Molecular Neurobiology https://doi.org/10.1007/s12035-018-1042-8
Baicalin Modulates APPL2/Glucocorticoid Receptor Signaling Cascade, Promotes Neurogenesis, and Attenuates Emotional and Olfactory Dysfunctions in Chronic Corticosterone-Induced Depression Chong Gao 1 & Qiaohui Du 1 & Wenting Li 1 & Ruixia Deng 1 & Qi Wang 2 & Aimin Xu 3 & Jiangang Shen 1 Received: 20 November 2017 / Accepted: 27 March 2018 # Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract Olfactory dysfunction is often accompanied with anxiety- and depressive-like behaviors in depressive patients. Impaired neurogenesis in hippocampus and subventricular zone (SVZ)-olfactory bulb (OB) contribute to anxiety- and depressivelike behaviors and olfactory dysfunctions. However, the underlying mechanisms of olfactory dysfunction remain unclear. Our previous study indicates that adaptor protein, phosphotyrosine interacting with PH domain and leucine zipper 2 (APPL2), could affect the activity and sensitivity of glucocorticoid receptor (GR) and mediate impaired hippocampal neurogenesis, which contribute the development of depression. In the present study, we further identified the roles of APPL2 in olfactory functions. APPL2 Tg mice displayed higher GR activity and less capacity of neurogenesis at olfactory system with less olfactory sensitivity than WT mice, indicating that APPL2 could be a potential therapeutic target for depression and olfactory deficits. We then studied the effects of baicalin, a medicinal herbal compound, on modulating APPL2/GR signaling pathway for promoting neurogenesis and antidepressant as well as improving olfactory functions. Baicalin treatment inhibited APPL2/GR signaling pathway and improved neurogenesis at SVZ, OB, and hippocampus in APPL2 Tg mice and chronic corticosterone-induced depression mouse model. Behavioral tests revealed that baicalin attenuated depressive- and anxiety-like behaviors and improve olfactory functions in the chronic depression mouse model and APPL2 Tg mice. Taken together, APPL2 could be a novel therapeutic target for improving depressantrelated olfactory dysfunctions and baicalin could inhibit APPL2-mediated GR hyperactivity and promote adult neurogenesis, subsequently releasing depressive and anxiety symptoms and improving olfactory functions for antidepressant therapy. Keywords Glucocorticoid receptor . APPL2 . Depression . Olfactory functions . Neurogenesis . Hippocampus . Olfactory bulb . Subventricular zone
Introduction Electronic supplementary material The online version of this article (https://doi.org/10.1007/s12035-018-1042-8) contains supplementary material, which is available to authorized users. * Jiangang Shen
[email protected] 1
School of Chinese Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 10 Sasssoon Road, Hong Kong, Hong Kong SAR, China
2
Institute of Clinical Pharmacology, Guangzhou University of Chinese Medicine, Guangzhou, Guangdong, China
3
Department of Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, 21 Sasson Road, Hong Kong, Hong Kong SAR, China
Chronic depressive disorder is a high prevalent mental disease not only affecting the emotional adaptability but also inducing olfactory dysfunctions [1–4]. The impairment of adult hippocampal neurogenesis contributes to the pathogenic process of chronic depression disorder [5, 6]. Neurogenesis includes multiple steps for the proliferation, differentiation, and migration of neural stem/progenitor cells (NSPCs) for brain repair. In hippocampus, NSPCs regulate the adaptability of animals in face of psychiatric stressors [7]. In subventricular zone (SVZ) of the lateral ventricle (LV), NSPCs could migrate to the ipsilateral OB along the rostral migratory stream (RMS) for maintaining the regular olfactory functions [8]. The effects of antidepressants are highly dependent on adult hippocampal
Mol Neurobiol
neurogenesis [9, 10]. Therefore, endogenous NSPCs become important sources for regeneration therapy to attenuate the depressive behavior and olfactory deficits in depression disorder. Glucocorticoid receptor (GR) is crucial therapeutic target for promoting neurogenesis to prevent depression. Cardinal symptoms of depression are mainly associated with hyperactivity status of hypothalamic–pituitary–adrenal (HPA) axis during stress response [11]. Chronic corticosterone (CORT) exposure induces depression with olfactory deficits and neurogenic impairment in SVZ and OB regions in experimental depressive model [12]. Chronic elevation of glucocorticoid causes multiple neurological impairments with the reduced neurotrophic factors and neuronal atrophy [13]. GR activation could inhibit neurogenesis during stress responses [14]. Intra-nuclear GR initiates the expression of glucocorticoid-regulated kinase 1 (SGK1) to form the feedback circle and regulate neurogenesis [14, 15]. Thus, GR signaling pathway could be a promising target for promoting adult neurogenesis and regulating mood behaviors and olfactory functions in chronic depressive disorders. Metabolic factors could modulate adult neurogenesis and affect neurological behaviors [16]. Many metabolic factors act as the mediator for adult NSPCs and affect the related neural behaviors [16]. For instance, metabolic factors like leptin, adiponectin, and insulin-like growth factor-1 (IGF-1) play different roles in regulating adult neurogenesis [17–19]. Adaptor protein APPL2 (adaptor protein, phosphotyrosine interacting with PH domain and leucine zipper 2) is a new identified metabolic factor for cell growth as well as insulinstimulated glucose uptake [20]. Our recent study showed that APPL2 transgenic mice had depressive-/anxiety-like behaviors in correlation with decreased capacity of hippocampal neurogenesis, showing the decreased proliferation, neurogenic fate choice as well as the maturation of NSPCs. The underlying mechanisms of APPL2 could be associated with the regulation of GR activity and the insensitivity of GR to stress stimulation [21]. Herein, we further explored the roles of APPL2 in modulating depression and olfactory dysfunctions and proposed APPL2 as therapeutic target for antidepressant therapy. Baicalin is a flavonoid isolated from the root of Scutellaria baicalensis, a commonly used medicinal herb for depression in Traditional Chinese Medicine (TCM). Baicalin has multiple biological functions including anti-inflammatory activity, antioxidant, and anti-apoptotic effects [22–24]. Baicalin could promote NSPCs toward neuronal differentiation [25] and prevented the depressive-like behaviors in chronic mild stress (CMS) animal model [26]. Baicalin downregulated SGK1 expression in the hippocampus and reversed corticosteroneinduced depressive-like behaviors [27]. However, whether baicalin modulates APPL2/GR signaling pathway for neurogenic-promotion and whether baicalin has promoting effects on olfactory function remain unknown. In the present study, we tested the hypothesis that baicalin could inhibit APPL2/GR signaling pathway and promote neurogenesis in
olfactory system, subsequently improving depressive and anxiety behaviors and attenuating olfactory deficits.
Material and Methods Depressive Animal Model Adult male C57BL/6N mice (8 weeks old) were obtained from laboratory animal unit (LAU), the University of Hong Kong. APPL2 Tg mice (male, 8 weeks) were raised in the minimal disease area (MDA) of LAU in HKU. All mice were maintained in the standard environment (12 h light/dark cycle, with lights on at 8:00 A.M., ad libitum access to dry food pellets and water) following the ethics roles of the University of Hong Kong (CULATR No. 3748-15; 4001-16). Genotyping procedure was conducted following the protocol as a previous report [20]. Depressive animal model was made by chronic administration of corticosterone (CORT) [12]. Briefly, CORT (4-pregnen-11bdiol-3 20-dione 21-hemisuccinate, Sigma-Aldrich) was dissolved in its vehicle (0.45% hydroxypropyl-cyclodextrin, -CD, Sigma-Aldrich) and was supplied ad libitum in drinking water (70 μg/ml, equivalent to 5 mg/kg/day) for 30 consecutive days (Supplementary Fig. 1). The mild stress that was induced by short period of CORT administration (10 days) was used to detect the olfactory resistance of the WT and APPL2 to stress.
