Deletion of CREB-Regulated Transcription Coactivator 1 Induces Pathological Aggression, Depression-Related Behaviors, and Neuroplasticity Genes Dysregulation in Mice

PRIORITY COMMUNICATION

Deletion of CREB-Regulated Transcription Coactivator 1 Induces Pathological Aggression, Depression-Related Behaviors, and Neuroplasticity Genes Dysregulation in Mice Lionel Breuillaud, Clara Rossetti, Elsa M. Meylan, Christophe Mérinat, Olivier Halfon, Pierre J. Magistretti, and Jean-René Cardinaux Background: Mood disorders are polygenic disorders in which the alteration of several susceptibility genes results in dysfunctional mood regulation. However, the molecular mechanisms underlying their transcriptional dysregulation are still unclear. The transcription factor cyclic adenosine monophosphate (cAMP) response element binding protein (CREB) and the neurotrophin brain-derived neurotrophic factor (BDNF) have been implicated in rodent models of depression. We previously provided evidence that Bdnf expression critically rely on a potent CREB coactivator called CREB-regulated transcription coactivator 1 (CRTC1). Methods: To further evaluate the role of CRTC1 in the brain, we generated a knockout mouse line and analyzed its behavioral and molecular phenotype. Results: We found that mice lacking CRTC1 associate neurobehavioral endophenotypes related to mood disorders. Crtc1–/– mice exhibit impulsive aggressiveness, social withdrawal, and decreased sexual motivation, together with increased behavioral despair, anhedonia, and anxiety-related behavior in the novelty-induced hypophagia test. They also present psychomotor retardation as well as increased emotional response to stressful events. Crtc1–/– mice have a blunted response to the antidepressant fluoxetine in behavioral despair paradigms, whereas fluoxetine normalizes their aggressiveness and their behavioral response in the novelty-induced hypophagia test. Crtc1⫺/⫺ mice strikingly show, in addition to a reduced dopamine and serotonin turnover in the prefrontal cortex, a concomitant decreased expression of several susceptibility genes involved in neuroplasticity, including Bdnf, its receptor TrkB, the nuclear receptors Nr4a1-3, and several other CREB-regulated genes. Conclusions: Collectively, these findings support a role for the CRTC1-CREB pathway in mood disorders etiology and behavioral response to antidepressants and identify CRTC1 as an essential coactivator of genes involved in mood regulation.

Key Words: Antidepressant, BDNF, CREB, fluoxetine, major depression, mood disorders, neuroplasticity genes regulation, transcriptional coactivator epressive and bipolar disorders, collectively referred to as mood disorders, are among the leading causes of disability worldwide, with a lifetime prevalence estimate of 20% in the US population (1). Approximately 60% of people committing suicide suffer from a mood disorder, and up to 10 million suicide attempts occur worldwide each year (2). In addition to depressive behavior, impulsiveness, hostility, and aggressiveness are traits associated with suicidal behavior, considered as a dangerous form of self-directed aggression (3). Of particular relevance, a significant proportion of patients with major depression present simultaneous symptoms of anger/aggression, which are reported as “depression with anger attacks,” “agitated depression,” or “irritable-hostile de-

D

From the Center for Psychiatric Neuroscience (LB, CR, EMM, CM, PJM, J-RC), Prilly-Lausanne; Service of Child and Adolescent Psychiatry (LB, CR, EMM, CM, OH, J-RC), Department of Psychiatry, University Medical Center, University of Lausanne; and the Laboratory of Neuroenergetics and Cellular Dynamics (PJM), Brain Mind Institute, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland. Authors CR, EMM, and CM contributed equally to this work. Address correspondence to Jean-René Cardinaux, Ph.D., Centre de Neurosciences Psychiatriques, Site de Cery, CH-1008 Prilly-Lausanne, Switzerland; E-mail: [email protected]. Received Jul 14, 2011; revised Mar 15, 2012; accepted Apr 7, 2012.

