Harmine produces antidepressant-like effects via restoration of astrocytic functions

Accepted Manuscript Harmine produces antidepressant-like effects via restoration of astrocytic functions

Fengguo Liu, Jingjing Wu, Yu Gong, Peng Wang, Lei Zhu, LiJuan Tong, Xiangfan Chen, Yong Ling, Chao Huang PII: DOI: Reference:

S0278-5846(17)30239-7 doi: 10.1016/j.pnpbp.2017.06.012 PNP 9134

To appear in:

Progress in Neuropsychopharmacology & Biological Psychiatry

Received date: Revised date: Accepted date:

26 March 2017 14 June 2017 14 June 2017

Please cite this article as: Fengguo Liu, Jingjing Wu, Yu Gong, Peng Wang, Lei Zhu, LiJuan Tong, Xiangfan Chen, Yong Ling, Chao Huang , Harmine produces antidepressantlike effects via restoration of astrocytic functions. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Pnp(2017), doi: 10.1016/j.pnpbp.2017.06.012

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Harmine produces antidepressant-like effects via restoration of astrocytic functions Fengguo Liua, Jingjing Wub, Yu Gongc,d, Peng Wangc,d, Lei Zhue, Li-Juan

a

PT

Tongc,d, Xiangfan Chenc,d, Yong Lingc,d, Chao Huangc,d,*

Department of Neurology, Danyang People’s Hospital, #2 Xinmin Western Road,

RI

Danyang 212300, Jiangsu, China d

SC

Department of Cardiology, Suzhou Kowloon Hospital of Shanghai Jiaotong

University School of Medicine, #118 Wansheng Street, Suzhou 215021, Jiangsu,

b

NU

China

Department of Pharmacology, School of Pharmacy, Nantong University, #19 Qixiu

c

MA

Road, Nantong 226001, Jiangsu, China

Key Laboratory of Inflammation and Molecular Drug Target of Jiangsu Province,

#19 Qixiu Road, Nantong 226001, Jiangsu, China e

ED

Department of Pharmacy, First People’s Hospital of Yancheng, Yulong Western Road,

*

EP T

Yancheng 224006, Jiangsu, China

Corresponding author at: Department of Pharmacology, School of Pharmacy,

Nantong University, #19 Qixiu Road, Nantong, Jiangsu, China 226001, Tel:

AC C

(+86)-513-85061297, Fax: (+86)-513-85061297; Email: [email protected]

Abbreviations: ALS, amyotrophic lateral sclerosis; ANOVA, analysis of variance; BDNF, brain-derived neurotrophic factor; CUS, chronic unpredictable stress; DCX, doublecortin; DMSO,

Dimethyl Sulphoxide; EAATs,

excitatory amino-acid

transporters; FITC, fluorescein isothiocyanate; FST, forced swimming test; GAPDH, glyceraldehydes-3-phosphate dehydrogenase; GFAP, glial fibrillary acidic protein; GLAST, glutamate/aspartate transporter; GLT-1, glutamate transporter-1; GS, glutamine synthetase; L-AAA, L-Alpha-Aminoadipic Acid; TST, tail suspension test

ACCEPTED MANUSCRIPT Abstract Depression is a world-wide disease with no effective therapeutic methods. Increasing evidence indicates that astrocytic pathology contributes to the formation of depression. In this study, we investigated the effects of harmine, a natural β-carboline alkaloid and potent hallucinogen, known to modulate astrocytic glutamate transporters, on chronic

PT

unpredictable stress (CUS)- induced depressive-like behaviors and astrocytic dysfunctions. Results showed that harmine treatment (10, 20 mg/kg) protected the

RI

mice against the CUS- induced increases in the immobile time in the tail suspension test (TST) and forced swimming test (FST), and also reversed the reduction in sucrose

SC

intake in the sucrose preference experiment. Harmine treatment (20 mg/kg) prevented the reductions in brain-derived neurotrophic factor (BDNF) protein levels and

NU

hippocampal neurogenesis induced by CUS. In addition, harmine treatment (20 mg/kg) increased the protein expression levels of glutamate transporter 1 (GLT-1) and

MA

prevented the CUS- induced decreases in glial fibrillary acidic protein (GFAP) protein expressions in the prefrontal cortex and hippocampus, suggesting that restoration of

ED

astrocytic functions may be a potential mechanism underlying the antidepressant-like effects of harmine. This opinion was proved by the results that administration of mice

EP T

with L-Alpha-Aminoadipic Acid (L-AAA), a gliotoxin specific for astrocytes, attenuated the antidepressant-like effects of harmine, and prevented the improvement effects of harmine on BDNF protein levels and hippocampal neurogenesis. These

AC C

results provide further evidence to confirm that astrocytic dysfunction contributes critically to the development of depression and that harmine exerts antidepressant-like effects likely through restoration of astrocytic functions.

Key words: harmine; depression; astrocyte; glutamate transporter 1; chronic unpredictable stress

ACCEPTED MANUSCRIPT 1. Introduction

Major depression is a common disease affecting numerous persons in the world-wide. At present, nearly all of the clinical antidepressants are developed out of the monoaminergic deficit hypothesis of depression, and these agents have been used

PT

as classical antidepressants for a very long time (Lanni et al., 2009; Kern et al., 2012). However, more and more studies and clinical reports show that these antidepressants

RI

exhibit numerous limitations (Fava, 2010; Fabbri et al., 2013; Sanchez et al., 2015). For example, the first application of traditional antidepressants is effective only in

SC

about one-third of patients, and approximately two-third of patients fail to achieve clinical improvements after trying several times (Schwartz et al., 2016). Thus, it is

NU

necessary to develop novel mechanism-based antidepressants in order to improve the present status of drug therapy of depression.

MA

Increasing evidence indicates that astrocytic dysfunction is actively involved in the pathogenesis of depression. Decreased numbers of hippocampal astrocytes have

ED

been observed in rodents treated with chronic stresses or maternal deprivation (Ye et al., 2011; Leventopoulos et al., 2007). Post- mortem studies of tissues from depressed

EP T

patients describe reduced numbers of astrocytes in the brain (Oh et al., 2012; Ménard et al., 2016), and these reductions are supported by the finding that the glial fibrillary acidic protein (GFAP)-expressing cells are reduced by chronic stresses in both

AC C

prefrontal cortex and hippocampus (Banasr and Duman, 2008; Ardalan et al., 2017). In other studies, researchers show that expressions of glial-specific excitatory amino-acid transporters (EAATs: EAAT1 /glutamate/aspartate transporter (GLAST), EAAT2 /glutamate transporter-1 (GLT-1)) and glutamine synthetase (GS) can be altered in brain tissues from depressed patients (Choudary et al., 2005; Bernard et al., 2011). Functionally, the astrocytic abnormality has been shown to induce a decrease in glutamate uptake and cycling, an accumulation of glutamates, as well as an impairment in brain-derived neurotrophic factor (BDNF) signals and hippocampal neurogenesis (Lutgen et al., 2016; Martin et al., 2012). Thus, inhibition of astrocytic dsyfunction may be a potential strategy for depression therapy.

ACCEPTED MANUSCRIPT Harmine is a

hallucinogenic alkaloid

found

in the seed of Peganum

harmala and Banisteriopsis caapi, both of which are traditionally used for ritual and medicinal preparations in the Middle East, Central Asia, and South America (Sourkes, 1999). In past years, a wide range of pharmacological effects of harmine, such as antioxidation (Moura et al., 2007; Kim et al., 2001), antigenotoxicity (Moura et al.,

PT

2007), and anti-diabeties (Waki et al., 2007), have been revealed. Preclinical findings show that harmine has potential antidepressant- like activities in acute and chronic

RI

depression models (Fortunato et al., 2009; Fortunato et al., 2010; Aricioglu and Altunbas, 2003; Farzin and Mansouri, 2006). More strong correlations between

SC

harmine and depression have been evidenced by recent studies in depressed patients, in which a single dose of ayahuasca, a harmine-containing hallucinogenic plant, has

NU

been reported to produce rapid and sustained antidepressant- like activities (Osório Fde et al., 2015; Sanches et al., 2016). Mechanistically, the antidepressant- like effects

MA

of harmine may be mediated by restoration of BDNF signals (Fortunato et al., 2009; Fortunato et al., 2010). An inverse-agonistic mechanism located in the benzodiazepine

ED

receptors may also mediate the antidepressant- like effects of harmine, as flumazenil, an inhibitor of the benzodiazepine receptor, has been shown to abrogate the

EP T

antidepressant-like effects of harmine in the forced swimming test (FST) (Aricioglu and Altunbas, 2003). Recently, several different studies show that harmine increases GLT-1 gene and protein expression as well as glutamate uptake activity in animal

AC C

models of amyotrophic lateral sclerosis (ALS) (Li et al., 2011) and cerebral ischemia (Sun et al., 2014), suggesting that harmine may exert neuroprotective effects via enhancement of GLT-1 functions. Within these contexts, we hypothesized that harmine exerts antidepressant-like effects likely through restoration of astrocytic functions. As anticipated, we showed that harmine not only prevented mice depressive- like behaviors, but also reversed the reductions in BDNF protein levels and hippocampal neurogenesis induced by CUS. Harmine also prevented CUS-induced decreases in GFAP protein expressions and up-regulated GLT-1 protein expressions in mice hippocampus and prefrontal cortex. Specific inhibition of astrocytic functions abrogated the protective effects of harmine

ACCEPTED MANUSCRIPT in depressed mice. Since astrocytes contribute critically to the integrity of neuronal functions (Ben Haim and Rowitch, 2017; Yang et al., 2015), our studies indicate that the astrocyte may be a potential target for the antidepressant- like effects of harmine.