Drug Treatment WT mice were randomly separated into five groups, including control, depression model, low and high dosage baicalin treatment, and fluoxetine positive control. Drugs were dissolved in the vehicle consisting with 30% ethanol, 30% polyethyleneglycol (PEG), and 40% H2O. Baicalin (3.35 mg/kg/day for low dose; 6.7 mg/kg/day for high dosage) or fluoxetine (18 mg/kg/day) or vehicle were orally administrated into the mice starting at day 15 for 7 consecutive days. APPL2 Tg mice were used to explore whether APPL2 signaling was involved in baicalin’s antidepressant effects. The APPL2 Tg mice were treated with 6.7 mg/kg/day of baicalin by oral administration and performed FST and olfactory sensitivity test before sacrificed by heart perfusion fixation with 4% paraformaldehyde (PFA). WT mice received the drug treatments with the same protocol.
Behavioral Tests We performed the behavioral tests including splash, tail suspension test, forced swim test, open field test, and novelty suppressed feeding for evaluating the depressive, anxiety, and olfactory functions of the animals (for details, refer to supplementary information). All drugs were continuously used during the behavioral tests (Supplementary Fig. 1).
Mol Neurobiol
Neurogenesis Study
Supplementary Information
NSPCs proliferation was identified by co-immunostaining of bromodeoxyuridine (5-bromo-2′-deoxyuridine, BrdU) and specific cellular markers (details seen in supplementary information). BrdU (50 mg/kg) was daily injected to wildtype mice and APPL2 Tg mice synchronized with baicalin and vehicle treatment. BrdU (50 mg/kg/day) was intraperitoneally injected into the mice for five consecutive days. Mice were then sacrificed at day 5 after the last injection (time line shown in Supplementary Fig. 1).
Methodology details including the behavioral tests, immunofluorescence, confocal image acquisition (dSTORM; 3D reconstruction), and western blot are shown in the supplementary information.
Cell Culture and Drug Treatment Human SH-SY5Y cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and maintained in high glucose Dulbecco’s modified Eagle medium (DMEM; Hyclone) supplemented with 10% fetal bovine serum (FBS; Gibco), 1% penicillin/streptomycin (PS; Gibco) and 1% L-glutamine (Gibco). Cells were maintained in 75cm2 flask (Falcon® 75cm2 Rectangular Canted Neck Cell Culture Flask with plug-seal cap; Corning Life Sciences, USA). The cells with the density of 2.0 × 105 cells/well were seeded into the 12-well Falcon™ polystyrene microplates (Corning) for dSTORM image staining. The cells with the density of 1.0 × 105 cells were seeded into the 24-well plate for LSM800 image detecting. Cells were treated with 1 μM of dexamethasone (DEX; Sigma-Aldrich, D4902, dissolved DMSO) for 1 h. Afterward, medium was changed and the cells were incubated with the medium containing baicalin (50 μM) for 3 h (Supplementary Fig. 1). Details for cell fixation and image preparation were seen in supplementary information.
Statistical Analysis All the data were presented with mean ± SEM. Data analysis was performed with Prism 6 (GraphPad Software, USA). Two-tailed student unpaired t test was applied for data statistical analysis for result comparison of the two groups. Before t test analysis, F test was performed to compare the variance of the two groups, and t test with Welch’s correction was performed when F test showed significance. Variables that passed the normality test were analyzed with ANOVA followed by Turkey’s multiple comparison test for multiple comparisons. Two-way ANOVA was used for analyzing multiple comparisons of the paired data. Sidak’s multiple comparisons were used for statistical analysis between two groups in two-way ANOVA. Turkey’s multiple comparison test was applied to analyze the difference over two groups in two-way ANOVA. A p value of < 0.05 was used as a cut-off for statistical significance. Statistical details were shown in Supplementary Table 2.
Results APPL2 Serves as the Negative Modulator to Inhibit Adult Neurogenesis in Olfactory System via Regulating GR Signaling Previously, we found that APPL2 Tg mice had hyperactivity of GR in hippocampus area under basic and stress conditions. APPL2 Tg mice had the impaired hippocampal neurogenesis and insensitivity of GR to environmental stress, inducing inadaptability emotional behavior to stress [21]. Olfactory bulb (OB) is one of the main regions to process olfactory information in adult brain, and subventricular zone (SVZ) is the main neurogenic region for neuroblasts formation in OB [28]. Whether APPL2 can regulate the olfactory behavior and neurogenesis in olfactory system remains unexplored. To answer this question, we investigated the olfactory sensitivity of APPL2 Tg mice in comparing with wild-type (WT) littermates. During the test, APPL2 Tg mice presented the significant shortened exploration time to odorant when compared with WT mice (Fig. 1a, two-way ANOVA, F (3, 48) = 0.8883, p = 0.454; 10−2: p = 0.0117, 10−3: p = 0.0316, Supplementary Table 2), indicating APPL2 could affect the olfactory function. We then addressed the roles of APPL2 in modulating GR and its downstream signaling serum/glucocorticoid-regulated kinase 1 (SGK1) in the OB. APPL2 Tg mice had significantly higher expression of GR phosphorylation and SGK1 protein in OB than the WT mice (Fig. 1b–d; student t test, p < 0.05, supplementary Table 2). This result indicates that APPL2 could regulate the activity of GR in OB region, which highly associated with the regulation of olfactory neurogenesis. Herein, we subsequently investigated neuronal differentiation of NSPCs by co-immunostaining BrdU with immature neuron marker doublecortin (DCX) at SVZ and OB in APPL2 Tg mice and WT littermates. BrdU (50 mg/kg/day) was injected into the mice for 7 days. Immunofluorescence (IF) showed that APPL2 Tg mice had significantly decreased DCX+/ BrdU+ cells in SVZ and the granular cell layer (GCL) of OB, indicating the APPL2 overexpression could impair adult neurogenesis in hippocampus and olfactory system (Fig. 1e, f, two-way ANOVA, F (1, 18) = 0.1342, p = 0.7184; SVZ: p = 0.0031; OB-GCL: p = 0.0098, Supplementary Table 2). To further explore the roles of APPL2 in affecting olfactory neurogenesis, we performed olfactory sensitivity test and investigated the olfactory exploration time to odorant of both
Mol Neurobiol
WT and APPL2 Tg mice under short-term stress. BrdU (50 mg/kg/day) was injected for late 7 days and the BrdU+/ DCX+ cells in OB and SVZ were detected for labeling immature neurons (Fig. 1g, h). After short-term administration of corticosterone (CORT, 70 μg/ml, equivalent to 5 mg/kg/day, 10 consecutive days), WT mice had no change in the BrdU+/ DCX+ cells at the olfactory system, whereas APPL2 Tg mice
showed significantly reduced BrdU+/DCX+ cells in OB and SVZ (Fig. 1j, k; two-way ANOVA, APPL2 Tg: F (1, 20) = 0.2262, p = 0.4483; WT: F (1, 16) = 0.6042, p = 0.4483; SVZ: p = 0.0408, OB-GCL: p = 0.0092; Supplementary Table 2). Taken together, we conclude that APPL2 could increase GR signaling and inhibit neurogenesis in OB and SVZ under short-term CORT-induced stress. Thus, APPL2 could be a
Mol Neurobiol
Fig.
