Receptor Activity Is Necessary for Physiological Brain Plasticity in Mice

Home , Sigma receptor

Sigma-1 (σ1) receptor activity is necessary for physiological brain plasticity in mice Lucie Crouzier, Simon Couly, Chloé Roques, Coralie Peter, Rislen Belkhiter, Maëva Arguel Jacquemin, Anna Bonetto, Benjamin Delprat, Tangui Maurice

To cite this version:

Lucie Crouzier, Simon Couly, Chloé Roques, Coralie Peter, Rislen Belkhiter, et al.. Sigma-1 (σ1) receptor activity is necessary for physiological brain plasticity in mice. European Neuropsychophar- macology, Elsevier, 2020, 10.1016/j.euroneuro.2020.08.010. hal-03020937

HAL Id: hal-03020937 https://hal.archives-ouvertes.fr/hal-03020937 Submitted on 24 Nov 2020

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Sigma-1 (σ1) receptor activity is necessary for physiological brain plasticity in mice

Lucie Crouziera, Simon Coulya, Chloé Roquesa, Coralie Petera, Rislen Belkhitera, Maëva

Arguel Jacquemina, Anna Bonettoa, Benjamin Delprata, Tangui Mauricea,*

a MMDN, Univ Montpellier, EPHE, INSERM, Montpellier, France

* Corresponding author at INSERM UMR_S1198, Université de Montpellier, CC105, place Eugène

Bataillon, 34095 Montpellier cedex 5, France. E-mail: [email protected] Abstract

The sigma-1 receptor (S1R) is a membrane-associated protein expressed in neurons and glia at mitochondria-associated endoplasmic reticulum (ER) membranes (MAMs). S1R interacts with different partners to regulate cellular responses, including ER stress, mitochondrial physiology and Ca2+ fluxes. S1R shapes cellular plasticity by directly modulating signaling pathways involved in inflammatory responses, cell survival and death.

We here analyzed its impact on brain plasticity in vivo, in mice trained in a complex maze, the Hamlet test. The device, providing strong enriched environment (EE) conditions, mimics a small village. It has a central agora and streets expanding from it, leading to functionalized houses where animals can Drink, Eat, Hide, Run, or Interact. Animals were trained in groups,

4 h/day for two weeks, and their maze exploration and topographic memory could be analyzed. Several groups of mice were considered: non-trained vs. trained; repeatedly administered with saline vs. NE-100, a selective S1R antagonist; and wildtype vs. S1R KO mice. S1R inactivation altered maze exploration and prevented topographic learning. EE induced a strong plasticity measured through resilience to behavioral despair or to the amnesic effects of scopolamine, and increases in S1R expression and bdnf mRNA levels in the hippocampus; increases in neurogenesis (proliferation and maturation); and increases of histone acetylation in the hippocampus and cortex. S1R inactivation altered all these parameters significantly, showing that S1R activity plays a major role in physiological brain plasticity. As S1R is a major resident protein in MAMs, modulating ER responses and mitochondrial homeostasy, MAM physiology appeared impacted by enriched environment.

Key words: sigma-1 receptor; Hamlet test; enriched environment; brain plasticity; neurogenesis; histone acetylation

2 1. Introduction

The σ1 receptor (S1R) is a membrane-bound protein expressed in numerous cell types such as neurons and glia (Alonso et al., 2000). It is considered as a protein involved in signal modulation, chaperoning or interacting with different partners that include inositol 1,4,5- trisphosphate receptors (IP3R), 78 kDa glucose-regulated protein (GRP-78; BiP), or inositol- requiring enzyme 1 (IRE1) (Hayashi et al., 2000; Hayashi & Su, 2007; Su et al., 2010; Mori et al., 2013). Highly expressed at mitochondria-associated endoplasmic reticulum (ER) membranes (MAMs) (Hayashi & Su, 2007), it dissociates following cellular stress or via agonist stimulation from IP3Rs, enhancing calcium entry into the mitochondria (Hayashi et al.,

2000; Mori et al., 2013). S1R also modulates NMDA receptors (Martina et al., 2007), voltage- dependent calcium channels (Zhang & Cuevas, 2002) and store-operated Ca2+ entry, a mechanism promoted by depletion of intracellular Ca2+ stores (Brailoiu et al., 2016).

Moreover, S1R activity shapes cellular plasticity also by directly modulating the activity of: (i) pleiotropic transcription factors, such as nuclear factor κB (NFκB), cyclic adenosine monophosphate (cAMP) response element-binding protein (CREB), c-fos; (ii) ion channels, like small conductance Ca2+-activated K+ current (SK channels); (iii) kinases, like calcium/calmoduline-dependent kinase IV (CaMKIV); (iv) G-protein coupled receptors; or (v) trophic factors, like brain-derived neurotrophic factor (BDNF) (Martina et al., 2007; Fujimoto et al., 2012; Kourrich et al., 2013; Navarro et al., 2013; Moriguchi et al., 2015). S1R consequently plays an important role in brain plasticity, response to stress, learning and memory, and neuroprotection.

S1R activity can be triggered by small molecules acting as S1R agonists. Numerous reports described beneficial effects of S1R agonists against cognitive deficits or motor impairments associated with aging (Maurice et al., 1996; Maurice, 2001), depression

(Matsuno et al., 1996; Urani et al., 2001; Fukunaga et al., 2017) and neurodegenerative diseases (for review, Maurice & Goguadze, 2017). Boosting S1R activity, which is

3 understood as restoring intracellular calcium homeostasis, stabilizing mitochondrial physiology and facilitating cellular adaptive responses (Su et al., 2010; Ryskamp et al.,

2019), has therefore been proposed to alleviate numerous aspects of neurodegenerative processes and sustaining brain plasticity. Interestingly, S1R has mainly be viewed as a modulating protein. S1R is devoid of effect in physiological conditions, but its activation results in marked cellular effects in case of impairment or overactivation of brain plasticity

(Maurice & Su, 2009; Su et al., 2010). The relative importance of S1R activity in brain plasticity in physiological conditions remains to be determined.

We therefore addressed S1R role in physiological brain plasticity in male mice, by analyzing its impact on the adaptive consequences of enriched environment (EE). We used the Hamlet test, a novel fully automatized behavioral analysis appliance (Viewpoint), to generate EE and test topographic memory in mice (Crouzier et al., 2018; Crouzier &

Maurice, 2018). The apparatus mimics a small village with a central agora and streets expanding from it, leading to functionalized houses (Drink, Eat, Hide, Run, Interact). EE is induced by training animals in the Hamlet, in groups of 6/8 individuals, during 4 h per day, for several weeks. Memory can be tested by depriving mice from water and testing their ability to locate the Drink house. We previously reported that training in the Hamlet increased hippocampal neurogenesis, cell proliferation and neuronal maturation, and modified the amnesic efficacy of muscarinic or nicotinic acetylcholine receptor antagonists (Crouzier et al.,

2018). We here analyzed the impact of EE on brain plasticity after S1R inactivation, induced by a repeated treatment with the selective S1R antagonist NE-100 or in wildtype vs. S1R knockout (KO) mice. The impact of S1R inactivation was first analyzed on behavioral responses in the Hamlet (exploration behavior and topographic memory testing). S1R expression and activity were analyzed after training. The impact of S1R inactivation on EE- induced brain plasticity was then analyzed first at the behavioral level; the behavioral despair, vulnerability to the amnesic drug scopolamine and the antidepressant or anti- amnesic effects of selective S1R agonists were measured. The impact of S1R inactivation on

EE-induced brain plasticity was then analyzed at the morphological and epigenetic levels:

4 hippocampal neurogenesis and histone acetylation were analyzed. Results showed that all aspects of EE-induced brain plasticity were blocked after pharmacological or genetic invalidation of S1R activity. This study provides the first in vivo demonstration of the importance of S1R activity in physiological adaptive plasticity.

5 2. Materials and methods

2.1. Animals

C57BL/6J mice (430 males and 12 females) from Elevage Janvier (St Berthevin, France) and

S1R KO (24 males) and C57BL/6J (16 males and 10 females) from Envigo (Bresso, Italy) were used at 7-9 weeks of age. All experiments were done in male mice. Females were used as interacting animals during Hamlet training. S1R KO mice, obtained through permission of

Esteve laboratories (Barcelona, Spain) were generated according to the described method

(Langa et al., 2003) and backcrossed onto the C57BL/6 background for more than 10 generations. Mice harboring the mutation were then bred to homozygosity. Mice were housed in plastic cages in groups of 6/8 individuals. They had free access to food and water except when specified and they were kept in a regulated environment (22 ± 1°C, 40-60% humidity) under a 12 h light/dark cycle (light on at 07:00 am). Experiments were carried out between 09:00 am and 06:00 pm, in an experimental room within the animal facility. All animal procedures were conducted in strict adherence to the European Union directive

2010/63 and the ARRIVE guidelines (Kilkenny et al., 2010).

2.2. Drugs and administration procedures

(R)-(+)-N-cyclopropylmethyl-α-ethyl-N-methyl-α-((2E)-3-phenyl-2-propenyl)benzene- methanamine hydrochloride (igmesine) was a gift from F.J. Roman (Pfizer, Fresnes, France).

(-)-Scopolamine hydrobromide (Scop), 2-(4-morpholinethyl)-1-phenylcyclohexanecarboxylate hydrochloride (PRE-084) and 4-methoxy-3-(2-phenylethoxy)-N,N- dipropylbenzeneethanamine hydrochloride (NE-100) were from Sigma-Aldrich (Saint-

Quentin-Fallavier, France). All drugs were solubilised in physiological saline (vehicle solution) and administered intraperitoneally (IP) or subcutaneously (SC), in a volume of 5 µl/g body

6 weight. NE-100 injections were performed b.i.d. at 4 h time interval —immediately before and after the session— on days 1-5 and 8-12 of Hamlet training. Other drugs were administered at specified times before the behavioral tests.

