Symptomatic and neurotrophic effects of GABAA receptor positive allosteric modulation in a mouse model of chronic stress

Abstract

Chronic stress is a risk factor for Major Depressive Disorder (MDD), and in rodents, it recapitulates human behavioral, cellular and molecular changes. In MDD and after chronic stress, neuronal dysfunctions and deficits in GABAergic signaling are observed and responsible for symptom severity. GABA signals predominantly through GABAA receptors (GABAA-R) composed of various subunit types that relate to downstream outcomes. Activity at α2-GABAA-Rs contributes to anxiolytic properties, α5-GABAA-Rs to cognitive functions, and α1-GABAA-Rs to sedation. Therefore, a therapy aiming at increasing α2- and α5-GABAA-Rs activity, but devoid of α1-GABAA-R activity, has potential to address several symptomologies of depression while avoiding side-effects. This study investigated the activity profiles and behavioral efficacy of two enantiomers of each other (GL-II-73 and GL-I-54), separately and as a racemic mixture (GL-RM), and potential disease-modifying effects on neuronal morphology. Results confirm GL-I-54 and GL-II-73 exert positive allosteric modulation at the α2-, α3-, α5-GABAA-Rs and α5-containing GABAA-Rs, respectively, and separately reduces immobility in the forced swim test and improves stress-induced spatial working memory deficits. Using unpredictable chronic mild stress (UCMS), we show that acute and chronic administration of GL-RM provide pro-cognitive effects, with mild efficacy on mood symptoms, although at lower doses avoiding sedation. Morphology studies showed reversal of spine density loss caused by UCMS after chronic GL-RM treatment at apical and basal dendrites of the PFC and CA1. Together, these results support using a racemic mixture with combined α2-, α3-, α5-GABAA-R profile to reverse chronic stress-induced mood symptoms, cognitive deficits, and with anti-stress neurotrophic effects.

Fig. 1. Electrophysiological, pharmacokinetic and behavioral profiles of GL-II-73 and GL-I-54.

Electrophysiological recordings were obtained from HEK-293T cells transfected with full-length cDNA for human GABAA receptor subtypes α1β3γ2, α2β3γ2, α3β3γ2, α4β3γ2 or α5β3γ2, in presence of GL-I-54 (A) or GL-II-73 (B), and in the presence of GABA in the medium (5 μM). Pharmacokinetic profile of GL-I-54 was also examined (C). Plasma, brain and free brain concentration-time profile of GL-I-54 after intraperitoneal cassette administration of 3 mg/kg dose in male C57BL/6 mice (n = 3 per time point). C_max, maximum concentration in plasma or brain; T_max, time of maximum concentration in plasma or brain; AUC_0–720, area under the plasma or brain concentration-time curve from 0 to 720 min; AUC_0-∞, area under the plasma or brain concentration-time curve from 0 to extrapolated infinite time; t_1/2, elimination half-life from plasma or brain; β, elimination constant rate from plasma or brain; K_p, brain-to-plasma partition coefficient (K_p = AUC_0-∞, brain/AUC_0-∞, plasma); K_p,uu, ratio of unbound brain to unbound plasma drug concentrations (K_p,uu = K_p × unbound fraction in brain/unbound fraction in plasma). GL-I-54 and GL-II-73 were then tested in the elevated plus maze and forced swim test at the dose of 5 mg/kg, in mice previously exposed to chronic restraint stress (CRS). Drugs were administered IP, 30 min prior to testing. Percent time spent in the open arms (D) and percent entries into open arm (E) showed no difference between groups. Time immobile (F) showed more time spent immobile in the CRS-Vehicle group, and less time immobile in animals receiving GL-I-54, compared to CRS-Vehicle. GL-II-73 did not show an effect. GL-I-54 was tested in the Y-maze task, assessing working memory (G). Animals subjected to CRS and receiving vehicle showed a significant decrease in alternation rate, suggesting a working memory deficit. Animal subjected to CRS and receiving the highest dose of GL-I-54 showed significant increase in alternation rate, suggesting reversal of working memory deficits induced by CRS. Finally, independent mice were tested in the rotarod (H; N = 5 Control Vehicle and N = 6 GL-I-54). Mice were trained to maintain themselves on a rotating rod (rotarod) for 3 trials. Then, they were injected with GL-I-54 at 10 mg/kg, and tested 5 min, 20 min and 60 min past injection time. Latency to fall from the rod was recorded, and showed significant reductions in latency to fall in mice receiving GL-I-54. All values are represented as mean ± standard error of the mean. *p < 0.05, **p < 0.01, ***p < 0.001 compared to 100% in A, B) or to Control, Control/Vehicle in D-G. ^θp < 0.05 compared to CRS-Vehicle.

