Harmine produces antidepressant-like effects via restoration of astrocytic functions

Major depression remains difficult to treat and many standard antidepressants, developed from the monoaminergic deficit hypothesis, have limited efficacy: about one-third of patients respond to first-line treatment and a substantial proportion remain unresponsive after multiple trials. Increasing evidence implicates astrocytic pathology in depression: reductions in astrocyte number and in astrocyte markers such as GFAP have been observed after chronic stress in animals and in post-mortem tissue from depressed patients, and changes in astrocyte-specific glutamate transporters (EAATs, including GLT-1/EAAT2 and GLAST/EAAT1) and glutamine synthetase have been reported. Functionally, astrocytic dysfunction reduces glutamate uptake, perturbs glutamate cycling, and can impair BDNF signalling and hippocampal neurogenesis, processes linked to mood regulation. Harmine, a naturally occurring β-carboline alkaloid present in Peganum harmala and in Banisteriopsis caapi (a component of ayahuasca), has demonstrated a range of pharmacological effects and prior preclinical antidepressant-like activity. Clinical observations of rapid antidepressant effects after ayahuasca intake further motivate investigation, although ayahuasca contains other active compounds. On this basis Liu and colleagues hypothesised that harmine’s antidepressant-like properties may arise from restoration of astrocytic functions. The study therefore tested whether harmine reverses depressive-like behaviours induced by chronic unpredictable stress (CUS) in mice and whether such effects are associated with changes in astrocyte markers, GLT-1 expression, BDNF levels and hippocampal neurogenesis; the role of astrocytes was probed using the astrocyte-specific gliotoxin L-AAA.

The investigators used male C57BL/6J mice subjected to a 5-week chronic unpredictable stress (CUS) protocol to induce depressive-like behaviour. The CUS procedure comprised two randomly assigned daily stressors drawn from a list that included cage shaking, overnight lights-on, cold exposure (4°C), restraint, cage tilt, wet bedding, flashing light, noise, and temporary water deprivation. Behavioural endpoints included the forced swim test (FST), tail suspension test (TST) and a sucrose preference test; immobility in FST and TST was scored by an experimenter blinded to treatment, and sucrose preference was measured after standard two-bottle choice acclimatisation and a 24 h fluid/food deprivation before testing. Harmine was administered intraperitoneally at 10 or 20 mg/kg daily for 10 days, prepared in 10% dimethyl sulphoxide; fluoxetine (20 mg/kg) served as a positive control. To test whether astrocytes are necessary for harmine's effects, L-Alpha-Aminoadipic Acid (L-AAA) was delivered intracerebroventricularly via osmotic minipumps (1 μg/μL, infusion rate 1 μL/h for chronic 10-day infusions) after stereotaxic cannula implantation. The extracted text does not clearly report the full experimental timeline for each group or all sample sizes for every assay, although some n values are provided for individual analyses. Biochemical and histological measures were performed immediately after behavioural testing. Hippocampal and prefrontal cortical BDNF protein levels were quantified by sandwich ELISA. Protein expression of astrocyte and glutamate-related markers (GFAP, GLT-1, GLAST, GS) and neurogenesis marker DCX were assessed by western blot; DCX-positive cells in the hippocampus were visualised and counted by immunofluorescence. Statistical analysis used two-way analysis of variance (ANOVA) with Bonferroni post hoc tests, reporting data as mean ± SEM and treating p < 0.05 as significant.

