Behavioural and neuroplastic effects of the new-generation antidepressant agomelatine compared to fluoxetine in glucocorticoid receptor-impaired mice.

HAL Id: hal-00863103

https://hal.archives-ouvertes.fr/hal-00863103

Submitted on 3 Feb 2021

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

new-generation antidepressant agomelatine compared to fluoxetine in glucocorticoid receptor-impaired mice.

Eleni Païzanis, Thibault Renoir, Vincent Lelièvre, Françoise Saurini, Maxette Melfort, Cecilia Gabriel, Nicholas Barden, Elisabeth Mocaër, Michel Hamon,

Laurence Lanfumey

To cite this version:

Eleni Païzanis, Thibault Renoir, Vincent Lelièvre, Françoise Saurini, Maxette Melfort, et al.. Be- havioural and neuroplastic effects of the new-generation antidepressant agomelatine compared to fluoxetine in glucocorticoid receptor-impaired mice.. International Journal of Neuropsychopharma- cology, Oxford University Press (OUP), 2010, 13 (6), pp.759-774. �10.1017/S1461145709990514�.

�hal-00863103�

Behavioural and neuroplastic effects of the new-generation antidepressant agomelatine compared to fluoxetine in glucocorticoid receptor-impaired mice

Eleni Paı¨zanis

, Thibault Renoir

, Vincent Lelievre

, Franc¸oise Saurini

, Maxette Melfort

, Cecilia Gabriel

, Nicholas Barden

, Elisabeth Mocae¨r

, Michel Hamon

and Laurence Lanfumey

1INSERM UMR S677-F-75013 Paris, France

2UPMC, Universite´ Paris 06, UMR S677, F-75013 Paris, France

3INSERM UMR S676-F-75019 Paris, France

4IRIS, Courbevoie, France

5Department of Neuroscience, CHUL, University Laval, Ste Foy, Quebec, Canada

Abstract

Major depression is associated with reduced hippocampal volume linked to stress and high glucocorticoid secretion. Glucocorticoid receptor-impaired (GR-i) mice, a transgenic model for affective disorders with hypothalamic–pituitary–adrenal (HPA) axis feedback control deficit, were used to assess the anti- depressant-like effects of the mixed melatonin receptor agonist/5-HT2Creceptor antagonist, agomelatine, compared to the selective 5-HT reuptake inhibitor (SSRI), fluoxetine, on hippocampal neurogenesis, GR and BDNF expression and antidepressant-responsive behaviour (tail suspension test, TST). GR-i and paired wild-type (WT) mice were given acute or chronic (21 d) treatment with these drugs. Both hippo- campal cell proliferation and BDNF mRNA expression were down-regulated in GR-i mice, and these alterations were reversed by chronic agomelatine and fluoxetine treatments, whereas GR mRNA down- regulation was reversed only by agomelatine. Furthermore, chronic agomelatine, but not fluoxetine, increased survival of newly formed cells in the ventral part of the hippocampus without changing their phenotypic differentiation into neurons. In the TST, the enhanced immobility of GR-i mice was reduced to WT level by acute (but not chronic) fluoxetine and chronic (but not acute) agomelatine. These results indicate that agomelatine reversed the neuroplastic changes and helpless behaviour associated with HPA axis alterations in GR-i mice, suggesting neurobiological and behavioural effects mostly similar to those typically seen with classical antidepressants such as fluoxetine, but through clearly distinct mechanisms.

Received 6 February 2009 ; Reviewed 1 April 2009 ; Revised 1 July 2009 ; Accepted 28 July 2009 ; First published online 24 September 2009

Key words: BDNF, hippocampus, hypothalamic–pituitary–adrenocortical axis, neurogenesis, transgenic mice.

Introduction

Neuroimaging and post-mortem studies recently showed that major depressive disorder (MDD) is most often associated with structural and functional brain alterations, notably a shrinkage of dendritic spines

(Zhou et al. 2004), a reduction of hippocampal volume possibly related to neuronal cell death and neuro- genesis deficiency (Rajkowska, 2000 ; Sheline, 2000), although no clinical studies have up to now supported the latter hypothesis (Czeh & Lucassen, 2007 ; Reif et al. 2006). Interestingly, such alterations have also been observed in various animal models of stress or depression (Fuchs & Flugge, 1998). In addition, various antidepressant treatments, including selective serotonin reuptake inhibitor (SSRIs), as well as antagonists at hormone receptors responsible for

Address for correspondence : Dr L. Lanfumey, INSERM UMR S677/

894, Faculte´ de Me´decine UPMC, Site Pitie´-Salpeˆtrie`re, 91 boulevard de l’Hoˆpital, F-75013, Paris, France.

Tel. : 33140779707 Fax: 33140779790 Email: laurence.lanfumey@upmc.fr

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

hypothalamic–pituitary–adrenal axis (HPA) acti- vation, can reverse these changes (Keilhoff et al. 2006 ; Malberg & Duman, 2003 ; Mayer et al. 2006).

In rodents, adult neurogenesis, which involves precursor cell proliferation, migration and differen- tiation, mainly occurs in two brain regions, the sub- granular zone (SGZ) of the dentate gyrus (DG) in the hippocampus and the subventricular zone in the forebrain. Many factors, including hormones and neurotransmitters, regulate hippocampal neurogen- esis (Paizanis et al. 2007). In particular, modifications in the expression of neurotrophins and growth factors like brain-derived neurotrophic factor (BDNF) have been reported to contribute to changes in hippocampal neurogenesis associated with stress-related disorders and antidepressant treatments (Hashimoto et al. 2004 ; Paizanis et al. 2007).

Compelling data also showed that MDD is often associated with HPA axis dysfunction. HPA axis ac- tivity is controlled by a feedback mechanism triggered mainly through the stimulation of brain glucocorticoid receptors (GR) (Dorio et al. 1993 ; Feldman &

Weidenfeld, 1999), and which reduced expression and function have been proposed to contribute to HPA axis alterations in depression (Pariante & Miller, 2001).

Accordingly, transgenic mice with impaired brain expression of GR (GR-i) have been generated, as a murine model presenting many of the neuroendocrine changes associated with depressive disorders, includ- ing tonic increase in HPA axis activity and resistance to suppression of corticosterone secretion by dexa- methasone (Barden et al. 1997 ; Froger et al. 2004 ; Pepin et al. 1992). In addition to having cognitive deficits (Montkowski et al. 1995), GR-i mice are characterized by a reduced CRH expression in the hypothalamus, maladaptive responses of the 5-HT system to stress (Froger et al. 2004), but no changes in ACTH and cor- ticosterone plasma levels at various time-points of the circadian cycle (Barden et al. 1997 ; Karanth et al. 1997).

HPA axis dysregulation in these mice is only apparent under stressful conditions (Barden et al. 1997 ; Froger et al. 2004) and can be reversed by antidepressant treat- ment (Barden, 1996 ; Montkowski et al. 1995).

Our study aimed at further characterizing the depression-like syndrome of these mice by analysing both their hippocampal neurogenesis capacity and their despair-like behaviour in the tail suspension test (TST), in comparison with paired wild-type (WT) controls. In addition, the antidepressant-like activity of the new compound, agomelatine, a potent agonist at melatonin MT

and MT

receptors (Audinot et al. 2003) and antagonist at the 5-HT

receptor (Millan et al.

2003), was also analysed in comparison with the SSRI

antidepressant fluoxetine, on hippocampal neurogen- esis, BDNF and GR mRNA expression and behaviour in GR-i and WT mice.

Materials and methods Animals

Transgenic mice (strain B6C3F1 ; line 1.3) (Pepin et al.

1992), heterozygous for the presence of the transgene- expressing rat GR antisense RNA, were mated together. Mice homozygous for the presence of the transgene (GR-i) or its absence (WT) were bred at UMR677 INSERM/UPMC animal facility, and housed under standard conditions [22 ¡ 1 x C, 60 % relative humidity, 12-h light/dark cycle (lights on 07:00 hours), food and water available ad libitum]. Behavioural and anatomical studies were performed on 10-wk-old male mice. Genotype was determined by PCR analysis as previously described (Froger et al. 2004).

Procedures involving animals and their care were conducted in accordance with directives of the European Community (Council directive 86/609) and the French Agriculture and Forestry Ministry (Council directive 87-848, October 19, 1987, permissions 75-116 to M.H. and 75-977 to L.L.).

Antidepressant treatments

On the basis of previous data in validated animal models of depression and anxiety (Bourin et al. 2004 ; Cryan & Lucki, 2000 ; Millan et al. 2005 ; Papp et al.

