Lateral habenula astroglia modulate the potentiating antidepressant-like effects of bright light stimulation in intractable depression

Abstract

Background: Beside image vision, light plays a pivotal role in regulating diverse non-visual functions, including affective behaviors. Recently, bright light stimulation (BLS) was revealed to be beneficial for treating non-seasonal depression, although its mechanism of action is not fully understood.

Methods: We developed a novel mouse model of refractory depression, induced through social isolation and chronic despair during the active (dark) phase of the animal, and we have tested if antidepressant treatments, including BLS, could protect against anxio-depressive-like behavior.

Results: We report that anxiety- and depressive-like behaviors are resistant to BLS as well as to both conventional and new antidepressants, including ketamine. Remarkably, we unveil that BLS potentiates the effect of antidepressants, and this beneficial effect is mediated via rod retinal photoreceptors. Furthermore, we demonstrate that both chemogenetic activation of lateral habenula (LHb) astroglia and serotonin (5-HT) depletion prevent the potentiating effect of BLS on chronic despair.

Conclusion: These results reveal, for the first time, that BLS enhances the efficacy of antidepressants through an unexpectedly circuit involving rods, LHb astroglia and 5-HT.

Keywords: animal model; bright light stimulation; chemogenetic; ketamine; refractory depression; rods.

FIGURE 1

CDMRD: a pharmaco-resistant model of depression: (A) Schematic overview of the experimental design: CDMRD paradigm, followed by the test phase. (B) CDMRD: 6-weeks old C57BL/6J male mice are exposed to a 12L/12D cycle, with light on (50 lux) at zeitgeber 0 (ZT0) and off at ZT12. Animals were forced to swim on 5 successive days [D] for 10 min at ZT14 and a passive stress coping (PSC) was performed every week during 8 weeks (n = 8) at ZT14. One-way repeated measures ANOVA with post hoc Wilcoxon Test, **p < 0.01 vs. D1. (C) Sucrose Preference Test (SPT): Naïve and stressed mice (48 h post-stress) were subjected to the sucrose preference test (n = 8). Elevated plus maze (EPM): naïve and stressed mice (1-week post-stress) were subjected to the elevated plus maze (n = 8–13). Spontaneous locomotor activity (SLA) of mice 24 h before the last PSC test (n = 6–7). Plasma corticosterone profile measure by radioimmunoassay (RIA) in naïve grouped mice (n = 5), prior stress (n = 10) or after CDMRD (n = 10). LSD test: *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001.

FIGURE 2

Lack of antidepressant action in CDMRD mice. (A) Following CDMRD, mice exposed to a 12L/12D cycle (50 lux during the light phase) were treated with the SSRI escitalopram (10 mg/kg i.p), 5 consecutive days from D10 to D15. A passive stress coping (PSC) test was then realized at ZT14, one-way ANOVA with post hoc LSD Test, **p < 0.01 vs. D1. Antidepressant-like effect of a sub-chronic treatment of escitalopram in non-stressed mice (CTL): naïve mice were treated with vehicle or the SSRI escitalopram (10 mg/kg) during 5 consecutive days. On day 5, they received the last injection 30 min before the PSC test. (n = 8) unpaired Student’s t-tests; ***p < 0.001. (B) Following CDMRD, mice exposed to a 12L/12D cycle (50 lux during the light phase), received an acute injection of ketamine (Ket, 3 mg/kg, i.p.) and the muscarinic antagonist scopolamine (Sco, 0.1 mg/kg, i.p) 4 weeks post-stress, at day 33 (D33). Antidepressant-like effect of a combination of Ketamine + Scopolamine in non-stressed mice (CTL): naïve mice were submitted to a PSC test. 24 h later, the mice were injected with vehicle or a combination of ketamine (3 mg/kg, i.p) and the muscarinic receptor antagonist scopolamine (0.1 mg/kg, i.p), 30 min before the second PSC test (n = 7–8), *p < 0.05 and ***p < 0.001 using unpaired Student’s t-tests. Data are expressed as means ± S.E.M.

FIGURE 3

BLS, through rods, potentiates the antidepressant response of a combination of ketamine and scopolamine: (A) The day after the last day of CDMRD induction protocol in wild-type (WT) animals, a bright white light stimulation (BLS, 1000 lux), 1 h per day from ZT11 to ZT12 (light on from ZT0 to ZT12) was administered to the CDMRD mice (A), (n = 6). One-Way repeated measures ANOVA with post hoc Wilcoxon test. *p < 0.05 vs. D1. CDMRD wild-type (WT) animals (A), (n = 7), melanopsin knockout (Opn4 ^−/− , (B), n = 7) and rod-deficient mice (c, Nrl ^−/− , (B), (n = 6) housed under a 12L/12D cycle, are exposed to an additional 1-h BLS from ZT11 to ZT12 every day until the end of the protocol. 4 weeks post-stress, on day 33 (D33), all genotypes received an acute injection of Ketamine (Ket, 3 mg/kg, i.p) and Scopolamine (Sco, 0.1 mg/kg, i.p). Note that the augmenting effect of BLS was absent in Nrl ^−/− mice. One-Way repeated measures ANOVA with post hoc Wilcoxon test. *p < 0.05; **p < 0.01 vs. D1.

FIGURE 4

Role of LHb astrocytes and 5-HT system in the antidepressant effect of BLS: (A) Following CDMRD, mice were intracerebrally injected with vehicle or a AAV hM3q⁽⁺⁾-GFAP-Gq-mCherry virus in the LHb. Coronal image and scheme of brain section showing m-Cherry labelling (red) in the LHb (white square). (B) Astroglial and neuronal staining in the LHb: Anti-GFAP (green) and anti-NeuN (blue) immunohistochemistry confirming expression of AAV hM3q⁽⁺⁾-GFAP-Gq-mCherry virus in astroglia but not in neurons. scale bar = 150µm, enlargement, scale bar = 75 µm. (C) CDMRD mice were intracerebrally injected with vehicle or a AAV hM3q⁽⁺⁾-GFAP-Gq-mCherry virus in the LHb. A 4-week BLS was then applied. At day 33 (D33), mice received an injection of Clozapine-n-oxide (CNO, 1 mg/kg, i.p.) 1 h before the passive stress coping test. Half an hour after the CNO injection, the combination of ketamine/scopolamine (Ket + Sco, respectively 3 and 0.1 mg/kg, i.p) was administered to mice (n = 6–16). ***p ≤ 0.001 vs. D1. On D33, the two groups (sham and AAV-injected) were compared using LSD tests. **p ≤ 0,01. (D) CDMRD mice were subjected to the same BLS paradigm previously described. Three days before the passive stress coping test at D33 (ZT14), mice (n = 16) were divided into 2 groups, one received a dose of 150 mg/kg/day (i.p) of para-chlorophenylalanine (PCPA) to reduce the levels of 5-HT whereas the control group received a 0.9% saline injection (n = 6–10). At D33, both groups received the combination of Ket + Sco (respectively 3 and 0.1 mg/kg, i.p). **p ≤ 0.01; ***p ≤ 0.001 vs D1. On day 33, the two groups (vehicle and PCPA) were compared using LSD tests.