WO2011141822A2

WO2011141822A2 - Methods for induction, predictive diagnosis, and treatment of affective behaviors by modulation of ppar and rxr receptors

Info

Publication number: WO2011141822A2
Application number: PCT/IB2011/001447
Authority: WO
Inventors: Wojtek Krezel
Original assignee: INSERM (Institut National de la Santé et de la Recherche Médicale); Centre National De La Recherche Scientifique (Cnrs); Université De Strasbourg
Priority date: 2010-05-11
Filing date: 2011-05-11
Publication date: 2011-11-17
Also published as: WO2011141822A3

Abstract

The present invention is directed to methods for modulating or detecting changes in peroxisome proliferator activated receptors (PPARs) and retinoid X receptors (RXR) to: 1) treat certain neuropsychiatric disorders, 2) identify susceptibility to such disorders in a clinical setting, and 3) screen novel pharmacological agents, including retinoids and retinoid derivatives for adverse psychiatric side-effects, and 4) generate certain neuropsychiatric disorders in preclinical conditions.

Description

METHODS FOR INDUCTION, PREDICTIVE DIAGNOSIS, AND TREATMENT OF AFFECTIVE BEHAVIORS BY MODULATION OF PPAR AND RXR RECEPTORS

FIELD OF THE INVENTION

[0001] The present invention is directed to methods for modulating or detecting changes in peroxisome proliferator activated receptors (PPARs) and retinoid X receptors (RXR) to: 1) treat certain neuropsychiatric disorders, 2) identify susceptibility to such disorders in a clinical setting, and 3) screen novel pharmacological agents, including retinoids and retinoid derivatives for adverse psychiatric side-effects, and 4) generate certain neuropsychiatric disorders in preclinical conditions.

BACKGROUND OF THE INVENTION

[0002] Depression covers a highly heterogeneous group of disorders, all of which share some common, core symptoms including affective abnormalities (depressed mood, despair, feelings of guilt or decreased interest in pleasurable (hedonic) stimuli such as food and sex), or more secondary cognitive symptoms, such as deficits in decision making, attention/working memory (WM). The variability in the phenomenology, comorbidity, and nosographic frontiers of depression has led to the classification of affective disorders into several types, including three major groups: major depression, dysthymia, bipolar disorder (BD). Although genetic factors contribute to the etiology of these diseases as it has been demonstrated for BD (Taylor et al., 2002), the epidemiology of these disorders suggest that environmental conditions constitute up to 50-60% of the risk factor in different types of depression (Fava and Kendler, 2000). The functional anatomy of affective disorders is not known. But on the evidence of human brain imaging studies, the dysfunction of prefrontal cortex-subcortical circuits and associated limbic regions were observed in BD, major and dysthymic depression (Drevets, 2001). In particular, increased activities of amygdala and nucleus accumbens were observed in functional MRI and PET neuroimaging of depressed patients (Anand and Shekhar, 2003; Drevets, 2003; Strakowski et al, 2005).

[0003] The observation that a majority of effective antidepressant treatments increase serotonergic and/or noradrenergic transmission, led to development of the monoaminergic hypothesis. An increase of serotoninergic and/or noradrenergic tone, which is a common point for the activities of classical antidepressants, appears to be effective only after several weeks of administration and is restricted to about 65% of patients, suggesting that monaminergic signaling may be only indirectly related to depression. Emerging hypothesis' of depression further suggest implications of abnormal signaling of CRF (Holsboer, 2000), AMPA receptors (Alt et al, 2006), G proteins and secondary messenger systems (Gould and Manji, 2002), or BDNF (Angelucci et al., 2005; Craddock et al., 2005). These hypotheses are not mutually exclusive since different signaling pathways are strongly inter-related and no specific genetic factor (including genes implicated in control of these signaling pathways) was identified as causal factor of depression.

[0004] Peroxisome proliferator activated receptors (PPARs) are ligand activated transcription factors, which belong to the superfamily of hormone nuclear receptors. There are three PPAR isotypes (PPARa, β, γ), which mediate signaling of dietary fatty acids in control of lipid homeostasis (Lee et al., 2003). All of the PPARs are expressed in the human and mouse brain suggesting that they can control specific CNS functions (Gofflot et al., 2007). Recently, PPAR signaling has been identified as a positive modulator of mnemonic processes (Campolongo et al., 2009; Mazzola et al., 2009), neurogenesis (Ramanan et al., 2009) or neuroprotection (Ramanan et al. , 2010; Deplanque et al., 2003; Bordet et al., 2006). The involvement of PPARa in reducing oxidative stress (Deplanque et al., 2003) or inflammatory responses via NF-kappaP and AP-1 pathways (Ramanan et al., 2008) are some of the mechanisms through which PPARa contributes to neuroprotection, whereas the mechanisms relevant to control of mnemonic functions are not known. And despite expression of PPARs in brain regions involved in emotional processing and affective behaviors, the role of PPARs in control of such functions has not been addressed, nor has their role as a potential therapeutic target for affective behaviors. Likewise, the relationship between retinoid X receptors— a an obligatory heterodimerisation partner of PPARs in gene transactivation— and PPARs has not been addressed as a potential therapeutic target for affective behaviors. [0005] Dopaminergic signaling— and in particular its mesolimbic pathway— also plays an important, reinforcing role in regulation of motivated/affective behaviors. Abnormally low dopaminergic signalling has been suggested to be involved in clinical depression (Millan, 2006; Nestler and Carlezon, 2006). Classical antidepressant treatments increase dopaminergic tone and its signalling via different dopamine receptor subtypes, which suggests a direct implication of dopamine in the efficiency of such treatments (Renard et al, 2001 ; Valentini et al, 2004; Willner et al, 2005). The dopaminergic reuptake inhibitor, Bupropion, has been found effective as an antidepressant treatment (Foley et al., 2006), although its actions also implicate noradrenergic transmission. Furthermore, some of the dopaminergic receptor ligands, such as bromocriptine and pergolide or pramipexole, are effective in the treatment of depression either as a monotherapy or as adjuvants (Corrigan et al., 2000; Mattes, 1997; Theohar et al., 1982). Although the dopamine receptor specificity of these agents is variable, all of them act as agonists of dopamine D2 receptor (D2R), which suggests that this receptor plays a particular role in the regulation of affective behaviors. This possibility is further supported by preclinical studies in rodent models used for research on depression. Thus, chronic mild stress leads to a reduction of D2R expression in the nucleus accumbens (Willner, 1997), and activation of D2R receptors has an antidepressant action in animal models of despair (Brocco et al., 2006; Siuciak and Fujiwara, 2004). Moreover, chronic antidepressant treatments including selective serotonin reuptake inhibitors (SSRIs), which primarily modulate serotonergic transmission, can also increase D2R expression in humans and rodents (Ainsworth et al., 1998; Dziedzicka- Wasylewska et al., 1997; Larisch et al., 1997). In line with such findings, an inhibition of D2R in human and animal models prevents antidepressant activities of fluoxetine (Prozac ®) and/or other antidepressants (Willner et al., 2005 and references therein).

[0006] The expression of D2R is modulated at the transcriptional level by retinoic acid (RA), an active form of vitamin A (Krezel et al., 1998; Samad et al., 1997). Such control implicates activities of retinoic acid receptors (RARa, β, γ) and retinoid X receptors (RXRa, β and γ), which in the form of heterodimers act as transcription factors and mediate RA signalling in vivo (Kastner et al., 1997). RAR and RXRy are the predominant retinoid receptors expressed in the striatum, including the nucleus accumbens (Krezel et al., 1999; Zetterstrom et al., 1999). Concomitant ablation of these receptors in RAR /RXRy double knockout mice leads to strong reduction of D2R expression in the dorsal and ventral striatum and marked locomotor deficits (Krezel et al., 1998). The involvement of murine retinoid receptors in the control of dopaminergic signaling in the striatum might suggest a potential role of the retinoid pathway in modulation of affective behaviours.

[0007] Such modulation is further suggested by clinical data on depression associated with altered retinoid signaling in acne vulgaris patients treated with isotretinoin (Bremner and McCaffery, 2007). A number of clinical reports describe cases of depression developed in patients who were treated for acne vulgaris with isotretinoin (Roaccutane® or Accutane®). In the US, 4992 cases of psychiatric side effects were recorded between 1982, when isotretinoin was introduced onto market and 2004, including 192 suicide attempts (Kontaxakis et al., 2009).

[0008] RXRs were also proposed to mediate genomic actions of n-3 polyunsaturated fatty acids (n-3 PUFAs) (de Urquiza et al., 2000; Lengqvist et al., 2004). Such functions of RXRs could be directly relevant for the pathology of affective disorders, as decreased n-3 PUFA signaling has been suggested to be associated with depression and use of n-3 PUFAs such as docosahexaenoic acid or eicosapentaenoic acid were reported beneficial in clinical conditions (Logan, 2004; Peet and Stokes, 2005) and in animal models used in research on depression (Carlezon et al., 2005; Naliwaiko et al., 2004).

[0009] Accordingly, there is a need in the art to better understand the role that RXRs play in the pathology of affective disorders, in order to identify: 1) modulators of RXRs and their potential use in treating and/or diagnosing affective disorders, 2) procedures to generate animal experimental models for use in research into neuropsychiatric disorders and 3) procedures to identify adverse neuropsychiatric effects of novel pharmacological treatments including in particular retinoid treatments. There is also a need in the art for further understanding the role that RXRs and PPARs play in the pathology of affective disorders, and a need to identify potential therapeutic targets to treat affective disorders based on PPAR/RXR signalling. There is likewise a need in the art for additional therapeutics capable of treating neuropsychiatric disorders and their symptoms by targeting non-monaminergic signaling modalities.

SUMMARY OF THE INVENTION

[0010] In one aspect, the present invention is directed to a methods of treating or ameliorating a neuropsychiatric disorder, including depression, by administering one or more PPAR modulators, one or more RXR modulators, or a combination of both. In one exemplary embodiment, the modulator is an agonist. In another exemplary embodiment, the agonist is a PPARa agonist. Exemplary PPARa agonist include fibrates, such as fenofibrate, bezafibrate, ciprofibrate, clofibrate, gemfibrozil, perfluorooctanoic acid, tetradecylthio acetic acid, N-Oleoylethanolamine, WY14643, CP-775146, CP-868388 or GW 7647 In another exemplary embodiment, the agonist is a PPARp agonist. Examplary PPAR β include GW501516 and 2-Bromohexadecanoic acid. In another exemplary embodiment the agonist is a pan-RXR agonist Exemplary pan-RXR agonist include retinoids and n-3 polyunsaturated acids. In one exemplary embodiment, the RXR agonist is selected from 9 cis-retinoic acid, docosahexaenoic acid (DHA), eicosapentaenoic acid, bexaroten, methoprene acid, oleic acid, phytanic acid, BMS649 called also UVI2108 or SR 1 1237, CD 3254.

[0011] In yet another aspect, the present invention is directed to a method for determining a subject's susceptibility to a neuropsychiatric disorder by assessing the subject's PPAR functionality, RXR functionality, or both, wherein a decrease in PPAR or RXR functionality indicates a susceptibility to a neuropsychiatric disorder. The receptor functionality may be assessed from a biological sample from the subject. In one exemplary embodiment, the neuropsychiatric disorder is major depressive disorder. The PPAR and RXR functionality may be assessed by looking at the subject's PPAR and/or RXR genotype, levels of PPAR and RXR, or PPAR and RXR activity. In one exemplary embodiment, the functionality is assessed by determining the subject's receptor genotype in order to detect single nucleotide polymorphisms or other mutations indicative of compromised PPAR and RXR functionality. In another exemplary embodiment, the functionality is assessed by determining the subject's receptor gene expression levels as compared to a standardized control, wherein a decrease in receptor gene expression levels as compared to the standardized control indicates an increased susceptibility to a neuropsychiatric disorder. In yet another embodiment, the functionality is assessed by determining the subject's receptor protein levels as compared to a standardized control, wherein a decrease in receptor protein levels as compared to the standardized control indicates an increased susceptibility to a neuropsychiatric disorder. In another exemplary embodiment, compromised PPAR or RXR signaling functionality is detected by determining the levels or activity of PPAR, RXR, or both and comparing those levels or activity before and after in vivo or ex vivo administration of PPAR and/or RXR specific ligands, wherein a failure to increase PPAR or RXR levels post-administration indicates an increased susceptibility to a neuropsychiatric disorder.

[0012] In another aspect, the present invention is directed to a method for screening adverse neuropsychiatric side effects of pharmacological agents by analyzing the functional, physiological and molecular effects of such agents on PPAR and RXR functionality. In one exemplary embodiment the method comprises administering the pharmacological agent to an experimental animal model or in vitro system and determining the effect on PPAR functionality, RXR functionality, or both. The effect on PPAR and RXR functionality can be determined at the molecular level using the same methods for assessing functionality as discussed in the above method for determining a subject's susceptibility to a neuropsychiatric disorder. The pharmacological agent's effect on PPAR and RXR functionality may be further assessed using a behavioral test such as a forced swim test, a sucrose preference test, Y-maze spontaneous alteration test, a delayed non-match to place test, operant paradigms or prepulse inhibition or comparable behavior test assessing despair and/or cognitive activity. In addition, in vitro systems including cells cultured in vitro and tissue explants, may be used to provide an assessment of RXR PPAR functionality. In one exemplary embodiment, the experimental animal model is a rodent strain carrying mutations compromising or inactivating PPAR or RXR signaling or a rodent strain in which PPAR or RXR signaling was abolished pharmacologically using an antagonist.. In another exemplary embodiment, the pharmacological agent being tested is a retinoid or retinoid derivative.

[0013] In yet another aspect, the present invention is directed to a method for identifying pharmacological agents useful in treating neuropsychiatric disorders comprising administering the pharmacological agent to an experimental animal or an in vitro system, wherein the experimental animal or in vitro systems has a pharmacologically and, or genetically compromised PPAR or RXR functionality and determining whether PPAR or RXR dependent functions are restored. The effect on PPAR and RXR signaling functionality can be assessed using a behavioral test such as a forced swim test, sucrose preference test, Y-maze spontaneous alteration test, a delayed non-match to place test, operant paradigms or prepulse inhibition or comparable behavior test assessing despair and/or cognitive activity, wherein a decrease in despair behaviors and/or increase in cognitive behaviors indicates restoration of PPAR and RXR dependent functionality. Alternatively, the effect on PPAR and RXR functionality can be determined by looking at downstream markers, such as transcriptional targets of PPAR and RXR signaling or transcriptional markers associated with depression. In one exemplary embodiment, restoration of RXR functionality can be assessed by determining the level of dopamine D2 receptors or dopamine D2 receptor activity. In one exemplary embodiment, the experimental animal is a RXR-PPAR double knock-out mouse. In another exemplary embodiment, the experimental animal is a RXRy knock-out animal. In yet another embodiment, the experimental animal is a PPAR knock-out animal. In yet another embodiment, the experimental animal is a RXRy +/- PPAR +/- animal. In another embodiment, the experimental animal is a wild type animal in which PPAR functionality, RXR functionality has been compromised by administration of a PPAR or RXR antagonist.

BRIEF DESCRIPTION OF THE FIGURES

[0014] Figure 1. Implication of RXRs in DHA and retinoid control of despair behaviors in mice. The immobility time in the forced swim test, used as a measure of despair behaviors was reduced in CBy mice after acute treatment with 30mg/kg of fluoxetine (A), or 5hrs after acute treatment with lmg/kg of DHA (n =10, n =10,

veh 0.1 n =10), or with lmg/kg of UVI2108 (n =10, n =5, n =10), but not BR1211 (n =4,

1 veh 0.1 1 veh ^η ₀ =5_> n =8) (B). The effects of DHA were mimicked by atRA (n ^=11, n_Q =5, n =13, n =12) but not by pan-RAR TTNPB (n =10, n =10, n =6, n =8) (C). The effects of

10 veh 0.1 1 10

DHA (lmg/kg) or atRA (5mg/kg) treatments were suppressed by lmg/kg of pan-RXR antagonist BR1211 (D). The numbers of animals tested in each group are indicated in corresponding histograms or in the legend. DHA, docosahexaenoic acid; atRA, all-trans retinoic acid; UVI2108, a pan-RXR agonist (named also BMS649), BR121 1 ; a pan-RXR antagonist; ***, p<0.001 with respect to the corresponding vehicle treated group.

[0015] Figure 2. Implication of RXRs in DHA and retinoid control of spontaneous alternation in mice. Administration of DHA in CBy mice (n =19, n =8, veh 0.1 n =13) facilitated working memory, measured as spontaneous alternation (SPA) in the Y- maze, in the dose dependent manner and such effects were mimicked by RXR agonists, UVI2108 (n =10, n =5, n =10) (A). The promnemonic effects of DHA were also veh 0.1 1

mimicked by atRA (n =12, n =6, n =14, n =6) treatment, but not by administration of

^J veh 1 5 10 ^{7 J}

TTNPB (n =10, n =10, n =9, n =7), a pan-RAR agonist (B). The promnemonic veh 1 5 10

activities of DHA (lmg/kg) and atRA (5mg/kg) were blocked by co -treatment with lmg/kg of pan-RXR antagonist, BR121 1 (C). The numbers of mice used in each group are indicated in corresponding bar graphs or in the legend. ***, p<0.001 , **, p<0.01 significantly different from vehicle treated group.

[0016] Figure 3. RXRy mediates DHA and retinoid effects on despair behaviors and spontaneous alternation. RXRy appeared critical for mediating activities of DHA and pan-RXR synthetic agonist, since increased despair behaviors (A) displayed by RXRy-/- mice, raised on 60%C57BL6J x 40%129SvEms/j background were not sensitive to DHA (lmg/kg) or UVI2108 (lmg/kg) treatment, whereas such treatments significantly improved performance of RXRy heterozygous mice (B). Similarly, working memory deficits (C) were reversed by the DHA or UVI2108 treatments only in RXRy+/-, but not in RXRy-/- mice (D) The numbers of mice used in each group are indicated in corresponding bar graphs. ***, p<0.001 ; **, p<0.01 ; *, p<0.05 significantly different from corresponding group of WT non-treated mice; ###, p<0.001 ; ##, p<0.01 ; #, p<0.05 different from WT mice in the same treatment group; §§§, p<0.001 different from WT and RXRy+/- mice in the same treatment group.

[0017] Figure 4. Implication of RXRs in DHA and retinoid control of working memory in delayed non-match to place task. The performance of RXRy-/- mice and their WT littermate controls were expressed as percent of correct choices. Following 10 days of training, ITIs were increased to test working memory. The phase of pharmacological tests was indicated as "Drug treatments" and was followed by tests of critical ITIs in non-treated mice (A). The effects of drug treatments in RXRy-/- and WT mice were expressed as percent of correct choices (B). *, p<0.05 as compared to corresponding group of WT, non-treated mice (A) and WT, vehicle treated mice (B).

[0018] Figure 5. Increased despair behavior in RXRy-/- null mutant mice is reversible by chronic antidepressant treatment. The time of immobility in the forced swim test was measured in naive n =6 and n =8, n =8 single and n

WT RARP-/- RXRy-/- ° RARp-/-RXRy-/-

=8 double mutant littermates (A). Chronic, 21 -day long treatment with fluoxetine (20 mg/kg/24hrs) reduced the immobility time in n =10 mice as compared to n =10

° ° ^J RXRy-/- RXRy-/- mice fed with control diet but not in n =8 mice as compared to n =1 1 mice fed with

WT WT

control diet (B). Data are presented as mean values ± SEM. p<0.001 with respect to non-treated (A) or vehicle -treated (B) WT mice; ##, p<0.01 as compared with vehicle- treated RXRy-/- mice.

[0019] Figure 6. Loss of sucrose preference in RXRy-/- mice is reversible by chronic antidepressant treatment. Sucrose preference was measured as percent of sucrose solution consumption with respect to total amount of liquid consumed during the night phase in n =1 1 and n =1 1 , n =7, and n =7 null mutant mice

WT RARP-/- RXRy-/- RARp-/-RXRy-/-

(A). Chronic, 19-day long treatment with fluoxetine (20 mg/kg/24hrs) normalised sucrose preference in n =18 mice as compared to n =22 mice fed with control diet but

RXRy-/- RXRy-/- not in n =13 mice as compared to n =15 mice fed with control diet (B). Data are

WT WT

presented as mean values ± SEM. *, p<0.05 significantly different from WT group; **, p<0.01 different from vehicle treated WT mice.

[0020] Figure 7. Abnormal serotonergic signaling is not sufficient to generate despair behavior in RXRy-/- mice. Total tissue levels of 5HT were standardised by tissue weight and compared for the hippocampus and striatum in n =6 and n =6

° WT RXRy-/- mice (A). For the same animals and structures the serotonin turnover was calculated as ratio of standardised measures of 5HIAA and 5HT (B). Acute treatment with fluoxetine (20mg/kg) did not alter immobility time in the forced swim test neither in _/ ⁼8 mice as compared to n =8 vehicle-treated mice nor in n =7 mice as compared to n =9

RXRy-/- WT WT vehicle -treated mice (C), whereas the same dose significantly reduced immobility time in control strain of n =6 mice as compared to n =6 vehicle -treated mice (D). Data

BALBc BALBc

are presented as mean values ± SEM. *, p<0.05 or **, p<0.01 for selected comparisons and p<0.001 in comparison with vehicle treated BALBc mice.