2003 ; Santarelli et al. 2003), mice received either acutely or chronically for 21 d, at 17:00 hours, i.e. 2 h before the onset of the dark period, agomelatine [50 mg/kg body weight intraperitoneally (i.p.)] dis- solved in 1 % hydroxy-ethyl-cellulose (HEC), or flu- oxetine (10 or 30 mg/kg i.p. for acute administration, and 10 mg/kg i.p. for chronic treatment) dissolved in saline (0.9 % NaCl). Control mice received vehicle only under identical conditions.

Immunohistochemistry

Immunoperoxidase staining

In order to quantify hippocampal cell proliferation, animals received a single i.p. injection of 5-bromo-2- deoxyuridine (BrdU, Sigma-Aldrich, France; 75 mg/kg dissolved in saline, 10 mg/ml) 30 min (acute con- dition) or 16 h (chronic condition) after injection of agomelatine, fluoxetine or their vehicles, and were sacrificed 2 h later, i.e. at 07:30 hours (acute condition) or 11:00 hours (i.e. 18 h after the last injection for chronic treatment). For assessing cell survival and differentiation, BrdU (75 mg/kg i.p.) was injected daily

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

during the first week of the 21-d antidepressant treat- ment, and mice were sacrificed 4 wk later (Malberg et al. 2000), i.e. after a 2-wk washout period of anti- depressant treatments.

At the selected appropriate time, mice were deeply anaesthetized with pentobarbital (60 mg/kg i.p.), then transcardially perfused with 75 ml of 0.15

NaCl supplemented with 15 m

NaNO

followed by 75 ml of 4 % paraformaldehyde in 0.1

phosphate buffered saline (PBS, pH 7.4) at 4 x C. Brains were removed and post-fixed in the same paraformaldehyde-PBS solution overnight at 4 x C. Serial coronal sections (40 mm) were cut throughout the entire hippocampus using a vibratome (VT1000S, Leica, Germany).

Immunohistochemistry for the immature neuron marker doublecortin (DCX) or for BrdU was then performed. Free-floating sections were rinsed in 0.1

PBS, incubated for 5 min in 3 % H

O

/10 % methanol in 0.1

PBS and then rinsed. For BrdU staining only, sections were incubated in 2

HCl+0.1 % Triton X-100-supplemented 0.1

PBS (30 min, 37 x C), and then immersed in 0.1

sodium tetraborate buffer (pH 8.6), for 30 min. Sections were then incubated for 1 h in 0.1

PBS supplemented with 5 % normal rabbit serum and 0.1 % Triton X-100 for BrdU staining, or in 3 % bovine serum albumin/0.3 % Triton X-100 in 0.1

PBS for DCX immunohistochemistry. Sections were sub- sequently transferred to the same but fresh mixture supplemented with the primary anti-BrdU-antibody (monoclonal-rat IgG, 1 : 100, Abcys, France) or anti- DCX-antibody (polyclonal-goat, 1 : 100, Santa Cruz Biotechnology, USA) for an overnight incubation at 4 x C. They were then exposed to the secondary anti- body (biotinylated rabbit anti-rat IgG for BrdU or biotinylated anti-goat for DCX, 1 : 200, Vector Labora- tories, France) for 2 h at room temperature, and stained using ABC staining system (Vectastain ABC Elite kit, Vector Laboratories). Peroxidase activity was revealed by incubating sections with 0.05 % 3,3k- diaminobenzidine and 0.01 % H

O

in 0.1

PBS for 5 min. After several rinses, sections were counter- stained with Cresyl Violet, mounted, dehydrated, and coverslipped in Eukit medium.

Quantitative analyses

Cell counts were performed on blind-coded slides examined under 10r and 20r objectives (Leica). DG in the dorsal (bregma x0.94 to x2.30 mm) and ventral (bregma x2.30 to x3.80 mm) extensions of the hippo- campus (Franklin & Paxinos, 1997) was analysed in every fourth section (160 mm apart) by counting all BrdU-immunopositive cells in the SGZ defined as a

two-cell bodies-wide zone of the hilus along the base of the granule cell layer (GCL), using a computer image analysis system (Mercator, Explora Nova, France) and a mounted video 3CCD camera (Sony, France). Results are presented as the total number of BrdU-immunopositive cells per hippocampus.

For the counting of DCX-positive cell numbers, three sections containing the DG were used per mouse.

Sections were taken from one-in-four series from the hippocampus, excluding the most rostral and caudal sections. The selected sections were highly compar- able between the different mice regarding their rostrocaudal location. Cells were counted throughout the entire thickness of the section, using 40r magni- fication. The average number of immunopositive cells/section was calculated for each mouse. DG volume was measured according to the Cavalieri principle (West & Gundersen, 1990).

Immunofluorescent labelling

Triple immunofluorescent labelling for BrdU, neuron- specific nuclear protein (NeuN), and glial fibrillary acidic protein (GFAP) was performed and analysed by confocal microscopy using rat anti-BrdU (1 : 100, Abcys), rabbit anti-GFAP (1 : 300, Dako, France), and mouse anti-NeuN (1 : 500, Chemicon International, Millipore, France) antibodies. The secondary anti- bodies used were a biotinylated rabbit anti-rat anti- body (1 : 200 ; Vector Laboratories) coupled with streptavidin-Alexa Fluor 488 (1 : 1600 ; Molecular Probes, USA), an Alexa Fluor 568-conjugated goat anti-rabbit IgG (1 : 1600 ; Molecular Probes) and a cy5-conjugated donkey anti-mouse IgG (1 : 1600 ; Jackson ImmunoResearch, USA), respectively. Cover- slips were finally mounted in Fluoromount-G solution (Clinisciences, France). For phenotype determination, 100 BrdU-positive cells were analysed for the co- expression of BrdU and NeuN (neurons) or BrdU and GFAP (astrocytes) in each mouse, and means ¡

.

.

. were calculated from 5–6 mice per group.

Cell death determination

Apoptotic cells in adjacent sections were detected using the terminal deoxynucleotidyltransferase (TdT)- mediated deoxyuridine triphosphate (dUTP) nick-end labelling (TUNEL) technique adapted from Bauer et al.

(2003). After 30-min preincubation in 0.1 % sodium citrate/0.1 % Triton X-100/PBS at 4 x C, for the block- ade of endogenous peroxidases by 3 % H

O

/PBS, sections were incubated with TUNEL dilution buffer for 15 min followed by the TdT reaction with biotiny- lated dUTP in a humid chamber for 1 h at 37 x C (Roche

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

Diagnostic, France). Sections were then incubated with the Vector ABC-Elite kit (Abcys), and peroxidase immunostaining was used to detect TUNEL-labelled cells. To obtain a positive control, some sections were treated for 15 min at 37 x C with 20 U/ml DNAse I (New England Biolabs, France) diluted in DNAse buffer prior to the enzymatic reaction. As a negative control, TdT was omitted in the TUNEL reaction mix added to some of the samples. For each mouse, the average number of immunopositive cells/section was calculated.

Real-time reverse transcription–PCR determinations of GR and BDNF mRNAs

Mice were killed 30 min after a single injection (acute experiments) or 16 h after the last injection for the 3-wk treatment with agomelatine, fluoxetine or their respective vehicle, and the hippocampus was rapidly dissected and frozen in liquid nitrogen. Total RNA was extracted using the ‘RNeasy Mini kit’ (Qiagen, France). RNA quality was assessed by capillary electrophoresis using RNA Labchip kits and Agilent Bioanalyser 2100 (Agilent Technologies, France).

Aliquots (900 ng) of each sample were used for reverse transcription (iScript kit ; Bio-Rad, USA). For the GR-encoding mRNA, a pair of primers was designed to selectively amplify a 93-bp fragment spanning the 1234–1326 region of the published mouse sequence (NR3C1, NM008173) while for BDNF-encoding mRNA, we used a primer set designed to amplifly an amplicon common to all the described splice variants of BDNF transcript (MMBDNFA) and corresponding to the fragments 697-782 of NM008361, 330-410 of NM008360, 197-279 of NM031168, 463-545 of NM010548, 548-624 of NM013693 (Husson et al. 2005).

b

-microglobulin was chosen as housekeeping gene and the primer set corresponding to nucleotides 99–185 of the published sequence of the mouse b

-microglobulin mRNA (MMB2MR) was used. Real- time PCR was set up on a myIQ

apparatus (Bio-Rad) using SYBR Green-containing supermix from Bio-Rad, for 50 cycles of a three-step procedure, including 20 s denaturation at 96 x C, 20 s annealing at 62 x C, and 20 s extension at 72 x C. Semi-quantitative determinations based on the difference in specific ratios (GR or BDNF/b

-microglobulin) between groups were calcu- lated for each experimental series. Specific products were determined as clear single peaks at their melt curves. All sample measurements were performed in duplicate. Gene expression in each sample was normalized to b

-microglobulin expression, i.e. : 2r (Ct

xCt

).