2.3. The Hamlet test

The Hamlet apparatus, designed by us but manufactured by Viewpoint (Lissieu, France), is shown in Fig. 1a (Crouzier et al., 2018). The device is 1.6 m in diameter and placed in a room with standard laboratory furniture and posters on the walls. The room was uniformly illuminated (200 Lux). The Hamlet has an agora at its centre and streets expanding in a star shape towards five functionalized compartments with roofs, called houses (Fig. 1b). The walls and walkways are made of white or black infrared-transparent Plexiglas. The agora served as a gathering area and as a start box for training and test trials. The functionalized houses contained a food dispenser (physiological function encoded: Eat), two water dispensers (Drink), a novomaze® (Viewpoint) (Hide), a running wheel (Run) or a compartment for a stranger mouse (Interact). A wire mesh screen separated the stranger mouse, whereas the other houses could be entered freely. The stranger mouse was a female

C57BL/6J mouse. Movements were recorded by an IR light camera set above the Hamlet and analyzed using a Videotrack® software (Viewpoint). Training was performed according to the published protocol (Crouzier & Maurice, 2018). In brief, animals were placed in the

Hamlet in groups of 6/8 from the same housing cage, for 4 h per day during a 2-weeks training period (Fig. 1c). During training number of entries and duration of presence in each house were measured. Global activity was determined from the measure of the time in sec of occupation of the five houses divided by the number of mice in the hamlet for each hour during the 4 h of a training session. Topographical memory was evaluated by testing twice animals, 3 and 4 days after the last training session, individually, after water-deprivation

(WD) or not (NWD) (Crouzier & Maurice, 2018). For WD condition, the drinking bottle was removed from the housing cage 15 h before testing. Each mice, in WD or NWD condition,

7 was placed in the agora and free to explore the Hamlet during a 10-min session. The exploratory behavior was videotracked and analyzed in terms of latency to reach the goal and other houses and number of errors (entries into a street not directing to the goal house).

Trained (T) mice were compared between NWD and WD conditions, and with non-trained

(NT) animals, naive to the Hamlet, which were also tested in WD or NWD condition.

2.4. Forced swim test

Behavioral despair was measured in a single session. Mice were forced to swim, in a 2 l becher (diameter 12 cm) half-filled with water at 22°C, for 6 min. The duration of immobility was analyzed during the last 5 min, using a Videotrack® software (Viewpoint) for quantification of movements. Igmesine was administered IP, 30 min before the test.

2.5. Spontaneous alternation in the Y-maze

Spontaneous alternation performance was measured in the Y-maze as an index of spatial working memory. Animals were allowed to explore the maze for 8 min and the percentage of alternation was calculated as (actual alternations / number of arm entries - 2) x 100. PRE-

084 was administered IP, 10 min before scopolamine, which was administered SC, 20 min before the session.

2.6. Step-through passive avoidance procedure

Long-term non-spatial memory was tested using a passive avoidance test. The step-through latency, i.e., the time spent to enter the dark compartment of the two-compartment box, and the level of sensitivity to the shock was recorded during training. Shock sensitivity was coded as 0 = no reaction, 1 = flinching reactions; 2 = flinching and vocalization. None of the treatments affected these parameters during training (data not shown). Retention was

8 measured after 24 h. Each mouse was placed again into the white compartment. The step- through latency was recorded up to 300 s. PRE-084 was administered IP 10 min before scopolamine, which was administered SC 20 min before the training session. No drug injection was performed before the retention session.

2.7. Tissue preparation

Animals were sacrificed and their brain dissected on ice. Frontal cortex and hippocampus were isolated, frozen in liquid nitrogen and stored at -80°C until use.

Histone protein extraction. Histones were extracted using a histone extraction kit (ref. ab113476, Abcam, Cambridge, UK). Briefly, samples were homogenized in a pre-lysis buffer using a Dounce homogenizer. The homogenate was centrifuged (1,000 g, 5 min at 4°C). The pellet was resuspended and incubated 30 min in 3 vol of lysis buffer at 4°C and then centrifuged at 15,000 g for 5 min at 4°C. Then, 0.3 vol of a balanced dithiothreitol buffer was added to the histone containing supernatant. Protein concentration was determined using a standard BCA kit (Uptima, Interchim, Montluçon, France), using bovine serum albumin as standard.

Total protein extraction. Tissues were sonicated with cold lysis buffer pH 6.8, containing 0.5

M Tris-HCl, with SDS 20%, glycerol, PhosStop 10x and Complete protease inhibitor 25x

(Roche, USA). The homogenate were centrifuged at 21,000 g for 15 min at 15°C and the supernatant was collected to analyze protein concentration using the BCA assay.

2.8. Western blot analyses

Protein extracts were mixed with Laemmli buffer pH 6.8, containing 240 mM Tris-HCl, 8%

SDS, 40% glycerol, 5% β-mercaptoethanol and 0.04% bromophenol blue, and denatured in boiled water during 5 min. Each well of the SDS-PAGE were loaded with 10 µg (for histone protein analyses) or 20 µg protein (for S1R analyses). Protein were separated at 100 V for 80

9 min and transferred to 0.20 µm PVDF membrane (Merck Millipore, Billerica, MA, USA) at 100

V for 1 h. Non-specific antibody sites were blocked by 0.1% TBS containing 5% fat milk, for

60 min at room temperature. Membrane incubation with primary antibody was performed overnight at 4°C. Polyclonal antibodies were: rabbit anti-β-actine (1:5000, ref. ab8227,

Abcam), rabbit anti-S1R (1:200, ref. HPA0180002, Sigma-Aldrich), mouse anti-histone H2B

(1:4000, ref. 05-1352, Millipore), rabbit anti-acetyl-histone H2B (1:20000, ref. 07-373,

Millipore), rabbit anti-histone H3 (1:3000, ref. ab1791, Abcam), rabbit anti-acetyl-histone H3

(1:5000, ref. 06-599, Millipore), rabbit anti-histone H4 (1:5000, ref. 07-108, Millipore) and rabbit anti-acetyl-histone H4 (1:10000, ref. 07-329, Millipore). After incubation with HRP- labeled anti-rabbit (1:2000, ref. ab6721, Abcam) or anti-mouse (1:2000, ref. A4416, Sigma-

Aldrich) secondary antibodies, bands were visualized with Luminata Crescendo Western

HRP® substrate (Millipore) using an Odyssey Fc® Imaging System (LI-COR Biosciences, Bad

Homburg, Germany) and quantified using the ImageJ software (NIH, Bethesda, MD, USA).

2.9. RNA extraction and real-time reverse transcription polymerase chain reaction

Cortex and hippocampus RNAs were extracted using the NucleoSpin® RNA Kit (Macherey-

Nagel, Hoerdt, France) and concentration and purity were evaluated using the Agilent RNA

6000 Nano® Kit (Agilent Technologies, Santa Clara, CA, USA). Total RNA was reverse transcribed into cDNA, 1 h at 37°C, in retrotranscription buffer containing M-MLV reverse transcriptase (Promega, Madison, WI, USA), reaction buffer containing 250 mM Tris-HCl,

375 mM KCl, 15 mM MgCl2, 50 mM dithiothreitol, pH 8.3 (Promega), random primers (500

µg/mL, Promega), dNTP mix (Promega). RNA sample (1 µg/µL) was denatured 5 min at

70°C and the amount have been measured by quantitative PCR, using LightCycler 480

(Roche). PCR cycling conditions were 95°C for 10 s, 45 cycles of 95°C for 10 s, 60°C for 10 s, 72°C for 10 s, 95°C for 5 s, 65°C for 1 min and 40°C for 30 s. β-actine was used as the housekeeping gene. The primers (Eurogentec, Liège, Belgium) were as follows : oprs1, 5’-

TCTGAGTACGTGCTGCTCTT-3’ (forward) and 5’-GTGCCCTCTTTCCATTGGTG-3’

10 (reverse); bdnf, 5'-GCCTTGTCCGTGGACGTTTA-3' (forward) and 5'-

CAGAGCAGCTGCCTTGATGTT-3' (reverse); β-actine, 5’-TCTGAGTACGTGCTGCTCTT-3’

(forward) and 5’-CTGTCCCTGTATGCCTCTG-3’ (reverse). The data were analyzed using

LightCycler 480 Software release 1.5.0 SP4.

2.10. Neurogenesis analyses

The brain sections were blocked with a solution containing 0.01 M PBS, 0.1% Triton X-100 and 10% normal goat serum solution for 30 min. They were subsequently incubated at 4°C overnight with primary antibodies diluted in PBS: rabbit anti-Ki-67 (dilution 1:100, ref. ab16667, Abcam) or rabbit anti-doublecortine (DCX, dilution 1:500, ref. 4604S, Cell

Signaling). After several washes, sections were incubated with biotinylated goat anti-rabbit secondary antibody (dilution 1:200, ref. BA-1000, Vector laboratories) and finally in avidin– biotin–peroxidase complex (PK-4000) and detected with diaminobenzidine. After dehydration through successive baths of alcohol, all sections were stained with Cresyl violet and cover slipped with mounting medium (Mountex). The immunoreactive cells in the granule cell layer and subgranular zone of the dorsal dentate gyrus (bregma -1.8 to 3.8) were manually counted using an optical microscope (DM 2500, Leica) under white light at a magnification of x400. Each DCX neuron having a single dendrite was measured within 40 μm of the soma.

An image of hippocampus cells was projected on a computer, with a X40 objective and a line segment was drawn along the middle of the cells. DCX dendrites crossing the segment were counted. The number of Ki67 or DCX positive cells was determined by combining cell counts per section for the whole DG and multiplying by the number of brain sections including the hippocampus. All data were expressed as percentage of NT group.