PT

2. Materials and methods

RI

2.1.Animals

SC

8−10 weeks old male C57BL/6J mice were housed five per cage under standard conditions (12-h light/dark cycle; lights on from 07:00 to 19:00; 23 ± 1°C ambient

NU

temperature; 55 ± 10% relative humidity) for 1 week with free access to food and water. Each experimental group consisted of 10 mice. Behavioral experiments were

MA

carried out during the light phase. Animal experiments were conducted in accordance with internationally accepted guidelines for the use of animals in toxicology as adopted by the Society of Toxicology in 1999) and approved by the University

EP T

2.2.Materials

ED

Animal Ethics Committee of Nantong University (Permit Number: 2110836).

AC C

Harmine and fluoxetine were purchased from MedChem Express (Princeton, NJ, USA). L-Alpha-Aminoadipic Acid (L-AAA) was the product of Sigma (Saint Louis, MO, USA). The doses of fluoxetine, harmine and L-AAA were chosen as previous studies (Zhu et al., 2010; Jiang et al., 2015). Harmine and fluoxetine were prepared in 10% dimethyl sulphoxide (DMSO) and were administered intraperitoneally (i.p.) in a volume of 10 mL/kg. L-AAA was dissolved in a phosphate buffer and was administered intracerebroventricularly (i.c.v.). Antibodies against doublecortin (DCX), GFAP, GS, GLAST, and glyceraldehydes-3-phosphate dehydrogenase (GAPDH) were purchased from Cell Signaling Technology (Beverly, MA, USA). Antibody against GLT-1 was the product of Abcam (Cambridge, MA, USA). Hoechst 33258

ACCEPTED MANUSCRIPT was purchased from Santa Cruz Biotechnology (Santa Cruz, California, USA).

2.3.Tail suspension test (TST) and FST

The TST and FST were performed according to the method s of Steru (Steru et al.,

PT

1985) and Porsolt (Porsolt et al., 1977), respectively. In the FST, the experimental mice were individually placed in a clear glass cylinder (height 25 cm, diameter 10 cm)

RI

filled to 10 cm with water at 25 ± 1°C for 6 min. In the TST, the experimental mice were suspended 50 cm above the floor for 6 min by adhesive tape placed

SC

approximately 1 cm from the tip of the tail 2 h after the last injection. The immobile time was recorded during the last 4 min by an investigator blind to the study. For FST,

NU

the immobile time was defined as the time spent by the mouse floating in the water without struggling and making only those movements necessary to keep its head

MA

above the water. For TST, the mice were considered immobile only when they hung passively and were completely motionless, and any mice that try to climb their tails

ED

were removed from statistical analysis. The concrete schematic diagram for the experimental timeline in the TST and FST in CUS models was presented in Figure

EP T

1A.

AC C

2.4.Chronic unpredictable stress (CUS)

The CUS, consisting of daily exposure to two of the following stressors in a random order over a 5-week period: cage shaking (1 h), lights on during the entire night (12 h), placement in cold room (4°C, 1 h), mild restraint in small cages (2 h), 45° cage tilt (14 h), lights-off during the daylight phase (3 h), wet cage (14 h), flashing light (6 h), noise in the room (3 h), and water deprivation during the dark period (12 h), was used to induce depressive-like behaviors in C57BL/6J mice.

2.5.Sucrose preference experiment

ACCEPTED MANUSCRIPT The sucrose preference experiment was performed at day 36. The experimental mice were given the choice to drink from two bottles in individual cages, one with 1% sucrose solution and the other with water (Yang et al., 2017). All were acclimatized for 2 days to two-bottle choice conditions, and the position of two bottles was changed every 6 h to prevent possible effects of side preference in drinking behavior.

PT

The experimental mice were then deprived of food and water for 24 h, and on day 39, the mice were exposed to pre-weighed bottles for 1 h with their position interchanged.

RI

Sucrose preference was calculated as a percentage of the consumed sucrose solution relative to the total amount of liquid intake. The schematic diagram for sucrose

SC

preference experiment was presented in Figure 1A.

NU

2.6.Chronic intracerebroventricular infusions of L-AAA

MA

In this experiment, L-AAA was used to block astrocytic functions. Briefly, C57BL/6J mice were anaesthetized with pentobarbital sodium and placed in a

ED

stereotaxic frame (Jiang et al., 2015a, Quintessential Stereotaxic Injector, STELING Corporation, DALE, IL, USA). The cannulas were implanted into the left lateral brain

EP T

ventricle (-0.2 mm anterior and 1.0 mm lateral relative to bregma and 2.3 mm be low the surface of the skull, Kleinridders et al., 2009) and connected to an osmotic minipump (Alzet model 2002 for chronic injections and Alzet model 1003D for acute

AC C

injections, Alza Corporation, Cupertino, CA, USA) according to manufacturer ’s protocols. Minipumps were filled with 1 μg/μL L-AAA or 1 μg/μL vehicle in artificial cerebrospinal fluid (ACSF) and were implanted subcutaneously in the interscapular region. The speed of chronic (10 days) injections was 1 μL/h. The total dosage of L-AAA in this experiment was within the range used in the literature (Takada and Hattori, 1986; Khurgel et al., 1996).

2.7. BDNF analysis

Immediately after the behavioral experiments mice were killed and the

ACCEPTED MANUSCRIPT hippocampal and prefrontal cortical tissues were dissected and stored at -70°C. The BDNF protein levels were measured by anti-BDNF sandwich-ELISA according to manufacturer’s protocols (Chemicon, USA). Briefly, brain tissues were homogenized in phosphate buffers containing phenylmethylsulfonyl fluoride (1 mM) and EGTA (1 mM). The 96-well plates were pre-coated for 24 h with the samples diluted (1:2) in

PT

sample diluent and standard curve ranged from 10 to 500 pg/mL of BDNF. The plates were then washed three times by sample diluent, and a monoclonal anti-BNDF

RI

antibody (1:500) prepared in sample diluents was added into each well and incubated for 3 h at room temperature. After washing, a peroxidase conjugated anti-rabbit

SC

secondary antibody (1:1000) was added into each well and incubated at room temperature for 1 h, and then the streptavidin-enzyme, substrate, and stop solution

NU

were added into the 96-well plates. The amounts of BDNF protein were determined by absorbance in 450 nm, and measured by Lowry’s method using bovine serum

MA

albumin as a standard, as previously described by Frey et al. (Frey et al., 2006). The standard curve demonstrates a direct correlation between Optical Density and BDNF

ED

protein concentrations.

EP T

2.8.Western blot

This experiment was conducted according to previous studies with some

AC C

modifications (Yao et al., 2016, Huang et al., 2017). Briefly, brain tissues were dissected and homogenized in lyses buffer containing 50 mM Tris-HCl (pH 7.4), 1 mM EDTA, 100 mM NaCl, 20 mM NaF, 3 mM Na3 VO4 , 1 mM PMSF, 1% NP-40 and protease inhibitor cocktail on ice for 30 min. The lysates were centrifuged at 12,000×g for 15 min, and the supernatants were harvested. After being separated in 10% SDS/PAGE, 30−50 μg of denatured protein were transferred to nitrocellulose membranes (Bio-Rad, Hercules, CA, USA). The milk-blocked nitrocellulose membranes were incubated overnight at 4°C with 1:500–1:1000 primary antibodies against DCX, GFAP, GS, GLAST, GLT-1 or GAPDH, and then further incubated for 2 h at room temperature with IRDye 680- labeled secondary antibodies (1:5000).