Behavioural studies

Mice were tested either 30 min (acute condition) or 16–20 h after the last injection (chronic treatment) of agomelatine, fluoxetine or their respective vehicle.

TST was performed using the ID-Tech-Bioseb (France) computerized device following a procedure pre- viously described (Renoir et al. 2008). The total duration of immobility during a 6-min test was deter- mined for each mouse.

Statistical analyses

Data (presented as means ¡

.

.

.) were analysed using unpaired two-tailed Student’s t test for com- parison of two groups with a Welsh’s correction when necessary (Prism 4.0, GraphPad software, USA). A two-way analysis of variance (ANOVA) followed by Bonferroni’s post-hoc test to determine specific group differences, was also used when necessary. In all cases, the significance level was set at pf0.05.

Results

Effects of impaired GR function on hippocampal cell proliferation, neurogenesis and cell death

Basal conditions

Cell proliferation. In GR-i and WT mice, BrdU-labelled cells located in the SGZ (Fig. 1a) showed irregularly shaped nuclei and diffuse patterns of BrdU staining and were frequently associated in clusters (Fig. 1b) as expected for proliferating neural stem cells (Malberg

& Duman, 2003). The total number of DG BrdU- positive cells was significantly lower in naive trans- genic GR-i compared to WT mice (about x40 %, p<0.001, Student’s t test), and this difference was noted also in the dorsal (about x35 %, p<0.001, Student’s t test) and ventral (about x50 %, p<0.01, Student’s t test) parts of the hippocampus analysed separately (Fig. 1 c). In contrast, DG volume estimate did not reveal any significant difference between GR-i and WT mice (Fig. 1d).

In GR-i and WT mice, age had a significant negative effect on cell proliferation (F

=22.84, p<0.0001), but the significant difference in BrdU labelling between GR-i and WT mice was maintained whatever the age of the animals from 5- to 10-wk-old (F

=36.57, p<0.0001) (data not shown).

DCX labelling. Study of growth and maturation of newly born cells in the DG, using DCX labelling (Rao

& Shetty, 2004), showed that the number of immature,

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

DCX-immunopositive neurons was significantly lower (about x30 %, p<0.05, Student’s t test) in GR-i com- pared to WT naive mice (Fig. 2 a–c).

Cell death. TUNEL staining was quite low in basal conditions in GR-i and WT mice (Fig. 3 a). As shown in Fig. 3 b, no significant difference in TUNEL staining in the hippocampus was observed between control and transgenic mice.

Effects of acute antidepressant treatments

Under acute conditions, neither agomelatine (50 mg/

kg i.p.) (F

=0.05, p>0.05) nor fluoxetine (30 mg/kg i.p.) (F

=2.51, p>0.05) significantly modified the total number of BrdU-labelled cells in the SGZ of the whole hippocampus, in GR-i and WT mice (Fig. 4a, b).

Furthermore, regional analysis did not reveal any changes in the number of BrdU-positive cells, in the dorsal (agomelatine : F

=0.05, p>0.05 ; fluoxetine : F

=10.57, p>0.05) or the ventral (agomelatine : F

=0.59, p>0.05 ; fluoxetine : F

=3.15, p>0.05) parts of the hippocampus after acute treatment with

antidepressant vs. vehicle in GR-i and WT mice (Fig. 4a, b).

Effects of chronic antidepressant treatments

Cell proliferation

A 3-wk treatment with agomelatine (50 mg/kg.d i.p.) differentially promoted cell proliferation in the whole (F

=40.42, p<0.001), the dorsal (F

=6.20, p<0.05) and the ventral (F

=21.95, p<0.001) parts of the hippocampus. In WT mice, agomelatine significantly increased the number of BrdU-labelled cells only in the whole DG (about +20 %, p<0.001, Bonferroni’s post- hoc test) and its ventral part (about +35 %, p<0.01, Bonferroni’s post-hoc test) whereas in GR-i mice, this effect was significant in the whole DG (about +26 %, p<0.001, Bonferroni’s post-hoc test), as well as the dorsal (about +24 %, p<0.05, Bonferroni’s post-hoc test) and ventral (about +34 %, p<0.05, Bonferroni’s post-hoc test) extensions (Fig. 5a). However, no hippo- campal volume modifications were associated with these effects (Fig. 5b).

SGZ GCL H

WT GR-i

Dorsal

Ventral (a)

(c)

50 µm

1000 800 600 400 200 0

Total

Number of BrdU-labelled cells per hippocampus

WT GR-i

***

600

400

200

0

Dorsal

WT GR-i

***

600

400

200

0

Ventral

WT GR-i

**

(d)

1.5

1.0

0.5

0.0

WT GR-i Dentate gyrus volume (mm3)

(b)

50 µm

Fig. 1.Hippocampal cell proliferation in glucocorticoid receptor-impaired (GR-i) and wild-type (WT) mice under basal conditions. (a) Photomicrographs showing the distribution of BrdU-positive cells in the dentate gyrus of the dorsal and ventral parts of the hippocampus in a WT and a GR-i mouse. Immunolabelled cells are located in the subgranular zone (SGZ) at the border between the granule cell layer (GCL) and the hilus (H) in the dentate gyrus. (b) BrdU-labelled cells are most often associated as clusters. (c) Numbers of BrdU-labelled cells counted in the entire dentate gyrus and its dorsal and ventral parts in GR-i and paired WT mice. The numbers of BrdU-labelled cells (per hippocampus) are significantly lower in GR-i mice compared to their WT controls. **p<0.01, ***p<0.001, mean¡_S.E.M. (n=8 mice per group), Student’sttest. (d) Dentate gyrus volume (in mm³) in GR-ivs. WT mice, mean¡S.^E.^M. (n=8 mice per group).

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

A 3-wk treatment with fluoxetine (10 mg/kg.d i.p.) also induced a significant increase in the number of BrdU-labelled cells in the whole DG (F

=5.30, p<0.05) but only in GR-i mice (about +60 % over saline-treated transgenic mice, p<0.001, Bonferroni’s post-hoc test). Regional analysis showed that this effect also reached statistical significance in the dorsal (about +54 %, p<0.05, Bonferroni’s post-hoc test) as well as the ventral (about +64 %, p<0.01, Bonferroni’s post- hoc test) parts of the hippocampus in GR-i but not WT mice (Fig. 5 c). On the other hand, fluoxetine treatment did not modify the hippocampal volume in GR-i and WT mice (Fig. 5d).

Cell survival

A significant effect of genotype was observed on the survival of newly generated cells in the hippocampus.

Indeed, in the HEC- and saline-treated groups, the number of labelled cells 4 wk after BrdU injections was lower in GR-i mice compared to WT mice

[whole DG : F

=7.40, p<0.05 ; dorsal DG : F

=4.36, p<0.05 ; ventral DG : F

=10.04, p<0.01, after HEC treatment (Fig. 6a) ; whole DG : F

=11.99, p<0.05 ; dorsal DG : F

=9.76, p<0.05 ; ventral DG : F

=9.94, p<0.05, after saline treatment (Fig. 6b)].

A 3-wk treatment with agomelatine significantly increased the number of BrdU-positive cells in the ventral DG of GR-i mice (about +55 %, p<0.05, Bonferroni’s post-hoc test) but not WT mice (Fig. 6a).

Conversely, no significant changes were noted in its dorsal extension and the whole DG (Fig. 6a) and a 3-wk treatment with fluoxetine (10 mg/kg i.p.) was ineffective in the dorsal, ventral and whole DG in GR-i and WT mice (Fig. 6b). Furthermore, the treatments did not modify hippocampal DG volumes in GR-i and WT mice (not shown).

Phenotype of surviving cells

To determine the phenotype of BrdU-positive sur- viving cells, triple immunofluorescent labelling was performed using markers for mature neurons (NeuN) and astrocytes (GFAP) in addition to BrdU (Fig. 6c).