1 APPL2 could be a modulator to olfactory sensitivity via promoting neurogenesis at olfactory system. a Statistical result indicates that APPL2 Tg mice showed the decreased olfactory sensitivity, which reflected by the decrease olfactory exploration to odorant (two-way ANOVA, *p < 0.05, Supplementary Table 2). b Western blot showing the phosphorylation of GR (S211) and the expression of SGK1 in the OB region of the WT and APPL2 Tg mice. c, d Statistical analysis showing APPL2 Tg mice had the over phosphorylation level of GR and the increased expression of SGK1 in the OB region compared with WT (student t test, *p < 0.05, Supplementary Table 2). e Immunofluorescent image showing the neurogenesis at OB/SVZ by labelling BrdU (Green), DCX (Red) and DAPI (Blue). f Statistical result showing the incorporation of BrdU with DCX was dramatically decreased in both SVZ and OB in APPL2 Tg mice (two-way ANOVA, **p < 0.01, Supplementary Table 2). g, h The experimental procedure of short-term CORT treatment induced mild stress. Immunofluorescence showing neurogenesis in olfactory system by labelling BrdU (Green), DCX (Red) and DAPI (Blue). i, j Statistical analysis indicates in WT mice, the short-term (10 days) CORT treatment did not affect the incorporation ratio of BrdU with DCX in olfactory system in WT mice (two-way ANOVA, WT: F (1, 16) = 0.6042; p = 0.4483; Supplementary Table 2). However, same treatment induced the dramatic decreased BrdU+/DCX+ cell density in both OBGCL and SVZ of APPL2 Tg mice (two-way ANOVA, F (1, 20) = 0.2262, SVZ: p = 0.0408, OB-GCL: p = 0.0092; Supplementary Table 2)
therapeutic target for improving depression-associated olfactory deficits on anti-depressive therapy.
Baicalin Improved the Adult Brain Neurogenesis in APPL2 Tg Mice and Depression Model via Regulating GR Signaling Previous study indicates that baicalin could promote NSPCs toward neuronal differentiation and prevent the depressive-like behaviors in chronic mild stress (CMS) animal model [26]. Baicalin also revealed to downregulate SGK1 expression in the hippocampus and reverse corticosterone-induced depressive-like behaviors [27]. In the present study, to elucidate whether baicalin could modulate APPL2/GR signaling, we investigated the effects of baicalin on APPL2 expression, GR phosphorylation and SGK1 expression in OB and hippocampus in both APPL2 Tg mice and CORT-induced chronic depressive mouse model. Both WT and APPL2 Tg mice were orally administrated with baicalin for 7 consecutive days (6.7 mg/ kg/day). Western blot analysis showed that baicalin treatment decreased the hyperactivity of GR and downregulated the expression level of SGK1 in both OB and hippocampus of the APPL2 Tg mice (Fig. 2a, student t test, Supplementary Table 2). To induce depressive animal model, WT mice were exposed to chronic CORT treatment (70 μg/ml/day, 30 days). The mice were orally administrated with baicalin at a dosage of 6.7 mg/kg/day for late 15 days (experimental procedure showed in Supplementary Fig. 1A). Notably, chronic CORT administration significantly upregulated the expression of APPL2
protein in both OB and hippocampus of the mice (Fig. 2b, one-way ANOVA, p < 0.01, Supplementary Table 2). Baicalin treatment significantly downregulated the expression of APPL2 in the OB and hippocampus in the WT mice exposed to CORT treatment (Fig. 2b, one-way ANOVA, p < 0.05, Supplementary Table 2). Consistently, the CORT-treated mice had significantly higher level of the phosphorylation of GR at Ser211 and the expression SGK1 in the OB and hippocampus than normal control m i c e ( F i g . 2 c ; o n e - w a y A N O VA , p < 0 . 0 0 0 1 ; Supplementary Table 2). Baicalin treatment significantly reduced the GR phosphorylation and SGK1 expression at the OB and hippocampus of the CORT mice, whose effects were similar to the fluoxetine (FLX)-treated mice as positive control (Fig. 2c; Supplementary Table 2). We next performed cell image experiments in SH-SY5Y cells and tested whether baicalin could affect GR nuclear translocation (Fig. 2d, e). The cells were incubated with GR agonist dexamethasone (DEX) with or without baicalin treatment (experimental procedure showed in Supplementary Fig. 1). By using dSTORM (direct stochastic optical reconstruction microscopy), we found that DEX induced GR signal translocation into the nucleus (Fig. 2d). With 3D imaging reconstruction for the co-staining of GR and DAPI, we confirmed that DEX induced the nuclear translocation (Fig. 2e; Movie S1~3). As expected, baicalin treatment (50 μM) reserved the nuclear translocation of GR. Those results indicate that baicalin could buffer the over-activity and prevent the nuclear transport of GR. To further detect the effects of baicalin to adult neurogenesis in APPL2 Tg mice, we labeled newly proliferated cells by BrdU injection (50 mg/kg/day, 7 days). The in vivo immunofluorescent staining study showed that baicalin treatment significantly increased dual positive staining of BrdU and DCX in the hippocampus, SVZ, and OB-GCL regions of the APPL2 Tg mice (Fig. 2f, g, two-way ANOVA, F (2, 30) = 4.282, p = 0.0231; hippocampus: p = 0.0258; SVZ: p < 0.0001; OBGCL: p = 0.0201; Supplementary Table 2), suggesting that baicalin could reserve adult neurogenic capacity in the APPL2 Tg mice. Interestingly, baicalin treatment did not enhance neurogenesis in the three main niches of adult brain in WT mice under basic condition (Fig. 2f, h, Supplementary Table 2). Overall, those results indicate that baicalin could inhibit APPL2/GR signaling and promote neurogenesis in APPL2 Tg mice and chronic depressive animal model.
Baicalin Prevents CORT-Induced Neurogenic Impairments in Hippocampus, Olfactory Bulb, and Subventricular Zone Hyperactive HPA axis mediates impairment of hippocampal neurogenesis, contributing to the pathological process of depression [29]. We firstly detected BrdU incorporation in the
Mol Neurobiol
hippocampal area and explored the effects of baicalin on hippocampal neurogenesis in the CORT-induced depression model. The CORT mice showed significant lower level of BrdU+ staining cells in the hippocampus than the normal
control mice, indicating the impairment of NSCs proliferation under CORT stimulation. Baicalin treatment dosedependently increased the newly produced cells with BrdU+ staining in the hippocampus (Fig. 3a, d; one-way ANOVA;
Mol Neurobiol
Fig. 2