2.11. Statistical analyses

11 Data were expressed as mean ± S.E.M. They were analyzed using two-way ANOVA (with training and WD condition as independent factors) for multiple comparisons (F values) followed by a Bonferroni's post-hoc test; a Student's or Mann-Whitney's test for two populations comparison (NT/T, NWD/WD); or using a one-column t-test, when compared to a reference level. The levels of statistical significance considered was p < 0.05. Statistical analyses were performed using the GraphPad Prism v5.0 software. For reading clarity, all statistical values are detailled in Table 1 in the Supplementary Materials.

12 3. Results

We first analyzed the impact of S1R invalidation, induced either pharmacologically using a repeated treatment with the S1R antagonist NE-100 or genetically in S1R KO mice, on exploration and topographic memory in the Hamlet.

Animal behavior during training was analyzed at the beginning (Day 1), middle (Day

6) and end (Day 10) of the training period. The global activity was analyzed per hour per mice (Figs. 2a, 2d) and results were shown as radar graphs for entries (Figs. 2b, 2e) and duration of presence in the houses (Figs. 2c, 2f). Noteworthily, groups withe the Saline- or

NE-100-treatment did not show a similar profile as WT or S1R KO groups. Activity levels were almost constant between days 1, 6 and 10 for the first groups (Fig. 2a), while activity at days 6 and 10 was significantly decreased as compared to day 1 for the second ones (Fig.

2d). Strain differences and the mild stress due to repeated injections in the treated groups may explain these differences in activity profiles. For the Saline- or NE-100-treated groups

(Fig. 2 left panel), two-way ANOVA analyses showed treatment and interaction effects

(Suppl. Table 1; Fig. 2a). On days 6 and 10, exploration activity was significantlty reduced in

NE-100-treated groups as compared to Saline-treated animals. Radar graphs showed a significantly diminished exploration in the Hamlet at day 6 for entries (Fig. 2a, 2b) and day 10 for both entries and duration (Figs. 2a-2c). Noteworthily, duration of presence in the houses showed a significant preference for the Hide house in both groups.

WT and S1R KO mice (Fig. 2 right panel) showed genotype and day effects (Suppl.

Table 1; Fig. 2d). Both groups showed a similar trend with a high activity during the first day of training and then lower activity during days 6 and 10. However, S1R KO mice explored significantly less the Hamlet at days 6 and 10 than WT controls (Fig. 2d). This was true for both entries and duration (Figs. 2e, 2f). Noteworthily, analyses of entries into the houses showed a significant preference of the animals for the Run house for WT and Interact house

13 for S1R KO (Fig. 2e). Duration of presence showed a significant preference for the Hide house in WT mice only (Fig. 2f).

Topographic memory was analyzed 3 days after the last training session for each group of mice. Saline-treated animals, non-trained (NT) in the Hamlet and therefore naïve to the maze configuration, spent about 100 s (Fig. 3a) and did about 34 errors (Fig. 3b) to find the Drink house, when tested in either WD or NWD condition. Trained (T) mice showed a similar performance in NWD condition. However, they very significantly improved their ability in WD condition, with a latency about 40 s (Fig. 3a) and less than 20 errors (Fig. 3b). These decreases between NWD and WD conditions indicated that T animals were able to mobilize their topographic memory to more efficiently reach the Drink house. After NE-100 treatment, both NT and T groups did not show any difference in the ability to reach the Drink house between WD and NWD conditions, contrarily to saline-treated T mice (Figs. 3a, b).

The same experiment was performed with WT and S1R KO mice (Figs. 3c, 3d). Only trained WT mice showed a significant decrease in latency (Fig. 3c) and number of errors

(Fig. 3d) to find the Drink house in WD condition as compared to NWD condition. This was not the case for non-trained (WT or S1R-KO) groups or for trained S1R KO mice (Figs. 3c,

3d). This indicated that S1R KO mice were unable to mobilize their topographic memory to efficiently locate the Drink house. Similar results were therefore observed after pharmacological or genetic invalidation of S1R activity, showing that S1R is necessary for topographic memory.

Interestingly, we also injected NE-100 only once, 20 min before the test, in T mice that did not receive any treatment during training, to see if the S1R antagonist could alter topographic memory retention (Figs. 3e, 3f). We observed that Saline-treated animals showed significant reductions in the latency and number of errors to reach the Drink house in

WD condition, but not NE-100-treated mice. In both protocols (repeated treatment in Figs.

3a, 3b or acute pre-test treatment in Figs. 3e, 3f), the NE-100 treatment resulted in a marked topographic memory impairment. This suggested that the drug interfered with training but is also able to provoke a retrograde amnesia in trained animals.

14 We previously reported that plasticity induced by training in the Hamlet leads to behavioral stimulation and brain plasticity (Crouzier et al., 2018). We therefore analyzed several behavioral responses in mice treated with saline or NE-100 during training: (1) the induction of behavioral despair, a physiological response to acute stress when mice are unescapably placed in a water tank; and (2) induction of learning and memory deficits induced by the muscarinic acetylcholine receptor antagonist scopolamine. In each procedure, the antidepressant or anti-amnesic efficacy of a reference S1R agonist was also tested.

First, we confirmed that T mice are more resistant than NT animals in the forced swim test by measuring a significant diminution of immobility duration (Fig. 4a). This significant decrease between NT and T conditions was not observed in NE-100-treated mice, although the two-way ANOVA failed to show a treatment effect (Suppl. Table 1; Fig. 4a). Interestingly, the antidepressant efficacy of the reference S1R agonist, igmesine (Matsuno et al., 12996;

Urani et al., 2001), was unchanged. The drug significantly decreased immobility duration compared to vehicle-treated animals at 30 mg/kg, but not 10 mg/kg, in both NT and T groups

(Fig. 4b).

Second, we analyzed scopolamine-induced amnesia, since we previously reported that training decreased the sensitivity of mice to the scopolamine-induced amnesic effect

(Crouzier et al., 2018). Scopolamine blocked spontaneous alternation ability at the dose of

0.5 mg/kg in NT animals, but not in T animals (Fig. 4c). In NE-100-treated T mice, scopolamine decreased alternation and the performance of Scop-treated animals was not significantly different from the 50% random level (Fig. 4c).

We examined the efficacy of the anti-amnesic response of a reference S1R agonist,

PRE-084, against scopolamine-induced learning impairments in the different treatment groups (Fig. 5). The dose-response effects of scopolamine and PRE-084 were analyzed for spontaneous alternation (Figs. 5a, 5c, 5e, 5g, 5i) and passive avoidance responses (Figs.

5b, 5d, 5f, 5h, 5j). In NT mice, scopolamine blocked spontaneous alternation and passive avoidance at 0.5, 1.5 and 5 mg/kg (Figs. 5a, 5b). A pre-treatment with PRE-084 significantly

15 prevented the scopolamine, 0.5 mg/kg, effect at 0.3 mg/kg (Figs. 5c, 5d). In T mice, scopolamine failed affect both alternation and passive avoidance responses at 0.5 mg/kg, but higher doses of 1.5 or 5 mg/kg were necessary (Figs. 5e, 5f) confirming previous results observed (Crouzier et al., 2018). PRE-084 was tested against 1.5 and 5 mg/kg of scopolamine and the drug significantly attenuated the memory deficits, but at higher doses: at 1 mg/kg against scopolamine 1.5 mg/kg (Figs. 5g, 5h) and at 3 mg/kg against scopolamine

5 mg/kg (Figs. 5i, 5j). Therefore, considering the dose ratio between the impairing dose of scopolamine and the anti-amnesic dose of PRE-084 among the NT/T conditions, the anti- amnesic efficacy of the S1R agonist was preserved in mice trained in the Hamlet.

We then analyzed the impact of Hamlet training on S1R expression and protein levels. We selected two brain structures of major importance for cognitive functions, the cortex and hippocampus (Fig. 6a-6f). First, S1R mRNA analysis showed that training progressively, after 3 days, 1 and 2 weeks, increased S1R mRNA levels, as compared with

NT animals, in the cortex (Fig. 6a) and hippocampus (Fig. 6b). Increases were significant after 2 weeks training. The repeated NE-100 treatment during training prevented the increases and differences at 2 weeks were significant (Figs. 6a, 6b). The western blot analysis of S1R protein levels showed that training increased S1R protein in the hippocampus after 1 and 2 weeks (Fig. 6d). No change was observed in the cortex at all timepoints analyzed (Fig. 6c), suggesting structure specificity in the long-term impact of training on S1R expression.

As brain-derived neurotrophic factor (BDNF) plays a major role in environmental enrichment (Ickes et al., 2000; Mosaferi et al., 2015; Gualtieri et al., 2017) and S1R activity

(Ovalle et al., 2002; Fujimoto et al., 2009; Dawaldi et al., 2017), we measured bdnf mRNA expression in trained animals. In both the cortex (Fig. 6g) and hippocampus (Fig. 6h).

Training in the Hamlet increased rapidly and significantly bdnf mRNA levels in both structures after 3 days and 1 week training. bdnf mRNA levels returned to basal value after 2 weeks training.

16 Environmental enrichment, and particularly Hamlet training, was previously shown to increase both newborn cell proliferation and neuronal maturation in the hippocampal dentate gyrus (Crouzier et al., 2018). S1R has indeed been involved in neurogenesis (Sha et al.,

2013) and S1R agonists are able to boost neurogenesis in pathological models (Lucas et al.,

2008; Li et al., 2010; Moriguchi et al., 2015). We analyzed the impact of S1R inactivation on neurogenesis after training. We confirmed that training increased newborn Ki67-positive cells

(+35%) (Figs. 7a, 7b) and premature neurons labelled with DCX (+49%) (Figs. 7c, 7d). The

NE-100 treatment significantly blocked both training-induced increases in proliferation (Figs.