Double BrdU- and NeuN- [Fig. 6c(i)] or BrdU- and GFAP- [Fig. 6c(ii)] labelled cells were counted throughout the DG. Ratios of BrdU-NeuN/BrdU- GFAP double-labelled cells were similar in the dorsal and ventral parts of the hippocampus, and therefore corresponding values were pooled for calculations for the whole hippocampus. Most of the surviving DG BrdU-positive cells were found to also express NeuN (of a total of 100 BrdU-immunoreactive cells counted per animal : HEC=88.0 ¡ 2.3 and 85.3 ¡ 1.8 ;

(b) (c)

(a)

200 150 100 50 Number of DCX-positive cells/section 0

WT GR-i

GR-i

* 50 µm WT

SGZ H GCL

50 µm

Fig. 2.Immunostaining of DCX in the hippocampus of glucocorticoid receptor-impaired (GR-i) and wild-type (WT) mice under basal conditions. (a) Low-magnification photomicrographs of doublecortin (DCX) immunoreactivity in the dentate gyrus in a WT and a GR-i mouse. Most cell bodies immunopositive for DCX are located in the subgranular zone (SGZ) of the dentate gyrus. (b) High- magnification photomicrograph of DCX-immunopositive cells in the granule cell layer (GCL) with extended ramified dendrites. (c) Numbers of DCX-immunopositive cells/section in GR-ivs. WT mice. Each bar is the mean¡_S.E.M. of data obtained in three mice per group. *p<0.05, Student’s ttest.

(b) (a)

1.5

0.5

0.0 Number of TUNEL positive nuclei/section

WT GR-i

50 µm

Fig. 3.TUNEL labelling in the hippocampus of glucocorticoid receptor-impaired (GR-i) and wild-type (WT) mice under basal conditions. (a) Low-magnification photomicrograph of a rare TUNEL staining (arrow) in the dentate gyrus of a GR-i mouse. The upper-right box illustrates positive control of TUNEL labelling after DNAse I treatment. (b) Numbers of TUNEL-positive nuclei/section in GR-ivs. WT mice. Each bar is the mean¡_S.E.M. of data obtained in 5–6 mice per group.

No significant difference, Student’sttest.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

saline=91.5 ¡ 4.7 and 89.5 ¡ 1.0, in WT and GR-i mice, respectively, means ¡

.

.

., n=5–6), and only a few of them (10–15 %) expressed GFAP. Chronic treatment with agomelatine or fluoxetine did not affect the dif- ferentiation of newly generated cells towards neuronal or glial phenotype (double BrdU-NeuN immuno- reactive cells : agomelatine=89.5 ¡ 1.2 and 89.5 ¡ 0.8 ; fluoxetine : 88.1 ¡ 1.1 and 90.5 ¡ 1.6, in WT and GR-i mice, respectively, means ¡

.

.

., n=5–6 ; not signifi- cantly different from data above, for vehicle-treated mice).

Hippocampal expression of GR and BDNF mRNAs in naive mice

As expected, hippocampal GR mRNA expression was significantly down-regulated in naive GR-i com- pared to WT mice (x21.0 ¡ 7.8 %, mean % ¡

.

.

., n=6, p<0.05, two-tailed Student’s t test). In addition, hippocampal levels of BDNF mRNA were signifi- cantly lower in naive GR-i compared to WT mice (x41.0 ¡ 2.2 %, mean % ¡

.

.

., n=6, p<0.001, two- tailed Student’s t test) (Fig. 7a, b).

Effects of chronic antidepressant treatments

GR mRNA expression

As shown in Fig. 8a, chronic agomelatine adminis- tration exerted a significant effect on GR mRNA ex- pression since down-expression of this transcript was reduced from x45 % in vehicle (HEC) to only x21 % in agomelatine-treated GR-i compared to paired WT mice (Bonferroni’s post-hoc test, p<0.001) [genotype : F

=623.1 ; agomelatine : F

=82.10 ; interaction : F

=82.10]. Conversely, fluoxetine treatment did not significantly modify hippocampal GR mRNA levels in GR-i mice (Fig. 8b) [genotype : F

=184.1 ; fluoxetine : F

=20.24 ; interaction : F

=20.24]. Furthermore, neither fluoxetine nor agomelatine significantly modi- fied GR mRNA levels in the hippocampus of WT mice (not shown).

BDNF mRNA expression

Chronic administration of agomelatine (Fig. 8c) as well as fluoxetine (Fig. 8 d) reversed the down-regulation of BDNF mRNA in the hippocampus of GR-i mice,

1000 800 600 400 200 0

Total Dorsal

Number of BrdU-labelled cells per hippocampus

* GR-i

WT WT GR-i

HEC 1%Ago 50 HEC 1%Ago 50 600 400 200 0

HEC 1%Ago 50 HEC 1%Ago 50

*

Ventral GR-i 600 WT

400 200 0

HEC 1%Ago 50 HEC 1%Ago 50

*

(b)

1000 800 600 400 200 0

Total

Number of BrdU-labelled cells per hippocampus

GR-i WT

Saline

Fluox 30 Saline Fluox 30

*

600 400 200 0

Dorsal GR-i WT

Saline

Fluox 30 Saline Fluox 30

*

600 400 200 0

Ventral GR-i WT

Saline

Fluox 30 Saline Fluox 30

*

Fig. 4.Effects of acute treatment with agomelatine (a) or fluoxetine (b) on hippocampal cell proliferation in glucocorticoid receptor-impaired (GR-i) and wild-type (WT) mice. (a,b) BrdU-labelled cells were counted within the entire dentate gyrus of the hippocampus and its dorsal and ventral parts after a single injection of either agomelatine (50 mg/kg i.p.), or fluoxetine (30 mg/kg i.p.) or their vehicles [hydroxy-ethyl-cellulose (HEC), saline, respectively]. The number of BrdU-labelled cells in each sub-area of the dentate gyrus is expressed per hippocampus. No significant proliferative effect of either drug was observed in the dorsal or the ventral part of the hippocampus in GR-i and WT mice. Each bar is the mean¡S.^E.^M. of 3–5 independent determinations. *p<0.05, GR-ivs. WT mice, Bonferroni’spost-hoctest.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

so that its levels in agomelatine-treated (p<0.05, Bonferroni’s post-hoc test) [genotype : F

=0.12 ; ago- melatine : F

=27.15 ; interaction : F

=27.15] and fluoxetine-treated (p<0.001, Bonferroni’s post-hoc test) [genotype : F

=2.78 ; fluoxetine : F

=78.99 ; inter- action : F

=78.99] treated GR-i mice no longer dif- fered from those in paired vehicle-treated WT mice. In contrast, no significant effects of either agomelatine or fluoxetine on hippocampal BDNF mRNA levels were observed in WT mice (not shown).

Additional experiments demonstrated that acute administration of agomelatine or fluoxetine did not modify the levels of GR and BDNF mRNA in GR-i mice (Table 1).

Behaviour in the TST

Basal conditions

Under basal conditions, i.e. without any treatment, naive GR-i mice showed a significantly longer

immobility time compared to WT mice (p<0.001, Student’s t test) during the 6-min observation period of the TST (Fig. 9 a).

Effects of antidepressant treatments

Acute treatment

Acute agomelatine (50 mg/kg i.p.) had no effect on immobility duration in GR-i and WT mice. In contrast, a two-way ANOVA revealed a significant effect of acute fluoxetine (F

=6.84, p<0.01). In WT mice, 30 mg/kg i.p. fluoxetine led to a significant decrease of immobility time (p<0.01, Bonferroni’s post-hoc test), whereas, in GR-i mice, fluoxetine at both 10 and 30 mg/kg i.p. induced this effect (p<0.01 and p<0.05, respectively, Bonferroni’s post-hoc test) (Fig. 9b).