Baicalin improves the adult brain neurogenesis in APPL2 Tg mice and chronic CORT-induced depression model via regulating APPL2/GR signaling. a Western blot showing the effects of short-term treatment of baicalin (6.7 mg/kg/day for 7 days) on GR signaling in APPL2 Tg mice. Baicalin treatment decreased the phosphorylation level of GR and the expression of SGK1 in both OB and hippocampal regions (student t test, *p < 0.05; **p < 0.01; Supplementary Table 2). b Western blot showing the expression of APPL2 at OB and hippocampal tissues in CORT-induced depression model in WT mice with/without baicalin treatment. Statistical analysis revealing a dramatic increase in expression of APPL2 in OB and hippocampus, which was induced by chronic administration of CORT, whose effect was reserved by treatment of baicalin (6.7 mg/kg/ day, one-way ANOVA, *p < 0.05, **p < 0.01; Supplementary Table 2). c Western blot showing that at OB and hippocampus, CORT administration induced the elevated phosphorylation of GR as well as the increased expression level of SGK1 (one-way ANOVA, details seen in Supplementary Table 2). Baicalin treatment significantly reduced the phosphorylation level of GR and the expression of SGK1 in OB and hippocampus. Same effects were shown in fluoxetine group (one-way ANOVA, details seen in Supplementary Table 2). d Cell image showing the cellular distribution of GR with dSTORM (up: wide field; down: dSTORM). Images showed that administration of DEX induced the recruitment of GR into nucleus at cultured SH-SY5Y neuroblastoma cells (green: GR; red: β-Actin). Treatment of baicalin normalized the distribution of GR in subcellular level. e Regular confocal single cell image (left) and its 3D reconstruction for confirming the nuclear translocation of GR (green) in Control, DEX and baicalin treatment groups. Position of the nucleus was detected with DAPI staining (blue). f Confocal image showing the neurogenesis at DG, SVZ and OB in WT and APPL2 Tg mice. Neurogenesis was detected by immunofluorescence (IF) to co-label the newborn immature BrdU (green) and DCX (red) positive staining cells. DAPI (blue) was used to label the cell nuclei. g, h Statistical analysis revealing that baicalin treatment enhanced the cell density of the newborn immature neurons (BrdU/DCX) in three neurogenic niches of the adult brain of APPL2 Tg mice (two-way ANOVA, *p < 0.05; **p < 0.01; Supplementary Table 2)
control vs CORT: p < 0.0001; high vs CORT p = 0.0018; Supplementary Table 2). We further identified newborn neurons by dual immunostaining of BrdU/NeuN. The results revealed that CORT stimulation significantly decreased the density BrdU+/NeuN+ in the hippocampus of the mice, indicating the decreased hippocampal neurogenesis. Baicalin treatment significantly increased the population of newborn neurons with BrdU+/NeuN+ staining in hippocampal DG region whose effects were similar with fluoxetine (Fig. 3a, e; one-way ANOVA, Supplementary Table 2). The results suggest that baicalin could promote hippocampal neurogenesis in the CORT-induced depressive mice. We then assessed the activation of NSCs by labeling the proliferative marker minichromosome maintenance proteins 2 (MCM2). In the hippocampal DG region, CORT stimulation remarkably reduced the density of MCM2+ staining cells (Fig. 3b, f; oneway ANOVA; p = 0.0004; Supplementary Table 2), suggesting that CORT could suppress the activation of NSCs in SGZ. The CORT mice that received high dosage of baicalin had significantly higher rates of MCM2+ cells than the CORT vehicle group (Fig. 4b, f; one-way ANOVA; p = 0.0067),
which resembled the effects of fluoxetine (Fig. 3b, f; oneway ANOVA; p = 0.0108; Supplementary Table 2). Furthermore, DCX was co-stained with the BrdU to identify the generation rate of immature neurons (Fig. 3c). Compared with normal control mice, the CORT mice showed significant lower population of BrdU+/DCX+ staining cells in the hippocampal DG region (Fig. 3c, g; one-way ANOVA; p = 0.0014; Supplementary Table 2). Treatment of baicalin at a dosage of 6.7 mg/kg/day had the increased BrdU+/DCX+ neurons in the same area of the mice, indicating the promoting effects of baicalin on neuronal differentiation (Fig. 3c, g; one-way ANOVA; high vs CORT: p = 0.0208; Supplementary Table 2). Baicalin had similar effects on neuronal differentiation to fluoxetine (Fig. 3c, g; one-way ANOVA; p = 0.0004; Supplementary Table 2). Those results indicate that baicalin could promote adult hippocampal neurogenesis in the CORTinduced depressive mice. We then investigated the density of BrdU + /NeuN + staining cells in different OB subregions including the granular cell layer (GCL) and rostral migration stream (RMS) (Fig. 4a). GCL is the OB region with the most newborn neurons migrating and developed into granular cells (~ 94%) [30]. The CORT group had a striking decrease of the BrdU+ neurons in the GCL and RMS subregions (Fig. 4a, c; two-way ANOVA, F (4, 50) = 6.531, p < 0.001; Supplementary Table 2). Baicalin treatment significantly increased the density of BrdU+/NeuN+ cells in the GCL areas whose effects were similar with fluoxetine (Fig. 4a, c; Supplementary Table 2). RMS is the primary path of the migration of newborn neurons from SVZ to OB. We tracked the newborn immature neurons in the RMS (Fig. 4b, d). Administration of CORT caused a decrease of the immature neuronal density (Fig. 4b, d; oneway ANOVA, p = 0.0006; Supplementary Table 2). In comparison with the CORT alone group, the baicalintreated mice (6.7 mg/kg/day) had significantly increased rates of BrdU+/DCX+ cells in the SVZ and OB (Fig. 4b, d; one-way ANOVA, p = 0.0249; Supplementary Table 2). The effects of baicalin were similar to fluoxetine treatment (Fig. 4b, e; one-way ANOVA, p = 0.0352; Supplementary Table 2). Those results indicate that baicalin could promote neurogenesis in OB, potentially contributing to the improvement of olfactory functions. We next assessed the neurogenesis by staining proliferative marker Ki67 for active NSPCs in SVZ. Overall, Ki67 positive cells were collectively counted from the dorsal, middle, and ventral part of SVZ together (Fig. 5a, n = 6 per group). The CORT mice had significant lower level of NSPCs proliferation in SVZ than the normal control mice (Fig. 5a, c; one-way ANOVA, p < 0.0001; Supplementary Table 2). In baicalin treatment groups with both high and low dosages, there was significantly increased density of Ki67+ cells in the SVZ (Fig. 6a, c; one-way ANOVA, p < 0.0001; Supplementary
Mol Neurobiol
BrdU
200 m
BrdU/NeuN
BrdU 50 m
BrdU/NeuN
B
D MCM2 MCM2/DAPI
100 m
BrdU+ Cell Density (/mm3)
**
***
20000 15000
5000
ol C
O
tr
BrdU+/NeuN+ Cell Density (mm3) MCM2+ Cell Density (/mm3)
40000
***
*
10000
35
C 3.
10000
35
7m g Fl uo /kg xe tin e
kg g/
R O
on
tr
ol
T
0 C
3.
**
8000
*
***
6000 4000 2000
3.
6. 7m g Fl uo /kg xe tin e
kg
m
g/
R O
C
C
on
tr
ol
T
0
35
G
DCX+/BrdU+ Cell Density (/mm3)
3.35 6.7 FLX
Merge
*
20000
20 m
DCX
**
m
50 m
C
100 m
***
30000
6.
BrdU
6. 7m g g Fl uo /kg xe tin e
O
g/ k
R
ol
T
0 C on tr
F
BrdU/DCX BrdU/DCX/DAPI BrdU/DCX BrdU/DCX/DAPI
*
5000
m
3.35 FLX
6.7
E 15000
3. 35 RT m g 6. /kg 7m g Fl uo /kg xe tin e
0
CORT
Control
C
**
10000
C on
CORT Control
A
Fig. 3 Baicalin improves adult hippocampal neurogenesis in chronic CORT induced depression model of WT mice. a Hippocampal dentate gyrus (DG) images showing the adult born neurons (red: NeuN; green: BrdU). b Images in DG to show the proliferating cells (green: MCM2; gray: DAPI). c Hippocampal DG image showing the newborn immature neurons (right section: enlarged image to show the morphology of DCX+/BrdU+ cells). d CORT administration induced the decreased density of BrdU+ cells (one-way ANOVA, ***p < 0.0001; Supplementary Table 2). High dosage of baicalin increased the BrdU+ cell number (one-way ANOVA, **p = 0.0018; Supplementary Table 2), whose effect was similar as treatment of fluoxetine (one-way ANOVA, **p = 0.0028; Supplementary Table 2). e CORT decreased the adult born neurons (one-way ANOVA, ***p = 0.0002; Supplementary Table 2).