7a, 7b) and maturation (Figs. 7c, 7d). Interestingly, a 2 weeks repeated treatment with PRE-

084 (1 mg/kg IP b.i.d.) was able to increase proliferation in NT mice (Figs. 7a, 7b), but not maturation (Figs. 7c, d). Moreover, S1R KO mice presented lower basal levels of proliferation

(-28%) and maturation (-25%) as compared with WT (Sal-treated NT) mice. In S1R KO mice, training failed to affect proliferation (Figs. 7a, 7b) and slightly but significantly increased maturation (+22%) (Figs. 7c, 7d). These results therefore showed that S1R activation therefore is involved in the impact of Hamlet training on hippocampal neurogenesis, particularly on cell proliferation.

Finally, we examined the impact of Hamlet training and NE-100 treatment on epigenetic remodelling and particularly histone acetylation levels. The acetylation level of histone H2B, H3 and H4 were analyzed by western blotting in chromatin preparations from the mouse hippocampus (Fig. 8) and cortex (Fig. 9). In the hippocampus, training increased, after 1 and 2 weeks, histone H2B acetylation (Figs. 8a, 8b) and H4 acetylation (Figs. 8e, f).

The level of H3 acetylation remained unchanged (Figs. 8c, 8d). The NE-100 treatment blocked both the training-induced increase in H2B (Figs. 8g, 8h) and H4 acetylation (Figs. 8i,

8j). Interestingly, the interaction factor of the two-way ANOVA was highly significant indicating that NE-100-treated T mice even showed a significant decreased in acetyl-H2B level as compared to saline-treated NT controls (Fig. 8h). Similar data were obtained in the cortex. Training increased, after 2 weeks, H2B acetylation (Figs. 9a, 9b) and, after 1 or 2 weeks, H4 acetylation (Figs. 9e, 9f). The level of histone H3 acetylation remained unchanged

17 (Figs. 9c, 9d). The NE-100 treatment after 2 weeks training blocked both the training-induced increase in H2B (Figs. 9g, 9h) and H4 acetylation (Figs. 9i, 9j). In both hippocampus and cortex expression of those histones, H2B, H3 and H4, did not change with training nor NE-

100 (Figs. 8a, 9c, 9e, 9g, 9i and Figs 9a, 9c, 9e, 9g, 9i).

18 4. Discussion

The S1R protein is as a ligand-operated intracellular chaperone that regulates numerous cellular signaling pathways. Although the numerous consequences of S1R activity in pathological conditions is increasingly documented, its involvement in physiological plasticity has been mainly suggested in cellular studies and through its impact on trophic factor systems. S1R overexpression enhanced neuritogenesis and neurite sprouting induced in PC12 cells by nerve growth factor (NGF) or epidermal growth factor (EGF) (Takebayashi et al., 2002; 2004). S1R has been repeatedly shown to regulate BDNF systems, particularly in forebrain structures (Ovalle et al., 2002; Kikuchi-Utsumi & Nakaki, 2008; Fujimoto et al.,

2012; Dalwadi et al., 2017). The first aim of the present study was therefore to establish the role of S1R in physiological brain plasticity. To address this question, we used a pharmacological invalidation of S1R activity, a repeated treatment with the S1R antagonist

NE-100, and a genetic inactivation model, S1R KO mice. Second, we stimulated brain plasticity by training mice in EE and compared the EE impact in normal vs. S1R deficient mice.

The Hamlet presents several advantages on classical EE procedures mainly based on cage enrichment. First, it relies on ethological traits, since the group of mice is displaced from its home cage into a wider space where basic needs (drink, eat, hide, run, interact) are differentially identified in specific spots as happens in nature. Second, quantitative information is directly recorded in each house during all the training period. Mouse activity throughout these different spots is therefore precisely analyzed and specific patterns are identified. Third, the Hamlet allows the analysis of topographic memory through the performance of T mice tested in WD vs. NWD condition and comparison with performances of NT mice (Crouzier et al., 2018). Topographic memory relies on both allocentric orientation strategy —involved, for instance, in spatial learning measured in water-maze procedures— and egocentric strategy relying on body-centered representation of the testing environment.

19 We previously reported that, as observed in imaging analyses performed in humans, topographic memory in the Hamlet involved the hippocampus-subiculum-parahippocampal gyrus axis and dopaminergic structures (Crouzier et al., 2018). The Hamlet test could therefore be considered as a pertinent test for complex orientation memories. In the present study, we used Hamlet training as a model of EE known to stimulate brain plasticity, on the one hand, and as a topographic memory test, on the other hand. Based on behavioral

(behavioral stimulation, memory abilities), morphological (hippocampal neurogenesis) and transcriptional readouts (histone acetylation), we determined the impact of S1R protein inactivation on brain plasticity and complex memories.

Present in numerous organs and cell types, S1R chaperones different partners like signaling proteins, ionophores, lipids and RNAs (Su et al., 2010; Delprat et al., 2019;

Ryskamp et al., 2019). Highly concentrated at MAMs, thus regulating Ca2+ influx or cholesterol and steroid intake into the mitochondria (Su et al., 2010), the protein is also present at the vicinity of plasma membranes, where it regulates several ion channels and neurotransmitter receptor activities (Su et al., 2010) and at the nucleus, where it facilitates transcription regulator translocation and activity (Tsai et al., 2015; Arun et al., 2019). S1R expression has been repeatedly shown to be regulated or altered in pathophysiological conditions. The most documented example of an increased S1R expression induced by an adaptive response is provided by the numerous reports examining the role of S1R in addiction. Repeated administration of cocaine increased S1R expression in the nucleus accumbens of rodents (Romieu et al., 2002) and consequently to immediate early genes, like

Fra-2, is responsible for cocaine-induced neuronal adaptations (Liu et al., 2005). Conversely,

S1R expression was diminished in pathological situations or its diminution resulted as an amplification of the pathological state. At the clinical level, the E102Q mutation of S1R resulted in a juvenile form of amyotrophic lateral sclerosis, and an aberrant cellular localization of the mutated protein (Al-Saif et al., 2011). In preclinical studies, S1R inactivation, in S1R KO mice, resulted in a moderate phenotype, but it amplified the development of neurodegenerative pathologies. S1R KO mice showed limited behavioral

20 alterations: male S1R KO mice developed increased immobility in the forced swimming test, but normal behavior in the elevated plus-maze and light/dark box tests and normal locomotor activity (Sabino et al., 2009; Chevallier et al., 2011). Females S1R KO mice showed age- related memory deficits after 12 month-of-age (Chevallier et al., 2011). This moderate impact of S1R inactivation on basic cellular physiology was confirmed by Snyder et al. (2016) who used electrophysiological approaches. They reported no change in pre-synaptic function, on glutamate-gated AMPA and NMDA receptors, but a small significant reduction in the magnitude of long-term potentiation in mutant compared to WT mice (Snyder et al., 2016).

In the present study, we characterized a novel behavioral response in NE-100-treated and S1R KO mice. Spontaneous exploration of the environment was markedly altered. Both

NE-100-treated and S1R KO mice showed a decreased presence in the most stimulating houses of the Hamlet that include Run, Drink, and Eat. S1R KO mice did not show a locomotor deficit but exploratory mobility was altered, particularly after the initial novelty- induced stimulation (i.e., on training days 6 and 10 but not 1). Moreover, previous experiments showed no locomotor deficit in S1R KO (Langa et al., 2003; Fontanilla et al.,

2009; Chevallier et al., 2011; Hong et al., 2017). We therefore conclude that S1R inactivation resulted in a significant alteration of the exploratory ability of mice. This observation identified a novel information on S1R KO mice behavioral phenotype, by revealing a marked ethological deficit.

Although S1R agonists have been repeatedly shown to have anti-amnesic activity, in pharmacological of pathological models of amnesia (Maurice & Goguadze, 2017), memory processes —and particularly the different steps distinguished as learning, consolidation, retention, forgetting, and reactivation— have repeatedly been shown to be unaffected by

S1R ligands in physiological conditions. Indeed, S1R agonists are not pro-mnesic drugs and

S1R antagonists are not amnesic drugs. Two notable exceptions were a deficit observed in the object recognition test in S1R KO mice, when the novel object was presented 24 h, but not 1 h, after the two familiar ones (Xu et al., 2017) and an age-related learning deficit identified in female S1R KO mice (Chevallier et al., 2011). We here brought a novel

21 information by showing in NE-100-treated or S1R KO mice that inhibition of S1R activity during memory formation resulted in a blockade of topographic memory. Topographic memory is encoded during the training period by latent learning in exploration in the Hamlet relying on complex information processus between the different hippocampus, subiculum and parahippocampal gyrus formations. S1R inactivation therefore resulted in a blockade of memory encoding. Moreover, we observed that an acute treatment with NE-100, 20 min before the memory test, also impaired topographic memory retention. This last observation was the first, and unexpected, observation that S1R inactivation altered memory retention. It also suggested that the impact of S1R activity on memory process could mainly be evidenced in complex memory tasks.

Long-lasting experience in EE condition generates behavioral resilience. This observation has been related to social and environmental enrichment in early period of life.

For instance, mice reared in a communal nest as pups —a model of social enrichment— spent longer time floating, a behavioral response inversely associated with display of endophenotypes of depression, than mice reared with their individual mother (D'Andrea et al., 2010). We previously observed behavioral resilience in mice trained in the Hamlet

(Crouzier et al., 2018). We confirmed here that, as compared to NT, T mice showed: (i) a decreased behavioral despair, with a significant reduction of the immobility duration in the forced swim test; and (ii) a decreased sensitivity to amnesic drugs, with a lack of alternation deficit or avoidance impairment induced by scopolamine at the usually amnesic dose.

Interestingly, we observed that S1R agonists are still able to produce an additional effect on behavioral despair and memory, although the effect was limited or the dose of compound had to be adapted. The neurophysiological substrata for depression and memory are different but showed together an important disadaptation after S1R inactivation.