Chronic treatment

A two-way ANOVA analysis showed a significant effect of chronic agomelatine treatment (50 mg/kg.d

(a) (b)

1000 800 600 400 200 0

Total

Number of BrdU-labelled cells per hippocampus

GR-i WT

HEC 1%Ago 50 HEC 1%Ago 50

##

###

*** ***

600 400 200 0

Dorsal GR-i WT

HEC 1%Ago 50 HEC 1%Ago 50

#

* 600

400 200 0

Ventral GR-i WT

HEC 1%Ago 50 HEC 1%Ago 50

##

#

2.0 1.5 1.0 0.5 0.0

Total GR-i WT

HEC 1%Ago 50 HEC 1%Ago 50 Dentate gyrus volume (mm3)Dentate gyrus volume (mm3)

(c) (d)

1000 800 600 400 200 0

Total

Number of BrdU-labelled cells per hippocampus

GR-i WT

Saline Fluox 10

Saline Fluox 10

Saline Fluox 10

Saline Fluox 10

***

### 600

400 200 0

Dorsal GR-i WT

#

***

Saline Fluox 10

Saline Fluox 10 600

400 200 0

Ventral GR-i WT

##

*

Saline Fluox 10

Saline Fluox 10 2.0

1.5 1.0 0.5 0.0

Total WT GR-i

Fig. 5.Effects of chronic treatment with agomelatine (a,b) or fluoxetine (c,d) on hippocampal cell proliferation in glucocorticoid receptor-impaired (GR-i) and wild-type (WT) mice. (a,c) BrdU-labelled cells were counted within the entire dentate gyrus (DG) of the hippocampus and its dorsal and ventral parts after 3-wk treatment with agomelatine (50 mg/kg i.p. daily), fluoxetine (10 mg/kg i.p. daily) or their vehicles [hydroxy-ethyl-cellulose (HEC), saline, respectively]. The number of BrdU-labelled cells in each sub-area of the DG is expressed per hippocampus. The proliferative effect of agomelatine was significant in the total DG and its dorsal and ventral parts in GR-i mice, but only in the total and ventral DG in WT mice. The proliferative effect of fluoxetine was significant in the total DG, and its dorsal and ventral parts in GR-i mice. Conversely, fluoxetine did not induce cell proliferation in the DG of WT mice. (b,d) Volume of the DG, in mm³. Volumes were determined by multiplying the sum of the DG area per section by the distance between sections (160mm). Each bar is the mean¡S.^E.^M. of 5–7 independent determinations.

*p<0.05, ***p<0.001, GR-ivs. WT mice, Bonferroni’spost-hoctest.^#p<0.05,^##p<0.01,^###p<0.001 agomelatinevs. HEC or fluoxetinevs. saline, Bonferroni’spost-hoctest.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

i.p., 21 d) (F

=6.871, p<0.05) with a significantly reduced immobility time in agomelatine-treated com- pared to HEC-treated GR-i mice (p<0.05, Bonferroni’s post-hoc test). However, the treatment had no effect in WT mice (Fig. 9 c). Compared to saline, chronic fluo- xetine treatment (10 mg/kg.d i.p., 21 d) did not affect immobility time in GR-i or WT mice (Fig. 9 c).

Discussion

Several recent studies have emphasized the role that HPA axis dysfunction may play in the physiopatho- logy of MDD and associated alterations in neuroplas- ticity and neurogenesis. In this context, the aim of our study was to further investigate the effects of the novel

1600 2000

1200 800 400 0

Total

Number of BrdU-labelled cells per hippocampus

GR-i WT

HEC 1%Ago 50 HEC 1%Ago 50

*

800 1000

600 400 200 0

Dorsal GR-i WT

HEC 1%Ago 50 HEC 1%Ago 50

*

800 1000

600 400 200 0

Ventral GR-i WT

HEC 1%Ago 50 HEC 1%Ago 50

**

#

(b)

(c) 1600 2000

1200 800 400 0

Total

Number of BrdU-labelled cells per hippocampus

GR-i WT

Saline

Fluox 10 Saline

Fluox 10 Saline

Fluox 10 Saline

Fluox 10 Saline

Fluox 10 Saline Fluox 10

*

800 1000

600 400 200 0

Dorsal GR-i WT

*

800 1000

600 400 200 0

Ventral WT GR-i

*

BRDU NeuN GFAP Merge _{20 µm}

20 µm

(i)

(ii)

Fig. 6.Effects of chronic treatment with (a) agomelatine or (b) fluoxetine on hippocampal cell survival and (c) differentiation in glucocorticoid receptor-impaired (GR-i) and wild-type (WT) mice. (a,b) BrdU-labelled cells were counted within the entire dentate gyrus (DG) and its dorsal and ventral parts in mice sacrificed 4 wk after the last injection of BrdU. The number of BrdU-labelled cells in each sub-area of the DG is expressed per hippocampus. (a) Agomelatine exerted a significant effect on cell survival in the ventral part of the DG in GR-i mice. (b) Fluoxetine did not exert a significant effect on cell survival in GR-i or WT mice. Each bar is the mean¡S.^E.^M. of 4–5 independent determinations. *p<0.05, **p<0.01, GR-ivs. WT mice, Bonferroni’spost-hoctest.^#p<0.05, agomelatinevs. hydroxy-ethyl-cellulose (HEC), Bonferroni’spost-hoctest. (c) Confocal microscopy illustration of BrdU (green)-, neuron-specific nuclear protein (NeuN) (blue)-, and glial fibrillary acidic protein (GFAP) (red)-labelled cells in the granule cell layer after 3-wk treatment with agomelatine. Arrows point at cells double-labelled (Merge) for BrdU and NeuN (i, top row of micrographs) or BrdU and GFAP (ii, bottom row of micrographs).

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

antidepressant agomelatine (Olie & Kasper, 2007), compared to fluoxetine, in the GR-i transgenic mouse, a murine model of depression with marked alterations in HPA axis function due to GR down-regulation in brain (Froger et al. 2004 ; Pepin et al. 1992).

Molecular, anatomical and behavioural characteristics of GR-i mice

As previously reported (Froger et al. 2004 ; Holsboer &

Barden, 1996 ; Pepin et al. 1992), genome incorporation of the transgene encoding GR antisense RNA down- regulated GR expression in GR-i mice. In addition, we also found that BDNF mRNA expression was down- regulated in the hippocampus of GR-i compared to WT mice. This finding is of particular relevance be- cause alterations in neurotrophic factors like hippo- campal BDNF have been frequently associated with depression-like symptoms in several animal models (Alfonso et al. 2006), and with MDD in humans (Karege et al. 2002). Neurotrophic factors are known to exert a key role not only in cell differentiation, but also in cell proliferation and survival at adult age (Lee et al.

2002 ; Sairanen et al. 2005) and the marked reduction of cell proliferation and survival together with the re- duced neurogenesis assessed by DCX expression that we observed within the dorsal and ventral parts of the DG might be causally related to the down-regulation of BDNF in GR-i mice. Interestingly, the decreased neurogenesis observed in GR-i mice was not linked to any detectable increase in apoptotic processes.

The dorsal hippocampus has been functionally associated with learning and memory whereas its ventral part is mostly concerned with fear-related behaviours and anxiety (Bannerman et al. 2004).

Interestingly, GR-i mice have been shown to display numerous abnormalities regarding spatial learning and short-term memory (Rousse et al. 1997), anxiety (Montkowski et al. 1995) and responses to stress (Froger et al. 2004 ; Linthorst et al. 2000). All these alterations might theoretically be linked to the deficits observed in hippocampal cell proliferation and sur- vival, probably causally correlated with the decreased BDNF expression.

Along with a deficit in hippocampal neurogenesis, GR-i mice exhibited a significant increase in immo- bility time in TST, indicating clear-cut despair-like behaviour similar to that reported in other animal models of affective disorders (Cryan et al. 2005).

Interestingly, these data differed from those obtained in the forced swim test (FST ; Barden et al. 2005), where GR-i mice appeared to be less immobile than paired WT controls. However, it should be emphasized that the latter test was shown to be less informative in mice than in rats (Borsini & Meli, 1988), and also to reflect some adaptive strategy unrelated to despair behaviour (Lucki, 2001 ; Thierry et al. 1984). Furthermore, TST and FST probably activate distinct neurochemical pathways underlying animal performance (Bai et al.

2001) and this might also account for the differences observed here between GR-i and WT mice.

Effects of antidepressant treatments

When given acutely, neither agomelatine nor fluox- etine enhanced cell proliferation in the hippocampal SGZ of GR-i and WT mice. These results are consistent with previous studies showing that chronic adminis- tration of antidepressants is necessary to increase cell proliferation and neurogenesis in the hippocampus (Banasr et al. 2006 ; Malberg et al. 2000 ; Santarelli et al.

2003).

However, fluoxetine at the acute dose of 30 mg/kg i.p. was effective in shortening immobility in the TST in GR-i and WT mice, in agreement with previous observations on other mouse strains (Cryan et al. 2005).

In contrast to fluoxetine, acute agomelatine did not affect immobility in the TST as also reported for the FST in Swiss male mice (Bourin et al. 2004).

Under chronic treatment conditions, agomelatine enhanced cell proliferation only in the ventral part of the hippocampus in WT mice. However, it should be noted that the number of proliferating cells in this re- gion was slightly lower in WT mice injected with HEC,

100

50

0 Relative gene expression level (GR/β2-microglobulin ratio, % WT)

*

WT GR-i (b)

100

50

0 Relative gene expression level (BDNF/β2-microglobulin ratio, % WT)

***

WT GR-i Fig. 7.Real-time RT–PCR quantification of glucocorticoid receptor (GR) and BDNF mRNA in the hippocampus of glucocorticoid receptor-impaired (GR-i)vs. wild-type (WT) mice. Bars represent (a) mean GR mRNA/b2-microglobulin mRNA or (b) BDNF mRNA/b2-microglobulin mRNA ratios¡_S.E.M. of 3–4 independent determinations expressed as percent of WT mice. GR and BDNF mRNA expressions were significantly down-regulated in GR-i compared to WT mice.

