A natural carotenoid crocin exerts antidepressant action by promoting adult hippocampal neurogenesis through Wnt/β-catenin signaling

Abstract

Introduction: Adult hippocampal neurogenesis (AHN) is acknowledged to play a critical role in depression. Emerging evidence suggests that the Wnt/β-catenin pathway can modulate hippocampal neurogenesis. Crocin, a natural carotenoid, possesses antidepressant property. Yet, how it affects neurogenesis and exerts antidepressant response remains unknown.

Objective: To explore the role of AHN and Wnt/β-catenin in the antidepressant action of crocin.

Methods: Depressive-related behaviors, including sucrose preference test (SPT), tail suspension test (TST), forced swimming test (FST), and sexual behaviors were performed following crocin treatment. Neurogenesis was characterized via immunohistochemistry, immunofluorescence, Golgi staining and electrophysiology approach. Wnt/β-catenin signaling was examined with western blot analysis. The role of AHN Wnt/β-catenin cascade in crocin's antidepressant response was assessed by conditional removal of glial fibrillary acidic protein (GFAP)-expressing newborn neural cells, temozolomide administration, microinfusion of Dkk1 or viral-mediated shRNA of Wnt3a.

Results: Crocin decreased the immobility duration in TST and FST without impairing the performance in sexual behaviors. Crocin boosted the proliferation and differentiation of progenitors, and promoted dendritic maturation and functional integration of hippocampal newborn neurons. Conditional removal of GFAP-expressing neural cells or temozolomide administration impaired the antidepressant response of crocin. Additionally, Wnt/β-catenin signaling was promoted following crocin treatment. In chronic unpredictable mild stress (CUMS) murine model, crocin treatment displayed antidepressant response in SPT, FST and TST, and restored the neurogenesis levels and Wnt/β-catenin signaling impaired by CUMS. Infusion of Dickkopf-1 (DKK1) or knockdown of Wnt3a in the hippocampus impaired the antidepressant response of crocin.

Conclusion: Crocin exerted antidepressant response, which was dependent on enhancement of AHN and activation of the Wnt/β-catenin pathway.

Keywords: Adult hippocampal neurogenesis; Crocin; Depression; Wnt/β-catenin.

Graphical abstract

Fig. 1

Crocin attenuated depressive-like behaviors without inducing adverse effect on sexual behavior. (A) Effects of crocin on the immobility time in FST and TST. (B-C) Effects of crocin on the spontaneous locomotor activity in OFT. (D-F) The mating duration, number of bouts and latency to mate in the sexual behavior test. Results are presented as mean ± SEM. One-way ANOVA followed by Bonferroni post hoc test. *p < 0.05; **p < 0.01; ns, no significance.

Fig. 2

Crocin improved adult neurogenesis in the hippocampus. (A) Time line of BrdU injection and drug treatment. (B) Dividing cells (arrows) were identified using Ki67 immunohistochemistry in crocin-treated or control group. Scale bar: 100 μm. (C) BrdU immunohistochemistry (gray-brown) was used for detecting 2-week survival cells (arrows), Scale bar: 100 μm. (D) Representative images of DG area showed four types of cells: BrdU⁺ cells (green) which co-express strong DCX (red) or NeuN (blue); BrdU unlabeled DCX⁺/NeuN⁺ immature neurons (denoted by solid arrows); BrdU unlabeled DCX^- /NeuN⁺ mature neurons (denoted by hollow arrow). (E-G) Quantification of Ki67⁺, total BrdU⁺ and NeuN⁺/DCX^-/NeuN⁺ cells. (H) The percentage of DCX^- /NeuN⁺ mature neurons in crocin-treated group or control group. Results are presented as mean ± SEM. Unpaired student’s t-test. *p < 0.05; ns, no significance.

Fig. 3

Crocin enhanced adult neurogenesis and synaptic plasticity in hippocampus. (A) Images of DCX⁺/BrdU⁺ immunohistochemistry after crocin treatment. Scale bar: 100 μm. (B) Sholl analysis of DCX⁺/BrdU⁺ neurons on crocin-treated or control group. (C-E) Chronic crocin administration significantly increased DCX⁺/BrdU⁺ cells, intersection number and dendritic length. (F-G) Dendritic spine density in DG region of hippocampus. Scale bar: 10 μm. (H) Crocin improved the ACSF–LTP recording of the last 10 min. Results are presented as mean ± SEM. Unpaired student’s t-test. *p < 0.05, **p < 0.01; ns, no significance.