EE induced —or restored, if we compare natural ethological conditions vs. research facility housing conditions…— brain plasticity with neurological consequences on dendrite spines and branching, synapse consolidation and formation, neural cell size and new neural cell generation (Nithianantharajah et al., 2006; Bayne, 2018). During Hamlet training, the

22 effects on brain morphology and physiology can be attributed to the enriching effects of the complexity and size of available space, the stimulatory effects of inanimated enrichments and increased social interactions. Gene expression in the brain is increased after periods of several days or weeks of EE and impacts particularly the expression of genes involved in protein synthesis, synaptic signaling, neuronal growth and structure, and protein processing

(Rampon et al., 2000). This is observed for trophic factors like BDNF (McQuaid et al., 2018), and we report here that bdnf mRNA levels are rapidly increased after 3 days and 1 week and oprs1 mRNA levels are increased after 1 and 2 weeks training in the Hamlet in forebrain structures. Noteworthily, whereas the first likely appears to be a direct physiological response to EE, the second, and the resulting increase in S1R protein, may rather be a delayed adaptive response to its recruitment by EE. Indeed, the NE-100 treatment completely blocked the training-induced change in bdnf mRNA level at early timepoints showing that

S1R activity is required to provoke the changes in BDNF contents induced by EE. S1R expression was however regulated at a later period of EE. This is an interesting observation and the mechanism by which EE can increase S1R expression remains to be determined.

Likely, increased energy needs in brain cells induced by EE could be related to increased mitochondrial functions, a cellular adaptation that could require S1R activity. Richetin et al.

(2017) identified a link between mitochondrial activity in brain cell and plasticity. They observed an increased mitochondrial respiration accompanying the stimulating effect of

Neurod1 overexpression on dendritic growth and spine formation in primary neurons. A pharmacological treatment with carbonyl cyanide chlorophenylhydrazone, inhibiting oxidative respiration, prevented Neurod1-dependent trophic effects. Therefore, manipulating the mitochondrial homeostasy in hippocampal neurons could sustain neuronal plasticity (Richetin et al., 2017).

EE increases neural cell proliferation, maturation and differentiation, vascularization of the dentate gyrus and dendritic complexity of hippocampal neurons (Bindu et al., 2007;

Schloesser et al., 2010). This enhanced neurogenesis is a direct consequence of an up- regulation of BDNF that increases the number and survival of newborn neurons and favors

23 their functionality (Scharfman et al., 2005; Bekinschtein et al., 2014). Moreover, voluntary physical exercise, a form of EE, increased expression of BDNF in the hippocampus

(Sølvsten et al., 2018), as well as neurogenesis (Fabel et al., 2009) and dendritic spine densities (Stranahan et al., 2009). We first confirmed that EE in the Hamlet increased both newborn cell proliferation and neuronal maturation (Crouzier et al., 2018). These effects were blocked in NE-100-treated or S1R KO mice, showing the importance of S1R activity in neurogenesis. Interestingly, PRE-084 augmented proliferation in a similar level as EE, but failed to boost maturation. This result confirmed previous observation that a continuous administration of SA4503, another S1R agonist dose-dependently enhanced the number of bromodeoxyuridine-positive cells in the subgranular zone of the hippocampus (Lucas et al.,

2008). We therefore confirmed the role of S1R in hippocampal neurogenesis and particularly in EE-induced neurogenesis. Noteworthily, the lack of effect of PRE-084 on maturation was more surprising, considering that S1R activity regulated gene transcription, neuritogenesis and dendrite growth in physiological as well as pathological conditions (Tsai et al., 2009;

Yamaguchi et al., 2018; Ryskamp et al., 2019).

Finally, we analyzed epigenetic markers since EE is known to increase histone acetylation, and particularly H3/H4 acetylation in the dorsal hippocampus of young rodents

(Fischer et al., 2007). Moreover, epigenetic regulations such as histone acetylation are one of the mechanisms by which EE translates in the brain into long-lasting changes and plasticity (Lopez-Atalaya et al., 2011; Neidl et al., 2016). We observed that EE in the Hamlet increased markedly H2 and H4 acetylation and that S1R inactivation in NE-100-treated mice completely prevented these effects. The mechanism by which S1R mediates these effects of

EE on histone acetylation remains to be explored in order to determine if it could involve a direct effect of S1R. Indeed, cellular studies showed that S1R could bind the nuclear envelope protein emerin and recruit chromatin-remodeling molecules, such as lamin A/C, barrier-to-autointegration factor (BAF) and histone deacetylase (HDAC) to form a macromolecular complex regulating transcriptional activity (Tsai et al., 2015). Furthermore,

S1R was shown to participate in the genesis of neuropathic pain through the epigenetic

24 modifications of histones, particularly H3 methylation, in the dorsal root ganglion (Wu et al.,

2017). Finally, S1R is a substrate of the histone acetylases p300/CBP-associated factor

(PCAF) and GCN5 (Yasui & Su, 2015). It is therefore possible that S1R main involvement in brain plasticity occurs through direct effect on histones resulting in a facilitation of transcriptional efficacy. Indeed, bdnf gene is highly regulated by chromatin acetylation and

EE facilitated transcriptional regulation through several mechanisms, including NF-kB- dependent gene transcription and bdnf exon I expression (Neidl et al., 2016). The role of

S1R on epigenetic regulations deserves further investigation.

In conclusion, we reported here that S1R activity plays a major role in EE-induced brain plasticity in mice, since pharmacologic or genetic S1R inactivation completely prevented: (i) the establishment of topographic memory in the Hamlet, (ii) behavioral resilience, (iii) hippocampal neurogenesis and (iv) epigenetic facilitation of transcription through increased histone acetylation. This study demonstrated in vivo the importance of

S1R activity in brain plasticity in physiological conditions.

25 Role of funding source

This work was supported by a SATT AxLR (Montpellier) maturation program. The funding source had no role in the design of the study or interpretation of the data.

Conflict of interest

TM is inventor and University of Montpellier, INSERM and EPHE owners of the patent FR15

57093, describing the Hamlet test. Other authors declare no conflict of interest.

Acknowledgements

The authors thank Marc Criton (SATT AxLR) for continuous support, Eric Batut and Jean-

Paul Robert (Viewpoint) for the development of the commercial prototype and Jose-Miguel

Vela, Daniel Zamanillo (Esteve) and Montse Del Castillo (Envigo) for supplying S1R KO mice. We thank the CECEMA-UM animal facility for animal housing services.

Author contributions

LC, SC, CR, CP, RB, MAJ, AB performed experiments and analyzed the data. LC, SC, BD advised the study and corrected, or wrote parts of, the manuscript. TM designed experiments, analyzed the data and drafted the manuscript. All authors reviewed and approved the final manuscript.

26 References

Al-Saif A, Al-Mohanna F, Bohlega S. A mutation in sigma-1 receptor causes juvenile

amyotrophic lateral sclerosis. Ann Neurol. 2011;70:913-919. doi: 10.1002/ana.22534

Alonso G, Phan V, Guillemain I, Saunier M, Legrand A, Anoal M, Maurice T.

Immunocytochemical localization of the sigma1 receptor in the adult rat central nervous

system. Neuroscience. 2000;97: 155-170. doi: 10.1016/s0306-4522(00)00014-2

Arun AS, Eddings CR, Wilson KL. Novel missense alleles of SIGMAR1 as tools to

understand emerin-dependent gene silencing in response to cocaine. Exp Biol Med

(Maywood). 2019;244:1354-1361. doi: 10.1073/pnas.1518894112

Bayne K. Environmental enrichment and mouse models: Current perspectives. Animal Model

Exp Med. 2018;1:82-90. doi: 10.1002/ame2.12015

Bekinschtein P, Kent BA, Oomen CA, Clemenson GD, Gage FH, Saksida LM, Bussey TJ.

Brain-derived neurotrophic factor interacts with adult-born immature cells in the dentate

gyrus during consolidation of overlapping memories. Hippocampus. 2014;24:905-911.

doi: 10.1002/hipo.22304

Bindu B, Alladi PA, Mansooralikhan BM, Srikumar BN, Raju TR, Kutty BM. Short-term

exposure to an enriched environment enhances dendritic branching but not brain-

derived neurotrophic factor expression in the hippocampus of rats with ventral subicular

lesions. Neuroscience. 2007;144:412-423. doi: 10.1016/j.neuroscience.2006.09.057

Brailoiu GC, Deliu E, Console-Bram LM, Soboloff J, Abood ME, Unterwald EM, Brailoiu E.

Cocaine inhibits store-operated Ca2+ entry in brain microvascular endothelial cells:

critical role for sigma-1 receptors. Biochem J. 2016;473:1-5. doi: 10.1042/BJ20150934

Chevallier N, Keller E, Maurice T. Behavioural phenotyping of knockout mice for the sigma-1

(σ1) chaperone protein revealed gender-related anxiety, depressive-like and memory

alterations. J Psychopharmacol. 2011;25: 960-975. doi: 10.1177/0269881111400648

27 Crouzier L, Gilabert D, Rossel M, Trousse F, Maurice T. Topographical memory analyzed in

mice using the Hamlet test, a novel complex maze. Neurobiol Learn Mem.

2018;149:118-34. doi: 10.1016/j.nlm.2018.02.014

Crouzier L, Maurice T. Assessment of Topographic Memory in Mice in a Complex

Environment Using the Hamlet Test. Curr Protoc Mouse Biol. 2018;8:e43. doi:

10.1002/cpmo.43

D'Andrea I, Gracci F, Alleva E, Branchi I. Early social enrichment provided by communal nest

increases resilience to depression-like behavior more in female than in male mice.

