Characterization of behavioral and endocrine effects of LSD on zebrafish

Subjects were 530 experimentally naïve adult short‑fin zebrafish (5–7 months old), male and female, acclimated at least 14 days and housed in mixed groups in 40‑L tanks at 25–27°C on a 12‑hour light cycle. Fish were fed twice daily and euthanised after testing for tissue collection. Behavioural testing occurred between 11:00 and 15:00 with water temperatures matched to the holding room. To avoid carry‑over effects each behavioural assay used a separate cohort of naïve fish. The study employed eight behavioural paradigms: the novel tank test (Experiments 1 and 2), observation cylinder (Experiment 3), light–dark box (Experiment 4), open field (Experiment 5), T‑maze (Experiment 6), social preference (Experiment 7) and shoaling (Experiment 8). Sample sizes varied by test (reported per experiment in the extracted text; typical group sizes were 10–15 fish per treatment group for individual assays, and shoals of eight fish per group). Trained observers manually scored endpoints (inter‑rater reliability >0.85) and sessions were also analysed by EthoVision XT7 video‑tracking to extract distance travelled, velocity and spatial coordinates. Behavioural measures included latency to top, time spent in top, number and duration of freezing bouts (defined as absence of movement except gills/eyes for ≥2 s), entries and durations in defined zones, arm entries in mazes, and inter‑fish distances for shoals. Track data were exported for per‑second analysis and subjected to 3D spatiotemporal reconstruction (X, Y, time on X‑, Y‑ and Z‑axes) to illustrate exploration topography. Ethograms were constructed from manually coded behavioural episodes—normal swimming, vertical drift, backward swimming, erratic bouts, bottom and surface freezing—to visualise activity frequencies and transitions. Pharmacological treatment employed multiple doses: the extracted text reports tests at 5, 25, 50, 75, 100 and 250 g/L, with the 250 g/L dose selected for the detailed characterisation because lower doses showed no or only non‑significant trends. In all experiments except Experiment 2, fish were pre‑exposed by submersion in 3‑L beakers containing 250 g/L LSD for 20 min prior to testing; Experiment 2 used a novel tank pre‑filled with 250 g/L LSD and observed animals for 30 min. The extracted text presents doses using the unit "g/L"; the methods summary here follows that reporting. Whole‑body cortisol was measured from fish in Experiments 1, 4 and 5. Individual whole‑body homogenates were extracted twice with diethyl ether, reconstituted and assayed by ELISA using a human salivary cortisol kit, with results expressed as relative cortisol concentration per gram body weight based on a 4‑parameter curve fit. Statistical comparisons used the non‑parametric Mann–Whitney U‑test, data are presented as mean ± SEM, and significance was set at P < 0.05.

Dose screening indicated that lower doses (5–100 g/L) produced no significant behavioural effects and 100 g/L showed only non‑significant trends; 250 g/L produced robust and consistent effects across tests and was therefore used for subsequent experiments. Sample sizes by experiment were reported in the methods (e.g. novel tank Experiment 1: n = 10–12 per group; Experiment 2: n = 15 per group; observation cylinder: n = 15 per group; light–dark: n = 12 per group; open field: n = 15 per group; T‑maze: n = 12 per group; social preference: n = 10 per group; shoaling: groups of 8 fish, with multiple groups per condition). In the 6‑min novel tank (Experiment 1) exposure to 250 g/L LSD produced significantly shorter latency to enter the top half, increased time spent in the top, more transitions, longer mean entry duration and substantially less freezing. Despite these changes in vertical distribution and freezing, neither total distance travelled nor average velocity differed significantly from controls. Representative 2D traces and 3D reconstructions showed a clear shift from predominant bottom dwelling in controls to active two‑dimensional swimming with frequent top dwelling in LSD‑treated fish. In Experiment 2 (30‑min novel tank filled with LSD), similar effects emerged rapidly within minutes, with consistently lower freezing throughout the session and no evidence of early anxiogenic inhibition. The observation cylinder test replicated the top‑dwelling phenotype and reduced freezing. Ethogram analyses revealed altered behavioural organisation in LSD‑treated fish, including additional behavioural elements and transitions not commonly observed in controls. In the light–dark box LSD increased activity in the white (light) compartment: treated fish exhibited a trend toward more time in the light and significantly longer average entry durations in the light half, yielding higher white:total time ratios (trends). In the open field LSD induced thigmotaxis, significantly reducing time spent in the central zone while leaving distance travelled and velocity unchanged, consistent with greater peripheral‑wall activity. In the T‑maze 250 g/L LSD tended to reduce the number of arm and centre entries and significantly decreased both frequency and duration of freezing episodes. The social preference test showed that LSD significantly reduced total arm entries, centre entries and both conspecific and empty arm entries, but did not change social preference ratios (conspecific:total or conspecific:empty) or time spent in the conspecific arm, indicating reduced spatial exploration without altered social preference per se. Shoaling behaviour was disrupted: LSD significantly increased average inter‑fish distance within groups, indicating reduced cohesion. Endocrinologically, whole‑body cortisol was significantly elevated in fish from Experiment 5 and showed similar upward trends in Experiments 1 and 4. Statistical testing used Mann–Whitney U comparisons with significance threshold P < 0.05; specific p‑values and confidence intervals are not reported in the extracted text.

Grossman and colleagues interpret the pattern of findings as evidence that adult zebrafish are behaviourally and physiologically sensitive to acute LSD exposure. The principal phenotypes were robust top dwelling in novel arenas, reduced freezing, altered spatiotemporal locomotor patterns (shown by ethograms and 3D reconstructions), increased thigmotaxis in an open field, altered light–dark behaviour, disrupted shoaling and elevated whole‑body cortisol. Despite reduced overall exploratory entries in some assays, social preference ratios were unchanged, suggesting that LSD affected spatial exploration and shoal cohesion more than the motivation to approach conspecifics. The authors note similarities and differences with earlier animal studies. Top dwelling parallels surface‑seeking responses reported in other fish species exposed to LSD, yet, unlike several early fish reports of immobility, zebrafish here remained actively swimming with minimal freezing, producing no change in gross locomotor measures such as distance travelled or velocity. Grossman and colleagues suggest serotonergic mechanisms may underlie aspects of the phenotype, citing similarities to zebrafish responses to fluoxetine and the known action of LSD at multiple 5‑HT receptor subtypes; they also highlight that the expected early anxiogenic phase seen in rodents and humans was not detected in these zebrafish assays, a discrepancy that may reflect species differences in receptor contributions or assay sensitivity. The discussion considers perceptual distortion as a possible explanation for some outcomes: increased occupancy of the light compartment without greater numbers of light entries could reflect altered sensory processing rather than hyperactivity. The thigmotaxis finding in zebrafish is noted to mirror effects of hallucinogens in rodents and may represent a conserved behavioural signature. With regard to social behaviour, the dissociation between unaltered social preference and disrupted shoaling indicates that different facets of sociality can be differentially modulated by LSD. Key limitations acknowledged by the authors include the need to characterise dose‑ and time‑dependent effects more fully (including chronic exposure and longer observation windows), to identify receptor mediation via selective agonists and antagonists, and to explore sex and strain differences in sensitivity. They also propose extending comparisons to other hallucinogens. Finally, the authors observe that increased cortisol parallels findings in other species and may reflect central modulation of the endocrine axis, though elevated cortisol here did not coincide with overt anxiogenic behaviour in the assays employed. Overall, the investigators conclude that zebrafish provide a useful model for studying hallucinogenic drugs, while noting several avenues for deeper mechanistic work.