[0021] Figure 8. Decreased expression of dopamine D2R receptor mRNA in the nucleus accumbens of RXRy-/- mice is reversed by chronic fluoxetine treatment.

D2R mRNA levels were measured by quantitative real-time RT-PCR and are presented as relative to the expression of the housekeeping gene 36B4, ⁿ _WT ⁼9, n^_R =10 (A). In situ hybridisation detection of D2R mRNA is shown in the whole striatum (left), and at high magnification in selected regions (boxed) of the caudate putamen (CPu), nucleus accumbens shell (NAcSh) and core (NAcCo) in WT and RXRy-/- mice (B). The numbers of D2R-positive cells in the shell and core of the NAc are presented as relative to D2R cell counts in the adjacent dorsal part of the CPu on the same section, for n =6 and

^J WT

n =6 mice (C). The effects of 19-days of chronic fluoxetine treatment on the relative

RXRy-/- ' ^J

numbers of D2R positive cells in NAcSh in n =3 and n =5 mice were compared

WT RXRy-/- with n =4 and n =4 mice fed control diet (D). Data are presented as mean values ±

WT RXRy-/-

SEM. **, p<0.01 ; ***, p<0.001 different from corresponding WT group or ##, p<0.01 for selected comparison.

[0022] Figure 9. Haloperidol induction of c-fos expression is impaired in the NAc shell of RXRy-/- mutants. The brain regions used for c-fos counts are schematized (A), c-fos positive cells were scored in selected regions of the dorsal striatum (striped area; CPu) (A), the nucleus accumbens shell (NAcSh) and core (NAcCo) (B). c-fos positive cells were counted for each structure, and are presented as means ± SEM (C) or as the ratio of haloperidol / vehicle induced c-fos cells for each genotype (D). Each experimental group consisted of n=6 animals. Data are presented as mean values ± SEM. **, p<0.01 , ###, p<0.001. [0023] Figure 10. Behavioral responses to haloperidol in RXRy-/- mice. Twenty minutes after treatment with saline (0) or haloperidol (0.1, 0.2, 1 or 2 mg/kg), locomotor activity was measured in WT and RXRy-/- mice in the open field test during 5 minutes. The number of animals tested in each genotype/treatment group was : n =7, n RXR_Y_/_

=7 , n =8 , n = 10, n =8 , n =7 , n =7 , n =7, n =7,

/0 WT/0.1 RXR -/-/O.I WT/0.2 RXRy-/-/0.2 WT/1 RXRy-/-/l WT/2 n =10 (A). Catalepsy was measured in the bar test 30 minutes after vehicle (0) or

RXR -/-/2 ^J

haloperidol (0.2 or 2mg/kg) injection (B). The number of animals tested in each genotype/treatment group was: n =8, n =8, n =10, n =7, n =8,

° ^{J i} ° ¹ WT/0 RXRy-/-/0 WT/0.2 RXRy-/-/0.2 WT/2 n =8. Data are presented as mean values ± SEM. * p<0.05 with respect to WT

RXRy-/-/2

haloperidol-treated group.

[0024] Figure 11. Re-expression of RXRy in NAc shell reverses depressive-like b e h av i o r s i n RXRy-/- mice through D2R dependent mechanism.

Immunohistochemical detection of RXRy in the dorsal caudate putamen (CPu) and nucleus accumbens (NAc) shell of WT non-infected mice (A, top row) or RXRy-/- mice after infection with AAV2 vector expressing GFP (A, middle row) or RXRy (A, bottom row). Re-expression of RXRy in the NAc shell reduced immobility in the forced swim test (B) and restored sucrose preference (C) in n =7 mice as compared with n

RXRy-/- RXRy-/-

=5 mice infected with AAV2-GFP. Acute inhibition of D2R signaling in the NAc shell after bilateral infusion of raclopride (5 μg/side) prevented AAV2-RXRy rescue of depressive-like behaviors in RXRy-/- mice as measured in the forced swim test for n =5 and n =7 infused mice (D). Locomotor activity in the open field test was raclopride ACSF

studied in the same mice 2 days after forced swim test (E). Data are presented as mean values ± SEM. *, p<0.05 in comparison with RXRy-/- mice infected with AAV2-GFP; **, p<0.01 in comparison with ACSF infused control group.

[0025] Figure 12. Expression of D2R in NAc shell reverses depressive-like behaviors in RXRy-/- mice. In situ hybridisation detection of D2R transcripts in the dorsal caudate putamen (CPu) and nucleus accumbens (NAc) shell of WT non-infected mice (A, top row) or in RXRy-/- mice after infection with AAV2 vector expressing GFP (A, middle row) or D2R (A, bottom row). Expression of D2R in the NAc increased locomotor activity during 30 min of the open field test (B). Antidepressant effects of D2R expression were evidenced by reduced immobility in the forced swim test (C) and restored sucrose preference (D) in RXRy-/- mice. The mean scores for n =7 and n XRY

AAV2-D2R =7 infected R

1 -/- mice were

in comparison with RXRy-/- mice infected with AAV2-GFP.

[0026] Figure 13. Null mutation of RXRy does not affect locomotor activity and anxiety.Locomotor activity was not affected by RXRyablation (genotype effect F[l ,651 ]=0.7, ns and genotype x iz^'meinteraction F[31 ,651]=0.93, ns) as measured in nwT= l l and nRXRy-/-= 12 mice during 32hrs in the actimetric cages, including 8hrs ofhabituation to the novel environment (1 l am-7pm) and activity in the familiar environment during entire darkand light phase (7pm- 7pm) (A). Locomotor reactivity to the novel environment in the open field test did not differ between

mice with respect to the distance covered during 30min and illustrated for 5min epochs (genotypeeffect F[l ,50]=0.27, ns and genotype x iz^'meinteraction F[5,50]=0.96, ns) (B). Anxiety evaluated by the percent of time spent in the central, anxiogenicpart of the open field during initial 10 minutes of the test was not affected by RXRy inactivation (t=-l .l , ns).

[0027] Figure 14. Expression of D1R is not affected in the striatum of RXRy-/- mice. Expression of dopamine D1R receptor mR A in the dorsal striatum (CPu) and nucleus accumbens (NAc) were measured by quantitative real-time RT-PCR and are presented as relative to the expression of the housekeeping gene 36B4,

nRXRy-/- =10.

[0028] Figure 15. Effects of AAV2 mediated RXRyor D2R expression on locomotor activity and anxiety in RXRy-/-mice.(A) Virus mediated expression of RXRydid not change locomotor activity of RXRy-/-mice during 32hrs in the actimetric cages, whereas expression of D2R altered activity of RXRy-/-mice as compared to GFP infected group (F[31 ,372]=1.75; p<0.01 , A OVA on repeated measures). Using student t-test post-hoc test we found that such difference is related to increased activity of AAV2- D2R injected mice during the first hour of the test (t=2.27; p<0.05). (B) Infection with AAV2-D2R, but not AAV2-RXRg increased also locomotor reactivity to novel environment of the open field (F[5,60]=3.74; p<0.01 , A OVA on repeated measures) as compared to AAV2-GFP infected mice and illustrated for 5min epochs of the 30minute- long test. (C) Anxiety evaluated by the percent of time spent inthe anxiogenic, central part of the open field during initial 10 minutes of the test wasnot affected by virus mediated expression of RXRyor D2R in RXRy-/-mice. nAAV2-GFP=7,

nAAV2- D2R=7. *, p<0.05 for comparison of AAV2-D2R and AAV2-GFP mice.

[0029] Figure 16. Increased PPAR signaling reduces despair behaviors in mice. Similarly to chronic, 21 -day antidepressant treatment with fluoxetine (10mg/kg/24hrs administered as a food supplement; A), fenofibrate (Fen)— a PPARa agonist reduced— immobility time in the forced swim test in a dose dependent manner in CBy mice (B). Subthreshold, lOmg/kg dose of fenofibrate synergized with subthreshold, O.lmg/kg dose of methoprene acid(MA), a RXR agonist, to reduce immobility time in CBy mice (C). p<0.01 and ***, p<0.001 as compared to respective vehicle group.

[0030] Figure 17. Antidepressant-like activities of fenofibrate in chronic stress model of depression— despair. Two weeks of chronic stress led to significant increase of despair behavior in C57BL6N mice (immobility in the forced swim test), which was normalized by intraperitoneal (IP) treatment with lOmg/kg of fluoxetine. Similar antidepressant effects displayed acute (IP) treatment with fenofibrate (50mg/kg) or chronic fenofibrate administration (in food) at the dose of 15mg/kg/24hrs. p<0.001 ; **, p<0.01, *p<0.05 as compared to chronically stressed, vehicle-treated mice.

[0031] Figure 18. Antidepressant-like activities of fenofibrate in chronic stress model of depression in mice— anhedonia. C57BL6N mice subject to the protocol of 5 weeks of chronic stress displayed anhedonia on evidence of significantly lower sucrose preference as compared to nonstressed mice. Such anhedonia was reversed by chronic (10 days) antidepressant treatment with 15mglkg/24hrs of fluoxetine or by 10 days of treatment with 15mg/kg/24hrs of fenofibrate. **, p<0.01 , *p<0.05 as compared to chronically stressed, non-treated mice.

[0032] Figure 19. Compromised PPARa and RXRy signaling synergise to increases despair behaviors in mice. Genetically compromised RXRy and PPARa signaling in RXRy+/-and PPARa+/-heterozygous mice synergise in increasing despair behaviors in mice as compared to compound PPARa or RXRy heterozygous mice. *, p<0.05; **, p<0.01 and ***, p<0.001 with respect to WT mice; #, p<0.05; ##, p<0.01 for selected comparisons.

[0033] Figure 20. Increased susceptibility of PPARa-KO mice to develop depressive-like behaviors under isotretinoin treatment— despair. Seven days of treatment with clinical dose of isotretinoin (lmg/kg) increased despair behaviors in PPARa-KO mice on evidence of significantly increased immobility time in the forced swim test as compared to isotretinoin treated WT mice or non-treated WT and PPARa-KO mice. p<0.001; **, p<0.01 for selected comparisons.

[0034] Figure 21. Increased susceptibility of PPARa-KO mice to develop depressive-like behaviors under isotretinoin treatment— anhedonia. Nine days of treatment with isotretinoin (lmg/kg) induced anhedonia in PPARa-KO mice on evidence of significantly decreased sucrose preference in the sucrose preference test as compared to isotretinoin treated WT mice or non-treated WT and PPARa-KO mice. *, p<0.05; **, p<0.01 for selected comparisons.

DETAILED DESCRIPTION OF THE INVENTION

[0035] The present invention is directed to the diagnosis and treatment of various neuropsychiatric disorders by identifying deficiencies in, and modulating the activity of, peroxisome proliferator activated receptors (PPARs) and retinoid X receptors (RXRs), either alone or in combination. In addition, the present invention is directed to methods for determining a subject's susceptibility to a neuropsychiatric disorder, for assessing a pharmacological agent's potential to cause an adverse neuropsychiatric side effects, and identifying new pharmacological agents useful for treating certain neuropsychiatric disorders.

[0036] Peroxisome proliferator activated receptors (PPARs) are ligand activated transcription factors, which belong to the superfamily of hormone nuclear receptors. There are three PPAR isotypes (PPARa, β, γ), which mediate signaling of dietary fatty acids in control of lipid homeostasis (Lee et al., 2003). All of the PPARs are expressed in the human and mouse brain suggesting that they can control specific CNS functions (Gofflot et al., 2007). RXRs are nuclear receptors that bind to a variety of ligands derived from cholesterol, fatty acids, and glucose and function as an obligatory heterdimerization partner with PPAR in regulating gene functions in certain parts of the brain, such as the nucleus accumbens (Wietrzych-Schindler et al. (2011). As used herein, PPAR and RXR functionality refers to the ability of PPAR and RXR to regulate gene transcription through their heterodimerization. Functionality can be measured by determining a subject or experimental animal's PPAR and RXR genotype, PPAR and RXR activity, PPAR and RXR levels, or a subject or experimental animal's response to behavioral test measuring despair or cognitive activities. Activity can be measured by biomarkers such as, but not limited to, beta-hydroxybutyrate for ketone body synthesis, which in starving conditions is reduced when PPAR activity is compromised, lipid metabolism associated markers, such as triglycerides, total cholesterol and its HDL and LDL fractions, which are generally reduced after activation of PPAR, and detection of direct and indirect PPAR transcriptional targets (Brun et al, Schmuth et ah, and Sheu et al). The term PPAR and RXR includes functional homologues and variants of the known isotypes of PPAR and RXR that retain PPAR and RXR activity. As used herein, levels of PPAR or RXR can refer to either the amount of a PPAR or RXR present in a cell or tissue, or the amount of PPAR or RXR gene expression.

Methods for Treating Neuropsychiatric Disorders

[0037] In one exemplary embodiment, the present invention provides methods of treating or ameliorating a neuropsychiatric disorder by administering a PPAR modulator, a RXR modulator, or combination thereof to a subject with a neuropsychiatric disorder or suffering from related psychiatric symptoms. As used herein, the terms "PPAR modulator" or "RXR modulator" include those compounds (i.e. chemicals, polypeptides, polynucleotides) that do one or more of the following, respectively: increase PPAR or RXR gene expression, increase PPAR or RXR levels in the cell, increase PPAR or RXR activity or signaling, facilitate PPAR and RXR dimerization and activation of gene transcription functions, or down regulate inhibitors of any of the above functions. The terms "treat", "treating" or "treatment" refer to the elimination or amelioration of, one or more, psychiatric symptoms associated with a neuropsychiatric disorder. [0038] Exemplary neuropsychiatric disorders that may be treated with methods of the present invention include, but are not limited to different forms of depression such as major depressive disorder, bipolar disorder, dysthymia, but also schizophrenia, attention- deficit hyperactivity disorder, body dysmorphic disorder, bulimia nervosa and other eating disorders, cataplexy, fibromyalgia, general anxiety disorder, impulse-control disorders, panic disorder, and post-traumatic stress disorder. In certain embodiments, the methods of the present invention may be used to treat certain symptoms of certain neuropsychiatric disorders including, but not limited to depressed mood, despair, increased feelings of guilt, decrease in interest in certain pleasurable (hedonic) stimuli such as food and sex, or secondary cognitive symptoms such as deficits in decision making, attention, and working memory (WM).

[0039] In certain embodiments, the present invention comprises administration of a pharmaceutically effective amount of a PPAR modulator to a subject with a neuropsychiatric disorder. The administration of the PPAR modulator may treat the neuropsychiatric disorders by eliminating or ameliorating certain psychiatric symptoms associated with a neuropsychiatric disorder. In one exemplary embodiment, the PPAR agonist of the present invention is administered to a subject with major depressive disorder, dysthymia, bipolar disorder, or schizophrenia. In another exemplary embodiment, the PPAR agonist is administered to a subject with a major depressive disorder.

[0040] In one exemplary embodiment the PPAR modulator is a PPAR agonist, or lactivator of PPAR activity. The PPAR agonist may activate PPAR activity by dissociating, blocking or inhibiting a negative regulator of PPAR activity. In another exemplary embodiment, the PPAR agonist may activate or facilitate association of the PPAR with its cognate heterodimerization partner. In another exemplary embodiment, the PPAR agonist may activate or facilitate PPAR's ability to associate with a retinoid X receptor (RXR).

[0041] In one exemplary embodiment, the PPAR agonist of the present invention is a PPARa, PPARp, or PPARy agonist. In another exemplary embodiment, the agonist is a PPARa agonist. In one exemplary embodiment the PPAR a agonist is a fibrate, GW501516, 2-Bromohexadecanoic acid. In another exemplary embodiment, the PPARa agonist is a Fenofibrate, Bezafibrate, Ciprofibrate, Clofibrate, Gemfibrozil, perfluorooctanoic acid, tetradecylthioacetic acid, N-Oleoylethanolamine, WY14643, CP- 775146, CP-868388 or GW7647, or a combination thereof. In another exemplary embodiment, the agonist is a PPAR β agonist. Exemplary PPAR β agonist include, but are not limited to GW501516 and 2-bromohexadecanoic acid.

[0042] In certain other embodiments, the present invention comprises administration of a pharmaceutically effective amount of a RXR modulator to a subject with a neuropsychiatric disorder. The administration of the RXR modulator may treat the neuropsychiatric disorder by eliminating or ameliorating certain psychotic symptoms associated with neuropsychiatric disorders. In one exemplary embodiment, the RXR modulator of the present invention is administered to a subject with major depressive disorder, dysthymia, bipolar disorder, or schizophrenia. In another exemplary embodiment, the RXR agonist is administered to a subject with major depressive disorder.

[0043] In one exemplary embodiment, the RXR modulator is a pan-RXR agonist, or activator of RXR activity. In another exemplary embodiment, the RXR agonist may activate RXR activity by dissociating, blocking or inhibiting a negative regulator of RXR activity. In another exemplary embodiment, the RXR agonist may activate or facilitate association of the RXR with its cognate heterodimerization partner. In another exemplary embodiment, the RXR agonist may activate or facilitate RXR's ability to associate with a peroxisome proliferator activated receptor (PPAR).

[0044] In one exemplary embodiment, the RXR agonist is a retinoid. In one exemplary embodiment, the RXR agonist is 9 cis-retinoic acid. In another exemplary embodiment, the RXR agonist is a n-3 polyunsaturated fatty acid. In another exemplary embodiment, the n-3 polyunsaturated fatty acid is docosahexaenoic acid, eicosapentaenoic acid, or a combination thereof. In yet another exemplary embodiment, the RXR agonist is selected from, but not limited to, the following: bexarotene, BMS649 also called UV12108, diphenylamine derivatives (Biol. Pharm. Bull. 1998, 21 :544-546; AGN194204 (Mol. Cell Biol. 1999, 19(5):3372-382); LGD 1069 (Boehm et al. J Med. Chem. 1994, 37:2930-41); LGI00268 (Bodim et al. J. Med. Chem. 1995, 38:3146-3155); PA024 (Honda et al. Am J Physiol. Endocrinol. Metab. 2002, 283:326-331); 9-cis- UAB30 (Kapetanovic et al. Intl. J. Tox. 2010, 29(2):157-164); NEt-3IP and Net-31B (Takamatu et al. ChemMedChem, 2008, 3(5):780-7); and LG100754, methoprene acid, oleic acid, phytanic acid, SR 11237, or CD 3254 .

[0045] The PPAR and RXR modulators may be administered alone, in a single combined formulation, sequentially, or concurrently. The PPAR and RXR modulators of the present invention may be administered by standard routes of administration including oral, parenteral, topical, intranasal, rectal or vaginal. The PPAR and RXR agonist of the present invention may be formulated with standard pharmaceutically acceptable carriers, preservatives, anti-oxidants, excipients, and flavoring agents. Pharmaceutical compositions suitable for delivery of modulators of the present invention and methods for their preparation will be readily apparent to those skilled in the art. Such compositions and methods for their preparation may be found, for example, in Remington's Pharmaceutical Sciences (19^th edition, Mack Publishing Company, 1995).

[0046] In one exemplary embodiment, the PPAR and RXR modulators of the present invention are administered in daily or sub-daily doses. In other exemplary embodiments the PPAR and RXR modulators are administered every 2 days, every 3 days, every 4 days, every 5 days, every 6 days. In yet another exemplary embodiment, the PPAR and RXR modulators are administered weekly.

[0047] The dose of the PPAR and RXR modulators of the present invention will depend on the formulation and route of administration as well as patient specific factors such as age, weight, sex, and the type and severity of the neuropsychiatric disorder or symptoms to be treated. The appropriate dosage may be readily determined by one of ordinary skill in the art taking into consideration the above factors. In one exemplary embodiment, a pharmaceutically effective dose of the PPAR and/or RXR is between about 1 mg and about 1 g; between about 1 mg and about 800 mg; between about lmg and about 700mg; between about 1 mg and about 600 mg; between about lmg and about 500 mg; between about 1 mg and about 400 mg; between about 1 mg and about 300 mg; between about lmg and about 200 mg; between about 1 mg and about 100 mg; between about 1 mg and about 75 mg; between about lmg and about 50 mg; between about 1 mg and about 25 mg; between about 1 mg and about 10 mg; between about 50 mg and about 100 mg; between about 50 mg and about lg; between about 50 mg and about 500 mg; or between about 50 mg and about 100 mg.

