Psilocin, LSD, mescaline, and DOB all induce broadband desynchronization of EEG and disconnection in rats with robust translational validity

Serotonergic psychedelics—principally tryptamine derivatives (for example LSD, psilocybin/psilocin, DMT) and phenethylamine derivatives (for example mescaline, DOB)—produce similar subjective effects in humans largely via 5-HT2A receptor agonism. Human neuroimaging and electrophysiology studies have repeatedly reported a broadband decrease in spectral power (most pronounced in the alpha band) and reductions in functional connectivity during psychedelic states, but comparable whole‑brain EEG data from awake, freely moving laboratory animals are lacking. Earlier animal EEG work was often done under anaesthesia or immobilisation, and recent preclinical recordings have mostly focused on local field potentials rather than multichannel cortical surface signals that approximate human resting‑state EEG. Vejmola and colleagues set out to characterise and compare the effects of four psychedelic compounds—psilocin (4 mg/kg), LSD (0.2 mg/kg), mescaline (100 mg/kg), and DOB (5 mg/kg)—on cortical EEG in freely moving rats. To maximise translational validity they recorded from a multichannel epidural array spanning frontal, sensorimotor and temporal/parietal cortices, analysed only epochs of behavioural inactivity as a resting‑state analogue, and quantified spectral power, topographic distributions and functional connectivity (cross‑spectral and phase‑lagged coherence) across four time windows before and after drug administration. The working hypothesis was that all substances would produce marked decreases in EEG power and connectivity.

Subjects were adult male Wistar rats (initially 12 animals per treatment group). Animals were randomly assigned, pair‑housed prior to surgery, and handled and tested once; ethical approval and EU guidance were followed. Following stereotactic implantation under isoflurane anaesthesia, gold‑plated epidural electrodes were positioned over defined frontal, motor, parietal and temporal cortical coordinates (twelve electrodes for regions of interest), with an additional reference electrode above the olfactory bulb and a subcutaneous ground—yielding the multichannel montage used for analysis. Postoperative analgesia and recovery procedures are described; rats were recorded seven days after surgery. Drugs were synthesised in‑house and administered subcutaneously at doses chosen for comparable behavioural potency in prior discrimination and locomotor studies: psilocin 4 mg/kg, LSD 0.2 mg/kg (freebase), mescaline hydrochloride 100 mg/kg, DOB 5 mg/kg, and saline (2 ml/kg) as control. All injections used 2 ml/kg volume. The investigator was not blinded to treatment. EEG was recorded in the home cage between c.07:30 and 13:00. A 10‑minute baseline was acquired immediately before dosing and recording continued for 90 minutes post‑injection. Data were sampled at 250 Hz with 16‑bit resolution and a ±500 μV dynamic range. Behaviour was scored in parallel (active versus inactive) by an observer blind to treatment; only wakeful inactive epochs were retained for analysis to approximate resting‑state EEG and reduce behavioural confounds. Preprocessing included bandpass filtering (0.5–45 Hz, linear FIR) and segmentation into four analysis epochs: baseline (0–10 min pre‑treatment) and three post‑dose windows (20–30, 50–60, and 80–90 min). Artifact‑free continuous segments of at least 50 s per epoch were selected and subjected to FFT to produce absolute power spectra from 0.5–40 Hz at 0.5‑Hz resolution. Frequency bands analysed were delta (1–4 Hz), theta (4–8 Hz), alpha (8–12 Hz), beta (12–25 Hz), high beta (25–30 Hz) and gamma (the extracted text does not clearly specify the numerical range for the gamma band in this analysis). Coherence analyses used cross‑spectral coherence and the imaginary part of coherency (phase‑lagged coherence) to mitigate volume conduction confounds; coherence was computed across 30 intra‑hemispheric and six inter‑hemispheric electrode pairs. Statistical analysis applied a Box‑Cox ratio transformation with λ = 0 to address interindividual variability, followed by Shapiro‑Wilk testing. Repeated‑measures ANOVA (RM‑ANOVA) with Greenhouse‑Geisser correction tested effects of treatment (between‑subjects) and time (within‑subjects) separately for each frequency band. Post hoc tests were Bonferroni‑corrected where interactions were significant. Global connectivity patterns were visualised via kernel density estimates of concatenated coherence changes and topographic maps were produced using 3D spline mapping; visualisations of coherence used line widths/colours to indicate effect magnitude and direction.