0006-3223/$36.00 http://dx.doi.org/10.1016/j.biopsych.2012.04.011

pression” (4 – 6). Comorbid depressive and aggressive symptoms have been reported for bipolar disorder and borderline personality disorders, which have been proposed to be part of the bipolar disorders spectrum (7,8). Moreover, a growing body of evidence suggests that mood disorders and obesity are related, worsening the outcome of these diseases (9). Comorbid depression, hostility, and anger predict increased risk for the metabolic syndrome (10), and the co-occurrence of these symptoms significantly increases suicide ideation and attempts (11). Altogether, these comorbidities suggest that depression, anger, and aggression as well as obesity have related neurobiological bases, and pathologies associating them might share a common etiological pathway. Animal models recapitulating these symptoms associated with mood disorders are thus critically needed to unravel their common molecular etiology. Twin and family studies have established that mood disorders are heritable, with a complex and highly polygenic etiology (12). Among the susceptibility genes involved, evidence from human and rodent studies points toward a dysregulation of the pleiotropic transcription factor cyclic adenosine monophosphate (cAMP) response element binding protein (CREB) and one of its target genes brain-derived neurotrophic factor (Bdnf) (13–15). CREB is involved in a wide range of neuroplasticity processes, including survival of neurons, dendritic growth, neuroprotection, neurogenesis, longlasting forms of synaptic plasticity, and long-term memory, mediated partly by the induction of Bdnf expression (16,17). Sustained hippocampal BDNF levels are necessary to cope with stress-induced depressive behaviors, and chronic antidepressant treatment and electroconvulsive seizure upregulate CREB activity, thus enBIOL PSYCHIATRY 2012;72:528 –536 © 2012 Society of Biological Psychiatry

BIOL PSYCHIATRY 2012;72:528 –536 529

L. Breuillaud et al. hancing the expression of Bdnf and its receptor tyrosine kinase B (TrkB) (13,18 –20). The current network hypothesis of depression points toward neuroplasticity-related malfunction in information processing within neural networks regulating mood (21,22). We previously provided evidence that cAMP- and calcium-induced Bdnf expression as well as hippocampal late-phase longterm potentiation critically rely on the potent and required CREB coactivator CREB-regulated transcription coactivator 1 (CRTC1), also known as transducer of regulated CREB activity 1 (TORC1) (23). To further evaluate the role of CRTC1 in the brain, we and others have generated a knockout mouse line, revealing the development of an obese phenotype in Crtc1–/– mice (24,25). Here, we show that Crtc1–/– mice have a behavioral profile that mirrors the combination of core mood disorders symptoms. They exhibit profoundly altered social-related behaviors, with pathological aggression as well as defective social and sexual interactions. Furthermore, altered emotional response to novel environments and stressful events are also present together with depressive-like symptoms revealed by an increased behavioral despair, anhedonia, and anxiety-related behavior in the novelty-induced hypophagia (NIH) test. Interestingly, Crtc1–/– mice exhibit a differential behavioral response to chronic fluoxetine treatment. Neurochemical analysis reveals reduced dopamine and serotonin turnover in the prefrontal cortex (PFC) of Crtc1–/– mice. Finally, the prefrontocortical-limbic system of mutant mice shows a decreased expression of numerous CREB-target neuroplasticity genes, including Bdnf and its receptor TrkB.