Behav Brain Res. 2010;215:71-76. doi: 10.1016/j.bbr.2010.06.030

Dalwadi DA, Kim S, Schetz JA. Activation of the sigma-1 receptor by haloperidol metabolites

facilitates brain-derived neurotrophic factor secretion from human astroglia. Neurochem

Int. 2017;105:21-31. doi: 10.1016/j.neuint.2017.02.003

Delprat B, Crouzier L, Su TP, Maurice T. At the crossing of ER stress and MAMs: A key role

of sigma-1 receptor? Adv Exp Med Biol. 2020;1131:699-718. doi: 10.1007/978-3-030-

12457-1_28

Fabel K, Wolf SA, Ehninger D, Babu H, Leal-Galicia P, Kempermann G. Additive effects of

physical exercise and environmental enrichment on adult hippocampal neurogenesis in

mice. Front Neurosci. 2009;3:50. doi: 10.3389/neuro.22.002.2009

Fischer A, Sananbenesi F, Wang X, Dobbin M, Tsai LH. Recovery of learning and memory is

associated with chromatin remodelling. Nature. 2007;447:178-182. doi:

10.1038/nature05772

Fontanilla D, Johannessen M, Hajipour AR, Cozzi NV, Jackson MB, Ruoho AE. The

hallucinogen N,N-dimethyltryptamine (DMT) is an endogenous sigma-1 receptor

regulator. Science. 2009;323:934-7. doi: 10.1126/science.1166127

Fujimoto M, Hayashi T, Urfer R, Mita S, Su TP. Sigma-1 receptor chaperones regulate the

secretion of brain-derived neurotrophic factor. Synapse. 2012;66:630-639. doi:

10.1002/syn.21549

28 Fukunaga K, Moriguchi S. Stimulation of the sigma-1 receptor and the effects on

neurogenesis and depressive behaviors in mice. Adv Exp Med Biol. 2017;964:201-11.

doi: 10.1007/978-3-319-50174-1_14

Gualtieri F, Brégère C, Laws GC, Armstrong EA, Wylie NJ, Moxham TT, Guzman R, Boswell

T, Smulders TV. Effects of environmental enrichment on doublecortin and BDNF

expression along the dorso-ventral axis of the dentate gyrus. Front Neurosci.

2017;11:488. doi: 10.3389/fnins.2017.00488

2+ Hayashi T, Maurice T, Su TP. Ca signaling via sigma1-receptors: novel regulatory

mechanism affecting intracellular Ca2+ concentration. J Pharmacol Exp Ther.

2000;293:788-798.

Hayashi T, Su TP. Sigma-1 receptor chaperones at the ER-mitochondrion interface regulate

Ca2+ signaling and cell survival. Cell. 2007;131:596-610. doi:

10.1016/j.cell.2007.08.036

Hong J, Wang L, Zhang T, Zhang B, Chen L. Sigma-1 receptor knockout increases α-

synuclein aggregation and phosphorylation with loss of dopaminergic neurons in

substantia nigra. Neurobiol Aging. 2017;59:171-183. doi:

10.1016/j.neurobiolaging.2017.08.007

Ickes BR, Pham TM, Sanders LA, Albeck DS, Mohammed AH, Granholm AC. Long-term

environmental enrichment leads to regional increases in neurotrophin levels in rat

brain. Exp Neurol. 2000;164:45-52. doi: 10.1006/exnr.2000.7415

Kikuchi-Utsumi K, Nakaki T. Chronic treatment with a selective ligand for the sigma-1

receptor chaperone, SA4503, up-regulates BDNF protein levels in the rat

hippocampus. Neurosci Lett. 2008;440:19-22. doi: 10.1016/j.neulet.2008.05.055

Kilkenny C, Browne W, Cuthill IC, Emerson M, Altman DG; NC3Rs Reporting Guidelines

Working Group. Animal research: reporting in vivo experiments: the ARRIVE

guidelines. Br J Pharmacol 2010;160:1577-9. doi: 10.1111/j.1476-5381.2010.00872.x

29 Kourrich S, Hayashi T, Chuang JY, Tsai SY, Su TP, Bonci A. Dynamic interaction between

sigma-1 receptor and Kv1.2 shapes neuronal and behavioral responses to cocaine.

Cell. 2013;152:236-247. doi: 10.1016/j.cell.2012.12.004

Langa F, Codony X, Tovar V, Lavado A, Giménez E, Cozar P, Cantero M, Dordal A,

Hernández E, Pérez R, Monroy X, Zamanillo D, Guitart X, Montoliu L. Generation and

phenotypic analysis of sigma receptor type I (sigma 1) knockout mice. Eur J Neurosci.

2003;18:2188-96. doi: 10.1046/j.1460-9568.2003.02950.x

Li L, Xu B, Zhu Y, Chen L, Sokabe M, Chen L. DHEA prevents Aβ25-35-impaired survival of

newborn neurons in the dentate gyrus through a modulation of PI3K-Akt-mTOR

signaling. Neuropharmacology 2010;59:323-33. doi:

10.1016/j.neuropharm.2010.02.009

Liu Y, Chen GD, Lerner MR, Brackett DJ, Matsumoto RR. Cocaine up-regulates Fra-2 and

sigma-1 receptor gene and protein expression in brain regions involved in addiction

and reward. J Pharmacol Exp Ther. 2005;314:770-9. doi: 10.1124/jpet.105.084525

Lopez-Atalaya JP, Ciccarelli A, Viosca J, Valor LM, Jimenez-Minchan M, Canals S, Giustetto

M, Barco A. CBP is required for environmental enrichment-induced neurogenesis and

cognitive enhancement. EMBO J. 2011;30:4287-98. doi: 10.1038/emboj.2011.299

Lucas G, Rymar VV, Sadikot AF, Debonnel G. Further evidence for an antidepressant

potential of the selective sigma1 agonist SA4503: electrophysiological, morphological

and behavioural studies. Int J Neuropsychopharmacol 2008;11:485-95. doi:

10.1017/S1461145708008547

Martina M, Turcotte ME, Halman S, Bergeron R. The sigma-1 receptor modulates NMDA

receptor synaptic transmission and plasticity via SK channels in rat hippocampus. J

Physiol. 2007;578:143-157. doi: 10.1113/jphysiol.2006.116178

Matsuno K, Kobayashi T, Tanaka MK, Mita S. σ1 Receptor subtype is involved in the relief of

behavioral despair in the mouse forced swimming test. Eur J Pharmacol.

1996;312:267-271. doi: 10.1016/0014-2999(96)00497-9

30 Maurice T, Goguadze N. Sigma-1 (σ1) receptor in memory and neurodegenerative diseases.

Handb Exp Pharmacol. 2017;244:81-108. doi: 10.1007/164_2017_15

Maurice T, Phan VL, Urani A, Kamei H, Noda Y, Nabeshima T. Neuroactive neurosteroids as

endogenous effectors for the sigma1 (s1) receptor: pharmacological evidence and

therapeutic opportunities. Jpn J Pharmacol. 1999;81:125-155. doi: 10.1254/jjp.81.125

Maurice T, Roman FJ, Su TP, Privat A. Beneficial effects of sigma agonists on the age-

related learning impairment in the senescence-accelerated mouse (SAM). Brain Res.

1996;733:219-230. doi: 10.1016/0006-8993(96)00565-3

Maurice T, Su TP. The pharmacology of sigma-1 receptors. Pharmacol Ther.

2009;124:195-206. doi: 10.1016/j.pharmthera.2009.07.001

Maurice T. Beneficial effect of the sigma1 receptor agonist PRE-084 against the spatial

learning deficits in aged rats. Eur J Pharmacol. 2001;431:223-7. doi: 10.1016/s0014-

2999(01)01436-4

McQuaid RJ, Dunn R, Jacobson-Pick S, Anisman H, Audet MC. Post-weaning environmental

enrichment in male CD-1 mice: Impact on social behaviors, corticosterone levels and

prefrontal cytokine expression in adulthood. Front Behav Neurosci. 2018;12:145. doi:

10.3389/fnbeh.2018.00145

Mori T, Hayashi T, Hayashi E, Su TP. Sigma-1 receptor chaperone at the ER-mitochondrion

interface mediates the mitochondrion-ER-nucleus signaling for cellular survival. PLoS

One. 2013;8:e76941. doi: 10.1371/journal.pone.0076941

Moriguchi S, Sakagami H, Yabuki Y, Sasaki Y, Izumi H, Zhang C, Han F, Fukunaga K.

stimulation of sigma-1 receptor ameliorates depressive-like behaviors in CaMKIV null

mice. Mol Neurobiol 2015;52:1210-22. doi: 10.1007/s12035-014-8923-2

Mosaferi B, Babri S, Mohaddes G, Khamnei S, Mesgari M. Post-weaning environmental

enrichment improves BDNF response of adult male rats. Int J Dev Neurosci.

2015;46:108-114. doi: 10.1016/j.ijdevneu.2015.07.008

Navarro G, Moreno E, Bonaventura J, Brugarolas M, Farré D, Aguinaga D, Mallol J, Cortés

A, Casadó V, Lluís C, Ferre S, Franco R, Canela E, McCormick PJ. Cocaine inhibits

31 dopamine D2 receptor signaling via sigma-1-D2 receptor heteromers. PLoS One.

2013;8:e61245. doi: 10.1371/journal.pone.0061245

Neidl R, Schneider A, Bousiges O, Majchrzak M, Barbelivien A, de Vasconcelos AP,

Dorgans K, Doussau F, Loeffler JP, Cassel JC, Boutillier AL. Late-life environmental

enrichment induces acetylation events and nuclear factor κB-dependent regulations in

the hippocampus of aged rats showing improved plasticity and learning. J Neurosci.