High dosage of baicalin (6.7mg/kg/d) increased the number of BrdU+ neurons (one-way ANOVA, *p = 0.0107; Supplementary Table 2), with the same effect of fluoxetine (one-way ANOVA, *p = 0.0236; Supplementary Table 2). f CORT induced the suppressed cell proliferation in DG (n = 6, one-way ANOVA, ***p = 0.0004; Supplementary Table 2). High dosage of baicalin increased the density of proliferation cells (one-way ANOVA, **p = 0.0067; Supplementary Table 2), with the same effect as fluoxetine (one-way ANOVA, *p = 0.0108). g CORT administration caused the decline of immature neuronal density (n = 6, oneway ANOVA, **p = 0.0014). High dosage of baicalin (one-way ANOVA, *p = 0.0208; Supplementary Table 2) as well as fluoxetine (one-way ANOVA, ***p = 0.0004; Supplementary Table 2) treatment upregulated the density of DCX+/BrdU+ neurons
Table 2). Baicalin had similar effects to fluoxetine (Fig. 5a, C; one-way ANOVA, p < 0.0001; Supplementary Table 2). Those results suggest that baicalin could promote the proliferation of NSPC in SVZ. We next identified neuronal differentiation of NSPCs by co-immunostaining BrdU with immature neuron marker DCX (Fig. 5b). Given that the spatial-segregated neuronal type is generated from the progenitors at SVZ, we separately analyzed the dorsal lateral and ventral subregions [31]. The
CORT mice had strikingly decreased newborn immature neurons (BrdU+/DCX+) at both dorsal and ventral subregions of SVZ (Fig. 5b, d, e; Supplementary Table 2). Compared with the CORT group, the mice treated with high dosage of baicalin (6.7 mg/kg/day) had significantly increased density of neuronal progenitors at both dorsal and ventral SVZ regions (Fig. 5b, d, e; Supplementary Table 2). However, fluoxetine only induced neuronal formation in dorsal part but had no effect on the ventral region (Fig. 5b, d, e; Supplementary
GCL
BrdU/NeuN/DAPI
100 m
FLX
6.7
3.35
GL GCL RMS
RMS
500 m
BrdU
DAPI
*** *** * **
5000 0
GCL
RMS
Merge BrdU
DAPI
D 100 m
***
40000
*
20000 10000
6.
7m
g/
Fl uo kg xe tin e
T
g/
R
m
3.
35
tr on
kg
0
C
FLX
*
30000
ol
Merge
6.7
DCX
50000
O
20 m
3.35
10000
Control CORT 3.35mg/kg 6.7mg/kg Fluoxetine
**
C
CORT Control
DCX
***
15000
DCX+/BrdU+ Cell Density (/mm3)
B
C BrdU+/NeuN+ Cell Density (mm3)
A
CORT Control
Mol Neurobiol
Fig. 4 Baicalin treatment enhances the neurogenesis in olfactory bulb (OB). a Confocal images showing the adult born neurons incorporated with BrdU at OB region in CORT treated depression model with WT mice (green: BrdU; red: NeuN). Image at left showing the whole view of the OB at one side, and the position of glomerular layer (GL), granular cell layer (GCL), and rostral migratory stream (RMS). b Confocal images showing the newborn immature neurons at RMS in OB. Images at the right showing the enlarged view to reflect the morphology of DCX+/BrdU+ cells in OB. c CORT treatment significantly decreased neuronal production at different OB subregions (two-way ANOVA, F
p = 0.0003; Supplementary Table 2). At GCL, Treatment of baicalin at both dosages (3.35mg/kg/d and 6.7mg/kg/d) reserved the declined neuronal production induced by CORT, with the same effects of fluoxetine (two-way ANOVA, details seen in Supplementary Table 2). d CORT induced the decreased DCX+/BrdU+ cells density at OB-RMS subregion (one-way ANOVA, ***p = 0.0006; Supplementary Table 2). Baicalin treatment at high dosage (6.7mg/kg/d) enhanced the immature neuronal density (one-way ANOVA, *p = 0.0249; Supplementary Table 2), whose effect was similar with fluoxetine (one-way ANOVA, *p = 0.0352; Supplementary Table 2)
Table 2). Those results indicate that baicalin could promote the neurogenesis in LV, potentially contributing to the neurogenesis in OB and maintenance of olfactory behaviors.
to the chronic CORT exposure (Fig. 6a, n = 10~12 per group). In splash test, the CORT group presented a significant shorter grooming time than normal control group (Fig. 6b; one-way ANOVA; p = 0.0044; Supplementary Table 2). Baicalin treatment at high dosage (6.7 mg/kg/day) showed a remarkably increased grooming time in the CORT-induced depressive mice. The effects of baicalin were similar with fluoxetine (Fig. 6b; one-way ANOVA; baicalin vs CORT: p = 0.0464; fluoxetine vs CORT: p = 0.0140; Supplementary Table 2). Tail suspension test (TST) and forced swim test (FST) were conducted to further confirm the antidepressant effects of baicalin by recording the immobile posture. The CORT group showed a significant decreased mobility (Fig. 6c, d; one-way
Baicalin Treatment Attenuates the CORT-Induced Depressive- and Anxiety-Like Behaviors and Olfactory Deficit To test whether baicalin could attenuate the depressioninduced emotional dysfunctions and olfactory deficits, we performed behavioral tests in the CORT-induced depression model. We first evaluated the effects of baicalin on depressive- and anxiety-like behaviors in WT mice subjected
(4, 50) = 6.531,
Mol Neurobiol
Cell Proliferation Dorsal
Middle
Ventral
Ki67/DAPI
50 m
B Dorsal
CORT
CORT Control
3.35
BrdU
DAPI
Ventral Merge
DCX
BrdU
DAPI
Merge
50 m
FLX
C: Cell Proliferation 15000
10000
5000
8000 6000 4000 2000
tin xe
g/ uo Fl
7m 6.
35 3.
e
kg
kg
T m
O
g/
R
ol C
tin xe
g/ Fl
uo
7m
g/ 6.
m
O
35 3.
e
kg
kg
T R
ol tr on C
tin xe
g/ uo Fl
7m 6.
35 3.
***
0
e
kg
kg
T m
O
g/
R
ol C
tr on
***
10000
0
0
C
*
tr
10000
*
on
20000
**
E: Ventral SVZ
C
*** *** ***
BrdU+/DCX+ Cell Density (mm3)
***
D: Dorsal SVZ
BrdU+/DCX+ Cell Density (mm3)
FLX
200 m
30000
6.7
6.7
3.35
Middle
Ventral
Ki67+ Cell Density (per mm3)
Dorsal DCX
C
Control
A
Fig. 5 Baicalin improves the neurogenesis in subventricular zone (SVZ) in CORT induced depression model in WT mice. a Confocal images showing cell proliferation through labeling proliferative marker Ki67 (green: Ki67; blue: DAPI). Right image showed the whole view of SVZ in single side. The Ki67+ cell density counting was concentrated on analyzing the dorsal, middle, and ventral parts of the SVZ. b Confocal images showing the immature newborn neurons in SVZ at dorsal and ventral parts. c CORT induced the striking decline of the cell proliferation (n = 6, one-way ANOVA, ***p < 0.0001; Supplementary Table 2). While treatment of baicalin at low and high dosages (3.35mg/ kg/d and 6.7mg/kg/d) enhanced the cell proliferative level (one-way ANOVA, ***p < 0.0001; Supplementary Table 2), whose effect was similar with fluoxetine (one-way ANOVA, ***p < 0.0001; Supplementary
Table 2). d In dorsal SVZ, CORT significantly decreased the density of immature neurons incorporated with BrdU (one-way ANOVA, **p = 0.0020; Supplementary Table 2). Baicalin at high dosage and fluoxetine increased the production of immature neurons in dorsal SVZ (one-way ANOVA, high dosage of baicalin vs CORT *p = 0.0130; fluoxetine vs CORT *p = 0.0260; Supplementary Table 2). e In ventral SVZ, CORT resulted the decline of immature neurons density (one-way ANOVA, ***p < 0.0001; Supplementary Table 2). While baicalin at high dosage (6.7mg/kg/d) increased the production of immature neurons in ventral SVZ (one-way ANOVA, ***p < 0.0001; Supplementary Table 2). Although fluoxetine did not show the significance (one-way ANOVA, p = 0.2046; Supplementary Table 2)