*p<0.05, ***p<0.001, Student’sttest. Range ofCtvalues were respectively 20.21–23.59 for GR, 20.55–24.90 for BDNF and 18.83–21.73 forb₂-microglobulin.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

the agomelatine vehicle, than in naive mice. Whether HEC, by inducing an unspecific stress, could reduce cell proliferation and whether agomelatine could re- verse this effect is an intriguing point to be considered.

However, the regional effect of agomelatine, that matches previous data in Wistar rats (Banasr et al.

2006), further emphasizes the critical implication of the ventral hippocampus in stress and response to antidepressants. In contrast, chronic fluoxetine was found to affect cell proliferation in WT mice in neither the dorsal nor the ventral part of the hippocampus.

These data were unexpected because earlier studies

brain-derived neurotrophic factor (BDNF) mRNAs in the hippocampus of GR-i mice – effects of acute agomelatine and fluoxetine treatments

Hydroxy-ethyl-

cellulose Agomelatine Saline Fluoxetine

GR 104.50¡20.37 117.80¡3.70 101.80¡13.62 99.19¡46.64 BDNF 102.00¡11.37 108.40¡1.89 102.10¡15.01 99.51¡11.79

Mean GR and BDNF mRNA/b₂-microglobulin mRNA ratios¡S.^E.^M. of three independent determinations.

(a) (b)

(c) (d)

125 150

100 75 50 25 Relative gene expression level (GR/β-microglobulin ratio, % WT)2 0

WT GR-i WT GR-i

HEC Agomelatine 125 150

100 75 50 25 0

WT GR-i WT GR-i Saline Fluoxetine

***

150 100 50 Relative gene expression level (BDNF/β-microglobulin ratio, % WT)2 0

WT GR-i WT GR-i HEC Agomelatine

* 150

100 50 0

WT GR-i WT GR-i Saline Fluoxetine

***

Fig. 8.Real-time RT–PCR quantification of glucocorticoid receptor (GR) and BDNF mRNA in the hippocampus of GR-impaired (GR-i) and wild-type (WT) mice – effects of chronic agomelatine and fluoxetine treatments. Effects of (a) agomelatine and (b) fluoxetine treatments on GR mRNA levels. Bars represent mean GR mRNA/b2-microglobulin mRNA ratios¡S.^E.^M. of six independent determinations expressed as percent of WT mice. Chronic agomelatine – but not fluoxetine – administration significantly reduced GR mRNA down-expression in GR-i mice compared to paired WT mice. ***p<0.001 agomelatinevs.

hydroxy-ethyl-cellulose (HEC) in GR-i mice, Bonferroni’spost-hoctest. Effects of (c) agomelatine and (d) fluoxetine treatments on BDNF mRNA levels. Bars represent mean BDNF mRNA/b2-microglobulin mRNA ratios¡S.E.M. of six independent determinations expressed as percent of WT mice. *p<0.05, ***p<0.001 agomelatine or fluoxetinevs. respective vehicle in GR-i mice, Bonferroni’spost-hoctest. Chronic administration of agomelatine (p<0.05, Bonferroni’spost-hoctest) as well as fluoxetine (p<0.001, Bonferroni’spost-hoctest) reversed the down-regulation of BDNF mRNA in GR-i mice. Range ofCtvalues were respectively 20.21–23.59 for GR, 20.55–24.90 for BDNF and 18.83–21.73 forb₂-microglobulin.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

have shown a promoting effect of chronic fluoxetine on cell proliferation in WT rodents (Kodama et al.

2004 ; Malberg et al. 2000 ; Santarelli et al. 2003).

However, other reports pointed out that chronic treatment with this SSRI exerted only a minor effect on hippocampal cell proliferation, and was even ineffec- tive in unstressed BALB/CJ mice (Alonso et al. 2004 ; Holick et al. 2008). These findings suggest that, at least in WT mice, fluoxetine-induced cell proliferation might be strain dependent.

In contrast to that noted in WT mice, the neurogenic effects of agomelatine and fluoxetine were highly sig- nificant in GR-i mice as administration of either drug up-regulated cell proliferation in the SGZ of the DG in these transgenic animals when administered chroni- cally. Similar data have been obtained in other rodent models of depression with predictive validity. In par- ticular, chronic treatment with agomelatine was found to significantly reverse the deficit in hippocampal cell proliferation, up to control levels, in prenatally

250 200 150 100 50 0

Immobility (s/6 min)

WT GR-i

***

(a)

250 200 150 100 50 0

Immobility (s/6 min)

(c) 250 200 150 100 50 0

Immobility (s/6 min)

(b)

HEC 1%Ago 50 HEC 1%Ago 50

HEC 1%Ago 50 HEC 1%Ago 50

WT GR-i

WT GR-i

250 200 150 100 50 0

Saline

Fluox 10Fluox 30 Saline Fluox 10Fluox 30

WT GR-i

##

##

#

#

250 200 150 100 50 0

Saline

Fluox 10 Saline Fluox 10

WT GR-i

Fig. 9.Tail suspension test (TST) in glucocorticoid receptor-impaired (GR-i) and wild-type (WT) mice – effects of acutevs.

chronic treatments with agomelatine and fluoxetine. (a) Immobility time (in seconds per 6 min) of naive GR-i and WT mice subjected to the TST. Each bar is the mean¡_S.E.M. of 10–12 mice in each group. ***p<0.001, Student’sttest. (b) Immobility time of mice treated acutely with agomelatine (50 mg/kg i.p.) or fluoxetine (10 or 30 mg/kg i.p.) 30 min before the test. Control GR-i and WT mice received the vehicle only [1 % hydroxy-ethyl-cellulose (HEC) or 0.9 % NaCl, 10 ml/kg i.p.] under the same conditions.

Data are expressed as immobility time (min). Each bar is the mean¡_S.E.M. of 6–15 independent determinations.^#p<0.05,

##p<0.01 fluoxetinevs. saline, Bonferroni’spost-hoctest. (c) Effect of 3-wk treatment with agomelatine (50 mg/kg.d i.p.) or fluoxetine (10 mg/kg.d i.p.) on immobility time in the TST. Data are expressed as immobility time (in seconds per 6 min).

Each value is the mean¡S.^E.^M. of 9–11 independent determinations.^#p<0.05, agomelatinevs. HEC, Bonferroni’spost-hoctest.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

stressed rats (Morley-Fletcher et al. 2003). On the other hand, repeated treatment with fluoxetine reversed the reduction of hippocampal cell proliferation produced by chronic mild stress in mice (Alonso et al. 2004) and by learned helplessness in rats (Chen et al. 2006 ; Malberg & Duman, 2003).

Interestingly, in GR-i mice, not only the pro- liferation of hippocampal cells but also their survival was promoted by chronic administration of agomela- tine. In accord with previous data (Kempermann &

Kronenberg, 2003), y85–90 % of newborn cells ma- tured and became neurons, as determined by triple labelling for BrdU and neuronal or glial specific mark- ers. By prolonging the antidepressant treatment after BrdU injection, we showed that agomelatine increased the survival of newly generated neurons in the ventral part of the hippocampus of GR-i mice only, in agree- ment with previous data in Sprague–Dawley rats (Maccari et al. 2005). In contrast, chronic fluoxetine did not increase hippocampal cell survival in GR-i and WT mice, also in accord with data in rats (Malberg et al.

2000).

Another difference that appeared between the two antidepressant drugs under study was the effec- tiveness of chronic agomelatine, but not fluoxetine, in increasing GR mRNA expression in GR-i mice. Al- though a positive influence of fluoxetine on GR ex- pression has been described in vitro (Pariante et al.

2003), SSRIs appeared to have variable effects in vivo (Bjartmar et al. 2000 ; Seckl & Fink, 1992). Whether, chronic agomelatine but not fluoxetine would restore the effectiveness of GR-mediated negative feedback control of HPA axis in GR-i mice is a relevant question to be addressed in future studies.