2016;36:4351-61. doi: 10.1523/JNEUROSCI.3239-15.2016

Nithianantharajah J, Hannan AJ. Enriched environments, experience-dependent plasticity

and disorders of the nervous system. Nat Rev Neurosci. 2006;7:697-709. doi:

10.1038/nrn1970

Ovalle S, Andreu F, Pérez MP, Zamanillo D, Guitart X. Effect of the novel sigma1 receptor

ligand and putative atypical antipsychotic E-5842 on BDNF mRNA expression in the rat

brain. Neuroreport. 2002;13:2345-2348. doi: 10.1097/00001756-200212030-00035-

Rampon C, Jiang CH, Dong H, Tang YP, Lockhart DJ, Schultz PG, Tsien JZ, Hu Y. Effects of

environmental enrichment on gene expression in the brain. Proc Natl Acad Sci USA.

2000;97:12880-12884. doi: 10.1073/pnas.97.23.12880

Richetin K, Moulis M, Millet A, Arràzola MS, Andraini T, Hua J, Davezac N, Roybon L,

Belenguer P, Miquel MC, Rampon C. Amplifying mitochondrial function rescues adult

neurogenesis in a mouse model of Alzheimer's disease. Neurobiol Dis. 2017;102:113-

124. doi: 10.1016/j.nbd.2017.03.002

Romieu P, Phan VL, Martin-Fardon R, Maurice T. Involvement of the sigma1 receptor in

cocaine-induced conditioned place preference: possible dependence on dopamine

uptake blockade. Neuropsychopharmacology. 2002;26:444-55. doi: 10.1016/S0893-

133X(01)00391-8

Ryskamp DA, Korban S, Zhemkov V, Kraskovskaya N, Bezprozvanny I. Neuronal sigma-1

receptors: signaling functions and protective roles in neurodegenerative diseases.

Front Neurosci. 2019;13:862. doi: 10.3389/fnins.2019.00862

32 Sabino V, Cottone P, Parylak SL, Steardo L, Zorrilla EP. Sigma-1 receptor knockout mice

display a depressive-like phenotype. Behav Brain Res. 2009;198:472-476. doi:

10.1016/j.bbr.2008.11.036

Scharfman H, Goodman J, Macleod A, Phani S, Antonelli C, Croll S. Increased neurogenesis

and the ectopic granule cells after intrahippocampal BDNF infusion in adult rats. Exp

Neurol. 2005;192:348-356. doi: 10.1016/j.expneurol.2004.11.016

Schloesser RJ, Lehmann M, Martinowich K, Manji HK, Herkenham M. Environmental

enrichment requires adult neurogenesis to facilitate the recovery from psychosocial

stress. Mol Psychiatry. 2010;15:1152-1163. doi: 10.1038/mp.2010.34

Sha S, Qu WJ, Li L, Lu ZH, Chen L, Yu WF, Chen L. Sigma-1 receptor knockout impairs

neurogenesis in dentate gyrus of adult hippocampus via down-regulation of NMDA

receptors. CNS Neurosci Ther 2013; 19: 705-13. doi: 10.1111/cns.12129

Snyder MA, McCann K, Lalande MJ, Thivierge JP, Bergeron R. Sigma receptor type 1

knockout mice show a mild deficit in plasticity but no significant change in synaptic

transmission in the CA1 region of the hippocampus. J Neurochem. 2016;138:700-709.

doi: 10.1111/jnc.13695

Sølvsten CAE, de Paoli F, Christensen JH, Nielsen AL. Voluntary physical exercise induces

expression and epigenetic remodeling of VegfA in the rat hippocampus. Mol Neurobiol.

2018;55:567-582. doi: 10.1007/s12035-016-0344-y

Stranahan AM, Lee K, Martin B, Maudsley S, Golden E, Cutler RG, Mattson MP. Voluntary

exercise and caloric restriction enhance hippocampal dendritic spine density and BDNF

levels in diabetic mice. Hippocampus. 2009;19:951-61. doi: 10.1002/hipo.20577

Su TP, Hayashi T, Maurice T, Buch S, Ruoho AE. The sigma-1 receptor chaperone as an

inter-organelle signaling modulator. Trends Pharmacol Sci. 2010;31:557-566. doi:

10.1016/j.tips.2010.08.007

Takebayashi M, Hayashi T, Su TP. Nerve growth factor-induced neurite sprouting in PC12

cells involves sigma-1 receptors: implications for antidepressants. J Pharmacol Exp

Ther. 2002;303:1227-37.

33 Takebayashi M, Hayashi T, Su TP. Sigma-1 receptors potentiate epidermal growth factor

signaling towards neuritogenesis in PC12 cells: potential relation to lipid raft

reconstitution. Synapse. 2004;5390-103. doi: 10.1002/syn.20041

Tsai SY, Chuang JY, Tsai MS, Wang XF, Xi ZX, Hung JJ, Chang WC, Bonci A, Su TP.

Sigma-1 receptor mediates cocaine-induced transcriptional regulation by recruiting

chromatin-remodeling factors at the nuclear envelope. Proc Natl Acad Sci USA.

2015;112:E6562-E6570. doi: 10.1073/pnas.1518894112

Tsai SY, Hayashi T, Harvey BK, Wang Y, Wu WW, Shen RF, Zhang Y, Becker KG, Hoffer

BJ, Su TP. Sigma-1 receptors regulate hippocampal dendritic spine formation via a free

radical-sensitive mechanism involving Rac1xGTP pathway. Proc Natl Acad Sci USA.

2009;106:22468-22473. doi: 10.1073/pnas.0909089106

Urani A, Roman FJ, Phan VL, Su TP, Maurice T. The antidepressant-like effect induced by

sigma1-receptor agonists and neuroactive steroids in mice submitted to the forced

swimming test. J Pharmacol Exp Ther. 2001;298:1269-79.

Wu HE, Su TC, Hung D, Su TP. Epigenetic regulation by sigma-1 receptor in dorsal root

ganglion: Link to neuropathic pain. Soc Neurosci Abstr. Online. 2017;580.14.

Xu Q, Ji XF, Chi TY, Liu P, Jin G, Chen L, Zou LB. Sigma-1 receptor in brain

ischemia/reperfusion: Possible role in the NR2A-induced pathway to regulate brain-

derived neurotrophic factor. J Neurol Sci. 2017;376:166-175. doi:

10.1016/j.jns.2017.03.027

Yamaguchi K, Shioda N, Yabuki Y, Zhang C, Han F, Fukunaga K. SA4503, A potent sigma-1

receptor ligand, ameliorates synaptic abnormalities and cognitive dysfunction in a

mouse model of ATR-X syndrome. Int J Mol Sci. 2018;19:2811. doi:

10.3390/ijms19092811

Yasui Y, Su TP. PCAF regulates the acetylation of Sigma-1 receptor chaperones Soc

Neurosci Abstr. Online. 2015;778.07.

34 Zhang H, Cuevas J. Sigma receptors inhibit high-voltage-activated calcium channels in rat

sympathetic and parasympathetic neurons. J Neurophysiol. 2002;87:2867-2879. doi:

10.1152/jn.2002.87.6.2867

35 Figure legends

Figure 1 The Hamlet test: apparatus and procedure. (a) The device in the animal facility experimental room. (b) Map of the Hamlet showing the streets expanding from the central agora and the 5 functionalized houses. (c) The protocol used with 2 weeks long training, 4 h/day and 5 days-a-week, and the memory test performed twice, once in water-deprived condition (WD) and once in non-water-deprived condition (NWD). The repeated injection of

NE-100 (3 mg/kg ip) was performed on each training day immediately before and after the presence in the Hamlet. The acute injection was performed 20 min before the WD memory test.

Figure 2 Training profiles of (left panel, a-c) Saline- or NE-100-treated mice and (right panel, d-f) WT or S1R KO mice in the Hamlet. (a, d) Global activity recorded as presence in all houses during the 4 h training period each day. (b, e) Entries and (c, f) durations (s) of presence in each house (Inter., Interaction), expressed per hour per mouse, are shown as radar graphs for Day 1 (upper row), Day 6 (middle row) and Day 10 of training (lower row).

*** p < 0.001 vs. Day 1, # p < 0.05, ## p < 0.01 vs. Saline or WT; Newmann-Keuls' test.

Figure 3 Topographic memory in (a, b) Saline- or NE-100-treated and (c, d) WT or S1R

KO mice trained in the Hamlet. Three days after the last training session, trained (T) or non- trained (NT, naive to the Hamlet) mice were tested during 10 min for their ability to reach the

Drink house. Animals were tested in water-deprived (WD) or, after 24 h, in non-water- deprived (NWD) condition. Data show the latency to reach the Drink house (a, c, e) and the number of errors (b, d, f). In (a,b), mice were tested in drug-free condition but after a repeated b.i.d. treatment with saline or NE-100 during the training period. In (c,d), WT or S1R

KO mice were tested. In (e, f), mice were trained in a drug-free condition but injected with

Saline or NE-100 once, 20 min before the memory test in WD or NWD condition. The

36 number of animals per group is indicated within the columns in (a, c, e). * p < 0.05, ** p <

0.01, *** p < 0.001; paired t-test.

Figure 4 Impact of Hamlet training and NE-100 treatment on behavioral responses. (a)

Behavioral despair of non-trained (NT) and trained (T) mice in the forced swim test. Animals were forced to swim during 6 min and the immobility measured by videotracking during the last 5 min. The number of animals per group is indicated within the columns. ** p < 0.01,

Bonferroni's test. (b) Effect of the S1R agonist igmesine in the forced swim test in NT and T mice. Igmesine was administered at 10 or 30 mg/kg IP, 30 min before the forced swim test session. * p < 0.05, ** p < 0.01 vs. same treatment NT group; # p < 0.05, ## p < 0.01 vs. V- treated group in the same NT/T condition; Bonferroni's test. (c) Scopolamine-induced deficit in alternation performance in Saline- or NE-100-treated NT or T animals. Scopolamine was administered at 0.5 mg/kg SC 30 min before the Y-maze session. ° p < 0.05, °° p < 0.01, °°° p < 0.001; one-column t-test vs. 50% chance level.