ANOVA, p < 0.0001; Supplementary Table 2). In TST, WT mice treated with high dosage of baicalin (6.7 mg/kg/day) had significantly higher mobility performance during suspension than vehicle control WT group (Fig. 6c; one-way ANOVA, p = 0.0021; Supplementary Table 2). While in FST, the CORT mice treated with baicalin at 3.35 and 6.7 mg/kg/day had significantly increased mobility in the swim tank than the CORT alone group (Fig. 6d; one-way ANOVA; low vs
C O RT: p = 0 . 0 3 7 1 ; h i g h v s C O RT: p < 0 . 0 0 0 1 ; Supplementary Table 2). As the positive control drug, fluoxetine also increased mobility time in the CORT mice (Fig. 3c, d; one-way ANOVA; Supplementary Table 2). Thus, baicalin could attenuate the CORT-induced depressive-like behaviors. Open field test (OFT) and novelty suppressed feeding (NSF) were performed for anxiety mood behavior study. In OFT, the CORT mice showed anxiety-like behavior with a
Mol Neurobiol
Fig. 6 Effects of baicalin to corticosterone (CORT)-induced anxiety- and depressive-like symptoms, as well as the olfactory deficits in WT mice. a In splash test, CORT resulted in a significant reduction of grooming behavior (one-way ANOVA, **p = 0.0044; Supplementary Table 2). High dosage baicalin (6.7 mg/kg/day) increased grooming time (oneway ANOVA, *p = 0.0464; Supplementary Table 2), whose effect was similar with fluoxetine (one-way ANOVA, *p = 0.0140; Supplementary Table 2). b In tail suspension test (TST), CORT induced the faster immobility compared with control (n = 12, one-way ANOVA, ***p < 0.0001; Supplementary Table 2). High dosage baicalin increased the mobility time (one-way ANOVA, **p = 0.0021; Supplementary Table 2), as the same effect of fluoxetine (one-way ANOVA, **p = 0.0027; Supplementary Table 2). c Forced swim test (FST) showing that CORT caused less mobility of the mice (n = 12, one-way ANOVA, ***p < 0.0001; Supplementary Table 2). While both dosages of baicalin (3.35mg/kg/d and 6.7mg/kg/d) attenuated CORT-induced immobility during the test (one-way ANOVA, low dosage *p = 0.037; high dosage ***p < 0.0001; Supplementary Table 2). Fluoxetine also had the same effects with baicalin (one-way ANOVA, ***p < 0.0001; Supplementary Table 2). d In open field test (OFT), CORT resulted the decreased exploration in central region, indicating the anxiety mood induced in depression (n = 12, one-way ANOVA, ***p = 0.0002; Supplementary Table 2). Both dosages of baicalin improved the exploration time in central region (one-way ANOVA, low dosage: **p = 0.0005; high dosage: ***p < 0.0001; Supplementary Table 2), as the same effect of fluoxetine (one-way ANOVA, ***p = 0.0002; Supplementary
Table 2). e Novelty suppressed feeding (NSF) showing the stressinduced anxiety. CORT induced the prolonged latency to feed (n = 10~12, one-way ANOVA, ***p < 0.0001; Supplementary Table 2) and decreased food consumption (one-way ANOVA, *p = 0.0109; Supplementary Table 2). Low dosage of baicalin (3.35 mg/kg/day) treatment resulted the shortening feeding latency (one-way ANOVA, **p = 0.0055; Supplementary Table 2) but did not show the significant change on food consumption (one-way ANOVA, p = 0.5106; Supplementary Table 2). While high dosage of baicalin (6.7mg/kg/d) induced the remarkable decrease of feeding latency (one-way ANOVA, ***p < 0.0001; Supplementary Table 2) together with the increased food consumption (one-way ANOVA, **p = 0.0028; Supplementary Table 2). Fluoxetine also showed the same effects with baicalin (one-way ANOVA, feeding latency ***p = 0.0012; food consumption *p = 0.0220; Supplementary Table 2). f Olfactory sensitivity test revealing the olfactory functions. Mice in all groups showed the significant different in three different dilutions (two-way AONVA, F ( 1 2 , 2 1 6 ) = 3.933, p < 0.0001; Supplementary Table 2). CORT shortened the exploration time to the odor in three dilutions (two-way ANOVA, p < 0.0001; Supplementary Table 2). While baicalin significantly reserved the decreased exploration time in different dilutions of odor, although low dosage baicalin failed at the highest dilution (two-way ANOVA, details can be seen in Supplementary Table 2). Fluoxetine also improved the olfactory functions, but also failed to reserve the olfactory sensitivity at the highest dilution (two-way ANOVA, p = 0.2128; Supplementary Table 2)
shorter exploring time staying at the central region than the vehicle mice (Fig. 6e; one-way ANOVA; p = 0.0002; Supplementary Table 2). Baicalin treatment (6.7 mg/kg/ day) significantly increased the exploration time in the central region in the CORT mice, similar to fluoxetine treatment (Fig. 6e; one-way ANOVA; low vs CORT: p = 0.0005; high vs CORT: p < 0.0001; fluoxetine vs CORT: p = 0.0002; Supplementary Table 2). In NSF, after food deprivation, the CORT mice presented the prolonged
latency to start the feeding and decreased food consumption (Fig. 6e, f; one-way ANOVA; Supplementary Table 2), indicating the impaired capability to cope with the conflict between will of feeding and fear to explore the novel environment. Baicalin treatment at both low and high dosages significantly reduced the feeding latency (Fig. 6e; one-way ANOVA; low vs CORT: p = 0.0055; high vs CORT: p < 0.0001; Supplementary Table 2). However, only high dosage of baicalin (6.7 mg/kg/day)
Mol Neurobiol
increased the food consumption (Fig. 6f; one-way ANOVA; p = 0.0028). Likewise, the CORT mice with fluoxetine treatment also showed the reduced feeding latency and the increased food consumption during the test (Fig. 6e, f; Supplementary Table 2). Those tests indicate that baicalin could alleviate the anxiety mood behaviors. High comorbidity of olfactory deficits is one of characters in depression patients [32]. Chronic CORT exposure causes olfactory dysfunctions in experimental animal model [12]. We then tested whether baicalin could restore the CORT impaired olfactory behaviors. Olfactory sensitivity test was used to assess olfactory function by explored to gradient dilution peanut oil solutions (n = 11~12 per group). The result showed a dramatic difference among different groups in each odor dilution (Fig. 6g; two-way ANOVA; F (12, 216) = 3.933, p < 0.001; Supplementary Table 2). Chronic CORT treatment decreased the olfactory sensitivity in all three dilutions (Fig. 6g; Supplementary Table 2). High dosages of baicalin treatment (6.7 mg/kg/day) significantly increased the exploration time to the peanut butter with three different dilutions (Fig. 6g; Supplementary Table 2). Low dosage of baicalin (3.35 mg/ kg/day) showed to elevate scent exploration under low and middle dilutions whose effects were similar with fluoxetine treatment (Fig. 6g; Supplementary Table 2). Therefore, baicalin could improve olfactory functions in the chronic depression mouse model.