Fig. 4

Effect of crocin on the differentiation and proliferation of stem neural cell in vitro. (A) Immunofluorescence staining of primary NSC cells with Ki67 antibodies. Cells were treated with 2 μM or 5 μM crocin. Scale bar: 50 μm. (B) Immunofluorescence staining of primary NSC cells with NeuN, DCX, DAPI antibodies. Cells were treated with 2 μM or 5 μM crocin. Scale bar: 50 μm. (C) The representative pictures of primary NSCs in concentric sholl radii. ells were treated with 2 μM or 5 μM crocin. Scale bar: 10 μm. (D-E) Quantification of Ki67⁺ and NeuN⁺/DCX^-/DAPI cells. (F) The dendritic length after differentiation in 2 μM crocin-treated, 5 μM crocin-treated or control group. (G) The intersection number after differentiation in 2 μM crocin-treated, 5 μM crocin-treated or control group. Results are presented as mean ± SEM. Statistical analysis: one-way ANOVA followed by Bonferroni post hoc test. **p < 0.01.

Fig. 5

Crocin-related effect on depressive-like behaviors and neurogenesis were blocked in GFAP-TK mice with VGCV treatment. (A) Timeline of GFAP-TK mice experiment. GFAP-TK mice were treated with VGCV for 6 weeks. Crocin treatment started from 12th week. (B) Young granule neurons (red) cells expressing DCX in DG of wild type mice. No DCX⁺ cells can be detected in DG of GFAP-TK mice. Scale bar: 100 μm. (C) Quantification of total number of DCX⁺ cells in DG under crocin treatment. (D-E) The effect of crocin on immobility time of GFAP-TK mice or wild type mice in TST and FST. Results are presented as mean ± SEM. Two-way ANOVA followed by Tukey post hoc test. *p < 0.05; **p < 0.01; ns, no significance.

Fig. 6

Crocin-related effect on depressive-like behaviors and neurogenesis were blocked in mice with TMZ treatment. (A) Experimental timeline for the impact of TMZ blockade of late-phase DCX⁺ progenitors on crocin’s antidepressant activity. (B) Sections of TMZ, Vehicle or crocin treated mice were immunolabled with DCX (red). Amplified pictures of the dash-line rectangle region were shown to observe DCX⁺ cell morphology alteration following TMZ treatment. Scale bar: 100 μm. (C-D) Quantification of dendritic length and total DCX⁺ cell number in DG (unpaired student’s t-test). (E-F) The effect of crocin on immobility time of TMZ-treated mice or vehicle control mice in TST and FST. (two-way ANOVA followed by Tukey post hoc). Results are presented as mean ± SEM. **p < 0.01; ns, no significance.

Fig. 7

Crocin-related effect on depressive symptoms and neurogenesis were reversed by Dkk1 treatment. (A) Timeline of Dkk1 experiment. (B) Cannula position and Hoechst33342 dye infusions into the DG of mice. (C-D) Mice pretreated with Dkk1 followed by crocin treatment did not exhibit reduced immobility time in TST and FST. (E-F) Crocin significantly increased the number of Ki67 positive cells and number of mature neurons in Vehicle group, but not in Dkk1 group. (G-J) Effects of crocin treatment on Wnt3a, p-β-catenin and p-GSK3β protein levels in the hippocampus of Dkk1 mice. Results are presented as mean ± SEM. Statistical analysis: two-way ANOVA followed by Tukey post hoc test. *p < 0.05; **p < 0.01.

Fig. 8

Effects of crocin on depressive-like behaviors and adult hippocampal neurogenesis in CUMS-induced mice. (A) Timeline of CUMS experiment. (B) The body weight of mice in control, CUMS, CUMS treated with Flx (20 mg/kg) or CUMS treated with crocin (25 mg/kg). Body weight was measured for 6 weeks. (C) Sucrose preference rate of mice in each group. (D-E) The immobility time (TST and FST) of mice in four groups were quantified. (F-I) Crocin reversed CUMS-induced decline of Ki67⁺ cells, BrdU⁺ cells, NeuN⁺/DCX⁺/BrdU⁺ cells and mature neurons (NeuN⁺/DCX^-/BrdU⁺) in DG area. (J-M) Effects of crocin treatment on Wnt3a, p-β-catenin and p-GSK3β protein levels in the hippocampus of CUMS mice. Results are presented as mean ± SEM. Results are presented as mean ± SEM. One-way ANOVA followed by Bonferroni post hoc test. CUMS vs Control, #p < 0.05; ##p < 0.01; Crocin vs CUMS, *p < 0.05; **p < 0.01.