ACCEPTED MANUSCRIPT Finally, immunoblots were visualized by scanning using the Odyssey CLx Western blot detection system. The band density was quantified by Image J software.

2.9.Immunofluorescence

PT

Animals were anesthetized with sodium pentobarbital (50 mg/kg) and perfused transcardially with 37°C saline and 4% paraformaldehyde in 0.01 M phosphate buffer.

RI

The brains were removed, frozen, and sectioned at 25 μm, and consecutive sections were collected in 24-well plates containing PBS. The sections were first

SC

permeabilized with 0.3% Triton X-100 for 30 min, then incubated with 3% bovine serum albumin in PBS for another 30 min at room temperature, and followed b y

NU

incubation in diluted goat anti-DCX (1:100) primary antibody in PBS containing 0.3% Triton X-100 and 1% BSA overnight at 4°C. After that, the primary antibody was

MA

removed by washing the sections three times in PBS. The sections were further incubated in fluorescein isothiocyanate (FITC)- labeled horse anti- rabbit IgG (1:50)

ED

for 2 h at room temperature. The cellular nuclei were marked with Hoechst 33258 (10 min). After being washed with PBS, sections were mounted on slides and

EP T

coverslipped and finally examined under a confocal fluorescence microscope (FV500; Olympus, Tokyo, Japan).

Statistical analysis

AC C

2.10.

All analyses were performed using SPSS 13.0 software (SPSS Inc., USA), and data are presented as mean ± SEM. Differences between mean values were evaluated using two-way analysis of variance (ANOVA), and the Bonferroni’s post hoc test was used to assess isolated comparisons. p < 0.05 was considered statistically significant.

3. Results

3.1. Chronic harmine treatment reverses the CUS-induced depressive-like behaviors

ACCEPTED MANUSCRIPT in C57BL/6J mice

To characterize the antidepressant-like effects of harmine, we employed a widely- used animal model of depression (Jiang et al., 2015b), CUS, in this study. TST and FST have been applied to detect the antidepressant- like activities of potential

PT

antidepressants (Steru et al., 1985; Porsolt et al., 1977). In the FST, two-way ANOVA revealed significant effects for stress [F(1, 72) = 39.70, p < 0.001] and drug treatment

RI

[F(3, 72) = 7.85, p < 0.001], but not for stress × drug treatment interaction [F(3, 72) = 0.01, p = 0.44] (Fig. 1B). Post hoc analysis showed that CUS markedly increased the

SC

immobile time of mice in the FST (Fig. 1B, n = 10, p < 0.05 vs. vehicle alone-treated group), and this increase was reversed by 10 days of harmine treatment (10, 20 mg/kg,

NU

Fig. 1B, n = 10, p < 0.05 or p < 0.01 vs. vehicle + CUS group), similar to fluoxetine treatment (Fig. 1B, n = 10, p < 0.05 vs. vehicle + CUS group). Chronic harmine

MA

treatment also reduced the immobile duration of naive mice in the FST (Fig. 1B, n = 10, p < 0.05 vs. vehicle alone-treated group).

ED

In the TST, two-way ANOVA revealed significant effects for stress [F(1, 72) = 35.47, p < 0.001] and drug treatment [F(3, 72) = 7.55, p < 0.001], but not for stress ×

EP T

drug treatment interaction [F(3, 72) = 0.15, p = 0.93] (Fig. 1C). Post hoc analysis showed that CUS robustly increased the immobile time of mice in the TST (Fig. 1C, n = 10, p < 0.01 vs. vehicle alone-treated group), and this increase was reversed by 10

AC C

days of harmine treatment (10, 20 mg/kg) (Fig. 1C, n = 10, p < 0.05 vs. vehicle + CUS group). Harmine treatment also reduced the immobile duration of naive mice in the TST (Fig. 1C, n = 10, p < 0.05 vs. vehicle alone-treated group). We then evaluated the antidepressant- like effects of harmine using the sucrose preference experiment. Two-way ANOVA revealed significant effects for stress [F(1, 72) = 44.64, p < 0.001], drug treatment [F(3, 72) = 4.32, p < 0.01], but not for stress × drug treatment interaction [F(3, 72) = 2.73, p = 0.05] (Fig. 1D). Post hoc analysis showed that CUS induced a significant decrease in sucrose consumption, compared with control group (Fig. 1D, n = 10, p < 0.01 vs. vehicle alone-treated group). While harmine produced no significant effects in naive mice, similar with fluoxetine

ACCEPTED MANUSCRIPT treatment, 10 days of harmine (10, 20 mg/kg) treatment caused a significant increase in sucrose intake in CUS-treated mice (Fig. 1D, n = 10, p < 0.05 vs. vehicle + CUS group). Taken together, these results suggest that harmine is able to reverse the CUS-induced depressive-like behaviors in mice.

PT

3.2. Harmine attenuates the CUS-induced reductions in BDNF protein levels and

RI

hippocampal neurogenesis

Adult normal mice were received daily injections of harmine (10, 20 mg/kg) for

SC

10 days, and the protein levels of BDNF in mice hippocampus and prefrontal cortex were detected. In the hippocampus, two-way ANOVA revealed significant effects for

NU

stress [F(1, 56) = 20.64, p < 0.001] and drug treatment [F(3, 56) = 4.83, p < 0.01], but not for stress × drug treatment interaction [F(3, 56) = 0.04, p = 0.99] (Fig. 2A). Post

MA

hoc analysis showed that the protein level of hippocampal BDNF was down-regulated by CUS (Fig. 2A, n = 8, p < 0.05 vs. vehicle alone-treated group), and this

ED

downregulation was reversed by harmine treatment at the doses of 10 and 20 mg/kg (Fig. 2A, n = 5, p < 0.05 vs. vehicle + CUS group). In accordance with previous

EP T

reports, chronic harmine treatment also increased the protein level of hippocampal BDNF in naive mice (Fig. 2A, n = 8, p < 0.05 vs. vehicle alone-treated group). In the prefrontal cortex, two-way ANOVA revealed significant effects for stress

AC C

[F(1, 56) = 18.45, p < 0.001] and drug treatment [F(3, 56) = 3.59, p < 0.05], but not for stress × drug treatment interaction [F(3, 56) = 0.25, p = 0.86] (Fig. 2B). Post hoc analysis showed that CUS reduced the protein level of BDNF in the prefrontal cortex (Fig. 2B, n = 8, p < 0.05 vs. vehicle alone-treated group), and this reduction was attenuated by harmine treatment (10, 20 mg/kg, Fig. 2B, n = 5, p < 0.05 vs. vehicle + CUS group). Chronic harmine treatment also increased the prefrontal cortical BDNF protein level in naive mice (Fig. 2B, n = 8, p < 0.05 vs. vehicle alone-treated group). Since hippocampal neurogenesis is closely implicated in the pathogenesis of depression and can be modulated by BDNF (Kim and Leem, 2016; Sakharkar et al., 2016), we explored whether harmine affects hippocampal neurogenesis. For DCX

ACCEPTED MANUSCRIPT positive cells, two-way ANOVA revealed significant effects for stress [F(1, 24) = 12.70, p < 0.01] and drug treatment [F(2, 24) = 7.72, p < 0.01], but not for stress × treatment interaction [F(2, 24) = 0.37, p = 0.70] (Fig. 2C, D). For DCX protein expression levels, two-way ANOVA revealed significant effects for stress [F(1, 24) = 39.06, p < 0.001] and drug treatment [F(2, 24) = 10.93, p < 0.001], but not for stress ×

PT

treatment interaction [F(2, 24) = 0.50, p = 0.61] (Fig. 2E, F). Post hoc analysis showed that CUS induced a significant decrease in the number of DCX positive cells

RI

in the hippocampus (Fig. 2D, n = 5, p < 0.01 vs. vehicle alone-treated group), and accordingly the expression level of DCX protein was also reduced (Fig. 2F, n = 5, p <

SC

0.05 vs. control). Chronic treatment of mice with harmine (20 mg/kg) substantially reversed the CUS- induced decreases in the number of DCX positive cells (Fig. 2C, D,

NU

p < 0.05 vs. vehicle + CUS group) as well as in the expression level of DCX protein

MA

in the hippocampus (Fig. 2E, F, p < 0.05 vs. vehicle + CUS group).