In contrast to the differences emphasized above, both agomelatine and fluoxetine, under chronic treat- ment conditions, were found to reverse the down- regulation of hippocampal BDNF mRNA in GR-i mice, in line with previous data in other rodent models (Castren et al. 2007 ; Duman & Monteggia, 2006 ; Vinet et al. 2004). In humans, post-mortem studies also showed that the expression of BDNF was signifi- cantly higher in patients treated with antidepressant (Klempin & Kempermann, 2007). According to Castren et al. (2007), chromatin remodelling by antidepressant treatment might be causally related to BDNF up- regulation. Indeed, HPA axis activation in mice, as that occurring after stress or in GR-i mice, produced a long-lasting methylation of histone-3 subunits corre- lated with suppression of BDNF gene transcription (Tsankova et al. 2006). Chronic antidepressant treat- ment may counteract the reduction in BDNF mRNA expression by inducing the acetylation of the same

histone subunit. Interestingly, an antidepressant treatment that induced histone acetylation in mice previously subjected to social stress was without effect in non-stressed controls (Tsankova et al. 2006), in line with our findings that BDNF mRNA was up- regulated in GR-i mice but not WT mice after chronic treatment with agomelatine or fluoxetine. The possi- bility therefore exists that increased BDNF signalling is required for the induction of the behavioural re- sponse to antidepressant drugs, specifically in animals with depression-like symptoms (Castren et al. 2007 ; Saarelainen et al. 2003).

In conclusion, our data showed that chronic ago- melatine and fluoxetine treatments were able to increase hippocampal cell proliferation and BDNF mRNA levels in GR-i mice, a validated model of depression (Froger et al. 2004 ; Pepin et al. 1992).

However, whether these effects are causally related to behavioural improvement is still a question to be addressed. If, in GR-i mice, the decreased immobility in the TST after chronic agomelatine seemed to occur concomitantly with the promotion of neurogenesis, this was not the case with fluoxetine. With this SSRI, behavioural improvement was observed only under acute treatment conditions whereas induction of cell proliferation occurred under chronic treatment con- ditions. These data are consistent with those recently published about the occurrence of possible mismatch between the effects of SSRIs on neurogenesis on the one hand and behaviour on the other (Holick et al.

2008). Nevertheless, our data provided further sup- port to the idea that increase in hippocampal neuro- genesis involves mechanisms, which include the regulation of growth factors such as BDNF, that could be involved in the therapeutic effects of anti- depressants.

Acknowledgements

This research was supported by INSERM (France), the European Community NewMood (LSHM-CT-2003- 503474) Program and IRIS (Courbevoie, France). We are grateful to pharmaceutical companies (IRIS, Eli Lilly) for generous gifts of drugs. E. Paı¨zanis and T. Renoir were recipients of fellowships from the Ministe`re de l’Education Nationale et de la Recherche (France) during performance of this work. We thank Mickael Decressac (CNRS UMR 6187 Poitiers, France) for his help with DCX and TUNEL staining protocol.

Statement of Interest None.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

References

Alfonso J, Frick LR, Silberman DM, Palumbo ML,et al.

(2006). Regulation of hippocampal gene expression is conserved in two species subjected to different stressors and antidepressant treatments.Biological Psychiatry59, 244–251.

Alonso R, Griebel G, Pavone G, Stemmelin J,et al.(2004).

Blockade of CRF(1) or V(1b) receptors reverses stress-induced suppression of neurogenesis in a mouse model of depression.Molecular Psychiatry9, 278–286.

Audinot V, Mailliet F, Lahaye-Brasseur C, Bonnaud A,et al.

(2003). New selective ligands of human cloned melatonin MT1 and MT2 receptors.Naunyn Schmiedeberg’s Archives of Pharmacology367, 553–561.

Bai F, Li X, Clay M, Lindstrom T, Skolnick P(2001). Intra- and interstrain differences in models of ‘behavioral despair’.Pharmacology, Biochemistry and Behavior70, 187–192.

Banasr M, Soumier A, Hery M, Mocaer E, Daszuta A(2006).

Agomelatine, a new antidepressant, induces regional changes in hippocampal neurogenesis.Biological Psychiatry 59, 1087–1096.

Bannerman DM, Rawlins JN, McHugh SB, Deacon RM, et al.(2004). Regional dissociations within the hippocampus – memory and anxiety.Neuroscience and Biobehavioral Reviews28, 273–283.

Barden N(1996). Modulation of glucocorticoid receptor gene expression by antidepressant drugs.Pharmacopsychiatry29, 12–22.

Barden N, Shink E, Labbe M, Vacher R,et al.(2005).

Antidepressant action of agomelatine (S 20098) in a transgenic mouse model.Progress in

Neuropsychopharmacology and Biological Psychiatry29, 908–916.

Barden N, Stec IS, Montkowski A, Holsboer F, Reul JM (1997). Endocrine profile and neuroendocrine challenge tests in transgenic mice expressing antisense RNA against the glucocorticoid receptor.Neuroendocrinology66, 212–220.

Bauer S, Rasika S, Han J, Mauduit C,et al.(2003). Leukemia inhibitory factor is a key signal for injury-induced neurogenesis in the adult mouse olfactory epithelium.

Journal of Neuroscience23, 1792–1803.

Bjartmar L, Johansson IM, Marcusson J, Ross SB,et al.

(2000). Selective effects on NGFI-A, MR, GR and NGFI-B hippocampal mRNA expression after chronic treatment with different subclasses of antidepressants in the rat.

Psychopharmacology(Berlin)151, 7–12.

Borsini F, Meli A(1988). Is the forced swimming test a suitable model for revealing antidepressant activity?

Psychopharmacology(Berlin)94, 147–160.

Bourin M, Mocaer E, Porsolt R(2004). Antidepressant-like activity of S 20098 (agomelatine) in the forced swimming test in rodents : involvement of melatonin and serotonin receptors.Journal of Psychiatry and Neuroscience29, 126–133.

Castren E, Voikar V, Rantamaki T(2007). Role of neurotrophic factors in depression.Current Opinion in Pharmacology7, 18–21.

Chen H, Pandey GN, Dwivedi Y(2006). Hippocampal cell proliferation regulation by repeated stress and

antidepressants.Neuroreport17, 863–867.

Cryan JF, Lucki I(2000). Antidepressant-like behavioral effects mediated by 5-Hydroxytryptamine(2C) receptors.

Journal of Pharmacology and Experimental Therapeutics295, 1120–1126.

Cryan JF, Mombereau C, Vassout A(2005). The tail suspension test as a model for assessing antidepressant activity : review of pharmacological and genetic studies in mice.Neuroscience and Biobehavioral Reviews29, 571–625.

Czeh B, Lucassen PJ(2007). What causes the hippocampal volume decrease in depression? Are neurogenesis, glial changes and apoptosis implicated?European Archives of Psychiatry and Clinical Neuroscience257, 250–260.

Dorio D, Viau V, Meaney MJ(1993). The role of the medial prefrontal cortex (cingulate gyrus) in the regulation of hypothalamic–pituitary–adrenal responses to stress.

Journal of Neuroscience13, 3839–3847.

Duman RS, Monteggia LM(2006). A neurotrophic model for stress-related mood disorders.Biological Psychiatry59, 1116–1127.

Feldman S, Weidenfeld J(1999). Glucocorticoid receptor antagonists in the hippocampus modify the negative feedback following neural stimuli.Brain Research821, 33–37.

Franklin KB, Paxinos G(1997).The Mouse Brain in Stereotaxic Coordinates. Toronto, San Diego : Academic Press.

Froger N, Palazzo E, Boni C, Hanoun N,et al.(2004).

Neurochemical and behavioral alterations in

glucocorticoid receptor-impaired transgenic mice after chronic mild stress.Journal of Neuroscience24, 2787–2796.

Fuchs E, Flugge G(1998). Stress, glucocorticoids and structural plasticity of the hippocampus.Neuroscience and Biobehavioral Reviews23, 295–300.

Hashimoto K, Shimizu E, Iyo M(2004). Critical role of brain-derived neurotrophic factor in mood disorders.

Brain Reserach Reviews45, 104–114.

Holick KA, Lee DC, Hen R, Dulawa SC(2008). Behavioral effects of chronic fluoxetine in BALB/cJ mice do not require adult hippocampal neurogenesis or the serotonin 1A receptor.Neuropsychopharmacology33, 406–417.

Holsboer F, Barden N(1996). Antidepressants and hypothalamic–pituitary–adrenocortical regulation.

Endocrine Reviews17, 187–205.

Husson I, Rangon CM, Lelievre V, Bemelmans AP,et al.

(2005). BDNF-induced white matter neuroprotection and stage-dependent neuronal survival following a neonatal excitotoxic challenge.Cerebral Cortex15, 250–261.

Karanth S, Linthorst AC, Stalla GK, Barden N,et al.(1997).