Methods for Assessing Susceptibility to a Neuropsychiatry Disorder

[0048] In another aspect, the present invention is directed to methods of determining a subject's susceptibility to a neuropsychiatric disorder by detecting compromised PPAR or RXR functionality. In one exemplary embodiment, compromised PPAR or RXR functionality is detected by determining the levels or activity of PPAR, RXR, or both, and comparing those levels to a standardized control representative of normal PPAR and RXR levels or activity, wherein a decrease in activity or levels over the standardized control indicates increased susceptibility to a neuropsychiatric disorder. In another exemplary embodiment, compromised PPAR or RXR signaling functionality is detected by determining the levels or activity of PPAR, RXR, or both and comparing those levels or activity before and after in vivo or ex vivo administration of PPAR and/or RXR specific ligands, wherein a failure to increase PPAR or RXR levels post-administration indicates an increased susceptibility to a neuropsychiatric disorder. In another exemplary, the method may be used in a subject treated by a retinoid. In another exemplary embodiment, compromised PPAR or RXR functionality is determined by assessing the subject's PPAR genotype, RXR genotype, or both. The receptor functionality may be determined from a biological sample from the subject. Biological samples suitable for use with the present invention include, but are not limited to, blood, saliva, urine, serum, mucus, tears, sweat, or a combination thereof. In one exemplary embodiment, the method comprises isolation and purification of the receptor, the receptor gene, or receptor mRNA from the biological sample. As used herein, the term "receptor" refers to a PPAR or RXR receptor.

[0049] Exemplary neuropsychiatric disorders for which an increased susceptibility can be detected with methods of the present invention include, but are not limited to, different forms of depression such as major depressive disorder, bipolar disorder, dysthymia, but also schizophrenia, attention-deficit hyperactivity disorder, body dysmorphic disorder, bulimia nervosa and other eating disorders, cataplexy, fibromyalgia, general anxiety disorder, impulse-control disorders, , panic disorder, and post-traumatic stress disorder. In certain embodiments, the methods of the present invention may be used to treat certain symptoms of certain neuropsychiatric disorders including, but not limited to depressed mood, despair, increased feelings of guilt, decrease in interest in certain pleasurable (hedonic) stimuli such as food and sex, or secondary cognitive symptoms such as deficits in decision making, attention, and working memory (WM). In one exemplary embodiment, the neuropsychiatric disorder is a major depressive disorder, dysthymia, bipolar disorder, or schizophrenia. In another exemplary embodiment, the neuropsychiatric disorder is a major depressive disorder or related psychotic symptoms.

[0050] In one exemplary embodiment, the method determines the level or activity of PPAR. In yet another exemplary embodiment, the PPAR may be PPARa and/or PPAR . In one exemplary embodiment, a decrease of between approximately 25% and 100% in PPAR levels or approximately 15% and approximately 100% of activity over the standardized control indicates an increased susceptibility to a neuropsychiatric disorder. In another exemplary embodiment, a decrease of at least 50% in levels and 30% in activity indicates an increase susceptibility to a neuropsychiatric disorder. In certain exemplary embodiments, the above decrease in PPAR levels is indicative of an increased susceptibility to a major depressive disorder, dysthymia, bipolar disorder, or schizophrenia. In another exemplary embodiment, the above decrease in PPAR levels indicative of an increased susceptibility to a major depressive disorder or related psychotic symptoms.

[0051] In one exemplary embodiment, the method determines the level or activity of RXR. The RXR may be RXRa, β, or γ. In yet another exemplary embodiment, the RXR is RXRy. In one exemplary embodiment, a decrease of between approximately 25% and 100% in RXR levels or approximately 15% and approximately 100% in RXR activity over the standardized control indicates an increased susceptibility to a neuropsychiatric disorder. In another exemplary embodiment, a decrease of at least 50% in RXR levels and 30% in RXR activity over the standardized control indicates an increased susceptibility to a neuropsychiatric disorder. In certain exemplary embodiments, the above decrease in RXR levels is indicative of an increased susceptibility to a major depressive disorder, dysthymia, bipolar disorder, or schizophrenia. In another exemplary embodiment, the above decrease in RXR levels is indicative of an increased susceptibility to a major depressive disorder or related psychotic symptoms.

[0052] Any standard protein or metabolite detection and identification assay known in the art may be used to determine receptor levels and activities. Exemplary protein detection and identification methods include Westerns, 2D-PAGE, high-performance liquid chromatography and mass spectroscopy with the two latter techniques used also for identification of signaling biomarkers of PPAR and RXR gene regulation activity,

[0053] In another exemplary embodiment, the method of the present invention comprises determining the genotype or gene expression levels of the receptors from a biological sample. Any standard genotyping or gene expression level assay known in the art may be used to determine the genotype or gene expression levels. Exemplary genotype or gene expression level assays include PCR based assays, DNA fragment analysis, allele specific oligonucleotides probe assays, DNA sequencing, and DNA microarrays. Exemplary DNA fragment analysis assays include restriction length polymorphism, terminal restriction fragment length polymorphism, amplified fragment length polymorphism, or muli-plex ligation-dependent probe amplification.

Methods for Screenine Pharmacoloeical Aeents for Adverse Neuropsychiatric Side Effects

[0054] In another aspect, the present invention is directed to methods for screening pharmacological agents for adverse neuropsychiatric side effects. In one exemplary embodiment, the method comprises administering the pharmacological agent to an experimental animal, or in vitro system, and determining the effect, if any, on PPAR or RXR functionality. An in vitro system may comprise cell cultured in vitro or tissue explants. In one exemplary embodiment, a decrease in PPAR levels, RXR levels, or both indicates the potential for adverse neuropsychiatric side effects. In another exemplary embodiment, inhibition of PPAR and RXR heterodimerization and transactivation of transcriptional targets indicates the potential for adverse neuropsychiatric side effects. In yet another exemplary embodiment, inhibition of PPARa and RXRy dimerization and/or transactivation of transcription targets indicates the potential for adverse neuropsychiatric side effects. In one exemplary embodiment, the transcription target includes, but is not limited to dopamine D2 receptor.

[0055] The effect on PPAR and RXR functionality can be determined at the molecular level using the same methods for assessing functionality as discussed above for determining a subject's susceptibility to a neuropsychiatric disorder. In addition, the effect on PPAR and RXR functionality may be further assessed using a behavioral test such as, but not limited to, a forced swim test, a Y-maze spontaneous alteration test, a delayed non-match to place test, an actimetric cage test, a sucrose preference test, an open field test, or comparable behavioral test assessing despair and/or cognitive activity, wherein an increase in despair behavior or decrease in cognitive behavior indicates a compromised PPAR or RXR functionality.

[0056] The experimental animal can be d RXR knock-out, a PPAR knock-out. , a RXRy PPAR double knock-out, RXRy +/- PPAR +/- heterozygous animal, or an normal experimental animal treated with a PPAR antagonist, RXR antagonist, or both. The in vitro assay may comprise cells cultured in vitro or tissue explants in which PPAR signaling is compromised due to genetic ablation, pharmacological inhibition through use of a PPAR antagonist, or both. Exemplary PPAR antagonists include GSK0660, GW6471 , and GSK3787. A suitable RXR antagonist is BR121 1. In one exemplary embodiment, the cultured cells or isolated lymphocytes.

[0057] In one exemplary embodiment, the present invention comprises a method of screening retinoid or retinoid derivative treatments for adverse neuropsychiatric side effects. The method comprises administering the retinoid or retinoid derivative to an experimental animal or an in vitro system and determining RXR activity, wherein the inability of the retinoid or retinoid derivative treatment to decrease RXR functionality indicates a treatment without the potential for adverse neuropsychiatric side effects. The method may be used to screen retinoid or retinoid derivatives intended for treating acne vulgaris or related skin disease and disorders for potential adverse neuropsychiatric effects. The experimental animal can be a PPAR knock-out experimental animal or a normal experimental animal treated with a PPAR antagonist. The in vitro system may comprise cells cultured in vitro or tissue explants in which PPAR signaling is compromised due to genetic ablation, pharmacological inhibition, or both.

Method for Identifying Novel Pharmacological Agents for Treatment of

Neuropsychiatric Disorders

[0058] In another aspect, the present invention is directed to a method for identifying pharmacological agents useful in treating neuropsychiatric disorders. The method comprises administering the pharmacological agent or agents to be tested to one or more experimental animals or an in vitro system. The in vitro system may comprise cells cultured in vitro and tissue explants. In one exemplary embodiment, the in vitro culture system comprises isolated lymphocytes. The experimental animals or in vitro system used in the present method have compromised PPAR or RXR functionality, or both. After administration of the pharmacological agent, test are then run to determine whether the deficit cause by the compromised PPAR or RXR functionality is restored upon treatment with the pharmacological agent. In the case of experimental animals, functionality can be assessed using behavioral test such as forced swim test, Y-maze spontaneous alteration test, a delayed non-match to place test, actimetric cage test, sucrose preference test, and open field test or comparable behavior test assessing despair and/or mnemonic activity. A decrease in despair activity and improvement in cognitive activity indicate the potential usefulness of the tested pharmacological agent in treating a neuropsychiatric disorder or symptom. Alternatively, or in addition to, the effect on PPAR and RXR activity can be determined by looking at downstream biomarkers of PPAR and RXR functionality, such as transcriptional targets of PPAR and RXR signaling. In one exemplary embodiment, restoration of RXR functionality can be assessed by determining the level of dopamine D2 receptors or dopamine D2 receptor activity, wherein an increase in dopamine D2 receptor levels or activity indicates a potentially useful pharmacological agent for treating a neuropsychiatric disorder.

[0059] The experimental animal can be a RXR knock-out, a PPAR knock-out, a RXR-PPAR double knock-out, RXRy +/- PPAR +/- heterozygous animal, or an normal experimental animal treated with a PPAR antagonist, RXR antagonist, or both. The in vitro system may comprise cells cultured in vitro or tissue explants in which PPAR signaling is compromised due to genetic ablation, pharmacological inhibition through the use of a PPAR antagonist, or both. Exemplary PPAR antagonists include GSK0660, GW6471 , and GSK3787. A suitable RXR antagonist is BR1211. In another exemplary, the compromised PPAR functionality is a PPARa and/or PPAR functionality. In another exemplary, the compromised RXR functionality is a RXRy.

[0060] Pharmacological agents for treating different forms of depression such as major depressive disorder, bipolar disorder, dysthymia, and also schizophrenia, attention- deficit hyperactivity disorder, body dysmorphic disorder, may be identified using the methods of the present invention. In addition pharmacological agents useful for treating certain symptoms of certain neuropsychiatric disorders including, but not limited to depressed mood, despair, increased feelings of guilt, decrease in interest in certain pleasurable (hedonic) stimuli such as food and sex, or secondary cognitive symptoms such as deficits in decision making, attention, and working memory (WM) may also be identified using the methods of the present invention.

[0061] This invention is further illustrated by the following examples, which are not to be construed in any way as imposing limitations upon the scope thereof. On the contrary, it is to be clearly understood that resort may be had to various other embodiments, modifications, and equivalents thereof which, after reading the description herein, may suggest themselves to those skilled in the art without departing from the spirit of the present invention and/or scope of the appended claims.

EXAMPLES

Example 1 - RXRy mediates docosahexanoic acid modulation of despair behaviors in working memory in mice

Materials and Methods

Animals

[0062] The 8 week-old BALBcByJ (CBy) male mice purchased from Charles River (Lyon, France) were housed in groups of 5 mice/cage and were tested at the age of 4-5 months. RXRy-/-, RXRy+/- and wild type control mice (WT) were raised on mixed genetic background (60% C57BL/6J and 40% 129SvEms/j) from heterozygous crosses as described (Krezel et al., 1996) and tested at the age of 4-5 months. All mice were housed in 7am- 7pm light/dark cycle in individually ventilated cages, type "MICE" (Charles River, France). Food and water were freely available. The numbers of mice used for each experiment were indicated in corresponding graphs or in figure legends. All experiments were carried out in accordance with the European Community Council Directives of 24 November 1986 (86/609/EEC) and in compliance with the guidelines of CNRS and the French Agricultural and Forestry Ministry (decree 87848).

Drug treatments

[0063] All-trans RA, TTNPB, DHA (Sigma), UVI2108 and BR121 1 were dissolved in absolute ethanol and then in sunflower oil, so that the final solution contained 3% of ethanol. All these substances were administered by intraperitoneal injections at volume/weight ratio 3 ml/kg. All treatments were carried out between 8-11 AM and 5 hrs before the test, unless indicated differently.

Forced swim test

[0064] The forced swim test (Dalvi and Lucki, 1999), was carried out between 1pm and 4pm in the 2-liter glass beaker half-filled with water at 22-23°C (the water depth was 17 cm). All mice were tested only once in this task. To this end, each mouse was lowered gently into the water and the time of immobility was scored during a 6-minute testing period. The mouse was judged immobile when it floated in an upright position and made only small movements to keep its head above the water. After 6 min, the mouse was taken out of the water, left to dry under the red light lamp and returned to its home cage. The immobility scores of each animal were used as an index of despair behavior.

Y-maze spontaneous alternation

[0065] The Y-maze spontaneous alternation paradigm is based on the natural tendency of rodents to explore novel environment. When placed in the Y-maze mice will explore the least recently visited arm, and thus tend to alternate visits between the three arms. For efficient alternation mice need to use working memory and thus they should maintain an ongoing record of most recently visited arms, and continuously update such a record. A mouse with an impaired working memory cannot remember which arm it has just visited and thus show decreased spontaneous alternation (Holcomb et al., 1999; Wall and Messier, 2002).

[0066] The Y-maze apparatus and the procedure were as previously described (Wietrzych et al., 2005). Briefly, each mouse was placed at the end of one arm and allowed to explore freely the apparatus for 5min, with the experimenter out of the animal's sight. Spontaneous alternation performance (SAP) was assessed visually by scoring the pattern of entries into each arm and expressed as percentage of total number of entries without the first two visits. Total entries were scored as an index of ambulatory activity in the Y-maze and non mouse had to be excluded due to low locomotor performance (score below 9 entries), described protocol (Wietrzych et al., 2005), with modifications to facilitate pharmacological analysis. Briefly, animals were habituated to 22hr water deprivation during two consecutive sessions with water accessible for two hours between 4 and 7pm. Such protocol of water deprivation was maintained throughout the entire experiment. To habituate animals to reinforcement, 25% sucrose was placed in home cages on the second day of habituation. Habituation to the apparatus and experimental conditions were carried out from the third day over 2 consecutive days. For each habituation session, mice were separated in single cages 30 min prior to the test, and then each mouse was placed in the middle of the T-maze and allowed to visit freely the maze during minimum of 5 min, with 10 min of cut-off period. During this time animal had to visit all arms and drink drops of sucrose solution (25%), which were dispensed in the trays positioned at the end of each arm.

Delayed non-match to place (DNMTP)

[0067] Behaviorally na^'ive groups of 6 WT and 6 RXRy null mutant mice (one WT and one RXRy-/- mouse were excluded from analyses since they did not move in the maze by the end of the training session) were tested in the DNMTP in the T-maze according to previously described protocol (Wietrzych et al., 2005), with modifications to facilitate pharmacological analysis. Briefly, animals were habituated to 22hr water deprivation during two consecutive sessions with water accessible for two hours between 4 and 7pm. Such protocol of water deprivation was maintained throughout the entire experiment. To habituate animals to reinforcement, 25% sucrose was placed in home cages on the second day of habituation. Habituation to the apparatus and experimental conditions were carried out from the third day over 2 consecutive days. For each habituation session, mice were separated in single cages 30 min prior to the test, and then each mouse was placed in the middle of the T-maze and allowed to visit freely the maze during minimum of 5 min, with 10 min of cut-off period. During this time animal had to visit all arms and drink drops of sucrose solution (25%), which were dispensed in the trays positioned at the end of each arm.

[0068] For DNMTP, the training consisted of 6 daily trials separated by 30- 40minutes. Each trial was composed of the acquisition phase followed by retention phase. At the beginning of the acquisition phase one drop of 25% sucrose was deposited in the wells placed at the ends of the two opposing arms. One of these arms was blocked and mouse was placed in the start box, which was always positioned at the base of the T- maze. After 15 seconds the mouse was released and allowed to consume the sucrose reward. During the retention phase both arms were opened and the animal was released from the start box and allowed to enter the arm of choice where it was blocked on entrance. A choice was rewarded and considered as correct if the animal entered the arm not visited during the acquisition phase. After consuming the sucrose, or after 30 seconds if the arm was not baited, the animal was returned to its isolation cage. In contrast to previously used protocol (Wietrzych et al. , 2005) the retention phase followed immediately the acquisition phase and the interval (ITI) between the two phases was minimized to the time necessary for replacing the animal back into the start box. The latency to leave the start box was measured for each animal and one WT and one RXRy- /- mouse were excluded from the training since their latency to choose the arm during the retention phase exceeded 3min in more than one trial/day on two consecutive days, which was considered as exclusion criterion. After 10 days of training, the ITIs between the acquisition and retention phase were increased semi-randomly to 180, 360 and 540 seconds so that each animal was tested 6 times with each interval during three consecutive days (days 1 1-13 in Figure 4). During the inter-trial interval each animal was placed in its cage. Starting form the day 13 mice were tested only twice a week on two consecutive days. On the first day animals were retrained using minimal ITIs to recall the procedural aspects of the test and to homogenize animal performance before pharmacological treatments tested on the second day. Only one pharmacological treatment was tested every 7 days. To test pro-mnemonic activities of DHA, atRA and UVI2108 we have used a minimal ΓΓΙ, at which mice displayed deficits and which was 540sec for the WT group and 180sec for RXRy-/- mice. To analyse amnesic effects of BR1211 in WT mice we used ITI=360sec, which was the longest ITI, at which mice did not display working memory deficits.

Actimetric cages

[0069] The spontaneous activity was measured in actimetric cages (Immetronic, Pessac, France) with the help of two arrays of infra-red beam photo-cells installed on the side walls in each individual cage. Unless otherwise indicated, mice were placed in the actimetric cages 2 hours after the administration of relevant substances and the activity was scored for 1 hour, starting at 5 hours after injection.

Statistical analysis

[0070] Dose response treatments in CBy mice were analysed using one-way ANOVA with treatment used as independent factor. The interactions between BR121 1 treatment and DHA or atRA treatment in CBy mice were compared by two-way ANOVA with respective treatments used as independent factors. The mnemonic performance in the DNMTP test was evaluated with ANOVA on repeated measured whereas the effects of different intervals on memory of WT and RXRy-/- mice were analysed using two-way ANOVA with two independent factors, genotype and interval. The drug treatments in RXRy mutant mice were analysed using two-way ANOVA with two between-subject factors (genotype and treatment), with exception to pharmacological study in DNMTP test, in which drug effects (studied at different ITIs) were analysed separately for in WT and RXRy-/- groups using one-way ANOVA with treatment defined as independent factor. The Bonferroni comparisons were used for post-hoc analysis, whereas student t- test was used for comparisons of two groups.

Results

Modulation of despair behaviour by administration of DHA in BALBcByJ mice is mediated by RXRs

[0071] To investigate the antidepressant activities of DHA, we used the CBy strain of mice, known for their high despair behaviors as compared to other strains of mice (Dulawa et al., 2004). High immobility in the forced swim test was reduced by 45% in ® response to acute treatment with 30 mg/kg of fluoxetine (Prozac ) (Figure 1A), confirming that this parameter of animal behavior can be used to predict the antidepressant activity of pharmacological agents. Similarly to effects of fluoxetine treatment, an acute administration of 1 mg/kg of DHA reduced the immobility time of CBy mice by 48%, but low 0.1 mg/kg dose of DHA did not affect animal performance (F[2,29]=42.9, pO.001 ; Figure IB). The effect of DHA was evident 5 hrs after treatment, whereas it was absent when measured 30 min after injection of 1 mg/kg of DHA (151.3±5.5sec for vehicle [n=10] and 159.7±5sec for DHA [n=9] group; 1=1.1 , ns), indicating that DHA could act through genomic mechanisms, which are known to have longer latencies than non-genomic effects.

[0072] Retinoid X receptors (RXRs), which act as ligand-dependent transcription factors, are potential mediators of DHA genomic activities in vivo, since DHA can bind and enhance transactivation by RXRs in vitro (de Urquiza et al., 2000; Lengqvist et al., 2004). To address such a hypothesis, we have studied activities of DHA in mice in which RXRs were inactivated pharmacologically using the selective pan-RXR antagonist BR1211 that have been shown to be active in mice (Calleja et al., 2006). The behavioral effects of DHA were completely suppressed by co-treatment with 1 mg/kg of BR121 1 as illustrated by significant interaction between DHA and BR121 1 treatment (F[l ,40]=25.9, p<0.001 ; Figure I D). The same dose of BR121 1 did not affect animal behavior (F[2,14]=0.22, ns; Figure IB). In agreement with these data, the pan-RXR agonist, UVI2108 (also named BMS649), mimicked the effects of DHA in the forced swim test and the acute treatment at the dose of 1 mg/kg reduced the immobility of CBy mice by 50%, whereas lower dose of O. lmg/kg was inactive in this test (F[2,22]=73,2, p<0.001 ; Figure IB). Furthermore, acute treatment with all-trans RA, which under pharmacological conditions is rapidly transformed into 9-cis RA, a known potent RXR and RAR agonist, also reduced immobility of CBy mice at the dose of 5mg/kg or lOmg/kg, but not at lower dose of lmg/kg (F[3,37]=31.4, p<0.001 ; Figure 1 C). Such activity of atRA was RXR specific as it was not reproduced by treatment with TTNPB (F[3,30]=0.26, ns; Figure 1C), a pan-RAR agonist, and beneficial effects of atRA treatment were abolished by co-treatment with pan-RXR antagonist, the BR1211 {atRA x BR1211 treatment, F[l ,40]=37.6, p<0.001 ; Figure ID). The reduction of immobility in the forced swim test induced by lmg/kg of DHA, UVI2108 or BR1211 , or 5mg/kg of atRA treatments cannot be attributed to an increased arousal/locomotor activity, since such treatments did not alter the locomotor activity as measured in actimetric cages 5 hrs after treatment (abs t value <1.6 for any of the treatments as compared to its vehicle group; ns).