3.3. Effects of CUS and harmine on expressions of brain GFAP, GLT-1, GS and

ED

GLAST

EP T

To determine whether astrocyte-associated markers are altered in this experiment, we investigated the effects of CUS and/or harmine on expressions of GFAP, GLAST, GLT-1, and GS protein in mice hippocampus and prefrontal cortex. For hippocampal

AC C

GFAP, two-way ANOVA revealed significant effects for stress [F(1, 16) = 10.46 p < 0.01], but not for drug treatment [F(1, 16) = 2.19, p = 0.16] and stress × drug treatment interaction [F(1, 16) = 4.65, p < 0.05] (Fig. 3A, B). For hippocampal GLT-1, two-way ANOVA revealed significant effects for drug treatment [F(1, 16) = 18.44, p < 0.001], but not for stress [F(1, 16) = 1.23, p = 0.28] and stress × drug treatment interaction [F(1, 16) = 0.002, p = 0.97] (Fig. 3A, C). For hippocampal GS and GLAST, two-way ANOVA revealed no significant effects for stress, drug treatment and stress × drug treatment interaction (Fig. 3A, D). For prefrontal cortical GFAP, two-way ANOVA revealed significant effects for stress [F(1, 16) = 5.36, p < 0.05] and stress × drug treatment interaction [F(1, 16) = 4.65, p < 0.05], but not for drug treatment [F(1,

ACCEPTED MANUSCRIPT 16) = 2.19, p = 0.16] (Fig. 3E, F). For prefrontal cortical GLT-1, two-way ANOVA revealed significant effects for drug treatment [F(1, 16) = 21.55, p < 0.001], but not for stress [F(1, 16) = 0.04, p = 0.84] and stress × drug treatment interaction [F(1, 16) = 0.01, p = 0.91] (Fig. 3E, G). For prefrontal cortical GS and GLAST, two-way ANOVA revealed no significant effects for stress, drug treatment and stress × drug

PT

treatment interaction (Fig. 3E, H). Post hoc analysis showed that CUS induced significant decreases in the protein expression levels of GFAP (Fig. 3A, B, n = 5, p <

RI

0.05 vs. vehicle alone-treated group) but not in other markers (Fig. 3A, C, D, n = 5). No significant changes of GFAP protein expressions were observed after harmine

SC

treatment in non-stressed mice (Fig. 3A, B, n = 5). However, chronic harmine treatment markedly prevented the CUS- induced decreases in GFAP protein

NU

expressions in mice hippocampus (Fig. 3A, B, n = 5, p < 0.05 vs. vehicle + CUS group) and prefrontal cortex (Fig. 3E, F, n = 5, p < 0.01 vs. vehicle + CUS group),

MA

demonstrating that the architecture and/or function of astrocytes may be altered by CUS and that chronic harmine treatment may block these changes. In addition, we

ED

found that harmine alone treatment induced significant increases in GLT-1 protein expressions in mice hippocampus (Fig. 3A, C, n = 5, p < 0.05 vs. vehicle

EP T

alone-treated group) and prefrontal cortex (Fig. 3E, G, n = 5, p < 0.05 vs. vehicle alone-treated group). No changes were observed in hippocampal (Fig. 3A, D, n = 5) and prefrontal cortex (Fig. 3E, H, n = 5) GLAST and GS expressions following

AC C

chronic harmine treatment in stressed and non-stressed mice. Increased GLT-1 protein expressions by harmine treatment were in accordance with previous studies and suggest that harmine treatment may attenuate glutamatergic hyperactivity.

3.4. Blockade of astrocytic functions abrogates the antidepressant-like effects of harmine in the FST, TST and sucrose preference experiment

To investigate the potential role of astrocytes in the antidepressant- like effects of harmine, a potent inhibitor of astrocytic functions, L-AAA (Takada and Hattori, 1986; Khurgel et al., 1996), was employed in the following experiment. The CUS-treated

ACCEPTED MANUSCRIPT mice were co-injected with harmine (20 mg/kg) and L-AAA for 10 days, with behavioral tests performed 2 h after the last injection. Results showed that chronic harmine treatment markedly prevented the CUS- induced increases in the immobile time in the FST (Fig. 4A, p < 0.05 vs. vehicle + CUS group) and TST (Fig. 4B, p < 0.05 vs. vehicle + CUS group), as well as the CUS-induced decrease in sucrose intake

PT

(Fig. 4C, p < 0.01 vs. vehicle + CUS group). L-AAA co-administration abrogated these effects of harmine (Fig. 4A–C). For FST, two-way ANOVA revealed significant

RI

main effects for stress [F(1, 54) = 31.15, p < 0.001] and drug treatment [F(2, 54) = 9.09, p < 0.001], but not for stress × drug treatment interaction [F(2, 54) = 0.24, p =

SC

0.78] (Fig. 4A). For TST, ANOVA revealed significant main effects for stress [F(1, 54) = 28.75, p < 0.001] and drug treatment [F(2, 54) = 4.87, p < 0.05], but not for

NU

stress × drug treatment interaction [F(2, 54) = 0.18, p < 0.83] (Fig. 4B). For sucrose preference, ANOVA revealed significant main effects for stress [F(1, 54) = 41.08, p <

MA

0.001] and drug treatment [F(2, 54) = 6.45, p < 0.01], but not for stress × drug

ED

treatment interaction [F(2, 54) = 3.16, p = 0.05] (Fig. 4C).

3.5. Blockade of astrocytic functions prevents the improvement effects of harmine on

EP T

brain BDNF protein levels and hippocampal neurogenesis

The harmine- induced increase in hippocampal BDNF protein levels (Fig. 5A, p <

AC C

0.05 vs. vehicle + CUS group) in stressed mice were blocked by L-AAA co-administration. In the prefrontal cortex of stressed mice, L-AAA co-administration also abolished the restoration effect of harmine on the protein level of BDNF (Fig. 5B, p < 0.05 vs. vehicle + CUS group). For hippocampal BDNF, two-way ANOVA revealed significant effects for stress [F(1, 42) = 20.04, p < 0.001] and drug treatment [F(2, 42) = 3.88, p < 0.05], but not for stress × drug treatment interaction [F(2, 42) = 0.13, p = 0.88] (Fig. 5B). For prefrontal cortical BDNF, two-way ANOVA revealed significant effects for stress [F(1, 42) = 23.69, p < 0.001] and drug treatment [F(2, 42) = 8.77, p < 0.001], but not for stress × drug treatment interaction [F(2, 42) = 1.34, p = 0.27] (Fig. 5B).

ACCEPTED MANUSCRIPT Further studies showed that the astrocyte was essential for the effect of harmine on hippocampal neurogenesis, as L-AAA co-administration blocked the increases in hippocampal neurogenesis (Fig. 5C, D, p < 0.05 vs. vehicle + CUS group) and DCX protein expressions (Fig. 5E, F, p < 0.05 vs. vehicle + CUS group) in harmine-treated mice. For DCX positive cells, two-way ANOVA revealed significant effects for stress

PT

[F(1, 24) = 104.10, p < 0.001] and drug treatment [F(2, 24) = 12.45, p < 0.001], but not for stress × drug treatment interaction [F(2, 24) = 0.15, p = 0.86] (Fig. 5D). For

RI

DCX protein levels, two-way ANOVA revealed significant effects for stress [F(1, 24) = 16.57, p < 0.001] and drug treatment [F(2, 24) = 5.85, p < 0.01], but not for stress ×

SC

drug treatment interaction [F(2, 24) = 0.12, p = 0.89] (Fig. 5F).

NU

4. Discussion

MA

One of the major contributions of this study is the identification of the antidepressant-like effects of harmine, which have been reported in previous studies

ED

(Fortunato et al., 2009; Fortunato et al., 2010; Aricioglu and Altunbas, 2003; Farzin and Mansouri, 2006). The major difference between our and other’s studies is that we

EP T

evaluated the antidepressant- like effects of harmine in stressed conditions, while the others did it in normal animals. On the face of it, there seems to be nothing new for this difference. However, the disease-dependent effect existing in the evaluation of

AC C

antidepressants suggests that evaluation of antidepressants in stressed conditions is more important and necessary than that in normal conditions. Fortunately, harmine, in normal conditions, has successfully been confirmed to have antidepressant- like effects, and thus our results provide further evidence to prove the antidepressant- like effects of harmine. But for others, the stress environment would help evaluate their antidepressant-like activities properly. For example, in untreated control mice, chronic treatment with fingolomod, the first oral drug approved for the treatment of relapsing-remitting multiple sclerosis, caused no changes in BDNF protein levels and even a significant reduction in BDNF mRNA levels in the hippocampal, whereas in mice treated with chronic stresses, fingolomod administration markedly prevented

ACCEPTED MANUSCRIPT depressive-like behaviors (di Nuzzo et al., 2015). Recently, harmine is becoming a star compound, as it has been selected from more than 100,000 different chemicals with β cell growth-promoting effects (Wang et al., 2015). However, harmine is also famous as one of the ingredients in the psychoactive mixture ayahuasca, and is traditionally used by some indigenous people for religious

PT

purposes. In this study, we used several different methods, including the FST, TST and sucrose preference experiment, to evaluate the antidepressant- like effects of

RI

harmine, and found that chronic daily injections of harmine (10 days) markedly improved the behavioral impairments induced by CUS, and the antidepressant-like

SC

effects of harmine were similar to that of fluoxetine. Harmine’s effects on major depression have been found in previous studies (Fortunato et al., 2009; Fortunato et

NU

al., 2010; Aricioglu and Altunbas, 2003; Farzin and Mansouri, 2006), and in those studies, only the FST was used to evaluate the antidepressant- like effects of harmine.