Group composition for the analysed EEG data differed from the nominal 12 per group because of exclusions for technical issues: 10 rats for saline, nine for LSD, eight for psilocin, 12 for mescaline, and 11 for DOB. Behavioural scoring showed no baseline differences between groups. Although RM‑ANOVA indicated a significant interaction between time and treatment for activity (F(8,86) = 2.29, P < 0.05), Tukey post hoc tests did not identify significant differences between control and drug groups at the specified times. Spectral power: RM‑ANOVA revealed a significant main effect of treatment across all analysed frequency bands: delta (F(8,86) = 4.95, P < 0.0001), theta (F(8,86) = 8.11, P < 0.00001), alpha (F(8,86) = 5.69, P < 0.00001), beta (F(8,86) = 4.83, P < 0.0001), high beta (F(8,86) = 4.76, P < 0.0001), and gamma (F(8,86) = 7.08, P < 0.00001). All psychedelics except mescaline produced an overall decrease in absolute power across the 1–40 Hz range. Post hoc comparisons showed the largest power reductions in the 50–60 min window, smaller effects at 20–30 min, and persistence of effects at 80–90 min only for LSD and DOB. Topography: 3D topographic maps indicated that decreases in activity were widespread but showed substance‑specific spatial patterns. Tryptamine derivatives (psilocin, LSD) produced marked reductions across large fronto‑parieto‑temporal regions while sparing midline structures; phenethylamines (mescaline, DOB) also affected frontal midline areas. Frequency‑specific patterns emerged: psilocin and LSD reductions were most evident in lower frequencies (1–25 Hz), whereas mescaline and DOB more strongly affected higher frequencies (25–40 Hz). Notably, a rebound increase in occipital theta (4–8 Hz) activity emerged at later timepoints after LSD and mescaline. Connectivity: Kernel density plots of coherence changes showed that all active compounds shifted global connectivity distributions toward negative values compared with control, reflecting reduced functional connectivity. Cross‑spectral coherence decreases spanned the full spectrum, while phase‑lagged (imaginary) coherence decreases were concentrated mainly in lower frequency bands. Topographic coherence maps revealed discrete reductions between specific electrode pairs; overall, phenethylamine derivatives elicited larger connectivity changes than tryptamines. The authors report no systematic increases in connectivity that would offset the global decreases.

Vejmola and colleagues interpret their findings as evidence that, in freely moving rats recorded during behavioural inactivity, psilocin, LSD, mescaline and DOB produce a pronounced broadband desynchronization (reduced absolute spectral power across 1–40 Hz) together with widespread decreases in functional connectivity. The time courses varied: psilocin (shorter biological half‑life) produced earlier and transient changes, whereas LSD, mescaline and DOB had longer‑lasting effects with DOB showing a slower onset and pronounced late effects. The authors highlight a biphasic pattern for LSD and mescaline, where a late occipital rebound in theta and partially alpha power emerged; they suggest this may relate to stimulant‑like effects or to secondary pharmacology (for example involvement of D2/D4 receptors in a later phase), consistent with prior behavioural and pharmacokinetic data. Comparisons with prior work show concordance with human EEG/MEG literature reporting broadband power decreases—particularly in alpha—and reduced network integrity, supporting translational validity. The discussion notes that some human studies report increased high‑frequency (gamma) power but that muscle artefact can confound scalp EEG; because the present recordings were epidural, the observed decreases in high beta and gamma bands are unlikely to reflect muscle contamination and therefore may represent genuine cortical effects. The coherence findings are also aligned with limited animal and human data showing connectivity disruption after psychedelics; the authors propose that global desynchronization may be followed by a secondary reduction in interregional coupling. Limitations and future directions are acknowledged: direct comparisons across species and with advanced human network measures are constrained by methodological differences, and the authors indicate that applying source‑localisation approaches in rodents would improve translational mapping. They also note that comparisons with other animal findings remain limited by sparse prior data. Finally, the investigators present this multichannel, resting‑state‑analogue rat EEG paradigm as a robust translational model for probing the neurobiology of psychedelic effects and for future mechanistic or therapeutic studies.

The study concludes that all four psychedelics produced consistent electrophysiological signatures in rats—global decreases in cortical spectral power and reductions in functional connectivity—which closely mirror findings from human EEG/MEG research. Only modest differences were observed between tryptamine‑ and phenethylamine‑derived compounds. The authors argue that their protocol offers strong translational validity for studying psychedelic neurobiology in preclinical settings.