RXRs mediate DHA modulation of working memory

[0073] To address the role of RXRs in mediating mnemonic activities of DHA we have first used spontaneous alternation task. This test, although specific to rodent models was used for a rapid screen to test the efficiency of DHA and RXR ligands in modulation of working memory and to determine minimal doses for such modulations. We show that administration of DHA to CBy mice could improve their working memory performance (Figure 2A). A single dose of lmg/kg of DHA increased spontaneous alternation in the Y-maze from 55.3 ± 2.2% in vehicle-treated to 70.4 ± 2.4% in DHA-treated CBy mice, whereas the dose of 0. lmg/kg did not display such activity (F[2,37]=13,2, p<0.001 ; Figure 2A). Pro-mnemonic effect of DHA was evident at 5hrs after treatment (Figure 2A), but not at 30min (60.1±3.2% for vehicle [n=8] and 59.5±3.4% for DHA [n=8] group; t=-0.12, ns), suggesting that DHA could act through genomic mechanisms.

[0074] To address the role of RXRs in mediating mnemonic activities of DHA we have used pharmacological approach similar to that applied in studying despair behaviors. We found that the co-treatment with lmg/kg of BR1211 completely abolished the beneficial effect of DHA on spontaneous alternation in the Y-maze (F[l ,41]=8.3, p<0.01 for DHA x BR1211 treatment interaction), whereas on its own the same dose of BR 121 1 was behaviourally inert as determined 5hrs after its administration (F[2,14]=0.35, ns; Figure 2A). The implication of RXRs in mediating DHA signal was further supported by effects of other RXR ligands, which displayed DHA-like activities. Thus acute treatment with atRA (5mg/kg or lOmg/kg) also improved the spontaneous alternation performance in the Y-maze in CBy mice (F[3,34]=10.2, p<0.001 ; Figure 2B). The mnemonic activity of such treatment was RXR-specific, as it was reproduced with administration of a selective pan-RXR agonist UVI2108 (F[2,22]=7,6, p<0.01 ; see lmg/kg in Figure 2A), but not of TTNPB (F[3,32]=0;54, ns; Figure 2B), a pan-RAR agonist. In addition, co-treatment with lmg/kg of BR121 1 completely abolished beneficial effects of retinoic acid (F[l ,44]=7.4, p<0.01 for atRA x BR1211 treatment interaction; Figure 2C). The latency to leave the start box and the total number of arm entries were not affected by any of the treatments (abs t <2 for any of the comparisons between vehicle and drug treatment, ns) and attained on average 19 ± 0.7 sec for latency and 19 ± 0.2 of entries for vehicle and DHA (lmg/kg), UVI (lmg/kg), atRA (5mg/kg), TTNPB (5mg/kg) or BR1211 (lmg/kg) treatments.

RXRg is important RXR in mediating DHA modulation of despair behaviours and working memory

[0075] To address the individual contribution of specific RXR isotypes in such RXR functions we studied the effects of DHA or RXR agonist in mice carrying null mutations for selected RXR isotypes (pharmacology specific to different RXR isotypes does not exist). To this end a study was done to determine whether genetic inactivation of RXR or RXRy (ablation of RXRa is embryonic-lethal) has adverse effects on despair behaviors and mnemonic functions, which could be expected in agreement with pro-mnemonic and antidepressant effects of RXR agonists. In order to facilitate identification of such adverse effects we have carried out these studies in mice raised on C57BL6J x 129SvEms/j mixed genetic background, which display low despair behaviors and high working memory performance. In agreement with our recent report (Krzyzosiak et ah, 2010) we found that ablation of RXRy led to increased despair behaviors, as in the forced swim paradigm RXRy-/- mice remained immobile for 122±7.9 sec, which was significantly longer than the 54.4±5.5 sec of immobility displayed by WT littermates (Figure 3A). Mice carrying only one functional allele of RXRy displayed an To address the individual contribution of specific RXR isotypes in such RXR functions we studied the effects of DHA or RXR agonist in mice carrying null mutations for selected RXR isotypes (pharmacology specific to different RXR isotypes does not exist). To this end we first studied whether genetic inactivation of RXR or RXRy (ablation of RXRa is embryonic-lethal) has adverse effects on despair behaviors and mnemonic functions, which we could expect in agreement with pro-mnemonic and antidepressant effects of RXR agonists. In order to facilitate identification of such adverse effects we have carried out these studies in mice raised on C57BL6J x 129SvEms/j mixed genetic background, which display low despair behaviors and high working memory performance. In agreement with our recent report (Krzyzosiak et ah, 2010)) we found that ablation of RXRy led to increased despair behaviors, as in the forced swim paradigm RXRy-/- mice remained immobile for 122±7.9 sec, which was significantly longer than the 54.4±5.5 sec of immobility displayed by WT littermates (Figure 3A). Mice carrying only one functional allele of RXRy displayed an functionally predominant RXR in mediating DHA modulation of despair behaviors and working memory. In support of such hypothesis, RXR null mutation did not affect working memory performance in the spontaneous alternation in the Y-maze or despair behaviors in the forced swim paradigm (data not shown).

[0076] In order to further confirm the implication of RXRs and RXRy in DHA control of working memory we investigated activities of DHA and RXR ligands in WT and RXRy-/- mice in delayed non-match to place task. To carry out these analyses null mutant mice and their WT littermate controls were first trained in the DNMTP over 10 days. RXRy-/- mice displayed deficit in acquiring the task (significant effect of genotype F[l ,98]=7.7; p<0.01), which reflected delay in acquisition of the DNMTP evident during th th th

the 4 and the 5 day of training (Figure 4A). Starting from the 6 day of training performance of RXRy-/- mice improved and was comparable to that of WT littermate controls until the end of the training suggesting that they have acquired the task. Increasing ITIs differentially affected WT and RXRy-/- group {genotype x ITI, F[l,3]=3.48, p<0.05), which was related to marked working memory deficits in RXRy-/- mice at HT=180sec and 360sec, whereas similar deficit was observed in WT group only at rri=540sec (Figure 4A). The minimal His at which mice displayed significant deficit of working memory (180sec for RXRy-/- and 540sec for WT animals) were selected to test pro-mnemonic effects of pharmacological treatments. DHA treatment similarly to atRA and UVI2108 significantly increased working memory performance of WT mice at doses, which were active in the spontaneous alternation task (main effect of treatment F[3,15]=5.53, p<0.01 , see left panel in the Figure 4B). The DHA and atRA effects were RXR dependent as they could be suppressed by BR121 1 cotreatment (F[4,20]=5.2, p<0.01 for main effect of treatment). Thus, lmg/kg of DHA or 5mg/kg of atRA facilitated working memory and increased animal performance to 79.8±3.2% (p<0.05) and 80±6.1% (p<0.05) of correct choices respectively as compared to 50±7.6% in vehicle treated group (Figure 4B). Cotreatment with lmg/kg of BR121 1 blocked such effects of DHA and atRA as respective groups displayed 64±3.4% and 63±6.2% of corrects choices, which was not different from 50±7.6% for vehicle treated group (p<0.05). Such effect was not induced by amnesic activity of BR1211 treatment, as lmg/kg of BR121 1 did not affect mice performance at OT=360 sec, which was the ITI sensitive to reveal potential amnesic effects in WT mice. Indeed, mice treated with lmg/kg of BR121 1 made on average 83.2±5.2% of correct choices, which was comparable to performance of vehicle treated WT group (90±6.6%) at the same ITI (t=0.8, ns). RXRy appeared functionally predominant RXR in control of working memory in DNMTP task since DHA and UVI2108 did not improve performance of RXRy-/- mice (main effect of treatment F[2,12]=0.002, ns; right panel in Figure 4B). The analysis of latency to choose the arm during the retention test did not reveal any significant effect of treatment (F[3,160]=2.79, ns), suggesting thus that animal performance in DNMTP was not confused by non-mnemonic activities of pharmacological treatments. This is further supported by lack of significant treatment x genotype (F[3,160]=0.38, ns) or treatment x genotype x trial (F[15,160]=0.42, ns) interactions. Finally, it is unlikely that present behavioral data result from cross-reaction of different treatments related to longitudinal study as the effect of genotype was maintained in non-treated animals (main effect of genotype F[2,27]=17.6, p<0.001) after a series of pharmacological tests (see performance at different ITIs after "Drug Treatments" in Figure 4A).

Summary

[0077] Although n-3 PUFAs have been reported to improve affective and mnemonic performance in depressed and schizophrenic patients (Logan, 2004; Peet and Stokes, 2005) as well as in pre -clinical animal models (Moriguchi and Salem, 2003; Naliwaiko et al., 2004; Tanabe et al., 2004; Carlezon et al., 2005; Hashimoto et al., 2005), their molecular mechanism of action remains elusive. The above experiments show that in experimental conditions DHA has anti-depressant effects in the forced swim test and facilitates working memory performance in the spontaneous alternation and delayed non- match to place tasks and such effects are mediated by RXRs. In addition, the similarity in effects of DHA and all-trans RA treatments and/or abolishment of such effects by BR1211 , a pan-RXR antagonist, revealed also that RXRs are converging point for anti- depressant and pro-mnemonic activities of DHA and vitamin A. The above experiments excluded the possibility that despair behaviors and mnemonic performance after DHA, UVI2108, BR1211 or TTNPB treatments were confused by non-specific effects of such treatments, since the latency to leave the start box and the total number of arm entries in the Y-maze or latency to choose the arm in DNMTP task as well as locomotor activity in the actimetric cages (the potential indicators of anxiety level and/or locomotor performance), did not differ significantly among any of the treatments.

[0078] Several lines of evidence suggest that RXRy is the predominant RXR isotype in mediating DHA regulation of despair behaviors and working memory: (i) genetic inactivation of RXRy led to increased despair behaviors (Figure 3A) and mnemonic deficits specific to working memory (Figure 3C and 4A); (ii) genetic inactivation of RXR did not affect despair behaviors and working memory performance (data not shown; RXRa null mutants could not be tested due to embryonic lethality; data not shown); (iii) in contrast to acute treatments with BR121 1 , which abolished activity of DHA and which alone did not affect despair behaviors and working memory performance, longer treatments with BR1211 (2 or 4 days) resulted in increased despair behaviors and working memory deficits in the spontaneous alternation in the Y-maze, which were similar to those exhibited by RXRy null mutants (data not shown); (iv) the RXR agonists, including DHA and UVI2108, a pan-RXR synthetic agonist, did not improve forced swim performance, spontaneous alternation and DNMTP performance in RXRy null mutants. The resistance of RXRy-/- mutants to DHA and RXR agonist treatments suggest also that RXRa and RXR , of which expression is not altered in RXRy-/- mice, cannot mediate DHA modulation of the despair behaviors and working memory, and therefore are not functionally redundant with RXRy in the control of these functions. Finally, RAR signaling appeared dispensable for RXR control of despair behaviors and working memory, as pan-RAR agonist did not modulate immobility time in the forced swim task and spontaneous alternation performance whereas antidepressant and pro-mnemonic activities of all-trans RA were blocked by pan-RXR antagonist. Such observation is further supported by unaltered despair behaviors and the absence of working memory deficits in any of the RAR isotype-specific KO mice (Wietrzych et al., 2005 and unpublished data), which might suggest that DHA and retinoic acid modulation of working memory is mediated by RXRy homodimers or heterodimers with non-RAR nuclear receptors known to interact with RXRs (Mangelsdorf and Evans, 1995).

EXAMPLE 2 - Retinoid X Receptor Gamma Control of Motivated Behaviors involves Dopaminergic Signaling in Mice

Materials and Methods

Animals

[0079] RAR -/- and RXRy-/- single mutant, and RAR V-RXRy-/- double mutant male mice as well as their wild type (WT) control mice were raised on a mixed genetic background (60% C57BL/6J and 40% 129SvEms/j) from heterozygous crosses as described (Krezel et al., 1996), and tested at the age of 4-5 months. All mice were housed in groups of 4-5 mice per cage in a 7am-7pm light/dark cycle in individually ventilated cages, type "MICE" (Charles River, France). Food and water were freely available. All experiments were carried out in accordance with the European Community Council Directives of 24 November 1986 (86/609/EEC) and in compliance with the guidelines of CNRS and the French Agricultural and Forestry Ministry (decree 87848).

Behavioral Procedures

Forced swim test

[0080] The forced swim paradigm (Dalvi and Lucki, 1999) was carried out between lpm and 4pm in a 2-litre glass beaker half-filled with water at 22-23°C (the water depth was 15 cm). All mice were tested only once in this task. To this end, each mouse was lowered gently into the water and the time of immobility was scored during a 6-minute testing period. The mouse was judged immobile when it floated in an upright position and made only small movements to keep its head above the water. After 6 min, the mouse was taken out of the water, left to dry under a red light lamp and returned to its home cage.

Sucrose preference test

[0081] This task, designed to measure hedonic behaviors in mice (Moreau, 1997; Nestler et al., 2002), is based on the palatable nature of sucrose observed in a number of mouse strains. Mice were first habituated to experimental conditions by an overnight housing in individual cages equipped with one bottle filled with water. On the first day of the test, sucrose-naive mice were placed at 5pm in the same individual cages with one bottle filled with water and another with 1% sucrose solution. Three hours later (8pm) the bottles were weighed to measure liquid consumption and were replaced in cages until morning to continue habituation to experimental conditions. Over two additional days animals were further habituated to sucrose solution in their home cages. The measures of an overnight consumption were then carried out from 5pm until 8am to evaluate sucrose preference. Mice were not water deprived at any moment, in order to measure spontaneous sucrose preference and exclude any potential emotional confounds induced by stress of water deprivation. The sucrose preference was expressed as the percent of sucrose solution consumed with respect to total liquid consumption.

Actimetr

[0082] Spontaneous activity was measured in actimetric cages (Immetronic, Pessac, France) with the help of two arrays of infra-red beam photo-cells installed on the side walls in each individual cage. Mice were placed in actimetric cages at 1 lam and their activity was recorded over 32 hrs including a habituation period between 1 lam and 7pm and a complete dark/light cycle until 7pm of the next day.

Open field

[0083] Mice were tested in parallel in 5 automated open-fields (44.3 x 44.3 x 16.8 cm) made of PVC with transparent walls and a black floor, covered with transparent PVC (Panlab, Barcelona, Spain). The open fields were placed in a room homogeneously illuminated at 150 Lux. Unless otherwise specified each mouse was placed in the periphery of the open field and allowed to explore freely the apparatus for 30 min, with the experimenter out of the animal's sight. Activity parameters including distance travelled over the test session were calculated automatically.

Catalepsy test

[0084] Mice were injected intraperitoneally with 0.2 or 2mg/kg of haloperidol (Sigma) and after 30min were placed in the test cage with their forelimbs on the wooden transversal bar fixed at a level of 3 cm above floor level. The latency to move out from the bar was scored and used as index of catalepsy.

Production and use of adeno-associated virus (AA V) vectors [0085] For generation of AAV vectors we used a vector plasmid containing an expression cassette, in which a human cytomegalovirus immediate-early promoter (CMV promoter) was followed by the first intron of the human growth hormone gene, the cDNA of interest, woodchuck hepatitis virus posttranscriptional regulatory element (WRPE; nucleotides 1093 to 1684, GenBank accession no . J04514) and simian virus 40 polyadenylation signal sequence. This expression cassette was inserted between the inverted terminal repeats (ITR) of the AAV-2 genome as described (Li et al., 2006). The viral vectors used for expression of RXRy (AAV2-RXRy), D2R (AAV2-D2R) and EGFP (AAV2-GFP) contained the entire cDNA sequences of RXRy (GenBank accession no. NM 009107), D2R (long isoform, GenBank accession no. NM 010077.2) or EGFP, respectively. We used two helper plasmids, pAAV-RC and pHelper, harbouring the AAV rep and cap genes, and the E2A, E4, VAl genes of the adenovirus genome, respectively (Agilent Technologies, Santa Clara, CA). HEK293 cells were co-transfected with pAAV- RC and pHelper plasmids using the calcium phosphate coprecipitation method. AAV particles were then harvested and purified by two sequential continuous iodoxale ultracentrifugations. The vector titer was determined by quantitative PCR of DNase-I-

10 12

treated vector stocks, and were estimated at 10 to 10 vector genome copies (vg).

[0086] For rescue experiments and D2R expression we used behaviorally naive RXRy-/- male mice (n=29) at the age of 8 months. Each animal was anaesthetised using ketamine (100 mg/kg)/xylasine (10 mg/kg) solution and 0.7μ1 of AAV2-RXRy, AAV2- D2R or AAV2-GFP suspension was injected bilaterally into the nucleus accumbens (bregma = + 1 ,5; lateral = +/- 0,7; ventral = + 4,2, the coordinates identified prior to experiments using dye injections and corresponding to bregma = + 1,3; lateral = +/- 0,5; ventral = + 4,0 position in the Mouse Brain Atlas, Paxinos and Franklin, 2001) using a stereotaxic apparatus (Precision Cinematographique, Paris, France). The injection was carried out at 50 nl/min using a Harvard Apparatus PHD 2000 pump (Holliston, USA) and the injectors were withdrawn from the brain 20 min after the end of the injection. After placing stitches each animal was left to awake in the temperature -conditioned cage. Mice were tested 4 weeks later and their brains were removed for post-hoc analyses.

Drug infustion procedure [0087] PvXRy-/- mice (n=15) aged between 4-5 months were infected with AAV2- PvXRy as described. Four weeks later mice were anaesthetized with ketamine/xylazine solution and 8-mm-long stainless-steel guide cannulas (0.4 mm external diameter; Cortat, Courrendlin, Switzerland), were positioned bilaterally 1 mm above the NAcSh (bregma = + 1 ,5 ; lateral = +/- 0,7; ventral = + 3,2) using stereotaxic apparatus (Precision Cinematographique, Paris, France). The cannulas were fixed to the skull with anchoring screws and dental cement. , 1.2 mM NaStainless steel stylet rods were inserted into the cannulas to prevent occlusion. On the day of the experiment raclopride (a D2/D3 specific antagonist soluble in aqueous solutions; Sigma) was dissolved in fresh artificial cerebrospinal fluid (ACSF, which consisted of 3 mM KC1, 140 mM NaCl, 2 mM glucose, 1.2 mM CaCl₂, 1 mM MgCl₂, 0.27 mM NaH PO HPO , pH 7.4) and infusions of 0.25μ1 of raclopride (5μg/side) or vehicle (ACSF) were performed at 100 nl/min using Harvard Apparatus PHD 2000 pump and stainless-steel injector needles (0.28 mm external diameter) that protruded from the cannula by 1 mm, into the NAcSh. Three minutes after injection injector needles were removed from the brain, the stylet rods were replaced in cannula guides and mice were transferred to their home cage for 5 min prior to the forced swim test. 48 hrs later mice were semi-randomly infused with raclopride or ACSF to test the locomotor effects of such treatments. For this experiment mice were placed in the open field immediately after removing injectors and placement of stylet rods and their activity was scored 5 min later during 5 min. Three out of 15 mice were excluded from analysis due to unilateral AAV2-RXRy infection or incorrect guide placement, and two mice could not be infused for the open field test since the stylets remained blocked.

Pharmacological treatments

[0088] Haloperidol (Sigma- Aldrich) was dissolved in acetic acid solution and pH was neutralised with NaOH. For c-fos expression studies, mice were injected intraperitonealy (IP) with saline or 1 mg/kg of haloperidol, 90 minutes prior to sacrifice, whereas for analysis of the open field behavior saline or haloperidol were injected 20 min prior to the test and animals were tested for 10 min. IP injection was also used for acute fluoxetine (Lilly France) treatment 30min prior to forced swim test. For chronic treatment, fluoxetine was added to the standard chow diet. Accordingly, we supplemented standard chow in powdered form with fluoxetine to attain the dose of 20 mg of fluoxetine per kg of body weight during 24 hrs. To calculate such a dose, the food consumption was first estimated experimentally to be 4 g of food pellets per 24 hrs per animal. Fluoxetine - supplemented food pellets were immediately lyophilized and stored at -20°C until use. For treatment, standard chow pellets were replaced by fluoxetine-supplemented food pellets and were provided ad libitum in standard home cages throughout treatment period. The consumption of fluoxetine-containing pellets did not differ from the consumption of non-supplemented food pellets in control cages. WT and RXRy-/- mice treated with th fluoxetine or fed control diet were all tested for sucrose preference on the 19 day of

St

treatment, and in the forced swim test on the 21 day of treatment, with the exception of mice used for evaluation of fluoxetine effects on D2R expression, which were all behavio rally naive.