MA

In the current study, two other experiments with a high predictive validity for antidepressant activities, TST and sucrose preference experiments, were employed to

ED

assess the effect of harmine in depression. These experiments would help pay the way to understand the characteristic of harmine in depression therapy. Recently, ayahuasca,

EP T

a harmine-containing plant, has been shown to produce a rapid and sustained antidepressant-like effect in depressed patients, establishing stronger correlations between harmine and depression (Osório Fde et al., 2015; Sanches et al., 2016).

AC C

However, since ayahuasca also contains other substances, such as dimethyltryptamine, an agonist of serotonin (5-HT) 2A receptor actively involved in the pathogenesis of depression, evaluation of the correlation between harmine and depression using ayahuasca should be cautious, and thus although our results is beneficial for the development of harmine as a novel antidepressant, these results just provide preclinical evidence to strengthen the antidepressant-like activities of harmine. The impairment of hippocampal neurogenesis is an important pathophysiological event in depression formation (Serafini et al., 2014; Rotheneichner et al., 2014). Most current clinical antidepressants have been shown to exert antidepressant- like effects via restoration of hippocampal neurogenesis (Pascual-Brazo et al., 2014; Tang et al.,

ACCEPTED MANUSCRIPT 2012). We showed here that chronic stress impaired hippocampal neurogenesis, and harmine treatment markedly reversed this impairment. BDNF is a critical molecule upstream of hippocampal neurogenesis (Jiang et al., 2015a; Kim and Leem, 2016; Sakharkar et al., 2016). Hippocampal BDNF has been found to be reduced in depressed patients (Schmidt and Duman, 2007). Some clinical and preclinical

PT

antidepressants regulate depressive- like behaviors via activation of BDNF signals (Jiang et al., 2015a; Alboni et al., 2017; Liu et al., 2016). Searching agents that

RI

increase BDNF expressions would help develop new antidepressants. Our results showed that harmine markedly reversed the CUS-induced decreases in hippocampal

SC

and prefrontal cortical BDNF protein levels, further underscoring the importance of BDNF in the antidepressant-like effects of harmine (Fortunato et al., 2009; Fortunato

NU

et al., 2010). However, how harmine increases BDNF protein levels remains to be

MA

determined in the future study.

Astrocytes are the most abundant form of cells in the brain. Numerous studies employing depressed patients and animals show that astrocyte pathology mediates the

ED

pathogenesis of depression. First, the decreases of astrocyte-expressing GFAP have been seen in genetically stress susceptible Wistar Kyoto rat (Gosselin et al., 2009), as

EP T

well as in mice treated with CUS (Ye et al., 2011) and maternal deprivation (Leventopoulos et al., 2007). Second, the integrity of hippocampal astrocytic

AC C

plasticity mediates the antidepressant- like effects of some clinical antidepressants (Zarei et al., 2014), and the rapid antidepressant-like effect of ketamine correlates astroglial plasticity in the hippocampus (Ardalan et al., 2017). Furthermore, local infusion of a gliotoxin in the prefrontal cortex resulting in the decrease in GLT-1 and astrocyte numbers is sufficient to induce depressive- like phenotypes similar to that observed in CUS conditions (Banasr and Duman, 2008). Third, strong evidences of stress-related impairments of astrocyte functions have been reported in recent years. For instance, the glial-specific EAATs and GS reflecting the glutamate clearance and metabolism ability of astrocytes are reduced in some regions of depressed brains (Choudary et al., 2005; Bernard et al., 2011). The impairment of astrocyte functions

ACCEPTED MANUSCRIPT induces the brain incapable of clearing extracellular glutamate, ultimately resulting in an altered ratio of synaptic to extra-synaptic glutamate content as well as an altered neurotransmission that is considered to contribute to the pathogenesis of neuropsychiatric disorders including major depression. In this study, we showed that CUS reduced GFAP protein expressions in the prefrontal cortex and hippocampus,

PT

and these reductions were reversed by harmine treatment. GFAP is an intermediate filament protein that is traditionally considered a marker of astrocytes. Its expression

RI

can be modulated by neurodegenerative and neuropsychiatric stimuli (Olivera-Bravo et al., 2011; Schreiner et al., 2015). Although whether the reduction of astrocytic

SC

GFAP in depressed individuals is directly associated with depression remains unclear, the decrease of GFAP to some extent reflects the impairment of astrocytic functions

NU

and behavioral deficits. Thus, the regulating effect of harmine on GFAP expressions indicates that the astrocyte may be involved in the antidepressant-like effects of

MA

harmine. This hypothesis was further supported by the up-regulation effects of harmine on the protein expression levels of hippocampal and prefrontal cortical

ED

GLT-1, though CUS itself did not alter GLT-1 and EAATs expressions. It is well-known that the micro-environmental glutamate is increased abnormally in the

EP T

brain in depressed animals and patients (Jia et al., 2015; Dávalos et al., 2000). Since the increased glutamate is postulated to mediate the pathogenesis of depression (Paul and Skolnick, 2003; Deutschenbaur et al., 2016), promotion of glutamate clearance

AC C

would be beneficial for depression therapy. Harmine, because of its GLT-1 upregulation effect, is a potential candidate for that purpose. In fact, previous studies have already provided several similar candidates for that purpose. For example, ceftriaxone, a clinical available drug that increases glutamate transport, has been reported to exhibit antidepressant- like properties in the FST, TST, and novelty suppressed feeding experiment (Mineur et al., 2007). Riluzole, a Food and Drug Administration-approved drug for the treatment of ALS, also ameliorates depressive-like behaviors via up-regulation of GFAP and GLT-1 protein expression and promotion of glutamate clearance (Banasr et al., 2010).

ACCEPTED MANUSCRIPT The involvement of brain astrocytes in the antidepressant- like effects of harmine was further ascertained by the astrocyte inhibition experiment, which showed that injection of mice with L-AAA, a blocker of astrocytic functions (Banasr and Duman, 2008), prevented the amelioration effects of harmine on depressive- like behaviors in the experiments of FST, TST and sucrose preference. This finding provided a direct

PT

evidence to uncover the potential role of astrocytes in the antidepressant-like effects of harmine, and would help understand the in-depth mechanism of action of harmine

RI

in major depression. The reason for that saying is that i). astrocyte dysfunction is now considered an important factor that determines depression formation (Ye et al., 2011;

SC

Leventopoulos et al., 2007); ii). some well-known targets for antidepressants, such as 5-HT receptors and monoamine oxidase (MAO), are tightly correlated with astrocyte

NU

functions. For example, the 5-HT1A (Eriksen et al., 2002) and 5-HT2A receptor (Maxishima et al., 2001) have been confirmed to be expressed in astrocytes. Their

MA

activation may protect neurons against injuries induced by transient global cerebral ischemia (Lee et al., 2015) or Parkinson’s disease (Miyazaki et al., 2013).