Fig. 7 Baicalin attenuates the depressive behavior and olfactory deficit in APPL2 Tg mice. a, b Short-term CORT administration did not affect the olfactory sensitivity in WT mice but resulted in the decreased exploration time in APPL2 Tg mice during exploring period to the odorant (two-way ANOVA, WT, vehicle vs CORT: F (3, 44) = 0.7767, p = 0.5133; APPL2 Tg: vehicle vs CORT: F (3, 56) = 2.267, 10−3: p = 0.0284; Supplementary Table 2). c FST showing that treatment of baicalin significant increased the mobility of the APPL2 Tg mice in water tank without affecting the
Baicalin Improves the Depressive-Like and Olfactory Behaviors in APPL2 Tg Mice For further assessing whether baicalin could attenuate the depression-induced olfactory dysfunction via APPL2/GR pathway, we conducted the behavioral tests in APPL2 Tg mice. Firstly, by comparing the olfactory sensitivity between the vehicle and short-term CORT treatment (10 days), we found CORT administration did not affect the olfactory sensitivity in WT mice but resulted in the decreased exploration time in APPL2 Tg mice during exploring period to the odorant (Fig. 7a, b; two-way ANOVA, WT, vehicle vs CORT: F (3, 44) = 0.7767, p = 0.5133; APPL2 Tg: vehicle vs CORT: F (3, 56) = 2.267, p = 0.0906; 10–3: p = 0.0284; Supplementary Table 2). This data indicates that APPL2 overexpression results in the compromised olfactory capacity maintenance under short-term CORT-induced mild stress. Both APPL2 Tg mice and WT littermates were received short-term treatments of baicalin at high dosage (6.7 mg/kg/day). FST showed that the short-term treatment of baicalin induced a significant increase of the mobility in the APPL2 Tg mice in water tank, indicating that baicalin could reserve the APPL2 overexpression-induced depressive behaviors (Fig. 7c, two-way ANOVA, F (1, 24) = 2.15, p = 0.1556; APPL2: p = 0.0249; Supplementary Table 2). Moreover, olfactory sensitivity test showed that baicalin treatment enhanced the olfactory sensitivity of
same performance in WT mice (two-way ANOVA, F (1, 24) = 2.15, APPL2: p = 0.0249; Supplementary Table 2). d, e Olfactory sensitivity test showing that baicalin treatment enhanced the olfactory sensitivity of APPL2 Tg mice (two-way ANOVA, F (3, 56) = 1.398; 10−2: p = 0.018; 10−3: p = 0.0230; Supplementary Table 2). WT mice treated with baicalin had no change in their olfactory behaviors (two-way ANOVA; Supplementary Table 2)
Mol Neurobiol
APPL2 Tg mice but did not affect the olfactory functions in WT mice (Fig. 7d, e, two-way ANOVA, APPL2: F (3, 56) = 2.636, p = 0.0585; 10−2: p = 0.018; 10−3: p = 0.0230; WT: F (3. 40) = 1.284, p = 0.3053; Supplementary Table 2). The result suggests that baicalin could prevent APPL2 overexpression-induced olfactory impairment. Taken together, baicalin could attenuate depressive- and anxiety-like behaviors and improve olfactory functions in the chronic depression mouse model and whose antidepressant and olfactory protective effects are associated with modulating APPL2/GR signaling pathway.
Discussion In present study, we identified the roles of APPL2 in affecting olfactory functions and found that APPL2 could regulate the olfactory sensitivity. APPL2 Tg mice exhibited the hyperactivity of GR and the impaired neurogenesis at olfactory system (Fig. 1). Those discoveries open a new glimpse on the roles of APPL2 as a potential therapeutic target for depression and olfactory deficits. Therefore, we further explored whether the antidepressant and neurogenic promoting mechanisms of baicalin are related to modulate APPL2/GR signaling pathway and investigated its effects on olfactory functions. Baicalin inhibited APPL2/GR signaling pathway and improved the neurogenesis in APPL2 Tg mice and chronic CORT-induced depression mouse model (Fig. 2). The in vitro study showed that baicalin reserved the DEX induced nuclear translocation. Behavioral tests revealed that baicalin could attenuate depressive- and anxiety-like behaviors and improve olfactory functions in the chronic depression mouse model. To our knowledge, this is the first evidence that APPL2 could modulate olfactory functions. By targeting APPL2/GR signaling pathway, baicalin promotes neurogenesis in hippocampus, OB, and SVZ; subsequently improves the olfactory functions; and attenuates the depressive- and anxiety-like behaviors in the chronic depression mouse model. Besides depressive mood and anxiety emotion, depressive patients usually display various accompanied symptoms such as sleep disorders and the change of eating habit (anorexia or bulimia) [33, 34]. Olfactory deficit is also associated with development of depression [35]. However, the molecular regulations of depressive olfactory deficit remain unclear. It is well known that HPA axis plays crucial roles in stress response [11]. Chronic stress leads hyperactivity of HPA axis that activates GR in brain, which causes the behavioral dysfunctions. CORT induces depressive behavior accompanied with olfactory deficits and neurogenic impairment in SVZ and OB in experimental depressive model [12]. GR activation contributes to neurogenic impairment in chronic stress responses [14]. As a metabolic modulating factor, APPL2 negatively regulates adult hippocampal neurogenesis and induce the inadaptability to stress via GRdependent manner [21]. In the present study, we confirmed that APPL2 could be a new therapeutic target for antidepressant
therapy and improving olfactory functions. APPL2 Tg mice had shorter exploration time to odorant than WT mice in olfactory sensitivity test. Moreover, higher levels of GR phosphorylation and SGK1 expression in OB and had higher rates of DCX+/BrdU+ cells in SVZ and OB were shown in APPL2 Tg mice than that of WT mice. Those results suggest that APPL2 overexpression could impair adult neurogenesis in hippocampus and olfactory system, impacting on olfactory dysfunctions. Baicalin is an active compound isolated from an antidepressive medicinal plant Scutellaria baicalensis. Baicalin could promote neuronal differentiation of NSPCs [25] and attenuate the depressive-like behaviors in chronic mild stress animal model [26]. Baicalin could downregulate the expression of SGK1, a GR downstream signaling, in the hippocampus and improve corticosterone-induced depressive-like behaviors [27]. Herein, we found that baicalin treatment significantly downregulated the expression of APPL2 protein at the OB and hippocampus in the CORT-induced depressive WT mice (Fig. 2). Meanwhile, baicalin treatment significantly inhibited GR phosphorylation and SGK1 expression at the OB and hippocampus of the CORT mice. These results indicate that baicalin could inhibit APPL2/GR signaling pathway. Given that APPL2/GR signaling could impair adult hippocampal neurogenesis and affect olfactory dysfunction, we logically explored whether baicalin could promote neurogenesis in hippocampus, OB, and SVZ. By using different neurogenic biomarkers, we found that baicalin treatment increased neurogenesis in hippocampus, OB, and SVZ in the APPL2 Tg mice and chronic CORT-induced depression mouse model (Figs. 3, 4, and 5). The effects of baicalin on improving hippocampal neurogenesis are consistent with the previous report [36]. Notably, we found that baicalin promoted olfactory neurogenesis in chronic CORT-induced depression mouse model (Figs. 3, 4, and 5). Behavioral tests further confirmed our hypothesis that baicalin treatment could attenuate the CORT-induced depressive-/anxiety-like behaviors and the depression associated olfactory deficits (Fig. 6). Those results propose the potentials of baicalin to be functioned as an antidepressant agent for releasing the depressive-/anxiety-like behaviors as well as improving olfactory symptoms in chronic depressive disorders. The behavioral studies on APPL2 Tg mice further support that baicalin could improve the olfactory sensitivity even with genetic increased APPL2 protein in the mice (Fig. 7). Therefore, APPL2/GR signaling could be an important therapeutic target for improving olfactory sensitivity in chronic depressive disorders, and the inhibition of APPL2/GR signaling might contribute to the effects of baicalin on promoting olfactory functions and releasing the depressive-/anxiety-like behaviors. Olfactory deficits could exacerbate depressive symptoms and increase psychiatric stress to form vicious mental circle in patients with major depression disorder (MDD). For rodents, loss of olfactory sensitivity and discrimination predicts the weaken defensive capability to predators, resulting in the
Mol Neurobiol
exacerbated depressive symptoms. Herein, we selected the olfactory sensitivity as the standard for evaluating olfactory functions. It is important to further explore whether baicalin could improve the other olfactory behaviors like olfactory memory. Whether APPL2 could affect olfactory memory is also worthwhile for further study. In conclusion, APPL2/GR signaling pathway could induce olfactory dysfunctions through impairing neurogenesis in OB. By targeting APPL2/GR signaling pathway and improving neurogenesis in hippocampus, OB, and SVZ, baicalin could be an antidepressant agent for improving depressive and anxiety behaviors and attenuating olfactory deficits in chronic depressive disorders. Acknowledgements We truly appreciate the technology supports of Dr. Guo Jing and Mr. Jacky HUNG in Core facility HKU, the technology on dSTORM and Imaris software processing. Funding This work was supported by RGC GRF grants from Hong Kong Research Grants Council (GRF No. 777313M, 776512M, J.S.), RGC AoE/P-705/16, and Contract Research Grant with Beijing Tong Ren Tang Chinese Medicine Co., LTd.
