Repeated lysergic acid diethylamide (LSD) reverses stress-induced anxiety-like behavior, cortical synaptogenesis deficits and serotonergic neurotransmission decline

Psychedelic compounds have re-emerged as candidate psychiatric therapeutics, with lysergic acid diethylamide (LSD) drawing attention for putative anxiolytic and antidepressant properties. Earlier clinical and preclinical work suggests LSD acts at serotonin (5-HT) receptors, producing prosocial and mood effects in humans and modulating serotonergic and dopaminergic systems in animals; however, the precise neurobiological mechanisms that might underlie therapeutic actions remain unclear. Prior studies by the group showed dose-dependent effects of LSD on 5-HT and dopamine systems, and a 7-day low-dose regimen was previously linked to increased sociability and mPFC mTORC1 activation, motivating further mechanistic inquiry into stress-related outcomes. De Gregorio and colleagues set out to test whether acute or repeated low-dose LSD alters anxiety- and depressive-like behaviours, cortical dendritic spine density, and dorsal raphe nucleus (DRN) 5-HT neuronal activity in male mice exposed to chronic restraint stress. The study tested multiple acute doses (5, 15, 30 and 60 µg/kg, i.p.) for immediate behavioural effects, and evaluated a 7-day repeated regimen (5, 15 and 30 µg/kg/day, administered during the last week of a 15-day chronic restraint stress paradigm) for protective effects on stress-induced behavioural and neurobiological alterations. The work aimed to link behavioural outcomes to synaptogenesis and electrophysiological changes in 5-HT neurotransmission, with particular attention to potential desensitisation of 5-HT1A autoreceptors.

Adult male C57BL/6N mice (8–12 weeks old, 25–30 g) were used; all procedures followed institutional animal care approvals. Acute LSD experiments employed intraperitoneal injections of 5, 15, 30 and 60 µg/kg, with immediate assessment of the head twitch response (HTR) for 10 minutes followed by a battery of behavioural tests (elevated plus maze, light–dark box, novelty suppressed feeding, open field, forced swim test) at specified intervals after injection. For repeated-treatment paradigms, LSD at 5, 15 or 30 µg/kg (i.p.) was given once daily for 7 days during the final week of a 15-day chronic restraint stress (CS) protocol (2 hours/day). Behavioural or electrophysiological experiments were performed 24 hours after the last LSD or vehicle injection. Detailed procedural elements were reported as being in the Supplementary Material. Electrophysiological assessment of DRN 5-HT neurons included in vivo single-unit recordings under anaesthesia. Acute experiments used cumulative injections of LSD to characterise dose-dependent effects on spontaneous firing; repeated-treatment recordings compared baseline firing after 7 days of LSD versus vehicle. Microiontophoretic methods with multi-barrel electrodes were used to eject the 5-HT1A agonist 8-OH-DPAT in increasing doses to assess somatodendritic autoreceptor sensitivity. To control for potential 5-HT7 receptor contributions to 8-OH-DPAT effects, some experiments administered the 5-HT7 antagonist SB269970 prior to systemic 8-OH-DPAT. Morphological analysis targeted apical dendritic spine density of mPFC pyramidal neurons; mice were sacrificed 24 hours after the final treatment and spine counts performed as described (details referred to Supplementary Material). Behavioural outcomes included standard indices of anxiety-like and depressive-like behaviour (EPM, LDB, NSFT, OFT, FST, sucrose preference test), and statistical analyses comprised ANOVA and post-hoc comparisons for behavioural data and appropriate tests for electrophysiological measures (including repeated-measures ANOVA and t-tests where reported). When sample counts were reported, neuron and mouse numbers were specified (for example, 14 neurons from 4 LSD-treated mice versus 10 neurons from 4 vehicle-treated mice for a subset of recordings).