ED

Reactive astrocytes can produce the inhibitory gliotransmitter γ-aminobutyric acid (GABA) by MAO-B, and the released GABA reduces spike probability of granule

EP T

cells by acting on presynaptic GABA receptors (Jo et al., 2014). Thus, it is reasonable speculate that the mutual interaction between astrocytes and antidepressant targets may co-ordinate neuronal functions during stress stimulation. Further investigation of

AC C

the mutual interaction between astrocytes and antidepressant targets would promote the understanding about the role of harmine and other hallucinogens, such as lysergic acid diethylamide (LSD) (Dos Santos et al., 2016) and psilocybin (Carhart-Harris et al., 2016), in depression regulation. Further analysis showed that the harmine- induced increases in hippocampal and prefrontal cortical BDNF protein levels in CUS-treated mice were blocked by L-AAA treatment, demonstrating that the BDNF signal may be necessary for the regulation of depressive-like behaviors by endogenous astrocytes. BDNF has been considered to regulate depressive- like behaviors through promotion of hippocampal neurogenesis

ACCEPTED MANUSCRIPT (Jiang et al., 2015a; Kim and Leem, 2016; Sakharkar et al., 2016). Our results showed that L-AAA co-administration abrogated the promoting effect of harmine on hippocampal neurogenesis. Previous studies also showed that specific overexpression of BDNF in hippocampal astrocytes promotes local neurogenesis and elicits anxiolytic activities (Quesseveur et al., 2013), suggesting that BDNF improves

PT

depressive-like behaviors and hippocampal neurogenesis likely through enhancement of astrocytic functions. Taken together, all these findings emphasize the importance of

SC

RI

brain astrocytes in harmine’s effects on depression.

NU

5. Conclusions

Our results are in consistent with a growing number of studies sho wing that

MA

astrocyte dysfunction contributes critically to the development of depression (Ye et al., 2011; Leventopoulos et al., 2007; Oh et al., 2002; Ménard et al., 2016). Harmine, a

ED

natural-derived drug reported to increase GLT-1 expression (Li et al., 2011; Sun et al., 2014), reversed the behavioral deficits induced by CUS. The potential involvement of

EP T

GLT-1 in harmine’s effects on depression is supported by previous studies, which showed that ceftriaxone (Mineur et al., 2007) and riluzole (Banasr et al., 2010), two compounds with GLT-1 increasing effects, ameliorates depressive-like behaviors in

AC C

depressed animals or patients. The antidepressant- like effects of harmine were also evidenced by the astrocyte- mediated restoration of BDNF protein levels and hippocampal neurogenesis. These mechanistic findings not only provide a further validation for the hypothesis of astrocyte dysfunction in depression formation, but also help understand the potential role of harmine in depression therapy.

Acknowledgements

This work was supported by the Natural Science Foundation of China (No.

ACCEPTED MANUSCRIPT 81571323), the Natural Science Foundation of Jiangsu Province (No. BK20141240), and the Science and Technology Project of Nantong City (No. MS12015050). Ethical Statements

All animal experiments in this study were conducted in accordance with

PT

internationally accepted guidelines for the use of animals in toxicology as adopted by the Society of Toxicology in 1999) and approved by the University Animal Ethics

RI

Committee of Nantong University (Permit Number: 2110836).

SC

References

NU

Alboni S, van Dijk RM, Poggini S, Milior G, Perrotta M, Dre nth T, et al. Fluoxetine effects on molecular, cellular and behavioral endophenotypes of depression are

MA

driven by the living environment. Mol Psychiatry 2017;22:552−61. Ardalan M, Rafati AH, Nyengaard JR, Wegener G. Rapid antidepressant effect of

2017;174:483−92.

ED

ketamine correlates with astroglial plasticity in the hippocampus. Br J Pharmacol

EP T

Aricioglu F, Altunbas H. Harmane induces anxiolysis and antidepressant- like effects in rats. Ann N Y Acad Sci 2003;1009:196−201. Banasr M, Chowdhury GM, Terwilliger R, Newton SS, Duman RS, Behar KL, et al.

AC C

Glial pathology in an animal model of depression: reversal of stress- induced cellular, metabolic and behavioral deficits by the glutamate-modulating drug riluzole. Mol Psychiatry 2010;15:501−11. Banasr M, Duman RS. Glial loss in the prefrontal cortex is sufficient to induce depressive-like behaviors. Biol Psychiatry 2008;64:863−70. Ben Haim L, Rowitch DH. Functional diversity of astrocytes in neural circuit regulation. Nat Rev Neurosci 2017;18:31−41. Bernard R, Kerman IA, Thompson RC, Jones EG, Bunney WE, Barchas JD, et al. Altered expression of glutamate signaling, growth factor, and glia genes in the

ACCEPTED MANUSCRIPT locus coeruleus of patients

with

major depression.

Mol Psychiatry

2011;16:634−46. Carhart-Harris RL, Bolstridge M, Rucker J, Day CM, Erritzoe D, Kaelen M, et al. Psilocybin with psychological support for treatment-resistant depression: an open-label feasibility study. Lancet Psychiatry 2016;3:619−27.

PT

Choudary PV, Molnar M, Evans SJ, Tomita H, Li JZ, Vawter MP, et al. Altered cortical glutamatergic and GABAergic signal transmission with glial

RI

involvement in depression. Proc Natl Acad Sci U S A 2005;102:15653−8. Dávalos A, Shuaib A, Wahlgren NG. Neurotransmitters and pathophysiology of

SC

stroke: evidence for the release of glutamate and other transmitters/mediators in animals and humans. J Stroke Cerebrovasc Dis 2000;9:2−8.

NU

Deutschenbaur L, Beck J, Kiyhankhadiv A, Mühlhauser M, Borgwardt S, Walter M, et al. Role of calcium, glutamate and NMDA in major depression and therapeutic

MA

application. Prog Neuropsychopharmacol Biol Psychiatry 2016;64:325−33. di Nuzzo L, Orlando R, Tognoli C, Di Pietro P, Bertini G, Miele J, et al.

2015;3:e00135.

mice. Pharmacol Res Perspect

ED

Antidepressant activity of fingolimod in

EP T

Eriksen JL, Gillespie R, Druse MJ. Effects of ethanol and 5-HT1A agonists on astroglial S100B. Brain Res Dev Brain Res 2002;139:97−105. Fabbri C, Marsano A, Balestri M, De Ronchi D, Serretti A. Clinical features and drug

AC C

induced side effects in early versus late antidepressant responders. J Psychiatr Res 2013;47:1309−18. Farzin D,

Mansouri N. Antidepressant- like effect of harmane and other

beta-carbolines in the mouse forced swim test. Eur Neuropsychopharmacol 2006;16:324−8. Fava M. Switching treatments for complicated depression. J Clin Psychiatry 2010;71:e04. Fortunato JJ, Réus GZ, Kirsch TR, Stringari RB, Fries GR, Kapczinski F, et al. Chronic administration of harmine elicits antidepressant- like effects and increases BDNF levels in rat hippocampus. J Neural Transm (Vienna)

ACCEPTED MANUSCRIPT 2010;117:1131−7. Fortunato JJ, Réus GZ, Kirsch TR, Stringari RB, Stertz L, Kapczinski F, et al. Acute harmine administration induces antidepressive- like effects and increases BDNF levels in the rat hippocampus. Prog Neuropsychopharmacol Biol Psychiatry 2009;33:1425−30.

PT

Frey BN, Andreazza AC, Ceresér KM, Martins MR, Valvassori SS, Réus GZ, et al. Effects of mood stabilizers on hippocampus BDNF levels in an animal model of

RI

mania. Life Sci 2006;79:281−6.

Gosselin RD, Gibney S, O'Malley D, Dinan TG, Cryan JF. Region specific decrease

SC

in glial fibrillary acidic protein immunoreactivity in the brain of a rat model of depression. Neuroscience 2009;159:915−25.

NU

Huang C, Hu W, Wang J, Tong L, Lu X, Wu F, et al. Methylene blue increases the amount of HSF1 through promotion of PKA- mediated increase in HSF1-p300

MA

interaction. Int J Biochem Cell Biol 2017;84:75−88. Jia M, Njapo SA, Rastogi V, Hedna VS. Taming glutamate excitotoxicity: strategic

ED

pathway modulation for neuroprotection. CNS Drugs 2015;29:153−62. Jiang B, Huang C, Chen XF, Tong LJ, Zhang W. Tetramethylpyrazine Produces

EP T

Antidepressant-Like Effects in Mice Through Promotion of BDNF Signaling Pathway.Int J Neuropsychopharmacol 2015a;18. Jiang B,

Huang C,

Zhu Q,

Tong LJ,

Zhang W. WY14643 produces

AC C

anti-depressant- like effects in mice via the BDNF signaling pathway. Psychopharmacology (Berl) 2015b;232:1629−42. Jo S, Yarishkin O, Hwang YJ, Chun YE, Park M, Woo DH, et al. GABA from reactive astrocytes impairs memory in mouse models of Alzheimer’s disease. Nat Med 2014;20:886−96. Kern N, Sheldrick AJ, Schmidt FM, Minkwitz J. Neurobiology of depression and novel antidepressant drug targets. Curr Pharm Des 2012;18:5791−801. Khurgel M, Koo AC, Ivy GO. Selective ablation of astrocytes by intracerebral injections of alpha-aminoadipate. Glia 1996;16:351–8.