Fig. 2. Effect of acute treatment of GL-RM on anxiety, emotionality and working memory deficits in mice subjected to chronic stress.

Male and female mice were tested at baseline in the Y-maze, the phenotyper and the sucrose consumption test prior to being subjected to 6 weeks of UCMS (A). Weekly, mice were tested in the phenotyper test, the sucrose consumption and their weight and coat state were measured. After 6 weeks of UCMS, acute injections were performed 30 min prior to behavioral testing. In the elevated plus maze, time spent (B) and entries (C) in the open arms were measured, with no significant effect of UCMS, treatment or interaction. In the novelty suppressed feeding test, latency to approach (D) and latency to bite (E) were assessed. Statistical analyses showed that acute GL-RM treatment reduces latency to approach in mice subjected to UCMS. In the phenotyper test (F), mice were placed in the box overnight, where a stressful stimulus was applied at 11 pm, for 1 h. Time spent in the shelter zone showed that mice subjected to UCMS spent more time in the shelter than control mice. Calculating a residual avoidance score (G), statistical analyses showed a significant increase in residual avoidance in mice subjected to UCMS. In the sucrose consumption test (H), mice subjected to UCMS showed a significant decrease in preference to sucrose. In the forced swim test (I), mice subjected to UCMS showed increased immobility, while mice treated with GL-RM showed a reduction in immobility. Combining the individual score into a global z-score (J), statistical analyses confirmed a significant impact of UCMS, with reduced effect of GL-RM. Finally, mice were tested in the Y-maze (K), where statistical analyses showed altered alternation with UCMS, which is reversed by acute GL-RM treatment. *p < 0.05, **p < 0.01, ***p < 0.001 effect of UCMS, ^#p < 0.05, ^##p < 0.01 effect of GL-RM.

Fig. 3. Effect of chronic treatment of GL-RM on anxiety, emotionality and working memory deficits in mice subjected to chronic stress.

Male and female mice were tested at baseline in the Y-maze, the phenotyper and the sucrose consumption test prior to being subjected to 6 weeks of UCMS (A). After 3 weeks of UCMS, chronic treatment with GL-RM in the drinking water was initiated, for a total of 4 weeks. Weekly, mice were tested in the phenotyper test, the sucrose consumption and their weight and coat state were measured. After 6 weeks of UCMS, and 3 weeks of treatment, mice were tested in the elevated plus maze. Time spent (B) and entries (C) in the open arms were measured, but did not show statistical differences. Mice were also tested in the open field for the time spent (D) and number of entries (E) in the inner zone. Again, statistical analyses did not reveal any effect of UCMS nor treatment. In the novelty suppressed feeding test, latency to approach (F) and latency to bite (G) were assessed. Mice were tested in the Phenotyper weekly, and the residual avoidance scores from week 3 to 6 were analyzed, since the treatment was onboard during these testing periods (H). In the sucrose consumption test (I), mice receiving chronic GL-RM showed a significant increase in preference to sucrose. In the forced swim test (J), statistical analyses did not reveal significant differences between groups. Combining the individual score into a global z-score (K), statistical analyses confirmed a significant impact of UCMS at increasing emotionality, with a trend level effect ofchronic GL-RM reducing emotionality. Finally, mice were tested in the Y-maze (L), where statistical analyses showed altered alternation with UCMS, which is reversed by acute GL-RM treatment. *p < 0.05, **p < 0.01, ***p < 0.001 effect of UCMS, ^#p < 0.05, ^###p < 0.001 effect of GL-RM.

Fig. 4. Chronic treatment with GL-RM reverses chronic-stress induced spine density reduction in the PFC and the CA1.

After completion of the behavioral screening, mice were euthanized and brains were stained with Golgi-Cox solution. Pyramidal neurons (N = 6 per mouse) from 4 mice per group (A–C) were analyzed for dendritic length, spine counts and spine density. Basal and apical spine densities were measured in the PFC (D) and the CA1 of the hippocampus (E). ANOVA in the basal and apical segments revealed significant differences between groups, in both brain regions. This difference was explained by a decrease in spine density in mice subjected to UCMS compared to Control mice that was partially reversed by chronic treatment with GL-RM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 compared to “Control”; ^$$p < 0.01, ^$$$p < 0.001, ^$$$$p < 0.0001 compared to “UCMS”. Scale bar in (A–C) represents 50 µm.