Qualitative RT-PCR

[0089] Mice were killed by cervical dislocation. Whole brains were extracted, fresh- frozen in OCT, and kept at -80°C until use. Tissue corresponding to the nucleus accumbens (NAc) was collected with 0.5 mm punch from three subsequent 300 μιη-thick cryosections. Similarly, dorsolateral striatum (CPu) was collected using 0.8 mm punch from four subsequent frozen sections of 300 μηι. The accurate location of these brain structures was based on visual inspection of each section using a stereomicroscope (Leica, Wild M715) and its comparison with the stereotaxic atlas of mouse brain Paxinos and Franklin, 2001). Tissue samples were placed on dry ice and kept at -80°C until use.

[0090] Total RNA extraction was carried out using the RNeasy Micro Kit protocol (Qiagen, France). Total RNA from each tissue sample was transcribed into cDNA using QuantiTect® Reverse Transcription Kit according to the manufacturer's recommendation. Briefly, the reaction was carried out at 42°C for 20min in a total volume of 20 μΐ and was inactivated at 95°C. 20-times -diluted cDNA was used as a template, and quantitative real-time PCR was run in a LightCycler 480 (Roche, Diagnostics, Mannheim, Germany) using LightCycler SYBR Green kit (Roche, Diagnostics) with cDNA and gene-specific primers ( 100 μΜ) following the manufacturer's instructions. All of the reactions were performed in triplicate with the following cycling protocol: 10 min of heat activation of the enzyme at 95°C, 45 cycles of denaturation at 95°C for 5 sec, annealing at 60°C for 30 sec, and extension at 72°C for 20 sec. Fluorescence detection was performed at 72°C. Gene-specific primers were designed using Primer3 software (primer3_www.cgi) to amplify fragments of 150-250bp. The transcript amounts evaluated for DIR and D2R were normalised for the quantity and quality of each sample by division by the amount of transcript of the housekeeping gene acidic ribosomal phosphoprotein P0 (Arbp or 36B4; NM 007475) in the same sample and such relative values were presented in Figure 4A and 2S. 36B4 transcript amount was quantified using primers.

In situ hybridization and analysis of expression levels

[0091] In situ hybridisation (ISH) was performed on 14μηι-ΐ1ικΛ frozen sections with digoxigenin-labeled riboprobes synthesized from a 1680bp D2R cDNA template and an enkephaline 800bp cDNA template as described (Krezel et al., 1998). Hybridisation conditions were as described previously (Krezel et al., 1998) and are available on the http://empress.har.mrc.ac.uk/ website. The amount of probe used for hybridisation and signal detection conditions was adapted to avoid saturation of the chromogenic labelling (see below). Expression patterns were documented using a macroscope (Leica M420) or microscope (DM4000B), both connected to a Photometries camera with the CoolSNAP (v. 1.2) software (Roger Scientific, Chicago, IL).

[0092] For the analysis of cell counts and expression levels of D2R, the images were transformed into grey scale and analysed using ImageJ software (Rasband, 1997-2007). The strongest signal observed for any of the neurons in any of the brain sections remained between 67-95 units in a 0 to 255 unit grey scale (0 corresponding to black), being thus 25-30% below full saturation conditions in order to enable quantitative analysis of signal intensity. For each animal, the cell number and intensity of cellular signal was evaluated within selected regions of CPu, NAcSh and NAcCo on the same sections (for region selection see Figure 5 A and B) at bregma 1.10 and 1.40 (Paxinos and Franklin, 2001 ). The mean values of cell counts or intensity for each region were calculated and compared as described in Results.

Immunohistochemistry and c-fos cell counts [0093] Coronal sections (14 μπι thick) from unfixed frozen brains of 4-month old RXRy-/- mice and their WT littermates were collected on super-frost slides and stored at -80°C until analysis. Sections were post-fixed in 4% paraformaldehyde and treated with 1% H₂0₂ to block endogenous peroxides. For detection of RXRy we used rabbit anti-

RXRy polyclonal antibody (SC555, batch Al l l, SantaCruz, US), whereas c-fos was detected using rabbit anti-c-fos polyclonal antibody (1 : 1000, Chemicon). Both primary antibodies were detected using the ABC system (Vector, USA) according to the manufacturer's manual. For each animal and section, corresponding brain regions were identified according to the mouse brain atlas (Paxinos and Franklin, 2001) and c-fos positive cells were counted from identical surfaces defined by region-corresponding auto- shape figures (Figure 5 A, B) at two levels of the striatum (bregma 1.10 and 1.40; Mouse Brain Atlas, Paxinos and Franklin, 2001). The mean cell counts for each brain region were calculated and compared as described in Results.

HPLC measure of serotonin levels and its metabolites in the brain tissue

[0094] The brain samples of n=6 WT and n=6 RXRy-/- male littermates (4months old) were weighed immediately after collection and frozen at -80°C until use. Before analysis, samples were thawed, and homogenized in 10 volumes (w/v) of 0.1 M HC104 containing the internal standard DHBA (125ng/ml). The homogenates were centrifuged at 12.000 rpm for 20 min at 4°C and supernatant was retained for analysis. Serotonin and its metabolite 5-hydroxyindoleacetic acid (5HIAA) were evaluated using high performance liquid chromatography (HPLC) with electrochemical detection. The chromatographic system consisted of a 25 cm x 4.6 mm Hypersyl CI 8 ODS column (particle size 5μπι, Biochrom, France). The column was kept at a constant temperature of 30°C. The flow rate was 1.2 ml / min with a back pressure of 1 ,500 psi (Waters instrumentation). The system was linked to a Waters model 460 electrochemical detector with a glassy-carbon electrode. Detector potential was maintained at 0.85 V (reference: Ag/AgCl electrode). The mobile phase consisted of 0.05 M NaH2P04 and O.lmM EDTA (pH adjusted to 4.85 with NaOH) in double-distilled water with methanol (6 %). The system was calibrated by injecting various amounts (3.4 pg - 34 ng) of standard solutions, containing 1.1 ng of internal standard DHBA (3-4 dihydroxybenzylamine 1 mM in HC104 0.1M). The supernatant of each sample was injected onto the column, and peak identification was performed by comparing retention times with the calibration solution. Results were expressed in ng/g ± SEM.

Statistical analysis

[0095] The comparisons of behavioral performance in RAR -/-/RXRy-/-, RAR -/- and RXRy-/- null mutant mice were carried out using the protected least significant difference (PLSD) Fischer test. The pharmacological data for the treatments in WT and RXRy-/- mice were analysed using 2-way analysis of variance (ANOVA) - with treatment and genotype as two independent factors and behavioral responses as dependent variables. Comparison of the evolution of locomotor performance in the open field or actimetric cages were evaluated using ANOVA on repeated measures. Global and post-hoc statistical analyses were performed using the PLSD Fischer test and for two- group comparisons using student t-test (see t values in the text). Significant differences are indicated in the corresponding figures.

Results

Increased despair behaviors in RXRynull mice can be normalized by anti-depressant treatment

[0096] To address the contribution of retinoid receptors in control of despair behaviors in loss of function of RAR and/or RXRy in mice on performance were studied in the forced swim test. Concomitant ablation of RAR and RXRy in RAR -/-RXRy-/- double null mutant mice led to a marked increase of the immobility time, which attained 129±4.3 sec and was significantly longer (p<0.001) than in wild-type (WT) control mice, which remained immobile for 71±6.2 sec (Figure 5A). The increased immobility in the double mutant mice was principally due to the loss of function of RXRy, since single RXRy-/- mutants displayed similar high immobility time of 1 17±4 sec, whereas inactivation of RAR did not affect immobility time in this task (64.9±6.8 sec; p>0.05). An abnormal locomotor behavior is unlikely to account for the increased immobility time of RXRy-/- mice in the forced swim test, since RXRy-/- mice did not differ from their WT littermates with respect to spontaneous locomotion in actimetric cages, novelty- induced locomotion in the open field test, or locomotor coordination in the rotarod task (Krezel et al. 1998, and supplementary Figure 13). [0097] Since despair behaviors belong to the core symptoms of depression, experiments exploring whether antidepressant treatment could improve the performance of RXRy-/- mice were conducted. In agreement with previous reports of SSRI activities in C57BL6J and 129SV mouse strains (Dulawa et al., 2004), a 21 -day chronic treatment with fluoxetine at the dose of 20 mg/kg/24hrs did not affect performance of WT mice in the forced swim test, but such treatment reversed the despair behavior in RXRy-/- mice (Figure 5B), as illustrated by a significant genotype x fluoxetine treatment interaction (F[l ,35]=23.2, p<0.001). Thus, high immobility of vehicle treated RXRy-/- mice was reduced in fiuoxetine-treated RXRy-/- animals, which behaved comparably to vehicle treated WT mice (Figure 5B).

Inactivation of RXRg leads to anhedonia, which can be normalized by antidepressant treatment

[0100] To better evaluate the involvement of retinoid receptors in the control of affective behaviors, experiments investigating hedonic behaviors in the sucrose preference test in RAR -/- and RXRy-/- single and compound mutant mice were conducted. During the active, night phase of the circadian cycle WT mice displayed clear preference for a 1% sucrose solution as compared to plain water, since sucrose represented 91.3±2.7% of total liquid consumption. In compound RAR -/-RXRy-/- mutant mice, such a preference was absent as sucrose consumption reached 58.1±10.6% of total liquid intake (Figure 6A), which was significantly less than preference in WT mice (p<0.05) and not different from a chance level of 50% (t=-0.1 , ns, one-group t-test). Such anhedonic behavior was due to ablation of RXRy, since RXRy single null mutants displayed similar loss of sucrose preference and consumed 57.8±10.8% of sucrose solution, whereas RAR -/- mice consumed sucrose solution at 92±2.1% and were indistinguishable from their WT controls. The total liquid intake during the sucrose preference test was not different between WT and mutant mice (4.6±0.2 g for WT, 4.9±0.3 for RAR -/-, 4.9±0.2 for RXRy-/- and 5.2±0.3 for RAR V-RXRy-/- mutants). The absence of a sucrose preference in RAR V-RXRy-/- and RXRy-/- mice is unlikely to result from gustative deficits since all groups preferred water to 1% sucrose on the first presentation of sucrose drink. Thus, during 3 hrs of the first testing session, in sucrose- naive WT mice sucrose solution constituted 43±2.5% of total liquid consumption, as compared to 44±2.1% for RAR -/-, 37±5% for RXRy-/- and 32.4±7.3% for RARP-/- RXRy-/- mice, which for all groups was significantly less than the chance level of 50% (t>3.5 for any of the comparisons, p<0.05, one-group t-test). Such avoidance of sucrose solution by sucrose-naive mice results from a natural tendency for reserved consumption of novel food/drink and provides evidence for recognition of 1 % sucrose taste. A chronic, 19-day antidepressant treatment with 20 mg/kg/24hrs of fluoxetine, reversed the sucrose preference deficits in RXRy-/- mice (Figure 6B), which is reflected by significant genotype x treatment interaction (F[l,64]=7.9, p<0.01). Thus, fluoxetine -treated RXRy-/- mice preferred sucrose solution to water similarly to WT control mice, and the percentage of sucrose solution consumed by fiuoxetine-treated RXRy-/- mice was significantly higher than in vehicle-treated mutant animals (p<0.001).

Abnormal serotonergic signaling is not sufficient to generate depressive behaviors in RXRy -I- mice

[0101] Efficiency of fluoxetine to reverse affective abnormalities could suggest that altered serotonergic signaling is at the origin of depressive-like behaviors in RXRy-/- mice. To address this possibility global evaluation of serotonergic signaling experiments focusing on hippocampus and striatum (including NAc), the two regions differentially innervated by dorsal and median raphe 5HT inputs, which are suggested to play a role in control of affect (Lechin et al., 2006), were conducted. HPLC measurements of 5HT and its metabolite 5HIAA in tissue homogenates revealed a significant increase of 5HT levels in the striatum of RXRy-/- mice (2.01±0.18 ng/g for RXRy-/- and 1.46±0.17 for WT mice; t=2.2, p=0.05), which was not accompanied by altered metabolism of serotonin (Figure 7A). Although there was no significant difference in 5HT levels in the hippocampus, RXRy-/- displayed strong tendency (t=l .92, p=0.08) for increased metabolism of 5HT in this region (Figure 7B). Such abnormalities in the distribution of 5HT did not correlate with abnormal expression of 5HTla receptor, prominently involved in control of 5HT tone and proposed to play a role in control of affective behaviors and in actions of SSRI antidepressants (Blier et al., 1998; Blier and Ward, 2003). Indeed, the relative 5HTla mRNA levels (standardised with respect to expression of the house keeping gene 36B4) were comparable between RXRy-/- and their WT controls in the NAc (2.2±0.4 for RXRy-/- and 2.7±0.3 for WT; t=-l , ns) and hippocampus (1.4±0.2 for RXRy-/-and 1.9±0.3 for WT mice; t=-1.5, ns).

[0102] To investigate functional relevance of abnormal serotonergic signaling for depressive-like behaviors in RXRy-/- mice experiment testing whether acute fluoxetine treatment can reverse increased despair in the forced swim test were conducted. In contrast to chronic treatment, acute administration of fluoxetine at 20mg/kg (IP, 30min prior to the forced swim test) did not alter increased immobility of RXRy-/- mice as there was no significant interaction for genotype x fluoxetine treatment (F[l ,28]=3E-4, ns) and the main effect of genotype (F[l ,28]=19.4, p<0.001) remained significant despite of the treatment (Figure 7C). To control for the efficiency of acute fluoxetine treatment used wild type BALBcByJ (BALBc) mice, the strain susceptible to reveal anti-depressant activities of fluoxetine (Lucki et al., 2001 ) were used, and it was determined that 20mg/kg of fluoxetine was sufficient to significantly reduce despair behaviors in this strain (Figure 7D). The inefficiency of acute fluoxetine treatment to modulate despair behaviors suggests that abnormal serotonergic signal is not sufficient to generate depressive behaviors and adaptive changes associated with chronic fluoxetine treatments might be at the origin of affective abnormalities in RXRy-/- mice.

Reduced dopamine D2 receptor signaling in the nucleus accumbens ofRXRy-/- mice is reversible by fluoxetine treatment

[0103] Despair behavior and anhedonia and their reversal by chronic antidepressant treatment in RXRy-/- mice, indicates that null mutation of RXRy leads to deficits resembling some of the core symptoms of depression. The abnormal function of the nucleus accumbens (NAc), the key structure implicated in the control of motivated behaviors and one of the primary sites of RXRy expression (Krezel et al., 1999), might be at the origin of the behavioral abnormalities in RXRy-/- mice. As dopaminergic signaling in the NAc is critically involved in the modulation of motivated behaviors and in the mechanisms of antidepressant activities including fluoxetine, studies examing the expression of dopaminergic D l and D2 receptors in RXRy-/- mice were conducted. Using real-time quantitative RT-PCR, a significant 32% reduction of D2R expression in the NAc of RXRy-/- mice (t=-3.16, p<0.01 ; Figure 8A) was observed. Interestingly, no such reduction (t=-0.22, ns) was observed in the dorsal caudate putamen (CPu). The inactivation of RXRy did not affect expression of D1R as D 1R RNA levels were not significantly different between WT and RXRy-/- mice in the NAc (1 1.7±1.0 vs 10.6±1.0 units; t=-0.76, ns) and dorsal striatum (37.2±2.2 vs 39.9±0.8 units; t=l .35, ns), as measured by RT-PCR and calculated relative to expression of a reference housekeeping gene (Figure 6S).

[0104] In order to further investigate the regionalisation and the origin of reduced levels of D2R mRNA in the NAc, in situ hybridisation (ISH) studies were carried out (Figure 8B). Comparisons of RXRy-/- mice and their WT controls revealed that the number of neurons expressing D2R was significantly reduced in the shell of the NAc (152±4 vs 206±6; t=-8, p<0.001) and the core of the NAc (294±6 vs 338±16; t=-2.44; p<0.05), but not in the dorsal striatum (798±17 vs 841±34; t=-l .l , ns). To minimize cell counting errors related to differences in the signal intensity between different sections, which may account for more discrete changes, expression of D2R in the dorsal striatum (CPu), which was not affected by ablation of RXRy-/- was used as an internal (intra- section) control to calculate relative changes in D2R cell numbers. To this end, cell counts were divided in the NAcSh or NAcCo by those obtained for the adjacent region of CPu on the same section. A strong (36%) reduction in relative cell number only in the NAcSh (0.16±0.01 for RXRy-/- as compared to 0.25±0.02 for WT; t=-4.7, p=0.001), but not in the NAcCo (0.37±0.01 for RXRy-/- vs 0.40±0.02 for WT; t=-1.5, ns), of RXRy-/- mice was observed(Figure 8C). The difference in the magnitude of changes in D2R expression, might be related to much weaker expression of RXRy in the NAcCo as compared to shell region (Krezel et al., 1999). Reduction of D2R in the NAcSh might be functionally relevant as chronic fluoxetine treatment, in addition to reversing depressive- like behaviors, increased also the relative number of D2R positive cells in the NAcSh of behaviorally naive RXRy-/- mutants (0.15±0.02 for non-treated RXRy-/- vs 0.23±0.01 for fluoxetine treated RXRy-/- mice; t=-3.4, p<0.01), but not WT mice (0.24± 0.01 for non- treated WT vs 0.27±0.02 for fluoxetine treated mice; t=-l .2, ns) (Figure 8D).

[0105] A decrease of D2R-positive cell number in the NAcSh of RXRy-/- mice is most probably related to reduced transcription of D2R, rather than to the loss of a subpopulation of D2R expressing neurons. Supporting this hypothesis, the number of cells expressing mRNA coding for enkephaline, a neuropeptide found predominantly in D2R expressing neurons, was not significantly reduced (124.3±2.8 enkephaline -positive cells in the NAc shell of RXRy-/- mice, as compared to 133.6±3 cells in WT animals; t=- 2.34, ns). Thus, the reduced number of D2R-expressing neurons could reflect a general decrease of D2R transcription in the NAc shell, with a reduction below the detection threshold level in neurons expressing low levels of D2R. Alternatively, it might be related to reduced transcriptional control of D2R restricted to a selected neuronal population. To address this issue the intensity of D2R expression in the ISH experiments, was quantified using the Image J software (see Materials and Methods). The mean intensity of the D2R signal in the NAc shell was not different between WT and RXRy-/- mice when comparing absolute mean values (122.9±2.2 for WT and 125.1±2 for RXRy-/- mice; t=0.7, ns), or when such measures were normalised with respect to the intensity of D2R expression in the dorsal striatum within the same brain section, where D2R expression was not affected by ablation of RXRy (1.05±0.01 for WT and 1.08±0.02 for RXRy-/- mice; t=l , ns). These data suggest that RXRy might control expression of D2R in a selected sub-population of D2R neurons.

[0106] Finally, to assess whether cell- and regional-specific reduction of D2R mRNA expression leads to abnormal D2R activities, neuronal activation in RXRy-/- mice in response to haloperidol, a D2 preferential antagonist, was investigated. To this end, the induction of c-fos protein expression, a molecular marker related to neuronal activity and plasticity, was studied. An acute treatment with haloperidol (1 mg/kg) increased the number of c-fos positive cells in various regions of the striatum, including the shell and core of the NAc and CPu in all tested mice (Figure 9 A and B). However, in the shell of the NAc the magnitude of this increase was significantly lower in the in RXRy-/- than in WT mice as reflected by the significant interaction between genotype and treatment (F[l ,14]=5.6, p<0.05) and PLSD Fischer post-hoc analysis (p<0.01). Such difference was not observed in the core of the NAc or in the dorsal CPu in the same sections (compare NAc-Sh with NAc-Co and CPu for WT-Hal and KO-Hal in Figure 9C). To study whether such a decrease reflects the action of haloperidol or is related to the stress inflicted during drug injection, the numbers of c-fos positive cells in saline-injected mice were evaluated, and found that these numbers were not significantly different between WT and RXRy-/- mice (compare WT-veh and KO-veh in Figure 9C). Thus, the haloperidol-specific induction of c-fos in the shell of the NAc, calculated as the ratio of c- fos positive cells in the haloperidol-treated mice with respect to saline-treated mice, was lower by 48% in RXRy-/- mice as compared to their WT controls (Figure 9D). To validate these findings functionally the locomotor effects of low, non-cataleptic doses of haloperidol, which have been proposed to involve post-synaptic dopamine D2 receptors in the nucleus accumbens (Messier et al., 1992; Millan et al., 2004; Pijnenburg et al., 1976), was tested. All mice treated with haloperidol displayed reduction of locomotor activity in the novel environment of the open field, although such reduction was different depending on the genotype (significant genotype x treatment interaction, F[4,68]=2.7; p<0.05). Post-hoc analysis revealed that the decrease of locomotion was significantly lower in RXRy-/- mice as compared to WT controls for haloperidol doses of 0.1 and 0.2 mg/kg (p<0.05; Figure 10A). These effects of haloperidol cannot be attributed to an altered susceptibility of RXRy-/- mice to develop catalepsy, since a 0.2 mg/kg dose did not induce catalepsy in WT and RXRy-/- mice, whereas a high dose of haloperidol (2 mg/kg) induced comparable degrees of catalepsy in both genotypes (Figure 10B).