Figure 5 Anti-amnesic effect of the S1R agonist PRE-084 against scopolamine-induced learning deficits in (left panel, a-d) NT mice and (right panel, e-j) T mice. (a, b, e, f) Dose- response effect of scopolamine and (c, d, g-j) effect of PRE-084 on scopolamine-induced deficits, in (a, c, e, g, i) the Y-maze test and (b, d, f, h, j) the passive avoidance test.

Scopolamine (0.5, 1.5, 5 mg/kg SC) or vehicle solution (V) was administered 20 min before the Y-maze test session or the passive avoidance training session, with retention being measured after 24 h. Dose-response effect of PRE-084 (0.1-3 mg/kg IP) administered 10 min before scopolamine (0.5-5 mg/kg SC), administered 20 min before the Y-maze test session or passive avoidance training session. The number of animals per group is indicated within the columns in (a, c, e, g, i). ° p < 0.05, °° p < 0.01, °°° p < 0.001 vs. 50% level, one column t- test in (a,c,e,g,i); ** p < 0.01, *** p < 0.001 vs. V-treated group; ## p < 0.01, ### p < 0.001 vs.

Scop-treated group; Dunn's test in (b,d,f,h,j).

37 Figure 6 Impact of Hamlet training on sigmar1 mRNA and S1R protein levels and bdnf mRNA levels. mRNA levels were analyzed by qPCR in the cortex (a, g) and hippocampus (b, h) and protein contents were analyzed by western blot in the cortex (c) and hippocampus (d) for non-trained (NT) mice or animals trained during 3 days (T 3d), 1 week (T 1w) and 2 weeks (T 2w). In (a, b, g, h), qPCR data were expressed as percentage of NT Sal-treated group. In (e, f), immunoreactive bands were normalized with b-actine expression and expressed as percentage of NT Sal-treated group in (c, d). * p < 0.05, ** p < 0.01, *** p <

0.001 vs. NT group, # p < 0.05, ## p < 0.01, ### p < 0.001 vs. Sal-treated group; Bonferroni's test.

Figure 7 Impact of the Hamlet training and S1R inactivation on hippocampal neurogenesis. (a, b) Cell proliferation. Newborn cells were visualized using Ki67 immunolabelling in the hippocampal dentate gyrus: (a) Typical micrographs, with Ki67- positive cells pointed out using arrowheads, and (b) quantification. (c, d) Cell maturation.

Newborn cell survival was visualized using DCX immunolabelling in the hippocampal dentate gyrus: (c) Typical micrographs showing DCX immunoreactivity and (d) quantification. Scale bar = 50 µm in (a), 2.5 µm in (c). * p < 0.05, ** p < 0.01, *** p < 0.001 vs. NT group; ## p <

0.01 vs. Sal-treated T group; @@ p < 0.01 vs. NT S1R KO group; Dunnett's test.

Figure 8 Impact of the Hamlet training and NE-100 treatment on acetylation levels of histones H2B, H3 and H4 in the mouse hippocampus. (a-f) Effect of training on acetylation levels of H2B, H3 and H4 histones; (g-j) impact of NE-100 treatment on training-induced increases in acetylation levels of H2B and H4 histones. Animals were sacrificed after 1 week or 2 weeks training (T 1w, T 2w) and compared to non-trained controls (NT) and histone acetylation levels were analyzed by western blot on chromatin preparation from the hippocampus. (a, c, e) typical blots and (b, d, f) quantifications of acetylated-to-total histone ratio for (a, b) H2B, (c, d) H3 and (e, f) H4. The number of animals per group is indicated

38 within the columns and all determinations was performed at least in duplicate. *** p < 0.001 vs. NT group; Bonferroni's test.

Figure 9 Impact of the Hamlet training and NE-100 treatment on acetylation levels of histones H2B, H3 and H4 in the mouse cortex. (a-f) Effect of training on acetylation levels of

H2B, H3 and H4 histones; (g-j) impact of NE-100 treatment on training-induced increases in acetylation levels of H2B and H4 histones. Animals were sacrificed after 1 week or 2 weeks training (T 1w, T 2w) and compared to non-trained controls (NT) and histone acetylation levels were analyzed by western blot on chromatin preparation from the cortex. (a, c, e) typical blots and (b, d, f) quantifications of acetylated-to-total histone ratio for (a, b) H2B,(c, d)

H3 and (e, f) H4. The number of animals per group is indicated within the columns and all determinations was performed at least in duplicate. * p < 0.05, ** p < 0.01 vs. NT group;

Bonferroni's test.

Supplementary Material

Supplementary Table 1. Statistical analyses, values and probabilities for each figure.

One column t-test vs. 50% Figure Group Stat. value Probability (ns: non-significant)

4c NT/Saline+V t(7) = 9.39 p < 0.0001 NT/Saline+Scop t(12) = 0.550 p > 0.05, ns T/Saline+V t(9) = 6.32 p < 0.0001 T/Saline+Scop t(9) = 3.64 p < 0.01 T/NE-100+V t(7) = 3.46 p < 0.05 T/NE-100+Scop t(7) = 2.11 p > 0.05, ns

One-way parametric ANOVA Figure Stat. value Probability

5a F(3,54) = 5.74 p < 0.01 5c F(3,55) = 5.53 p < 0.01 5e F(3,65) = 10.3 p < 0.0001 5g F(4,63) = 6.76 p < 0.001 5i F(4,73) = 7.57 p < 0.0001 6c F(3,44) = 0.349 p > 0.05, ns 6d F(3,39) = 6.71 p < 0.001 7b F(5,62) = 25.6 p < 0.0001 7d F(5,51) = 8.26 p < 0.01 8b F(2,14) = 5.97 p < 0.05 8d F(2,15) = 0.474 p > 0.05, ns 8f F(2,13) = 4.45 p < 0.05 9b F(2,14) = 4.46 p < 0.05 9d F(2,15) = 2.12 p > 0.05, ns 9f F(2,21) = 6.04 p < 0.01

Kruskal-Wallis non-parametric ANOVA Figure Stat. value Probability 5b H = 34.4 p < 0.0001 5d H = 34.6 p < 0.0001 5f H = 18.2 p < 0.001 5h H = 27.4 p < 0.0001 5j H = 22.1 p < 0.001

Two-way parametric ANOVA Figure Group Stat. value Probability Parameter

2a F(1,156) = 5.55 p < 0.05 treatment F(2,156) = 0.197 p > 0.05, ns day F(2,156) = 8.35 p < 0.001 interaction F(78,156) = 15.2 p < 0.0001 matching (repeated measure) 2d F(1,156) = 4.92 p < 0.05 genotype F(2,156) = 49.1 p < 0.0001 day F(2,156) = 1.59 p > 0.05, ns interaction F(78,156) = 6.67 p < 0.0001 matching (repeated measure) 3a Saline F(1,88) = 31.9 p < 0.0001 training F(1,88) = 6.05 p < 0.05 water-deprivation F(1,88) = 8.49 p < 0.01 interaction 3a NE-100 F(1,60) = 0.723 p > 0.05, ns training F(1,60) = 4.88 p < 0.05 water-deprivation F(1,60) = 0.585 p > 0.05, ns interaction 3b Saline F(1,88) = 29.5 p < 0.0001 training F(1,88) = 5.49 p < 0.05 water-deprivation F(1,88) = 13.3 p < 0.001 interaction 3b NE-100 F(1,60) = 9.81 p < 0.01 training F(1,60) = 6.87 p < 0.05 water-deprivation F(1,60) = 1.64 p > 0.05, ns interaction 3c WT F(1,36) = 33.3 p < 0.0001 training F(1,36) = 4.63 p < 0.05 water-deprivation F(1,36) = 2.40 p > 0.05, ns interaction 3c SIR KO F(1,44) = 16.6 p < 0.001 training F(1,44) = 0.017 p > 0.05, ns water-deprivation F(1,44) = 0.719 p > 0.05, ns interaction 3d WT F(1,36) = 20.1 p < 0.0001 training F(1,36) = 14.1 p < 0.001 water-deprivation F(1,36) = 0.430 p > 0.05, ns interaction 3d SIR KO F(1,44) = 4.46 p < 0.05 training F(1,44) = 0.983 p > 0.05, ns water-deprivation F(1,44) = 1.02 p > 0.05, ns interaction 4b F(1,74) = 17.7 p < 0.0001 training F(2,74) = 8.14 p < 0.001 igmesine treatment F(1,74) = 1.09 p > 0.05, ns interaction 6a F(3,36) = 3.18 p < 0.05 training F(1,36) = 4.90 p < 0.05 NE-100 treatment F(3,36) = 0.737 p > 0.05, ns interaction 6b F(3,52) = 2.94 p < 0.05 training F(1,52) = 2.00 p > 0.05, ns NE-100 treatment F(3,52) = 1.85 p > 0.05, ns interaction 6g F(3,78) = 1.74 p > 0.05, ns training F(1,78) = 38.0 p < 0.0001 NE-100 treatment F(3,78) = 20.5 p < 0.0001 interaction 6h F(3,81) = 15.1 p < 0.0001 training F(1,81) = 22.4 p < 0.0001 NE-100 treatment F(3,81) = 15.7 p < 0.0001 interaction 8h F(1,26) = 1.99 p > 0.05, ns training F(1,26) = 14.4 p < 0.001 NE-100 treatment F(1,26) = 28.9 p < 0.0001 interaction 8j F(1,23) = 7.27 p < 0.05 training F(1,23) = 2.09 p > 0.05, ns NE-100 treatment F(1,23) = 0.498 p > 0.05, ns interaction 9h F(1,25) = 4.09 p > 0.05, ns training F(1,25) = 5.66 p < 0.05 NE-100 treatment F(1,25) = 7.43 p < 0.05 interaction 9j F(1,29) = 4.22 p < 0.05 training F(1,29) = 9.81 p < 0.01 NE-100 treatment F(1,29) = 4.17 p < 0.05 interaction