether the resting-state fMRI acquisition aligned with the late onset phase or peak psychoactive effects. The magnitude of observed effects may thus differ according to the asssessed temporal phase. Finally, an empirical question persists: how does one best compare 'apples and oranges' in psychedelic neuroscience? The subjective nature and variable duration of these experiences, coupled with a complex underlying polypharmacology beyond 5-HT 2A agonism, raises significant challenges for defining in vivo equivalences. As demonstrated here and by others, a perhaps ecologically sound approach involves matching single doses across relevant behavioral or cardiovascular outcomes. Dose-ascending pharmaco-imaging studies with numerous active controls are likely to be costly and burdensome for participants while further accentuating the risk of order effects due to the persisting effects of psychedelics (98). Moving forward, intelligent comparative study design will be key to evaluating the relative benefits and drawbacks of different psychedelics for therapy. In conclusion, this 7T rsfMRI study demonstrates that 2C-B induces significant multivariate disruptions in the brain's spatiotemporal organisation, expanding our understanding of classic psychedelics. A direct within-subject fMRI comparison of two 5-HT 2A agonists, our findings show secondary pharmacodynamics may actively shape psychedelic effects and highlight new ventures for psychedelic drug development.
Create a free account to open full-text PDFs.
Linguiti, S., Vogel, J. W., Sydnor, V. J. et al. · Neuroscience and Biobehavioral Reviews (2023)
Girn, M., Rosas, F. E., Daws, R. E. et al. · Trends in Cognitive Sciences (2023)
Kwan, A. C., Olson, D. E., Preller, K. H. et al. · Nature Medicine (2022)
Wall, M. B., Harding, R., Zafar, R. et al. · Molecular Psychiatry (2023)
Sessa, B., Fischer, F. M. · Drug Science Policy and Law (2015)
Mallaroni, P., Mason, N. L., Reckweg, J. T. et al. · Clinical Pharmacology and Therapeutics (2023)
Rickli, A., Moning, O. D., Hoener, M. C. et al. · European Neuropsychopharmacology (2016)
Rickli, A., Luethi, D., Reinisch, J. et al. · Neuropharmacology (2015)
Ray, T. S. · PLOS ONE (2010)
Liechti, M. E., Baumann, C., Gamma, A. et al. · Neuropsychopharmacology (2000)
Deco, G., Cruzat, J., Cabral, J. et al. · Current Biology (2018)
Singleton, S. P., Luppi, A. I., Carhart-Harris, R. L. et al. · Nature Communications (2022)
Pokorny, T., Preller, K. H., Kraehenmann, R. et al. · European Neuropsychopharmacology (2016)
McCulloch, D. E-W., Knudsen, G. M., Barrett, F. S. et al. · Neuroscience and Biobehavioral Reviews (2022)
McCulloch, D. E-W., Olsen, A. S., Ozenne, B. et al. · MedRvix (2023)
Studerus, E., Gamma, A., Vollenweider, F. X. · PLOS ONE (2010)
Nour, M. R., Evans, J., Nutt, D. J. et al. · Frontiers in Human Neuroscience (2016)
Timmermann, C., Roseman, L., Haridas, S. et al. · PNAS (2023)
Doss, M. K., Považan, M., Rosenberg, M. D. et al. · Translational Psychiatry (2021)
Doss, M. K., May, D. G., Johnson, M. W. et al. · Scientific Reports (2020)
Barrett, F. S., Doss, M. K., Sepeda, N. D. et al. · Scientific Reports (2020)
Carhart-Harris, R. L., Muthukumaraswamy, S., Roseman, L. et al. · PNAS (2016)
Madsen, M. K., Fisher, P. M., Burmester, D. et al. · Neuropsychopharmacology (2019)
Holze, F., Dolder, P. C., Ley, L. et al. · Neuropsychopharmacology (2020)
Girn, M., Roseman, L., Bernhardt, B. et al. · NeuroImage (2022)
Roseman, L., Leech, R., Feilding, A. et al. · Frontiers in Human Neuroscience (2014)
Tagliazucchi, E., Roseman, L., Kaelen, M. et al. · Current Biology (2016)
Preller, K. H., Burt, J. B., Adkinson, B. et al. · eLife (2018)
Lord, L. D., Expert, P., Atasoy, S. et al. · NeuroImage (2019)
Mortaheb, S., Fort, L. D., Mason, N. L. et al. · Biological Psychiatry (2023)
Carhart-Harris, R. L., Friston, K. J. · Pharmacological Reviews (2019)
Vollenweider, F. X., Preller, K. H. · Nature Reviews Neuroscience (2020)
Ort, A., Smallridge, J. W., Sarasso, S. et al. · iScience (2023)
Mediano, P. A. M., Rosas, F. E., Timmermann, C. et al. · ACS Chemical Neuroscience (2024)
Juliani, A., Chelu, V., Graesser, L. et al. · Biorxiv (2024)
Mallaroni, P., Mason, N. L., Kloft, L. et al. · NeuroImage (2024)
Tolle, H. M., Farah, J. C., Mallaroni, P. et al. · Network Neuroscience (2024)