RXRy in the nucleus accumbens is critical for control of despair and hedonic behaviors and modulation of D2R expression

[0107] To investigate whether RXRy expression in the nucleus accumbens shell plays a role in the control of depressive-like behaviors and expression of dopamine D2R, functional rescue experiments using stereotaxic injection of adeno-associated virus (AAV2) expressing RXRy in the NAc of RXRy null mutants were carried out. Using immunohistochemical analysis, the expression of RXRy protein in the WT non-injected mice was clearly detected (top panels in Figure 11A) or in the NAcSh of RXRy-/- mice infected with AAV2-RXRy vector (bottom panels in Figure 1 1 A), but not in RXRy-/- animals infected with AAV2-GFP virus (middle panels in Figure 1 1 A). The virus mediated expression of RXRy in RXRy-/- mice was detectable bilaterally at bregma 1.1 and 1.4 and specifically in the NAcSh in 5 (out of 10) mice injected with AAV2-RXRy whereas for AAV2-GFP infected mice, such pattern of GFP expression was identified in 7 (out of 10) animals. In the remaining animals (n=5 for AAV2-RXRy and n=3 for AAV2-GFP) viral infection was unilateral or not restricted to the NAc (e.g. spreading into ventral septum) and these mice were excluded from the analysis of behavioral data. The infection of RXRy null mutant mice with the AAV2-RXRy expressing vector led to a significant decrease of despair behaviors (t=2.8, p<0.05) and anhedonia (t=-2.7, p<0.05) as compared to RXRy-/- mice infected with the GFP-expressing virus (Figure 1 IB and C). Such behavioral effects of RXRy expression in the NAc were not confounded by altered locomotor activity, as spontaneous locomotion in the actimetric cages or novelty induced locomotion in the open field test were comparable between the groups (supplementary Figure 15). During the sucrose preference test the total amount of liquid consumed during the testing session was not different among the groups and was on average 4.7±0.4 mg/night.

[0108] In addition to reversal of behavioral deficits, re-expression of RXRy in the nucleus accumbens of RXRy-/- mice led to an increase of the number of D2R expressing neurons. In the NAc shell of RXRy-/- infected with AAV2-RXRy 190.2±1 1.1 D2R- positive neurons were identified, which was significantly more than 153.2=1=11.2 neurons in AAV2-GFP infected mutant mice (t=-2.34, p<0.05), which was also reflected by relative measures of D2R positive cell numbers with respect to adjacent CPu region (0.23±0.01 for AAV2-RXRy as compared to 0.18±0.01 in AAV2-GFP infected RXRy-/- mice; t=-3.1 , p<0.05). To address whether such an increase of D2R expression in the NAcSh is relevant to depressive-like behaviors in RXRy-/- mice, D2R signaling in the NAcSh was blocked by bilateral infusion of the D2R antagonist raclopride (5 μg/side) in AAV2-RXRy rescued RXRy-/- mice. Blocking D2R signaling compromised antidepressant effects of AAV2 -mediated re-expression of RXRy in RXRy-/- mice since such animals remained immobile in the forced swim test for 130.6=1=13.4 sec, which was significantly longer (t=4.1 , p<0.01) than RXRy-/- mice which were infected with AAV2- RXRy and infused with ACSF vehicle (57.9±9.7 sec; Figure 1 ID). Although raclopride infusion into the NAc led to a slight tendency to reduce general locomotor activity as measured in the open field (Figure 1 IE), such reduction was not significant (t=-0.87, p=0.4) and cannot account for increased immobility in the forced swim test.

AAV2 mediated expression of D2R in the nucleus accumbens reverses depressive- like behaviors in RXRy -/- mice

[0109] In order to further address the role of a reduction of D2R expression in the control of depressive-like behaviors in RXRy-/- mice, D2R signaling in the nucleus accumbens was increased by AAV2 mediated expression of D2R. Seven out of nine injected RXRy-/- mice were retained for statistical analysis as they displayed bilateral D2R expression revealed by increased number of D2R positive neurons in the NAc (210.7±10.5 in AAV2-D2R mice as compared to 153.2±1 1.2 in AAV2-GFP infected RXRy-/- mice; t=-4.3, p<0.01). The increase of D2R positive neurons was specific to the shell of NAc and was not present in the adjacent, dorsal part of the striatum on the same sections (Figure 12A). Such expression was functionally relevant as it increased locomotor activity in the open field (Figure 12B), which attained 122.3±10 m for AAV2- D2R mice as compared to 96.9±3.7 m of distance covered by AAV2-GFP mice (t=2.35, p<0.05). Increased activity resulted from abnormal reactivity to a novel environment and could be further demonstrated by increased locomotion in the actimetric cages, which

St

was evident during the 1 h of the 32 h test (Figure 15), thus reflecting enhanced D2R signaling (Ouagazzal and Creese, 2000; Zhang et al., 1996). In the forced swim test RXRy-/- mice infected with AAV2-D2R remained immobile for 58±11.1 sec, which was significantly less (t=-3.5, p<0.01) than 127.9±16.7 sec for GFP-AAV2 mice (Figure 12C). AAV2 mediated D2R expression also normalised anhedonia of RXRy-/- mice. Indeed, RXRy-/- mice infected with AAV2-D2R preferred sucrose to water and consumed 73.8±6.9% of sucrose as opposed to significantly lower (t=2.62, p<0.05), 53.3±4.3% for AAV2-GFP control mice (Figure 12D).

Summary

[0110] The experimental results provided here offer the first evidence that a specific retinoid receptor is implicated in the control of affective behaviors in mice. The results show that null mutation of RXRy leads to increased despair behavior in the forced swim test and anhedonia, the key symptom of depression as measured in the sucrose preference paradigm. The studies of single and compound RXRy-/- and RARP-/- mutant mice also provide evidence that RAR might not be the heterodimerisation partner of RXRy in control of affective behaviors. Although we cannot exclude some functional redundancy between RARs in their interactions with RXRy, it is unlikely that RARa or RARy, the two other RAR isotypes, may functionally compensate for the loss of RARP since these receptors display very limited coexpression with RXRy (Krezel et al., 1999). [0111] The behavioral abnormalities displayed by RXRy-/- mice are of particular relevance for research on depression, as they resemble some of the core symptoms specific to depressive disorders and they could be reversed by chronic fluoxetine treatment. Such functions of RXRy are specific to central control of affective behaviors, since RXRy null mutant mice do not present dysfunction of the peripheral nervous system or muscles and with the exception of compromised working memory (Wietrzych et al., 2005), they do not display any other apparent abnormalities (Krezel et al. 1996, Krezel et al., 1998). Thus, considering efficiency of chronic treatment with fluoxetine to reverse anhedonia and despair we speculated that abnormal serotonin signal might be at the origin of depressive-like phenotype in RXRy-/- mice. Although such a hypothesis could be further supported by increased 5HT tissue levels in the striatum and a strong tendency (p=0.08) for increased 5HT turnover in the hippocampus of knockout mice, the observation that acute modulation of serotonergic signaling by single treatment with fluoxetine did not affect performance of RXRy-/- mice in the forced swim test, suggest that altered serotonin signaling is not sufficient to generate depressive-like behaviors in these mice. In consequence, adaptive changes associated with chronic fluoxetine treatment might be relevant for beneficial effects of this antidepressant treatment and for the mechanisms of depressive-like behaviors in RXRy-/- mice. Therefore, dopaminergic signaling, known to be modulated by chronic fluoxetine treatment and the key neurotransmission pathway involved in the control of motivated behaviors was investigated. The dopamine D2 receptor has been suggested to be particularly relevant to such regulations, and its potential implication in depressive disorders and role in antidepressant therapies have been investigated (Dailly et al., 2004; Millan, 2006; Nestler and Carlezon, 2006). In addition, D2R is known to be a direct transcriptional target of retinoid receptors (Krezel et al., 1998; Samad et al., 1997). The above results show that the inactivation of RXRy led to a significant reduction in D2R mRNA expression specifically within the nucleus accumbens, whereas the expression of D1R was not affected in any part of the striatum. Interestingly, in situ hybridisation analysis suggested that reduced D2R expression may concern only a subpopulation of neurons in the shell of NAc, since in this region: (i) the number of enkephaline positive neurons, a distinct marker of D2R neurons was not altered in RXRy-/- mice, (ii) the intensity of D2R signal was not reduced in D2R positive neurons of RXRy-/- mice indicating that reduced D2R expression is not generalised, whereas (iii) chronic fluoxetine treatment increased the number of D2R positive neurons in RXRy-/- mice, but not in WT mice.

[0112] At the moment, it is not clear why reduction of D2R expression was not observed in the dorsal striatum of RXRy-/- mice, which together with the shell of NAc are the brain regions with the most prominent expression of RXRy (Krezel et al., 1999). One possible explanation is that transcriptional control of D2R expression in the dorsal striatum is subject to marked functional redundancy between RXRy and RXRa and/or RXR . Such a hypothesis is supported by an overall reduction of striatal D2R expression and severe locomotor deficits displayed by RXR /RXRy double null mutants, these defects being absent in the corresponding single null mutants (Krezel et al., 1998).

[0113] The reduction of D2R mRNA expression in the NAc of RXRy-/- mice is relevant for D2R functions, as c-fos induction by haloperidol treatment, a D2R antagonist was blunted in mutant mice. In concordance with the topography of deficits in D2R expression, reduced activation of c-fos expression was observed in the shell of the NAc, indicating that the reduction of D2R functions might be restricted to this region of the ventral striatum. Such observations are further supported by a blunted locomotor response of RXRy-/- mice to low, non-cataleptic doses of haloperidol. Thus, RXRy null mutant mice were less prone to reduction of locomotor activity in response to haloperidol treatment, suggesting compromised D2R responsiveness in the ventral striatum, the region strongly implicated in the control of horizontal locomotion (Amalric and Koob, 1993; Messier et al, 1992; Pijnenburg et al, 1976; Zhang et al, 1996).

[0114] Reduced D2R signaling in RXRy-/- mice might be directly related to depressive-like deficits displayed by these mice. In line with this hypothesis, that chronic fluoxetine reversal of depressive-like behaviors was accompanied by an increase of D2R expression in the NAcSh of RXRy-/- mice. To further explore the behavioral relevance of compromised D2R signaling in RXRy-/- mice and the implication of RXRy in such control, rescue experiments by virus mediated re-expression of RXRy in RXRy-/- mice were carried out. Re-expression of RXRy in the shell of NAc is critical for modulation of D2R expression and affective behaviors. This was illustrated by an increase of the number of D2R expressing neurons in the NAc, and a reversal of behavioral deficits in the forced swim and sucrose preference tests following AAV2 mediated re-expression of RXRy in the NAc of RXRy-/- mice. An increase of D2R expression in NAc appeared to play a critical role in the antidepressant-like activities of RXRy, since infusion of raclopride, a D2R D3R antagonist, prevented the rescue of despair behaviors in RXRy-/- mice infected with AAV2-RXRy. Furthermore, a long-lasting increase of D2R signaling by AAV2 mediated expression of D2R in the NAcSh of RXRy-/- mice reversed both despair behaviors in the forced swim and anhedonia in the sucrose preference test. The functionality of viral expression of D2R was confirmed by an increased number of D2R neurons, but also by an increase in novelty induced locomotion as tested in the open field or actimetric cages, which is in agreement with stimulating locomotor effects D2R activation in NAc (Ouagazzal and Creese, 2000; Zhang et al., 1996). Interestingly, such increased locomotion was not observed following re-expression of RXRy in RXRy-/- mice even though it also increased expression of D2R. Such difference might be related to quantitative and qualitative differences in D2R expression, which might have been stronger and display distinct, cell type specific activities after infection with AAV2-D2R as compared to mice infected with AAV2-RXRy. Although such increased activity may confound results of the forced swim test, it also suggest that reduced immobility, induced by AAV2 -mediated expression of D2R in RXRy-/- mice reflects antidepressant activities since: (i) inhibition of D2R signaling by raclopride, which prevented AAV2-RXRy rescue of despair behaviors in RXRy-/- mice, was devoid of non-specific behavioral effects on locomotion as measured in the open field test; (ii) viral expression of D2R also normalised anhedonia in RXRy-/- mice, a distinct measure of depressive-like behaviors, not affected by locomotor side -effects of AAV2-D2R infection. Finally, considering that AAV2 infections lead to low levels of retrograde transduction (Paterna et al., 2004), the data on antidepressant effects of AAV2-D2R infection of RXRy-/- mice suggest the role of post-synaptic D2R in NAc in control of affective behaviors.

[0115] In conclusion, this study provides the first evidence that the loss of RXRy signaling leads to depressive-like behaviors in mice, and indicates that decreased dopamine D2R signaling in the shell of the NAc plays a critical role in RXRy control of affective behaviors. Considering that retinoids or n-3 PUFAs (de Urquiza et al., 2000; Lengqvist et al., 2004; Wietrzych et al., 2011) can modulate RXR activities in vitro and in vivo, the present data might be of direct relevance for anti-depressant activities of n-3 PUFAs reported in clinical conditions (Logan, 2004; Peet and Stokes, 2005) or depression associated with isotretinoin treatment (Bremner and McCaffery, 2007; Kontaxakis et al., 2009). In addition, mnemonic deficits specific to working memory, which were described in RXRy-/- mice (Wietrzych et al., 2005) might be relevant to cognitive deficits associated with depression. Such deficits, although not considered as the core symptoms of depression, are found in most forms of clinical depression. Consequently, our data suggest that RXRy is a potential novel target for antidepressant treatments. Unlike conventional neuropharmacology, treatments targeting retinoid receptor(s) could modulate availability of specific neurotransmitter receptors by fine, transcriptional control of their expression. Thus, RXR ligands such as bexaroten

®

(Targretin ), used so far in cancer treatment, might be potentially interesting for clinical trials in treatment of depressive disorders.

EXAMPLE 3 - Dysfunction of PPAR signaling is susceptibility factor to develop depression under retinoid treatment

Materials and Methods

Animals

[0116] The C57BL6J and BALBcByJ (CBy) male mice were purchased from Charles River (Lyon, France) at the age of 5-8 weeks and were housed in groups of 5 mice/cage throughout experiments. The behavioral tests were done on adult mice at 3-5 months of age. RXRg mice were generated as described (Krezel et al., 1996), whereas PPARa heterozygouse and/or knockout mice were generated as previously described (Lee and Gonzalez, 1996). The genetic background of tested mice was 50% C57BL/6J and 50% 129SvEv. All mice were housed in 7am-7pm light/dark cycle in individually ventilated cages, type "MICE" (Charles River, France). Food and water were freely available. All behavioral tests were performed between 8am and 4pm, with exception to sucrose preference test, which were carried out between 5pm and 8h30pm in order to study spontaneous liquid intake. All experiments were carried out in accordance with European Community Council Directives of 24 November 1986 (86/609/EEC) and with the guidelines of CNRS and the French Agricultural and Forestry Ministry (decree 87848).

Forced swim test

[0117] The forced swim paradigm of despair behaviors has a high degree of pharmacological validity for research on depression, which is reflected by its sensitivity to major classes of antidepressants, including tricyclic or SSRI (Selective Serotonin Reuptake Inhibitors) antidepressants (Dalvi and Lucki, 1999).

[0118] The test was carried out between 1pm and 4pm in the 2-liter glass beaker half- filled (at least 15cm of depth) with water at 22-23°C. All mice were tested only one time in this task. To this end each mouse was lowered gently into the water and the time of immobility was scored during 6 min of the total testing period. The mouse was judged immobile when it floated in an upright position and made only small movements to keep its head above water. After 6-min, the mouse was taken out from the water, let to dry under the red light lamp and returned to its home cage. Floating scores of each animal were used as an index of despair behaviors.

Sucrose preference

[0119] Sucrose preference test used to assess anhedonia as a symptom of depressive- like behaviors in addition to its high face validity it is also sensitive to treatments with classical antidepressant in mice and rats (Muscat et al., 1992; Willner, 1997).

[0120] Before testing all mice were habituated to 1% sucrose solution for either at least 1 session separated or for two days. Sucrose preference test was carried out between 5pm and 9am. To this end, mice were isolated in individual cages equipped with water at 1 lam and sucrose bottle was presented at 5pm and left through the night until 9am. On the day of test, mice were placed individually in cages equipped with two bottles at the front of the cage and containing water and 1 % sucrose solution, respectively. Water and sucrose consumption were evaluated by weight and the sucrose preference was estimated by percent of consumed sucrose with respect to total consumption of both, sucrose and water. The value 50% of sucrose consumption corresponds to lack of sucrose preference. Food was freely available during the test. After each session animals were placed in home cages. The measures of sucrose preference were performed before forced swim test. For the analysis of effects of isotretinoin treatment, mice which did not display preference for 1 % sucrose solution were tested for preference for 2% solution in order to attain minimum of 70% of preference for each mouse prior to isotretinoin treatment.

Open Field test

[0121] Mice were tested in parallel in 5 automated open-fields (44.3 x 44.3 x 16.8 cm) made of PVC with transparent walls and a black floor, covered with transparent PVC (Panlab, Barcelona, Spain). The open fields were placed in a room homogeneously illuminated at 150 Lux. Each mouse was placed in the periphery of the open field and allowed to explore freely the apparatus for 30 min, with the experimenter out of the animal's sight. The distance travelled over the test session was calculated automatically.

Chronic unpredictable stress

[0122] A modified protocol of chronic unpredictable stress described previously by Willner et al., (1987) and Moreau et al., (1992, 1993) was used. Briefly, all animals were housed in groups of 5/cage until the CUS. On the first day of the CUS all mice were isolated in individual cages and were subject to two stressors during the day and to one stressor during the night in unpredictable order. The stressors used in this protocol included: tilted cage (45°), immobilization for lhr, wet cage, novel object in the cage, dark/light cycle reversal, wet bedding in the in the cage, change of saw-dust bedding to glass marbles. Two distinct groups of mice were used in this protocol, one underwent CUS protocol for 2 weeks whereas the second for 5 weeks.

Pharmacological treatments

[0123] The pan-RXR agonist, methoprene acid (Sigma) or PPARa agonist, fenofibrate (Sigma) were dissolved in absolute ethanol and then in sunflower oil, so that final solution contained less than 5% of ethanol. All these substances were administrated by intra-peritoneal injections at volume/weight ration 3ml/kg. All treatments with exception to chronic fluoxetine, fenofibrate, or isotretinoin administration were carried out between 8-1 l am in the morning and 5hrs before the test in order to study transcriptional activities of respective substances. For chronic fluoxetine, fenofibrate, or isotretinoin treatment, the substances were supplemented in standard chow diet (D04 diet, SAFE, France) as additive. To establish the dose, we estimated food consumption to be 4g of food pellets during 24hrs for one animal. According to such measure, we have supplemented standard food in powdered form with fluoxetine to attain the dose of lOmg/kg of body weight/24hrs of fluoxetine, fenofibrate to attain the dose of 15mg/mg/24hrs, or for isotretinoin lmg/kg/24hrs. Such preparation was used to form food pellets which were immediately lyophilized and stored at -20°C until use. During treatment animals were placed in cages with supplemented food pellets freely available throughout the experiment. The consumption of fluoxetine, fenofibrate, or isotretinoin containing pellets did not vary from consumption of non-supplemented food in control cages.

Statistical analysis

[0124] Pharmacological data for treatments of mice with different genotypes were analysed using 2-way analysis of variance (A OVA) with treatment and genotype as two between-subjects factors and behavioral responses as independent variables. Post-hoc analyses and analyses of behavioral effects of RXR modulators in C57BL6J and CBy mice were performed using student t-test or for more than two groups using PLSD Fischer test. For the chronic unpredictable stress experiments, the effects of treatments were evaluated using student t-test for two-group analysis or by one-way analysis of variance (ANOVA) followed by post-hoc PLSD Fischer test if multiple groups were studied.

Results

Fenofibrate reduces despair behaviors in the forced swim test similarly to antidepressant or RXR agonist treatments.