10.
11.
12.
13. 14.
15.
16.
17.
Compliance with Ethical Standards Conflicts of Interest The authors declare that there are no conflicts of interest in this study.
References Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE (2005) Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 62(6):593–602 2. Croy I, Negoias S, Symmank A, Schellong J, Joraschky P, Hummel T (2013) Reduced olfactory bulb volume in adults with a history of childhood maltreatment. Chem Senses 38(8):679–684 3. Clepce M, Gossler A, Reich K, Kornhuber J, Thuerauf N (2010) The relation between depression, anhedonia and olfactory hedonic estimates—a pilot study in major depression. Neurosci Lett 471(3): 139–143 4. Gusmao ID, Monteiro BM, Cornelio GO, Fonseca CS, Moraes MF, Pereira GS (2012) Odor-enriched environment rescues long-term social memory, but does not improve olfaction in social isolated adult mice. Behav Brain Res 228(2):440–446 5. Fenton EY, Fournier NM, Lussier AL, Romay-Tallon R, Caruncho HJ, Kalynchuk LE (2015) Imipramine protects against the deleterious effects of chronic corticosterone on depression-like behavior, hippocampal reelin expression, and neuronal maturation. Prog Neuro-Psychopharmacol Biol Psychiatry 60:52–59 6. Kohman RA, Rhodes JS (2017) The contribution of adult hippocampal neurogenesis to the progression of psychiatric disorders. Mod Trends Pharmacopsychiatry 31:124–151 7. Eisch AJ, Petrik D (2012) Depression and hippocampal neurogenesis: a road to remission? Science 338(6103):72–75 8. Whitman MC, Greer CA (2009) Adult neurogenesis and the olfactory system. Prog Neurobiol 89(2):162–175 9. Schloesser RJ, Orvoen S, Jimenez DV, Hardy NF, Maynard KR, Sukumar M, Manji HK, Gardier AM et al (2015) Antidepressant-
18.
19.
1.
20.
21
22
23
24
25
26
like effects of electroconvulsive seizures require adult neurogenesis in a neuroendocrine model of depression. Brain Stimul 8(5):862–867 Santarelli L et al (2003) Requirement of hippocampal neurogenesis for the behavioral effects of antidepressants. Science 301(5634): 805–809 Gregus A, Wintink AJ, Davis AC, Kalynchuk LE (2005) Effect of repeated corticosterone injections and restraint stress on anxiety and depression-like behavior in male rats. Behav Brain Res 156(1):105– 114 Siopi E et al (2016) Anxiety- and depression-like states lead to pronounced olfactory deficits and impaired adult neurogenesis in mice. J Neurosci 36(2):518–531 Altar CA (1999) Neurotrophins and depression. Trends Pharmacol Sci 20(2):59–61 Egeland M, Zunszain PA, Pariante CM (2015) Molecular mechanisms in the regulation of adult neurogenesis during stress. Nat Rev Neurosci 16(4):189–200 Anacker C et al (2013) Role for the kinase SGK1 in stress, depression, and glucocorticoid effects on hippocampal neurogenesis. Proc Natl Acad Sci U S A 110(21):8708–8713 Gao C, Wang Q, Chung SK, Shen J (2017) Crosstalk of metabolic factors and neurogenic signaling in adult neurogenesis: Implication of metabolic regulation for mental and neurological diseases. Neurochem Int 106:24–36 Moon HS, Dincer F, Mantzoros CS (2013) Amylin-induced downregulation of hippocampal neurogenesis is attenuated by leptin in a STAT3/AMPK/ERK-dependent manner in mice. Diabetologia 56(3):627–634 Carlson SW, Madathil SK, Sama DM, Gao X, Chen J, Saatman KE (2014) Conditional overexpression of insulin-like growth factor-1 enhances hippocampal neurogenesis and restores immature neuron dendritic processes after traumatic brain injury. J Neuropathol Exp Neurol 73(8):734–746 Zhang D, Wang X, Lu XY (2016) Adiponectin exerts neurotrophic effects on dendritic arborization, spinogenesis, and neurogenesis of the dentate gyrus of male mice. Endocrinology 157(7):2853–2869 Cheng KK et al (2014) The adaptor protein APPL2 inhibits insulinstimulated glucose uptake by interacting with TBC1D1 in skeletal muscle. Diabetes 63(11):3748–3758 Gao C, Chen X, Xu A, Cheng K, Shen J (2017) Adaptor protein APPL2 affects adult antidepressant behaviors and hippocampal neurogenesis via regulating the sensitivity of glucocorticoid receptor. Mol Neurobiol. https://doi.org/10.1007/s12035017-0785-y Li BQ, Fu T, Gong WH, Dunlop N, Kung HF, Yan Y, Kang J, Wang JM (2000) The flavonoid baicalin exhibits anti-inflammatory activity by binding to chemokines. Immunopharmacology 49(3):295– 306 Hwang JM, Wang CJ, Chou FP, Tseng TH, Hsieh YS, Hsu JD, Chu CY (2005) Protective effect of baicalin on tert-butyl hydroperoxideinduced rat hepatotoxicity. Arch Toxicol 79(2):102–109 Jung SH, Kang KD, Ji D, Fawcett RJ, Safa R, Kamalden TA, Osborne NN (2008) The flavonoid baicalin counteracts ischemic and oxidative insults to retinal cells and lipid peroxidation to brain membranes. Neurochem Int 53(6–8):325–337 Li Y, Zhuang P, Shen B, Zhang Y, Shen J (2012) Baicalin promotes neuronal differentiation of neural stem/progenitor cells through modulating p-stat3 and bHLH family protein expression. Brain Res 1429:36–42 Li YC, Shen JD, Li J, Wang R, Jiao S, Yi LT (2013) Chronic treatment with baicalin prevents the chronic mild stressinduced depressive-like behavior: involving the inhibition of cyclooxygenase-2 in rat brain. Prog Neuro-Psychopharmacol Biol Psychiatry 40:138–143
Mol Neurobiol 27
28 29
30 31
Li YC et al (2015) Baicalin decreases SGK1 expression in the hippocampus and reverses depressive-like behaviors induced by corticosterone. Neuroscience 311:130–137 Alvarez-Buylla A, Garcia-Verdugo JM (2002) Neurogenesis in adult subventricular zone. J Neurosci 22(3):629–634 Brummelte S, Galea LA (2010) Chronic high corticosterone reduces neurogenesis in the dentate gyrus of adult male and female rats. Neuroscience 168(3):680–690 Lledo PM, Valley M (2016) Adult olfactory bulb neurogenesis. Cold Spring Harb Perspect Biol 8(8):a018945 Sequerra EB (2014) Subventricular zone progenitors in time and space: generating neuronal diversity. Front Cell Neurosci 8:434
32
33 34
35 36
Yuan TF, Slotnick BM (2014) Roles of olfactory system dysfunction in depression. Prog Neuro-Psychopharmacol Biol Psychiatry 54: 26–30 Tsuno N, Besset A, Ritchie K (2005) Sleep and depression. J Clin Psychiatry 66(10):1254–1269 Lauer CJ, Krieg JC, Riemann D, Zulley J, Berger M (1990) A polysomnographic study in young psychiatric inpatients: major depression, anorexia nervosa, bulimia nervosa. J Affect Disord 18(4): 235–245 Li Q et al (2015) Reduced amount of olfactory receptor neurons in the rat model of depression. Neurosci Lett 603:48–54 Zhang K et al (2016) Baicalin promotes hippocampal neurogenesis via SGK1- and FKBP5-mediated glucocorticoid receptor phosphorylation in a neuroendocrine mouse model of anxiety/depression. Sci Rep 6:30951