Acute LSD effects in non-stressed mice: Acute intraperitoneal LSD produced a dose-dependent head twitch response, with significant increases at 30 µg/kg and more robustly at 60 µg/kg (one-way ANOVA F(4,40)=52.71, p<0.0001). Despite this pharmacological action, none of the tested acute doses (5–60 µg/kg) altered anxiety-like measures in the elevated plus maze, light–dark box, or novelty-suppressed feeding tests; corresponding ANOVAs were non-significant (for example, EPM percent open time F(4,39)=0.6857, p=0.6058). In open field testing there was a main effect on total distance (F(4,39)=4.045, p=0.007) but post-hoc comparisons did not reveal clear group differences; the 60 µg/kg dose produced increased grooming and rearing (stereotypies) and was therefore excluded from further experiments. Forced swim immobility time was unchanged across acute doses. Repeated LSD prevents stress-induced anxiety-like behaviour: The chronic restraint stress (15 days, 2 h/day) decreased time spent in the centre of the open field and reduced centre entries (stress main effects F(1,69)=163.7, p<0.001 and F(1,69)=162.6, p<0.001 respectively). Repeated LSD at 30 µg/kg/day for 7 days (last dose 24 hours before testing) prevented the stress-induced reduction in number of centre entries, with CS mice treated with 30 µg/kg showing significantly higher entries than CS-vehicle mice (post-hoc p<0.001). Lower repeated doses (5 and 15 µg/kg) did not prevent the stress effect. Repeated LSD did not alter total locomotion, grooming duration, or rearing frequency. Across other anxiety assays, the 30 µg/kg regimen normalised transitions in the light–dark box and prevented increased latency in the novelty suppressed feeding test, but it did not produce anxiolytic-like effects in the elevated plus maze. The chronic restraint stress paradigm used did not produce depressive-like behaviour: forced swim test immobility and sucrose preference were unaffected by CS or LSD. Cortical spine density: Fifteen days of chronic restraint stress reduced apical dendritic spine density in mPFC pyramidal neurons. Repeated LSD at 30 µg/kg/day prevented this stress-induced spine loss and additionally increased spine density in non-stressed (vehicle) mice, indicating a synaptogenic effect of the repeated regimen. Electrophysiology and 5-HT1A sensitivity: Acute cumulative injections of LSD (5, 15 and 30 µg/kg) decreased the spontaneous firing activity of DRN 5-HT neurons in non-stressed mice (repeated-measures ANOVA F(1.985,7.941)=34.41, p=0.0001), with 30 µg/kg completely suppressing firing in that preparation. By contrast, after 7 days of repeated LSD (30 µg/kg/day), the mean spontaneous firing frequency of DRN 5-HT neurons was significantly increased compared with vehicle (14 neurons from 4 LSD-treated mice versus 10 neurons from 4 vehicle mice; t=3.751, df=26, p<0.001). Repeated LSD also increased the number of spike events in 200 s (t=6.504, df=26, p<0.001) and the coefficient of variation of interspike intervals (t=3.170, df=26, p<0.01), indicating a more irregular firing pattern; burst discharge percentage was unchanged. Microiontophoretic application of the 5-HT1A agonist 8-OH-DPAT produced a current-dependent inhibition of DRN firing in vehicle-treated mice, whereas 7-day LSD-treated mice showed a statistically attenuated inhibitory response (current factor F(2.353,61.17)=46.30, p<0.001; treatment factor F(1,26)=11.89, p<0.001; interaction F(3,78)=11.89, p<0.001), consistent with reduced 5-HT1A autoreceptor sensitivity. Systemic co-administration of the 5-HT7 antagonist SB269970 did not alter the attenuated response to 8-OH-DPAT in LSD-treated animals, supporting the interpretation that 5-HT1A desensitisation underlies the changed responsiveness. Repeated LSD prevents stress-induced decline of DRN firing: Chronic restraint stress decreased spontaneous DRN 5-HT firing relative to controls (stress main effect F(1,156)=22.87, p<0.001). Importantly, mice exposed to CS but treated with repeated LSD (30 µg/kg/day) showed higher DRN firing rates than CS-vehicle mice (post-hoc p<0.001), indicating that the repeated regimen prevented the stress-induced decline in 5-HT neuronal activity.

De Gregorio and colleagues interpret their findings as evidence that repeated, low-dose LSD confers protection against stress-induced anxiety-like behaviour and associated neurobiological deficits, whereas acute dosing has minimal behavioural impact in naïve mice. Behaviourally, the 7-day 30 µg/kg LSD regimen prevented several measures of stress-exacerbated anxiety (notably entries in the open field centre, light–dark transitions, and novelty suppressed feeding latency) but did not affect elevated plus maze performance or produce antidepressant-like effects in the forced swim or sucrose preference tests; the authors note the chronic restraint stress model reliably elicits anxiety-like but not necessarily depressive-like phenotypes. At a cellular level, repeated LSD increased apical dendritic spine density in mPFC pyramidal neurons and prevented stress-induced spine loss, supporting a synaptogenic plasticity mechanism. Electrophysiological data showed a dichotomy between acute and repeated exposure: acute LSD decreased DRN 5-HT firing via 5-HT1A autoreceptor action, whereas repeated administration increased basal 5-HT firing and produced an attenuated inhibitory response to the 5-HT1A agonist 8-OH-DPAT, consistent with autoreceptor desensitisation. The increased firing and restored firing under stress are therefore proposed as a mechanism by which repeated LSD may enhance serotonergic neurotransmission, paralleling how selective serotonin reuptake inhibitors initially reduce firing but ultimately raise serotonin output after autoreceptor desensitisation. The authors situate their results within heterogeneous prior literature: some preclinical reports found anxiolytic or antidepressant trends with LSD or other psychedelics, while others did not, and species, dose, regimen and stress-model differences likely contribute to inconsistencies. They highlight converging evidence for psychedelics and related agents to promote synaptogenesis, increase BDNF, and engage mTORC1 signalling, all of which could mediate sustained behavioural effects. Key limitations acknowledged by the investigators include the restricted behavioural battery and the fact that chronic restraint stress did not induce depressive-like behaviour in their hands, limiting conclusions about antidepressant effects. They also note the possibility that stress and LSD effects could interact with memory processes rather than anxiety per se, the absence of sex-difference analyses, and the short follow-up period that leaves the persistence of protective effects unknown. Finally, the authors call for clinical studies to define minimal effective acute and repeated LSD regimens, the duration of effects, and to test repeated LSD in people with anxiety disorders not related to terminal illness, while emphasising that further preclinical work should examine other stress models and include both sexes.