[0125] To analyze antidepressant activities of pharmacological modulation of PPARa and RXR signaling, CBy mice— the strain proposed as animal model of depressive-like behaviors (Dulawa et al., 2004; Crowley et al., 2005)— were used. The first experiment tested whether classical antidepressant treatment can reduce despair behaviors in CBy mice under the present experimental conditions. The experiments confirmed previous observations (Dulawa et al., 2004) that chronic, 20-day treatment with 10mg/kg/24hrs of fluoxetine (Prozac®) efficiently reduced despair behaviors in this strain. Thus fluoxetine treated CBy mice displayed 100±4 sec of immobility in the forced swim test, which was significantly less than 172=1=6 sec in vehicle -treated, control cohort (Fig. 16a). Fenofibrate treatment, a synthetic PPARa agonist, reduced animal immobility in a dose -dependent manner (Fig. 16b). Thus, the dose of 50mg/kg significantly reduced animal immobility in the forced swim test as measured 5hrs after treatment (p<0.01) whereas lOmg/kg of fenofibrate did not affect animal performance in this test. Such behavioral effects of fenofibrate were similar to pan-RXR agonists including BMS649 (also called UVI2108; Wietrzych et al, 2011) and methoprene acid (MA; Fig. 16c - a distinct pan-RXR agonist (Harmon et al., 1995; Svensson et al., 2003). Accordingly, whereas low doses of MA (O.lmg/kg) did not affect mouse performance, higher dose of lmg/kg significantly (p<0.001) reduced immobility in the forced swim test as tested 5hrs after injection (Fig. 16c). The similarity of behavioral effects of fenofibrate and MA reflect most probably activation of the same signaling pathway. In agreement with this hypothesis, subthreshold doses of fenofibrate (lOmg/kg) and methoprene acid (O. l mg/kg) synergized in reducing immobility in the forced swim test (Fig. 16c). Notably, it is unlikely that fenofibrate-dependent reduction of immobility in the forced swim test reflects its non-specific behavioral activities on locomotion, as motor activity in the open field was not affected by fenofibrate treatment as tested for an active dose of 50mg/kg (t<1.7, non-significant, data not shown).

Chronic or acute fenofibrate treatment has antidepressant activities in chronic, unpredictable stress model of depression

[0126] To further investigate antidepressant activities of PPARa activation a chronic unpredictable stress model, one of the most relevant models used in research into depression (Willner, 1997), was used. The C57BL6N males who underwent the protocol of two weeks of chronic unpredictable stress displayed significant (F[4,24]=4.35, p<0.01) increase of despair behaviors as compared to non-stressed mice and such increase was normalized by acute treatment with 50mg/kg of fenofibrate or by chronic fenofibrate treatment at the dose of 15mg/kg/24hrs (Fig. 17). These behavioral effects of fenofibrate reflect antidepressant activities of fenofibrate as they were comparable to the effects of antidepressant treatment with fluoxetine at lOmg/kg (Fig. 17). It is unlikely that fenofibrate effects in the force swim test could be confused by increased locomotor activity, since acute of chronic treatments with fenofibrate did not affect locomotion as tested in the open field test (F[4,20]=0.9, ns; one-way Anova).

Chronic fenofibrate treatment reverses anhedonia in chronic unpredictable stress model of depression

[0127] To test whether fenofibrate antidepressant effects are not limited to despair behaviors, a long term (5 weeks) CUS protocol was used. The main effect of the group (F[3,28]=3.2, p<0.05) indicated the presence of significant differences between the different treatments. (Fig. 18). Using post-hoc analysis, C57BL6N mice displayed marked deficits in sucrose preference as compared to non-stressed mice (p<0.01 , PSLD) after CUS . Such deficit could be attributed to anhedonia, since antidepressant pretreatment of stressed mice with 10mg/kg/24hrs of fluoxetine for 10 days prevented such deficits. The effects of fluoxetine were reproduced by 10 days of chronic fenofibrate treatment at 15mg/kg/24hrs indicating its antidepressant effects on anhedonia. (Fig. 18). Reduced sucrose preference in stressed mice and its reversal by pharmacological treatments were not affected by abnormal drinking behaviors in any of the groups as total liquid consumption did not differ among the groups and attained on average 4.5ml of water/night (F[3,28]=2.9, ns; one-way Anova).

Genetically compromised PPARa and RXRy signaling synergize in generation of despair behaviors

[0128] In order to further identify interactions between RXRy and PPARa signaling pathways in control of affective behaviors, a loss of function approach was utilized. In these studies, the dose effect of RXRy null mutation was used as a control of despair behavior in the forced swim test, since genetic inactivation of only one allele of RXRy in heterozygouse (RXRy+/-) mice led to an increase of the immobility in the forced swim test (p<0.05 as compared to WT mice), which was significantly less than in RXRy-KO mice (p<0.05, as compared to RXRy-KO mice; Fig. 19 and Wietrzych et al., annexl). Considering that immobility in the forced swim test is a sensitive marker of RXRy functions, immobility was used to investigate functional synergies between concomitantly compromised signaling of RXRy and PPARa, its potential heterodimerisation partner. In agreement with such a hypothesis, genetically compromised PPARa and RXRy signaling in RXRy+/-PPARa+/- mice synergized in generation of despair behaviors, on evidence of significantly higher immobility scores in double heterozygous as compared to single heterozygouse RXRy+/- (p<0.05) or PPARa+/- (p<0.05) or WT control mice (p<0.01) (Fig. 19).

[0129] Importantly, double heterozygouse mice did not display anhedonia in sucrose preference test as they consumed 75.9±9.6% of sucrose as compared to 88.8±2.8% sucrose in WT mice. Thus RXRy+/-PPARa+/- did not develop the entire depressive-like phenotype similar to RXRy-KO mice suggesting some degree of functional redundancy between PPARa and other potential heterodimerisation partner(s) of RXRy.

PPARa-KO mice display susceptibility to develop depressive-like behaviors under isotretinoin treatment.

[0130] Despite functional interactions between PPARa and RXRy signaling in control of despair behaviors, ablation of PPARa in PPARa-KO mice did not lead to depressive-like behaviors, including despair behaviors (Fig. 20) and anhedonia (compare PPARa-KO and WT control groups in Fig. 21). Loss of PPARa function may constitute a susceptibility factor to develop depressive symptoms under treatment with 13-cis retinoic acid, known to induce depression in some cases in clinical conditions. 9-1 1 days of treatment with isotretinoin as diet supplement at the dose of lmg/kg/24hr led to strong increase of despair behaviors in PPARa-KO mice, but not in wild type control mice fed the same diet, which can be illustrated by significant interaction between genotype and treatment (F[l ,39]=4.6; p<0.05; Fig. 20). Accordingly, PPARa-KO mice fed with isotretinoin enriched diet remained immobile for 122.4±1.4 sec, which was significantly less that 63.1±6.6 sec scored for isotretinoin fed WT mice (p<0.001) or 61.8±6.6 sec for PPAR-KO mice fed with control diet (p<0.01). In addition to increased despair behaviours, isotretinoin treatment led to anhedonia in PPARa-KO mice, but not in WT mice (F[l ,33]=4.43; p<0.05; Fig. 21). Thus PPAR-KO mice fed with isotretinoin displayed significantly lower sucrose preference than WT mice fed with the same diet (p<0.05) or WT and PPARa-KO mice fed control diet (p<0.01).

Summary [0131] The association of 13-cis retinoic acid (isotretinoin) used for treatment of acne vulgaris with clinical depression stimulated recently a number of studies aiming at deciphering molecular mechanisms of depression and mechanisms of the susceptibility to develop depression under isotretinoin treatment. Long-term treatments with isotretinoin were reported to increase despair behaviour (O'Reilly et al., 2006) and suppress neurogenesis in the hippocampus in rodents (Crandall et al., 2004). These behavioural effects were observed in all wild type animals that were injected chronically with isotretinoin for several weeks. Such results and experimental models are not adequate to clinical condition, since only a minor fraction of acne patients develop depression, suggesting existence of genetic susceptibility factor to develop depression under isotretinoin rather than generalized side effect of this treatment. In addition the modes of administration and thus possibly the nature and the dose of active metabolites of isotretinoin were different between human treatments (oral or topical administration) and animal studies (chronic intraperitoneal injections).

[0132] In order to identify the genetic factor of the susceptibility to develop depression under isotretinoin treatment, the role of retinoid receptors in control of affective behaviors, the molecular mediator of isotretinoin activities, was investigated. The above studies of the gain and loss of function using pharmacogenetic approaches revealed that RXRy is the key retinoid receptor involved in control of affective behaviours in mice, providing thus the first and the only molecular landmark associating retinoid signaling pathway with depression (Wietrzych et al., annexl ; Krzyzosiak et al., annex 2). In particular, RXRy inactivation led to depressive-like symptoms, whereas activation of RXRy in wild type mice had antidepressant activities. Although RXRy could not be considered as susceptibility factor, such a role could be played by modulators of RXRy activities including its heterodimerisation partners. In the search of functional partner(s) the following functional screens for: (i) synergy in generation of antidepressant activities between pharmacological activation of RXRy and its potential partner and (ii) since no pharmacological tools exist for a number of potential RXRy heterodimerisation partners, the dose effect of RXRy null mutation in control of despair behaviors was screened for functional synergies in generation of despair behaviors between compromised signaling of RXRy and its potential heterodimerisation partner in double heterozygouse mice.

[0133] The experimental results show that activation of PPARa by a selective agonist, fenofibrate, has antidepressant activities as evidenced by: (i) reduced despair behaviors in the forced swim test in BALBc strain of mice, a genetic model for study of despair behaviors (ii) normalised increased despair behaviors and (iii) reversed anhedonia in C57BL6N mice subjected to chronic unpredictable stress, a mouse model of depressive-like behaviors. Such antidepressant activities of fenofibrate were comparable

®

with effects of fluoxetine (Prozac ) treatment, a reference antidepressant therapeutic. Interestingly fenofibrate displayed its antidepressant activities in the forced swim test, after acute treatment with 50mg/kg, which was below the doses used for metabolic studies (min. lOOmg/kg). Furthermore, even lower doses could be used in chronic treatments. Administration of 15mg/kg/24hrs for 10 days as food supplement efficiently prevented development of depressive-like symptoms, such as despair and anhedonia in chronic stress animal model. It is unlikely the antidepressant activities of fenofibrate are due to non-specific effects on locomotion or anxiety states, since fenofibrate treatment did not affect total distance (12.0±0.2 for vehicle treated and 12.2±0.7 meters for fenofibrate treated mice) or anxiety related parameter of percent of time spent in the centre (6.16±0.79% for vehicle treated 7.0±0.7% for fluoxetine treated mice) in the open field test.

[0134] The experimental results also show PPARa is the functional partner of RXRy in modulation of despair behaviours since: (i) fenofibrate, an agonist of PPARa displayed the same antidepressant effects in the forced swim test as different pan-RXR agonists, including BMS649 and methoprene acid, which were all comparable to classical antidepressant treatment with fluoxetine (Prozac) (ii) sub-threshold doses of fenofibrate synergized with methoprene acid, to decrease despair behaviors in the forced swim test, (iii) genetically compromised PPARa and RXRy signaling in double heterozygous mice synergized in generation of pro-depressive effects leading to significant increase of despair behaviors. Such synergy suggests that PPARa is functionally predominant PPAR interacting with RXRy in control of despair behaviors, which is further supported by our unpublished data on the absence of synergy in generation of despair behaviors in RXRy+/-PPARp+/- and RXRy+/-PPARy+/- double heterozygous mice (data not shown). Depressive-like phenotype displayed by RXRy+/-PPARa+/- mice did not reproduce the full spectrum of depressive-like phenotype of RXRy-KO mice, since double heterozygouse mice did not display anhedonia in sucrose preference test. Such data suggest some degree of functional redundancy between PPARa and different RXRy heterodimerisation partner in control of depressive-like symptoms. This hypothesis is further supported by absence of depressive-like behaviours in single PPARa-KO mice. In order to identify additional partners of RXRy screens for synergies in generation of despair and anhedonia in double heterozygouse mice for RXRy and its potential partners (RARa, RARp, RARg, TRa, TRb or NGFI-b) expressed in the nucleus accumbens, the key region implicated in RXRy control of depressive behaviors. None of the studied double heterozygouse mice displayed depressive-like behaviors (data not shown), thus the nature of nuclear receptor(s) redundant with PPARa in the interactions with RXRy remain to be identified.

[0135] Considering the role of PPARa as modulator of RXRy functions, we speculated that null mutation of PPARa might be a susceptibility factor to develop depression in response to isotretinoin treatment. In agreement with such hypothesis we found that chronic isotretinoin oral treatment (food supplement) at clinical dose of lmg/kg led to dramatic increase of despair behaviours in the forced swim test in PPARa- KO mice, but not in WT control mice fed with the same diet. In addition, isotretinoin treatment led also to marked anhedonia in PPARa-KO mice, but not in WT control group. This depressive-like phenotype was a phenocopy of RXRg-KO mice and was not confused by non-specific behavioral effects, such as abnormal locomotor activity or anxiety (data not shown). The mechanism through which isotretinoin leads to depressive behaviors in PPARa-KO mice is currently under investigation and is based on working hypothesis that isotretinoin might modulate signalling of RARa, β and y (all three isotypes are expressed in the nucleus accumbens) to: (i) compete with PPARa as alternative heterodimerisation partners of RXRy to control gene expression in opposite manner to PPARa, or (ii) compete with PPARa as alternative partners of RXRy to control expression of distinct pool of genes, with opposite functional relevance to those controlled by RXRy/PPARa heterodimer, (iii) sequester the pool of RXRy available to interact with heterodimerisation partner functionally redundant with PPARa (if distinct from RARs). In conclusion, present data provide evidence that PPARa is the susceptibility factor to develop depression in response to isotretinoin treatment in experimental conditions.

REFERENCES

Ainsworth K., Smith S.E., Zetterstrom T.S., Pei Q., Franklin M., and Sharp T. (1998) Effect of antidepressant drugs on dopamine Dl and D2 receptor expression and dopamine release in the nucleus accumbens of the rat. Psychopharmacology (Berl.) 140:470-77.

Alt A., Nisenbaum E.S., Bleakman D., Witkin J.M. (2006) A role for AMPA receptors in mood disorders. Biochem Pharmacol. 71(9): 1273 -88.

Amalric M., and Koob G.F. (1993) Functionally selective neurochemical afferents and efferents of the mesocorticolimbic and nigrostriatal dopamine system. Prog. Brain. Res. 99:209-26.

Anand A., Shekhar A. (2003) Brain imaging studies in mood and anxiety disorders: special emphasis on the amygdala. Ann. N.Y. Acad. Sci. 985:370-88.

Angelucci F., Brene S., Mathe A.A. (2005) BDNF in schizophrenia, depression and corresponding animal models. Mol Psychiatry. 10(4):345-52.

Barcelo-Coblijn G., Hogyes E., Kitajka K., Puskas L.G., Zvara A., Hackler L., Jr., Nyakas C, Penke Z., Farkas T. (2003). Modification by docosahexaenoic acid of age- induced alterations in gene expression and molecular composition of rat brain phospholipids. Proc. Natl. Acad. Sci. U.S.A. 100:11321-26.

Blier P., Pineyro G., el Mansari M., Bergeron R., and de Montigny C. (1998) Role of somatodendritic 5-HT autoreceptors in modulating 5-HT neurotransmission. Ann. N.Y. Acad. Sci. 861 :204-216.

Blier P., and Ward N.M. (2003) Is there a role for 5-HT1A agonists in the treatment of depression? Biol Psychiatry 53:193-203.

Bremner J.D., and McCaffery P. (2008) The neurobiology of retinoic acid in affective disorders. Prog. Neuropsychopharmacol Biol. Psychiatry. 32(2):315-31.

Brocco M., Dekeyne A., Papp M., and Millan M.J. (2006) Antidepressant-like properties of the anti-Parkinson agent, piribedil, in rodents: mediation by dopamine D2 receptors. Behav. Pharmacol. 17:559-72.

Bordet R., Gele P., Duriez P., Fruchart J.C. (2006) PPARs: A new target for neuroprotection. J. Neurol. Neurosurg. Psychiatry. 77:285-87.

Calleja C, Messaddeq N., Chapellier B., Yang H., Krezel W., Li M., Metzger D., Mascrez B., Ohta K., Kagechika H., Endo Y., Mark M., Ghyselinck N.B., Chambon P. (2006) Genetic and pharmacological evidence that a retinoic acid cannot be the RXR- activating ligand in mouse epidermis keratinocytes. Genes Dev. 20:1525-38.

Calon F., Lim G.P., Yang F., Morihara T., Teter B., Ubeda O., Rostaing P., Triller A, Salem N., Jr., Ashe K.H., Frautschy S.A., Cole G.M. (2004) Docosahexaenoic acid protects from dendritic pathology in an Alzheimer's disease mouse model. Neuron 43:633-45.

Campolongo P., Roozendaal B., Trezza V., Cuomo V., Astarita G., Fu J., McGaugh J.L., Piomelli D. (2009) Fat-induced satiety factor oleoylethanolamide enhances memory consolidation. Proc. Natl. Acad. Sci. U.S.A. 106:8027-31.

Carlezon W.A., Jr., Mague S.D., Parow A.M., Stoll A.L., Cohen B.M., Renshaw P.F. (2005) Antidepressant-like effects of uridine and omega-3 fatty acids are potentiated by combined treatment in rats. Biol. Psychiatry 57:343-50.

Cocco S., Diaz G., Stancampiano R., Diana A., Carta M., Curreli R., Sarais L., Fadda F. (2002) Vitamin A deficiency produces spatial learning and memory impairment in rats. Neuroscience 1 15:475-82.

Craddock N., O'Donovan M.C., Owen M.J. (2005) The genetics of schizophrenia and bipolar disorder: dissecting psychosis. J Med Genet. 42(3): 193-204.

Crandall J., Sakai Y., Zhang J., Koul O., Mineur Y., Crusio W.E., McCaffery P. (2004) 13-cis-retinoic acid suppresses hippocampal cell division and hippocampal-dependent learning in mice. Proc. Natl. Acad. Sci. U.S.A. 101 :51 11 -16.

Crowley J.J., Blendy J.A., Lucki I. (2005) Strain-dependent antidepressant-like effects of citalopram in the mouse tail suspension test. Psychopharmacology (Berl.) 183:257-64.

Corrigan, M.H., Denahan, A.Q., Wright, C.E., Ragual, R.J., and Evans, D.L. (2000) Comparison of pramipexole, fluoxetine, and placebo in patients with major depression. Depress. Anxiety 1 1 :58-65.

Dailly, E., Chenu, F., Renard, C.E., and Bourin, M. (2004) Dopamine, depression and antidepressants. Fundam. Clin. Pharmacol. 18:601-07.

Dalvi, A., and Lucki, I. (1999) Murine models of depression. Psychopharmacology (Berl.) 147:14-16.

Dalvi A., Lucki I. (1999) Murine models of depression. Psychopharmacology (Berl) 147:14-16.

de Urquiza A.M., Liu S., Sjoberg M., Zetterstrom R.H., Griffiths W., Sjovall J., Perlmann T. (2000) Docosahexaenoic acid, a ligand for the retinoid X receptor in mouse brain. Science. 290:2140-44.

DeMar J.C., Jr., Ma K., Bell J.M., Igarashi M, Greenstein D., Rapoport S.I. (2006) One generation of n-3 polyunsaturated fatty acid deprivation increases depression and aggression test scores in rats. J. Lipid. Res. 47:172-80. Deplanque D., Gele P., Petrault O., Six I., Furman C, Bouly M., Nion S., Dupuis B., Leys D., Fruchart J.C., Cecchelli R., Staels B., Duriez P., Bordet R. (2003) Peroxisome proliferator-activated receptor-alpha activation as a mechanism of preventive neuroprotection induced by chronic fenofibrate treatment. J. Neurosci. 23:6264-6271.

Dolle P., Fraulob V., Kastner P., Chambon P. (1994) Developmental expression of murine retinoid X receptor (RXR) genes. Mech. Dev. 45:91-104.

Drevets W.C. (2001) Neuroimaging and neuropatho logical studies of depression: implications for the cognitive-emotional features of mood disorders. Curr Opin Neurobiol. l l (2):240-9.

Drevets W.C. (2003) Neuroimaging abnormalities in the amygdala in mood disorders. Ann N Y Acad Sci. 985:420-44.

Dulawa S.C., Holick K.A., Gundersen B., and Hen R. (2004). Effects of chronic fluoxetine in animal models of anxiety and depression. Neuropsychopharmacology 29: 1321-30.

Dziedzicka-Wasylewska M., Rogoz R., Klimek V., and Maj J. (1997). Repeated administration of antidepressant drugs affects the levels of mRNA coding for Dl and D2 dopamine receptors in the rat brain. J. Neural. Transm. 104:515-24.

Etchamendy N., Enderlin V., Marighetto A., Pallet V., Higueret P., Jaffard R. (2003) Vitamin A deficiency and relational memory deficit in adult mice: Relationships with changes in brain retinoid signalling. Behav. Brain Res. 145:37-49.

Etchamendy N., Enderlin V., Marighetto A., Vouimba R.M., Pallet V., Jaffard R., Higueret P. (2001) Alleviation of a selective age-related relational memory deficit in mice by pharmacologically induced normalization of brain retinoid signaling. J. Neurosci. 21 :6423-29.

Fava M., and Kendler K.S. (2000). Major depressive disorder. Neuron 28:335-41.

Fedorova I., Salem N., Jr. (2006) Omega-3 fatty acids and rodent behavior. Prostaglandins Leukot Essent. Fatty Acids 75:271-89.