ACCEPTED MANUSCRIPT Kim DH, Jang YY, Han ES, Lee CS. Protective effect of harmaline and harmalol against dopamine- and 6-hydroxydopamine- induced oxidative damage of brain mitochondria and synaptosomes, and viability loss of PC12 cells. Eur J Neurosci 2001;13:1861−72. Kim DM, Leem YH. Chronic stress- induced memory deficits are reversed by regular

PT

exercise via AMPK-mediated BDNF induction. Neuroscience 2016;324:271−85. Kleinridders A, Schenten D, Konner AC, Belgardt BF, Mauer J, Okamura T, et al.

RI

MyD88 signaling in the CNS is required for development of fatty acid- induced leptin resistance and diet-induced obesity. Cell Metab 2009;10:249–59.

SC

Lanni C, Govoni S, Lucchelli A, Boselli C. Depression and antidepressants: molecular and cellular aspects. Cell Mol Life Sci 2009;66:2985−3008.

NU

Lee CH, Ahn JH, Won MH. New expression of 5-HT1A receptor in astrocytes in the

Neurol Sci 2015;36:383−9.

MA

gerbil hippocampal CA1 region following transient global cerebral ischemia.

Dos Santos RG, Osório FL, Crippa JA, Riba J, Zuardi AW, Hallak JE. Antidepressive,

ED

anxiolytic, and antiaddictive effects of ayahuasca, psilocybin and lysergic acid diethylamide (LSD): a systematic review of clinical trials published in the last

EP T

25 years. Ther Adv Psychopharmacol 2016;6:193−213. Leventopoulos M, Rüedi-Bettschen D, Knuesel I, Feldon J, Pryce CR, Opacka-Juffry J. Long-term effects of early life deprivation on brain glia in Fischer rats. Brain

AC C

Res 2007;1142:119−26.

Li Y, Sattler R, Yang EJ, Nunes A, Ayukawa Y, Akhtar S, et al. Harmine, a natural beta-carboline alkaloid, upregulates astroglial glutamate transporter expression. Neuropharmacology 2011;60:1168−75. Liu Z, Qi Y, Cheng Z, Zhu X, Fan C, Yu SY. The effects of ginsenoside Rg1 on chronic stress induced depression- like behaviors, BDNF expression and the phosphorylation of PKA and CREB in rats. Neuroscience 2016;322:358−69. Lutgen V, Narasipura SD, Sharma A, Min S, Al- Harthi L. β-Catenin signaling positively regulates glutamate uptake and metabolism in astrocytes. J Neuroinflammation 2016;13:242.

ACCEPTED MANUSCRIPT Martin JL, Brown AL, Balkowiec A. Glia determine the course of brain-derived neurotrophic factor- mediated dendritogenesis and provide a soluble inhibitory cue to dendritic growth in the brainstem. Neuroscience 2012;207:333−46. Maxishima M, Shiga T, Shutoh F, Hamada S, Maeshima T, Okado N. Serotonin 2A receptor- like

immunoreactivity

is

detected

in

astrocytes

but

not

in

PT

oligodendrocytes of rat spinal cord. Brain Res 2001;889:270−3. Ménard C, Hodes GE, Russo SJ. Pathogenesis of depression: Insights from human

RI

and rodent studies. Neuroscience 2016;321:138−62.

Mineur YS, Picciotto MR, Sanacora G. Antidepressant-like effects of ceftriaxone in

SC

male C57BL/6J mice. Biol Psychiatry 2007;61:250−2.

Miyazaki I, Asanuma M, Murakami S, Takeshima M, Torigoe N, Kitamura Y, et al.

NU

Targeting 5-HT(1A) receptors in astrocytes to protect dopaminergic neurons in Parkinsonian models. Neurobiol Dis 2013;59:244−56.

MA

Moura DJ, Richter MF, Boeira JM, Pêgas Henriques JA, Saffi J. Antioxidant properties of beta-carboline alkaloids are related to their antimutagenic and

ED

antigenotoxic activities. Mutagenesis 2007;22:293−302. Oh DH, Son H, Hwang S, Kim SH. Neuropathological abnormalities of astrocytes,

EP T

GABAergic neurons, and pyramidal neurons in the dorsolateral prefrontal cortices of patients with major depressive disorder. Eur Neuropsychopharmacol 2012;22:330−8.

AC C

Olivera-Bravo S, Fernández A, Sarlabós MN, Rosillo JC, Casanova G, Jiménez M, et al. Neonatal astrocyte damage is sufficient to trigger progressive striatal degeneration in a rat model of glutaric acidemia-I. PLoS One 2011;6:e20831. Osório Fde L, Sanches RF, Macedo LR, Santos RG, Maia-de-Oliveira JP, Wichert-Ana L, et al. Antidepressant effects of a single dose of ayahuasca in patients with recurrent depression: a preliminary report. Rev Bras Psiquiatr 2015;37:13−20. Pascual- Brazo J, Baekelandt V, Encinas JM. Neurogenesis as a new target for the development of antidepressant drugs. Curr Pharm Des 2014;20:3763−75.

ACCEPTED MANUSCRIPT Paul IA, Skolnick P. Glutamate and depression: clinical and preclinical studies. Ann N Y Acad Sci 2003;1003:250−72. Porsolt RD, Bertin A, Jalfre M. Behavioral despair in mice: a primary screening test for antidepressants. Arch Int Pharmacodyn Ther 1977;229:327−36. Quesseveur G, David DJ, Gaillard MC, Pla P, Wu MV, Nguyen HT, et al. BDNF

PT

overexpression in mouse hippocampal astrocytes promotes local neurogenesis and elicits anxiolytic- like activities. Transl Psychiatry 2013;3:e253.

RI

Rotheneichner P, Lange S, O'Sullivan A, Marschallinger J, Zaunmair P, Geretsegger

relations. Neural Plast 2014;2014:723915.

SC

C, et al. Hippocampal neurogenesis and antidepressive therapy: shocking

Sakharkar AJ, Vetreno RP, Zhang H, Kokare DM, Crews FT, Pandey SC. A role for

NU

histone acetylation mechanisms in adolescent alcohol exposure- induced deficits in hippocampal brain-derived neurotrophic factor expression and neurogenesis

MA

markers in adulthood. Brain Struct Funct 2016;221:4691−703. Sanchez C, Asin KE, Artigas F. Vortioxetine, a novel antidepressant with multimodal review

2015;145:43−57.

of

preclinical

and

clinical

data.

Pharmacol

Ther

ED

activity:

EP T

Sanches RF, de Lima Osório F, Dos Santos RG, Macedo LR, Maia-de-Oliveira JP, Wichert-Ana L, et al. Antidepressant Effects of a Single Dose of Ayahuasca in Patients With Recurrent Depression: A SPECT Study. J Clin Psychopharmacol

AC C

2016;36:77−81.

Schmidt HD, Duman RS. The role of neurotrophic factors in adult hippocampal neurogenesis, antidepressant treatments and animal models of depressive-like behavior. Behav Pharmacol 2007;18:391−418. Schreiner B, Romanelli E, Liberski P, Ingold-Heppner B, Sobottka-Brillout B, Hartwig T, et al. Astrocyte Depletion Impairs Redox Homeostasis and Triggers Neuronal Loss in the Adult CNS. Cell Rep 2015;12:1377−84. Schwartz J, Murrough JW, Iosifescu DV. Ketamine for treatment-resistant depression: recent developments and clinical applications. Evid Based Ment Health 2016;19:35−8.

ACCEPTED MANUSCRIPT Serafini G, Hayley S, Pompili M, Dwivedi Y, Brahmachari G, Girardi P, et al. Hippocampal neurogenesis, neurotrophic factors and depression: possible therapeutic targets? CNS Neurol Disord Drug Targets 2014;13:1708−21. Sourkes TL. "Rational hope" in the early treatment of Parkinson's disease. Can J Physiol Pharmacol 1999;77:375−82.

PT

Steru L, Chermat R, Thierry B, Simon P. The tail suspension test: a new method for screening antidepressants in mice. Psychopharmacology (Berl) 1985;85:367−70.

RI

Sun P, Zhang S, Li Y, Wang L. Harmine mediated neuroprotection via evaluation of glutamate transporter 1 in a rat model of global cerebral ischemia. Neurosci Lett

SC

2014;583:32−6.