Hypothalamic–pituitary–adrenocortical axis changes in a transgenic mouse with impaired glucocorticoid receptor function.Endocrinology138, 3476–3485.

Karege F, Perret G, Bondolfi G, Schwald M,et al.(2002).

Decreased serum brain-derived neurotrophic factor levels in major depressed patients.Psychiatry Research109, 143–148.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

Keilhoff G, Becker A, Grecksch G, Bernstein HG, Wolf G (2006). Cell proliferation is influenced by bulbectomy and normalized by imipramine treatment in a region-specific manner.Neuropsychopharmacology31, 1165–1176.

Kempermann G, Kronenberg G(2003). Depressed new neurons – adult hippocampal neurogenesis and a cellular plasticity hypothesis of major depression.Biological Psychiatry54, 499–503.

Klempin F, Kempermann G(2007). Adult hippocampal neurogenesis and aging.European Archives of Psychiatry and Clinical Neuroscience257, 271–280.

Kodama M, Fujioka T, Duman RS(2004). Chronic olanzapine or fluoxetine administration increases cell proliferation in hippocampus and prefrontal cortex of adult rat.Biological Psychiatry56, 570–580.

Lee J, Duan W, Mattson MP(2002). Evidence that brain- derived neurotrophic factor is required for basal neurogenesis and mediates, in part, the enhancement of neurogenesis by dietary restriction in the hippocampus of adult mice.Journal of Neurochemistry82, 1367–1375.

Linthorst ACE, Flachskamm C, Barden N, Holsboer F, Reul JM(2000). Glucocorticoid receptor impairment alters CNS responses to a psychological stressor : an in vivo microdialysis study in transgenic mice.European Journal of Neuroscience12, 283–291.

Lucki I(2001). A prescription to resist proscriptions for murine models of depression.Psychopharmacology153, 395–398.

Maccari S, Morley-Fletcher S, Mairesse J, Viltart O,et al.

(2005). Chronic treatment with agomelatine reversed the decrease in hippocampal cells neurogenesis and survival in prenatally stressed adults rats. Program No. 566.8. 2005 Abstract Viewer/Itinerary Planner. Washington, DC : Society for Neuroscience 2005. Online.

Malberg JE, Duman RS(2003). Cell proliferation in adult hippocampus is decreased by inescapable stress : reversal by fluoxetine treatment.Neuropsychopharmacology28, 1562–1571.

Malberg JE, Eisch AJ, Nestler EJ, Duman RS(2000).

Chronic antidepressant treatment increases neurogenesis in adult rat hippocampus.Journal of Neuroscience20, 9104–9110.

Mayer JL, Klumpers L, Maslam S, de Kloet ER,et al.(2006).

Brief treatment with the glucocorticoid receptor antagonist mifepristone normalises the corticosterone-induced reduction of adult hippocampal neurogenesis.Journal of Neuroendocrinology18, 629–631.

Millan MJ, Brocco M, Gobert A, Dekeyne A(2005).

Anxiolytic properties of agomelatine, an antidepressant with melatoninergic and serotonergic properties : role of 5-HT2Creceptor blockade.Psychopharmacology(Berlin)177, 448–458.

Millan MJ, Gobert A, Lejeune F, Dekeyne A,et al.(2003).

The novel melatonin agonist agomelatine (S20098) is an antagonist at 5-hydroxytryptamine2C receptors, blockade of which enhances the activity of frontocortical

dopaminergic and adrenergic pathways.Journal of Pharmacology and Experimental Therapeutics306, 954–964.

Montkowski A, Barden N, Wotjak C, Stec I,et al.(1995).

Long-term antidepressant treatment reduces behavioural deficits in transgenic mice with impaired glucocorticoid receptor function.Journal of Neuroendocrinology7, 841–845.

Morley-Fletcher S, Mairesse J, Mocaer E, Darnaude´ry M, et al.(2003). Chronic treatment with agomelatine increases hippocampal cell proliferation in prenatally stressed rats.

Program No. 506.20 Abstract Viewer/Itinerary Planner.

Washington, DC : Society for Neuroscience 2003. Online.

Olie JP, Kasper S(2007). Efficacy of agomelatine, a MT1/MT2 receptor agonist with 5-HT2Cantagonistic properties, in major depressive disorder.International Journal of Neuropsychopharmacology10, 661–673.

Paizanis E, Kelai S, Renoir T, Hamon M, Lanfumey L(2007).

Life-long hippocampal neurogenesis : environmental, pharmacological and neurochemical modulations.

Neurochemical Research32, 1762–1771.

Papp M, Gruca P, Boyer PA, Mocaer E(2003). Effect of agomelatine in the chronic mild stress model of depression in the rat.Neuropsychopharmacology28, 694–703.

Pariante CM, Kim RB, Makoff A, Kerwin RW(2003).

Antidepressant fluoxetine enhances glucocorticoid receptor function in vitro by modulating membrane steroid transporters.British Journal of Pharmacology139, 1111–1118.

Pariante CM, Miller AH(2001). Glucocorticoid receptors in major depression : relevance to pathophysiology and treatment.Biological Psychiatry49, 391–404.

Pepin MC, Pothier F, Barden N(1992). Impaired type II glucocorticoid-receptor function in mice bearing antisense RNA transgene.Nature355, 725–728.

Rajkowska G(2000). Postmortem studies in mood disorders indicate altered numbers of neurons and glial cells.

Biological Psychiatry48, 766–777.

Rao MS, Shetty AK(2004). Efficacy of doublecortin as a marker to analyse the absolute number and dendritic growth of newly generated neurons in the adult dentate gyrus.European Journal of Neuroscience19, 234–246.

Reif A, Fritzen S, Finger M, Strobel A,et al.(2006).

Neural stem cell proliferation is decreased in

schizophrenia, but not in depression.Molecular Psychiatry 11, 514–522.

Renoir T, Paizanis E, El Yacoubi M, Saurini F,et al.(2008).

Differential long-term effects of MDMA on the

serotoninergic system and hippocampal cell proliferation in 5-HTT knock-out vs. wild-type mice.International Journal of Neuropsychopharmacology11, 1149–1162.

Rousse I, Beaulieu S, Rowe W, Meaney MJ,et al.(1997).

Spatial memory in transgenic mice with impaired glucocorticoid receptor function.Neuroreport8, 841–845.

Saarelainen T, Hendolin P, Lucas G, Koponen E,et al.

(2003). Activation of the TrkB neurotrophin receptor is induced by antidepressant drugs and is required for antidepressant-induced behavioral effects.Journal of Neuroscience23, 349–357.

Sairanen M, Lucas G, Ernfors P, Castren M, Castren E (2005). Brain-derived neurotrophic factor and antidepressant drugs have different but coordinated

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021

effects on neuronal turnover, proliferation, and survival in the adult dentate gyrus.Journal of Neuroscience25, 1089–1094.

Santarelli L, Saxe M, Gross C, Surget A,et al.(2003).

Requirement of hippocampal neurogenesis for the behavioral effects of antidepressants.Science301, 805–809.

Seckl JR, Fink G(1992). Antidepressants increase glucocorticoid and mineralocorticoid receptor mRNA expression in rat hippocampus in vivo.Neuroendocrinology 55, 621–626.

Sheline YI(2000). 3D MRI studies of neuroanatomic changes in unipolar major depression : the role of stress and medical comorbidity.Biological Psychiatry48, 791–800.

Thierry B, Steru L, Chermat R, Simon P(1984).

Searching-waiting strategy : a candidate for an

evolutionary model of depression?Behavioral and Neural Biology41, 180–189.

Tsankova NM, Berton O, Renthal W, Kumar A,et al.(2006).

Sustained hippocampal chromatin regulation in a mouse model of depression and antidepressant action.Nature Neuroscience9, 519–525.

Vinet J, Carra S, Blom JM, Brunello N,et al.(2004). Chronic treatment with desipramine and fluoxetine modulate BDNF, CaMKKalpha and CaMKKbeta mRNA levels in the hippocampus of transgenic mice expressing antisense RNA against the glucocorticoid receptor.

Neuropharmacology47, 1062–1069.

West MJ, Gundersen HJ(1990). Unbiased stereological estimation of the number of neurons in the

human hippocampus.Journal of Comparative Neurology 296, 1–22.

Zhou Q, Homma KJ, Poo MM(2004). Shrinkage of dendritic spines associated with long-term depression of

hippocampal synapses.Neuron44, 749–757.

Downloaded from https://academic.oup.com/ijnp/article/13/6/759/690425 by SCD - Universite Toulouse III user on 03 February 2021