Harmine reversed CUS-induced behavioural deficits. Two-way ANOVA for sucrose preference showed significant effects of stress [F(1,72)=44.64, p<0.001] and drug treatment [F(3,72)=4.32, p<0.01], with post hoc tests indicating that 10-day harmine treatment at 10 and 20 mg/kg increased sucrose intake in CUS-exposed mice (reported n=10, p<0.05 versus vehicle+CUS). The extracted text states that chronic harmine (20 mg/kg) markedly prevented CUS-induced increases in immobility in the FST and TST and attenuated the CUS-induced reduction in sucrose intake, with group-level ANOVAs for FST [stress F(1,54)=31.15, p<0.001; drug F(2,54)=9.09, p<0.001] and TST [stress F(1,54)=28.75, p<0.001; drug F(2,54)=4.87, p<0.05]. Exact immobility times and some group means are not clearly reported in the extracted text. Harmine restored BDNF protein levels and hippocampal neurogenesis. In the hippocampus, two-way ANOVA revealed main effects of stress [F(1,56)=20.64, p<0.001] and drug [F(3,56)=4.83, p<0.01]; CUS reduced hippocampal BDNF and this reduction was reversed by harmine at 10 and 20 mg/kg (post hoc p<0.05). Chronic harmine also increased hippocampal BDNF in non-stressed (naïve) mice. The investigators assessed hippocampal neurogenesis by counting DCX-positive cells and measuring DCX protein: DCX-positive cell counts showed strong main effects for stress [F(1,24)=104.10, p<0.001] and drug [F(2,24)=12.45, p<0.001], and DCX protein levels showed effects for stress [F(1,24)=16.57, p<0.001] and drug [F(2,24)=5.85, p<0.01]; harmine reversed stress-induced reductions in both measures. Astrocyte-associated markers were modified by harmine. CUS reduced GFAP protein expression in the hippocampus and prefrontal cortex, and harmine treatment prevented these reductions. Although CUS did not reliably alter GLT-1 or EAAT expressions in this extraction, harmine increased GLT-1 protein expression in hippocampus and prefrontal cortex, consistent with prior reports linking harmine to enhanced GLT-1 and glutamate uptake. Astrocyte blockade with L-AAA abolished harmine’s beneficial effects. Co‑administration of intracerebroventricular L-AAA with harmine (20 mg/kg) for 10 days prevented harmine’s reversal of CUS-induced immobility in FST and TST and its improvement of sucrose preference (ANOVAs for sucrose preference: stress F(1,54)=41.08, p<0.001; drug F(2,54)=6.45, p<0.01). L-AAA also blocked harmine-induced increases in hippocampal neurogenesis and DCX protein expression (post hoc p<0.05 versus vehicle+CUS+harmine). Importantly, L-AAA co-treatment prevented harmine’s restoration of hippocampal and prefrontal cortical BDNF protein levels. The extracted text does not report adverse events or non-neurological safety data.

Liu and colleagues interpret their findings as supporting an astrocyte-mediated mechanism for harmine’s antidepressant-like effects. They note that unlike prior studies performed in non-stressed animals, their use of a CUS paradigm demonstrates efficacy in a disease-relevant model and argue that evaluation under stress conditions is important because some compounds show disease-dependent effects. Behavioural improvements produced by harmine were reported to be similar to those of fluoxetine in these assays. At the molecular level, harmine reversed CUS-induced reductions in hippocampal and prefrontal BDNF and restored hippocampal neurogenesis; because L-AAA blockade of astrocytes prevented these biochemical and cellular effects as well as the behavioural improvements, the investigators propose that astrocytic restoration underlies harmine’s actions. They emphasise harmine’s up-regulation of GLT-1 as a plausible means to enhance glutamate clearance, thereby counteracting glutamatergic dysregulation reported in depression. The authors relate these findings to other agents that increase GLT-1 (for example, ceftriaxone and riluzole), which have shown antidepressant-like properties. The authors acknowledge limitations evident in the extracted text: mechanistic details remain uncertain, notably how harmine increases BDNF protein levels, and translation from preclinical models to clinical efficacy requires caution. They also advise caution when inferring harmine’s role from clinical ayahuasca data, since ayahuasca contains additional active compounds such as DMT. Finally, they suggest further investigation into interactions between astrocytes and conventional antidepressant targets (for example 5-HT receptors and monoamine oxidase) and exploration of related hallucinogens in depression regulation. Overall, the study is presented as preclinical evidence that restoring astrocytic functions—particularly GLT-1 expression and BDNF-mediated neurogenesis—may contribute to harmine’s antidepressant-like effects.

The study concludes that harmine reverses behavioural deficits induced by chronic unpredictable stress in mice and that these antidepressant-like effects are associated with restoration of astrocytic markers (notably GFAP), up-regulation of GLT-1, recovery of BDNF protein levels, and promotion of hippocampal neurogenesis. Astrocytic inhibition with L-AAA abrogated harmine's behavioural and molecular effects, supporting a critical role for astrocyte function in harmine's mechanism of action. The authors frame these findings as further validation of the hypothesis that astrocyte dysfunction contributes to depression and as evidence that harmine is a candidate for further investigation in depression therapy. The extracted text summarises three key points: harmine restores astrocyte marker alterations in CUS-treated mice; astrocytic inhibition blocks harmine's antidepressant-like effects; and astrocytic inhibition blocks harmine's effects on BDNF and neurogenesis.