Takada M, Hattori T. Fine structural changes in the rat brain after local injections of

NU

gliotoxin, alpha-aminoadipic acid. Histol Histopathol 1986;1:271–5. Tang SW, Helmeste D, Leonard B. Is neurogenesis relevant in depression and in the

Psychiatry 2012;13:402−12.

MA

mechanism of antidepressant drug action? A critical review. World J Biol

ED

Waki H, Park KW, Mitro N, Pei L, Damoiseaux R, Wilpitz DC, et al. The small molecule harmine is an antidiabetic cell-type-specific regulator of PPARgamma

EP T

expression. Cell Metab 2007;5:357−70. Wang P, Alvarez-Perez JC, Felsenfeld DP, Liu H, Sivendran S, Bender A, et al. A high-throughput chemical screen reveals that harmine- mediated inhibition of

AC C

DYRK1A increases human pancreatic beta cell replication. Nat Med 2015;21:383−8.

Yang R, Wang P, Chen Z, Hu W, Gong Y, Zhang W, et al. WY-14643, a selective agonist

of

peroxisome

lipopolysaccharide- induced

proliferator-activated depressive- like

receptor-α,

behaviors

by

ameliorates preventing

neuroinflammation and oxido-nitrosative stress in mice. Pharmacol Biochem Behav 2017;153:97−104. Yang X, Sheng W, Ridgley DM, Haidekker MA, Sun GY, Lee JC. Astrocytes regulate α-secretase-cleaved soluble amyloid precursor protein secretion in

ACCEPTED MANUSCRIPT neuronal cells: Involvement of group IIA secretory phospholipase A2. Neuroscience 2015;300:508−17. Yao W, Huang L, Sun Q, Yang L, Tang L, Meng G, et al. The inhibition of macrophage foam cell formation by tetrahydroxystilbene glucoside is driven by suppressing vimentin cytoskeleton. Biomed Pharmacother 2016;83:1132−40.

PT

Ye Y, Wang G, Wang H, Wang X. Brain-derived neurotrophic factor (BDNF) infusion restored astrocytic plasticity in the hippocampus of a rat model of

RI

depression. Neurosci Lett 2011;503:15−9.

Zarei G, Reisi P, Alaei H, Javanmard SH. Effects of amitriptyline and fluoxetine on

SC

synaptic plasticity in the dentate gyrus of hippocampal formation in rats. Adv

NU

Biomed Res 2014;3:199.

MA

Figure legends

Fig. 1. Effects of harmine on CUS- induced behavioral abnormalities. (A) The

ED

schematic diagram for the experimental timeline and tissue preparation in the present study. (B, C) Quantitative analysis showing the inhibitory effects of harmine (10 or 20

EP T

mg/kg) on CUS- induced increases in the immobile time in the FST (B, n = 10, *p < 0.05, **p < 0.01 vs. vehicle alone-treated group; #p < 0.05, ##p < 0.01 vs. vehicle + CUS group) and TST (C, n = 10, *p < 0.05, **p < 0.01 vs. vehicle alone-treated group;

AC C

#p < 0.05 vs. vehicle + CUS group). (D) Quantitative analysis showing the inhibitory effect of harmine (10 or 20 mg/kg) on CUS- induced decrease in sucrose intake (n = 10, **p < 0.01 vs. vehicle alone-treated group; #p < 0.05 vs. vehicle + CUS group). The fluoxetine administration (20 mg/kg) was used as a positive control, and all data were shown as mean ± SEM.

Fig. 2. Effects of harmine on CUS-induced reductions in BDNF protein levels and hippocampal neurogenesis. (A, B) Statistical analysis showing the restoration effects of harmine (10, 20 mg/kg) on CUS- induced decreases in BDNF protein levels in mice hippocampus (A) and prefrontal cortex (B) (n = 8, *p < 0.05 vs. vehicle alone-treated

ACCEPTED MANUSCRIPT group; #p < 0.05 or ##p < 0.01 vs. vehicle + CUS group). (C, D) Representative images (C) and quantitative analysis (D) showing the restoration effect of harmine (20 mg/kg) on CUS-induced decrease in hippocampal DCX+ cells (n = 5, *p < 0.05, **p < 0.01 vs. vehicle alone-treated group; #p < 0.05 vs. vehicle + CUS group). (E, F) Representative images (E) and quantitative analysis (F) showing the restoration effect

PT

of harmine on CUS-induced decrease in hippocampal DCX protein expressions (n = 5, *p < 0.05, **p < 0.01 vs. vehicle alone-treated group; #p < 0.05 vs. vehicle + CUS

RI

group). The fluoxetine administration (20 mg/kg) was used as a positive control, and

SC

all data were shown as mean ± SEM.

Fig. 3. Effects of harmine treatment on GFAP, GLT-1, GS, and GLAST protein

NU

expressions with or without CUS treatment. (A–D) Representative images and quantitative analysis showing the effects of harmine (20 mg/kg) on the protein

MA

expression levels of GFAP (A, B), GLT-1 (A, C), GS (A, D), and GLAST (A, D) in mice hippocampus with or without CUS treatment (n = 5, *p < 0.05 vs. vehicle

ED

alone-treated group; #p < 0.05 vs. vehicle + CUS group). (E–H) Representative images showing the effects of harmine (20 mg/kg) on the protein expression levels of

EP T

GFAP (E, F), GLT-1 (E, G), GS (E, H), and GLAST (E, H) in mice prefrontal cortex with or without CUS treatment (n = 5, *p < 0.05 vs. vehicle alone-treated group; #p <

AC C

0.05 or ##p < 0.01 vs. vehicle + CUS group). All data were shown as mean ± SEM.

Fig. 4. Effects of L-AAA, a specific inhibitor of astrocytic functions, on CUS- induced behavioral abnormalities. (A, B) Quantitative analysis showing that L-AAA co-administration attenuated the restoration effects of harmine (20 mg/kg) on CUS-induced increases in the immobile time in the FST (A, n = 10, *p < 0.05 or **p < 0.01 vs. vehicle alone-treated group; #p < 0.05 vs. vehicle + CUS) and TST (B, n = 10, *p < 0.05 vs. vehicle alone-treated group; #p < 0.05 vs. vehicle + CUS). (C) Quantitative analysis showing that L-AAA co-administration attenuated the restoration effect of harmine (20 mg/kg) on CUS- induced decrease in sucrose intake (n = 10; **p < 0.01 vs. vehicle alone-treated group; ##p < 0.01 vs. vehicle + CUS).

ACCEPTED MANUSCRIPT All data were shown as mean ± SEM.

Fig. 5. Effects of L-AAA, a specific inhibitor of astrocytic functions, on CUS- induced reductions in BDNF protein levels and hippocampal neurogenesis. (A, B) Statistical analysis showing that L-AAA co-administration abrogated the prevention effects of

PT

harmine (10, 20 mg/kg) on CUS-induced decreases in BDNF protein levels in mice hippocampus (A) and prefrontal cortex (B) (n = 8, *p < 0.05 vs. vehicle alone-treated

RI

group; #p < 0.05 vs. vehicle + CUS group). (C, D) Representative images (C) and quantitative analysis (D) showing that L-AAA co-administration abrogated the

SC

prevention effects of harmine (20 mg/kg) on CUS- induced decrease in hippocampal DCX+ cells (n = 5, **p < 0.01 vs. vehicle alone-treated group; #p < 0.05 vs. vehicle +

NU

CUS group). (E, F) Representative images (E) and quantitative analysis (F) showing that L-AAA co-administration abrogated the prevention effects of harmine on

MA

CUS-induced decrease in DCX protein expressions in mice hippocampus (n = 5, *p < 0.05, **p < 0.01 vs. vehicle alone-treated group; #p < 0.05 vs. vehicle + CUS group).

AC C

EP T

ED

All data were shown as mean ± SEM.

AC C

EP T

ED

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC C

EP T

ED

MA

Figure 1

AC C

EP T

ED

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 2

AC C

EP T

ED

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 3

AC C

Figure 4

EP T

ED

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC C

EP T

ED

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 5

ACCEPTED MANUSCRIPT Highlights 

Harmine prevents depressive-like behaviors induced by CUS.



Harmine restores BDNF and hippocampal neurogenesis decrease in CUS-treated mice. Harmine restores the alteration of astrocyte markers in CUS-treated mice.



Astrocytic inhibition blocks the antidepressant- like effect of harmine.



Astrocytic inhibition blocks the effect of harmine on BDNF and neurogenesis.

AC C

EP T

ED

MA

NU

SC

RI

PT