Foley K.F., DeSanty K.P., and Kast R.E. (2006). Bupropion: pharmacology and therapeutic applications. Expert Rev. Neurother 6:1249-65.

Gofflot F., Chartoire N., Vasseur L., Heikkinen S., Dembele D., Le Merrer J., Auwerx J. (2007) Systematic gene expression mapping clusters nuclear receptors according to their function in the brain. Cell 131 :405-18.

Gould T.D., Manji H.K. (2002) Signaling networks in the pathophysiology and treatment of mood disorders. J Psychosom Res. 53(2):687-97.

Harmon M.A., Boehm M.F., Heyman R.A., Mangelsdorf D.J. (1995) Activation of mammalian retinoid X receptors by the insect growth regulator methoprene. Proc. Natl. Acad. Sci. U.S.A. 92:6157-60.

Hashimoto M., Tanabe Y., Fujii Y., Kikuta T., Shibata H., Shido O. (2005) Chronic administration of docosahexaenoic acid ameliorates the impairment of spatial cognition learning ability in amyloid beta-infused rats. J. Nutr. 135:549-55. Holcomb L.A., Gordon M.N., Jantzen P., Hsiao K., Duff K., Morgan D. (1999) Behavioral changes in transgenic mice expressing both amyloid precursor protein and presenilin-1 mutations: lack of association with amyloid deposits. Behav. Genet. 29:177- 85.

Kastner P., Mark M., Ghyselinck N., Krezel W., Dupe V., Grondona J.M., and Chambon, P. (1997) Genetic evidence that the retinoid signal is transduced by heterodimeric RXR/RAR functional units during mouse development. Development 124:313-26.

Kitajka K., Sinclair A.J., Weisinger R.S., Weisinger H.S., Mathai M., Jayasooriya A.P., Halver J.E., Puskas LG (2004) Effects of dietary omega-3 polyunsaturated fatty acids on brain gene expression. Proc. Natl. Acad. Sci. U.S.A. 101 :10931-36.

Kontaxakis V.P., Skourides D., Ferentinos P., Havaki-Kontaxaki B.J., and Papadimitriou, G.N. (2009). Isotretinoin and psychopathology: a review. Ann. Gen. Psychiatry 8:2.

Krezel W., Ghyselinck N., Samad T.A., Dupe V., Kastner P., Borrelli E., and Chambon P. (1998). Impaired locomotion and dopamine signaling in retinoid receptor mutant mice. Science 279:863-67.

Krezel W., Kastner P., Chambon P. (1999) Differential expression of retinoid receptors in the adult mouse central nervous system. Neuroscience 89:1291-1300.

Krezel W., Dupe V., Mark M., Dierich A., Kastner P., Chambon P. (1996) RXR gamma null mice are apparently normal and compound RXR alpha +/-/RXR beta -/-/RXR gamma -/- mutant mice are viable. Proc. Natl. Acad. Sci. U.S.A. 93:9010-14.

Krzyzosiak A., Szyszka-Niagolov M., Wietrzych M., Gobaille S., Muramatsu S., Krezel W. (2010) Retinoid X Receptor Gamma Control of Motivated Behaviours involves Dopaminergic Signalling in Mice. Neuron, 66(6):908-20.

Lee S.S., and Gonzalez F.J. (1996) Targeted disruption of the peroxisome proliferator- activated receptor alpha gene, PPAR alpha. Ann. N.Y. Acad. Sci. 804:524-29.

Lane M.A., Bailey S.J. (2005) Role of retinoid signalling in the adult brain. Prog. Neurobiol. 75:275-93.

Larisch R., Klimke A., Vosberg FL, Loffler S., Gaebel W., and Muller-Gartner H.W. (1997). In vivo evidence for the involvement of dopamine-D2 receptors in striatum and anterior cingulate gyrus in major depression. Neuro image 5:251-60.

Lee C.H., Olson P., Evans R.M. (2003) Minireview: lipid metabolism, metabolic diseases, and peroxisome proliferator-activated receptors. Endocrinology 144:2201-07.

Lengqvist J., Mata De Urquiza A., Bergman A.C., Willson T.M., Sjovall J., Perlmann T., and Griffiths W.J. (2004). Polyunsaturated fatty acids including docosahexaenoic and arachidonic acid bind to the retinoid X receptor alpha ligand-binding domain. Mol. Cell Proteomics 3:692-703.

Li X.G., Okada T., Kodera M., Nara Y., Takino N., Muramatsu C, Ikeguchi K., Urano F., Ichinose FL, Metzger D., Chambon P., Nakano I., Ozawa K., Muramatsu S. (2006) Viral-mediated temporally controlled dopamine production in a rat model of Parkinson disease. Mol. Ther. 13:160-66. Lim G.P., Calon F., Morihara T., Yang F., Teter B., Ubeda O., Salem N., Jr., Frautschy S.A., Cole G.M. (2005) A diet enriched with the omega-3 fatty acid docosahexaenoic acid reduces amyloid burden in an aged Alzheimer mouse model. J. Neurosci 25:3032- 40.

Logan A.C. (2004). Omega-3 fatty acids and major depression: a primer for the mental health professional. Lipids Health Dis 3:25.

Lopez A.D., Mathers CD., Ezzati M., Jamison D.T., Murray C.J. (2006) Global and regional burden of disease and risk factors, 2001 : systematic analysis of population health data. Lancet 367:1747-57.

Mangelsdorf D.J., Evans R.M. (1995) The RXR heterodimers and orphan receptors. Cell 83:841-50.

Marszalek J.R., Lodish H.F. (2005) Docosahexaenoic acid, fatty acid-interacting proteins, and neuronal function: breastmilk and fish are good for you. Annu. Rev. Cell Dev. Biol. 21 :633-57.

Mattes J.A. (1997). Pergolide to augment the effectiveness of antidepressants: clinical experience and a small double -blind study. Ann. Clin. Psychiatry 9:87-88.

Mazzola C, Medalie J., Scherma M., Panlilio L.V., Solinas M., Tanda G., Drago F., Cadet J.L., Goldberg S.R., Yasar S. (2009) Fatty acid amide hydrolase (FAAH) inhibition enhances memory acquisition through activation of PPAR-alpha nuclear receptors. Learn Mem. 16:332-37.

Messier C, Mrabet O., and Destrade C. (1992). Locomotor bias produced by intra- accumbens injection of dopamine agonists and antagonists. Pharmacol. Biochem. Behav. 41 :177-82.

Millan M.J. (2006). Multi-target strategies for the improved treatment of depressive states: Conceptual foundations and neuronal substrates, drug discovery and therapeutic application. Pharmacol. Ther. 110:135-70.

Millan M.J., Seguin L., Gobert A., Cussac D., and Brocco M. (2004). The role of dopamine D3 compared with D2 receptors in the control of locomotor activity: a combined behavioral and neurochemical analysis with novel, selective antagonists in rats. Psychopharmacology (Berl.) 174:341-57.

Moreau, J.L. (1997). Reliable monitoring of hedonic deficits in the chronic mild stress model of depression. Psychopharmacology (Berl) 134, 357-358; discussion 371-357.

Moriguchi T, Salem N, Jr. (2003) Recovery of brain docosahexaenoate leads to recovery of spatial task performance. J. Neurochem. 87:297-309.

Muscat R, Papp M, Willner P (1992) Reversal of stress-induced anhedonia by the atypical antidepressants, fluoxetine and maprotiline. Psychopharmacology (Berl) 109:433-438.

Naliwaiko K., Araujo R.L., da Fonseca R.V., Castilho J.C., Andreatini R., Bellissimo M.I., Oliveira B.H., Martins E.F., Curi R., Fernandes L.C., and Ferraz A.C. (2004). Effects of fish oil on the central nervous system: a new potential antidepressant? Nutr. Neurosci. 7:91-99. Nestler E.J., and Carlezon W.A., Jr. (2006). The mesolimbic dopamine reward circuit in depression. Biol. Psychiatry. 59:1 151-59.

Nestler E.J., Gould E., Manji H., Buncan M., Duman R.S., Greshenfeld H.K., Hen R., Koester S., Lederhendler I., Meaney M., Robbins T., Winsky L., Zalcman S. (2002). Preclinical models: status of basic research in depression. Biol. Psychiatry 52:503-28.

O'Donnell J (2003) Overview of existing research and information linking isotretinoin (accutane), depression, psychosis, and suicide. Am. J. Ther. 10:148-59.

O'Reilly K.C., Shumake J., Gonzalez-Lima F., Lane M.A., Bailey S.J. (2006) Chronic administration of 13-cis-retinoic acid increases depression-related behavior in mice. Neuropsychopharmaco lo gy 31 : 1919 -27.

Ouagazzal A.M., and Creese I. (2000) Intra-accumbens infusion of D(3) receptor agonists reduces spontaneous and dopamine-induced locomotion. Pharmacol. Biochem. Behav. 67:637-45.

Paterna J.C., Feldon J., and Bueler H. (2004). Transduction profiles of recombinant adeno-associated virus vectors derived from serotypes 2 and 5 in the nigrostriatal system of rats. J Virol 78:6808-17.

Paxinos G., and Franklin K.B.J. (2001). The mouse brain in stereotaxic coordinates (Academic Press San Diego, 1997, ISBN Number 0-12-26607-6; Library of Congress: QL937.F72).

Peet M., and Stokes C. (2005). Omega-3 fatty acids in the treatment of psychiatric disorders. Drugs 65:1051-59.

Pijnenburg A.J., Honig W.M., Van der Heyden J.A., and Van Rossum J.M. (1976). Effects of chemical stimulation of the mesolimbic dopamine system upon locomotor activity. Eur. J. Pharmacol. 35:45-58.

Puskas L.G., Kitajka K., Nyakas C, Barcelo-Coblijn G., Farkas T. (2003) Short-term administration of omega 3 fatty acids from fish oil results in increased transthyretin transcription in old rat hippocampus. Proc. Natl. Acad. Sci. U.S.A. 100:1580-85.

Ramanan S., Zhao W., Riddle D.R., and Robbins M.E. (2010) Role of PPARs in Radiation-Induced Brain Injury. PPAR Res. 2010:234975.

Ramanan S., Kooshki M., Zhao W., Hsu F.C., Robbins M.E. (2008) PPARalpha ligands inhibit radiation-induced microglial inflammatory responses by negatively regulating NF- kappaB and AP-1 pathways. Free Radic. Biol. Med. 45:1695-1704.

Ramanan S., Kooshki M., Zhao W., Hsu F.C., Riddle D.R., Robbins M.E. (2009) The PPARalpha agonist fenofibrate preserves hippocampal neurogenesis and inhibits microglial activation after whole-brain irradiation. Int. J. Radiat. Oncol. Biol. Phys. 75:870-77.

Renard C.E., Fiocco A.J., Clenet F., Hascoet M., and Bourin M. (2001) Is dopamine implicated in the antidepressant-like effects of selective serotonin reuptake inhibitors in the mouse forced swimming test? Psychopharmacology (Berl) 159:42-50.

Samad T.A., Krezel W., Chambon P., and Borrelli E. (1997) Regulation of dopaminergic pathways by retinoids: activation of the D2 receptor promoter by members of the retinoic acid receptor-retinoid X receptor family. Proc. Natl. Acad. Sci. U.S.A. 94:14349-54.

Siuciak J.A. and Fujiwara RA. (2004). The activity of pramipexole in the mouse forced swim test is mediated by D2 rather than D3 receptors. Psychopharmacology (Berl.) 175:163-69.

Stoll A.L., Damico K.E., Daly B.P., Severus WE, Marangell L.B. (2001) Methodological considerations in clinical studies of omega 3 fatty acids in major depression and bipolar disorder. World Rev. Nutr. Diet 88:58-67.

Strakowski S.M., Adler CM., Holland S.K., Mills N.P., DelBello M.P., Eliassen J.C. (2005) Abnormal FMRI brain activation in euthymic bipolar disorder patients during a counting Stroop interference task. Am J Psychiatry. 162(9): 1697-705.

Svensson S., Ostberg T., Jacobsson M., Norstrom C, Stefansson K., Hallen D., Johansson I.C., Zachrisson K., Ogg D., Jendeberg L. (2003) Crystal structure of the heterodimeric complex of LXRalpha and RXRbeta ligand-binding domains in a fully agonistic conformation. Embo J. 22:4625-33.

Tanabe Y., Hashimoto M., Sugioka K., Maruyama M., Fujii Y., Hagiwara R., Hara T., Hossain S.M., Shido O. (2004) Improvement of spatial cognition with dietary docosahexaenoic acid is associated with an increase in Fos expression in rat CA1 hippocampus. Clin Exp Pharmacol Physiol 31 :700-03.

Tanskanen A., Hibbeln J.R., Tuomilehto J., Uutela A., Haukkala A., Viinamaki H., Lehtonen J., Vartiainen E. (2001) Fish consumption and depressive symptoms in the general population in Finland. Psychiatr. Serv. 52:529-31.

Taylor L., Faraone S.V., Tsuang M.T. (2002) Family, twin, and adoption studies of bipolar disease. Curr Psychiatry Rep. 4(2):130-3.

Theohar C, Fischer-Cornelssen K., Brosch H., Fischer E.K., and Petrovic D. (1982). A comparative, multicenter trial between bromocriptine and amitriptyline in the treatment of endogenous depression. Arzneimittelforschung 32:783-87.

Valentini V., Frau R., and Di Chiara G. (2004). Noradrenaline transporter blockers raise extracellular dopamine in medial prefrontal but not parietal and occipital cortex: differences with mianserin and clozapine. J Neurochem 88:917-27.

Wall P.M., Messier C. (2002) Infralimbic kappa opioid and muscarinic Ml receptor interactions in the concurrent modulation of anxiety and memory. Psychopharmacology (Berl.) 160:233-44.

Wietrzych- Schindler M, Szyszka-Niagolov M, Ohta K, Endo Y, Perez E, de Lera AR, Chambon P, Krezel W. (2011) Retinoid X Receptor Gamma Is Implicated in Docosahexaenoic Acid Modulation of Despair Behaviors and Working Memory in Mice. Biol Psychiatry. Apr 15;69(8):788-94

Wietrzych M., Meziane H., Sutter A., Ghyselinck N., Chapman P.F., Chambon P., Krezel W (2005) Working memory deficits in retinoid X receptor gamma-deficient mice. Learn Mem 12:318-26.

Willner P. (1997) Validity, reliability and utility of the chronic mild stress model of depression: a 10-year review and evaluation. Psychopharmacology (Berl.) 134:319- 29.

Willner P. (1997). The mesolimbic dopamine system as a target for rapid antidepressant action. Int Clin Psychopharmacol 12 Suppl 3, S7-14.

Willner P., Hale A.S., and Argyropoulos S. (2005). Dopaminergic mechanism of antidepressant action in depressed patients. J Affect Disord 86:37-45.

Wolf G. (2006) Is 9-cis-retinoic acid the endogenous ligand for the retinoic acid-X receptor? Nutr Rev 64:532-38.

Zetterstrom R.H., Lindqvist E., Mata de Urquiza A., Tomac A., Eriksson U., Perlmann T., Olson L. (1999) Role of retinoids in the CNS: differential expression of retinoid binding proteins and receptors and evidence for presence of retinoic acid. E. J. Neurosci 11 :407-16.

Zhang M., Ouagazzal A., Sun B., and Creese I. (1996). Regulation of motor behavior by dopamine receptor subtypes: an antisense knockout approach. (Humana Press, Totowa, New Jersey).

Claims

1. A method of treating a neuropsychiartic disorder comprising modulating peroxisome proliferator activated receptors (PPAR), retinoid X receptors (RXR), or both, by administering to a subject with a neuropsychiartic disorder a composition comprising a pharmaceutically effective amount of a PPAR modulator, a RXR modulator, or a combination thereof.

2. The method of claim 1 , wherein the PPAR modulator is a PPAR agonist.

3. The method of claim 2, wherein the PPAR agonist is a PPAR a agonist.

4. The method of claim 3, wherein the PPAR β agonist.

5. The method of claim 2, wherien the PPAR agonist is a fibrate, GW501516, 2- bromohexadecanoic acid, perfiuorooctanoic acid, tetradecylthio acetic acid, N- Oleoylethanolamine, WY14643, CP-775146, CP-868388 or GW7647..

6. The method of claim 5, wherien the fibrate is a feno fibrate, bezafibrate, ciprofibrate, clofibrate, or a combination thereof.

7. The method of claim 1 , wherein the modulator is a RXR agonist.

8. The method of claim 7, wherein the RXR agonist is a retinoid, a n-3 polyunsaturated fatty acid, bexarotene, BMS649, diphenylamine derivatives, AGN194204, LGD1069, LGI00268, PA024, 9-cis-UAB30, Net-3IP, Net-31B, LGl 00754, methoprene acid, oleic acid, phytanic acid, CD3254, or a combination thereof.

9. The method of claim 8, wherein the RXR agonist is a retinoid.

10. The method of claim 9, wherein the retinoid is 9-cis retinoic acid.

11. The method of claim 8, wherein the RXR agonist is a n-3 polyunsaturated fatty acid.

12. The method of claim 11 , wherein the n-3 polyunsaturated fatty acid is docosahexaenoic acid (DHA).

13. The method of claim 11 , wherein the n-3 polyunsaturated fatty acid is eicosapentaenoic acid.

14. The method of any one of claims 1 to 13, wherein the modulator is adminstered orally, parenterally, or topically.

15. The method of claim 14, wherein oral administration comprises administration of the composition in the form of a tablet, a soft or hard capsule, lozenges, chews, gels, fast dispersing dosage forms, films, ovules, sprays, buccal/mucoadhesive patches, or diet additive.

16. The method of claim 14, wherein parenteral administration comprises administration of the composition intravenouslly, intra-arterially, intraperitoneally, intrathecally, intraventricularlly, or subcutaneously.

17. The method of claim 14, wherein topical administration comprises administration of the composition in the form of a gel, hydrogel, lotion, solution, cream, ointment, dusting powder, dressing, foam, film, skin patch, wafer, implant, sponge, fiber, bandage, or microemulsion.

18. A method of determining a subject's susceptibility to a neuropsychiatric disorder comprising sampling a biological sample from the subject and detecting a decrease in a PPAR or RXR functionality, wherein a decrease in functionality indicates an increased susceptibility to a neuropsychiatric disorder. Sous le traitement par un retinoide

19. The method of claim 18, wherein detecting a decrease in PPAR or RXR functionality is detected by determining the subject's PPAR genotype, RXR genotype, or both.

20. The method of claim 18, wherein detecting a decrease in PPAR or RXR functionality comprises determining the level or activity of PPAR, RXR, or both in the sample, wherein a decrease in PPAR or RXR levels or activity as compared to a standardized control indicates an increased susceptibility to a neuropsychiatric disorder.

21. The method of claim 18, wherein detecting a decrease in PPAR or RXR functionality comprises determining the level or activity of PPAR, RXR or both prior to administering one or more PPAR or RXR ligands, and determining the level or activity of PPAR, RXR, or both after administering the one or more ligands, wherein a failure to increase PPAR or RXR levels or activities indicates an increased susceptibility to a neuropsychiatric disorder.

22. The method of any one of claims 19 to 21 , wherein the PPAR or RXR funcationality is a PPARa and/or PPAR or RXRy functionality.

23. The method of any one of claims 19 to 22, wherein the neuropsychiatric disorder is major depressive disorder. Bipolar disorder, dysthymia, psychotic symptom

24. A method for identifying pharmacological agents useful in treating neuropsychiatric disorders comprising administering the pharmacological agent to an experimental animal, or cells in an in vitro system, wherein the animal or cells have a compromised PPAR functionality, RXR functionality, or both, and wherein the pharmacological agent's ability to restore deficits in the PPAR or RXR functionality indicates the pharmacological agent as useful in treating neuropsychiatric disorders.

25. The method of claim 24, wherein the experimental animal is a RXR PPAR double knock-out or double heterozygouse RXR +/- PPPAR +/- animal, a RXR knockout animal, a PPAR knock-out animal, a wild type animal treated with RXR antagonist(s), wild type animal treated with PPAR antagonist(s), wild type animal treated with RXR and PPAR antagonists.

26. The method of claims 24, wherein the PPAR or RXR functionality is genetically ablated or pharmacologically inhibited or combination of both.

27 The method of any one of claims 24 to 26, wherein the compromised PPAR functionality is a PPARa and/or PPAR functionality.

28 The method of any one of claims 24 to 26, wherein the compromised RXR functionality is a RXRy functionality. A method for screening pharmacological agents for adverse neuropsychiatric side effects comprising administering the pharmacological agent to an experimental animal according to the claim 25 or an in vitro system and detecting decreased PPAR or RXR functionality, wherein decreased PPAR or RXR functionality indicates the potential for adverse neuropsychiatric side effects.

The method of claim 31 , wherein the pharmacological agent is a retinoid or retinoid derivative and wherein PPAR functionality has been genetically or pharmacologically compromised and wherein the inability of the retinoid or retinoid derivative to decrease RXR functionality indicates a treatment without the potential for adverse side effects.

The method of claim 33, wherein the retinoid or retinoid derivative is for treating acne vulgaris or other related skin disorder.