Methods and Materials A detailed Methods and Materials section is available in Supplement 1. Animals and Behavioral Procedures Crtc1–/– mice and wild-type (WT) littermates were generated and genotyped as previously described (24). Male mice were used throughout the study except for the maternal aggression test. Mice were housed under a 12-hour light-dark cycle with ad libitum access to water and standard rodent chow diet. All behavioral tests were carried out in the dark phase of the reverse light cycle according to standard procedures. Activity parameters were measured with the Ethovision video tracking software (Noldus, Wageningen, The Netherlands). Experiments were conducted in accordance with the Swiss Federal Veterinary Office guidelines and were approved by the Cantonal Veterinary Service. Details of the behavioral tests are available in Supplement 1. Gene Expression Analysis Standard protocols for RNA purification, complementary DNA synthesis, and quantitative real-time polymerase chain reaction with SYBR Green detection were used as described (24). Chronic Fluoxetine Administration Fluoxetine hydrochloride was purchased from LKT Laboratories (St. Paul, Minnesota). Mice received fluoxetine hydrochloride in the drinking water at a concentration of 80 mg/L, which corresponded to approximately 18 mg/kg of body weight/day. Fluoxetine solution was replaced every week, and concentration was adjusted depending on the weight and average daily water consumption of the mice. Statistical Analysis We determined statistical differences for one factor between two groups with an unpaired Student t test. In multiple compar-

isons, data were analyzed by two-way analysis of variance with a Fisher Least Significant Difference post hoc test or by two-way analysis of variance for repeated measures with a Newman post hoc test for the comparison of the minute-by-minute moved distance in the open field and the elevated plus maze or for the fear conditioning experiments, when more than two groups were compared.

Results Crtc1–/– Males and Lactating Females Exhibit Enhanced Aggressive Behavior As previously described (24), Crtc1–/– mice are viable, fertile, and show neither obvious developmental defects nor gross neurological alterations. The lacZ reporter cassette introduced in the Crtc1 gene allowed us to characterize the pattern of Crtc1 expression in the brain. Analysis of Crtc1–/– mice brain revealed a widespread expression in cortical and limbic structures, which was confirmed by immunofluorescence analysis of WT mouse brain (Figure S1 in Supplement 1). We noticed, besides the obesity of the adult Crtc1⫺/⫺ mice (24), that Crtc1 null males engaged in chronic aggressive behaviors increasing from 6 to 8 weeks of age. Although they were housed with littermates from the time of weaning, the persistent delivery of bite wounds to the rump, back, and tail of their cage-mates led to severe wounding, thus requiring single housing before the aggression onset. Therefore, we quantified the aggressive behavior in Crtc1–/– male mice with the standard resident-intruder (RI) test after a 2-week isolation period. The Crtc1–/– resident males were significantly more aggressive toward intruders than WT littermates, as measured by their shorter first attack latency and the higher number and duration of their attacks (Figure 1A–C). Because female mice aggression dramatically increases when they are raising offspring, we evaluated the aggressive behavior of Crtc1–/– lactating females toward a harmless female intruder in a maternal aggression test 4 –5 days after parturition (Figure 1D– F). Strikingly, none of the WT females attacked the female intruder, whereas Crtc1 null females exhibited fierce aggression toward the intruder. Crtc1–/– Mice Have Altered Social-Related Behaviors and Increased Depressive-Like Symptoms A closer examination of Crtc1–/– males during the RI test revealed a dramatic reduction in social interaction (Figure 2A). Indeed, Crtc1-mutant males engaged rapidly and frequently in impulsive aggression without the normal investigatory and threatening sequence of acts and postures. This impulsive pathological aggression suggested that CRTC1-deficient males could be impaired in the detection and/or interpretation of socially relevant cues and/or that they try to avoid social contacts. Accordingly, Crtc1–/– males spent less time exploring a male stimulus mouse in a partition test (Figure 2B). Because mouse social behavior depends on olfactory social cues, we evaluated their recognition with a bedding preference test. The reduction in social investigation of Crtc1–/– males in the RI and the partition test was not the result of an impaired recognition of male-derived olfactory cues, because both Crtc1–/– and WT males explored preferentially male-soiled bedding over clean bedding (Figure 2C). In contrast, Crtc1 null males did not spend more time chemoinvestigating female-soiled bedding when presented simultaneously with male-soiled bedding (Figure 2D) or clean bedding (Figure 2E). This absence of preference was not due to a major alteration of the main olfactory system, as revealed by the hidden cookie test (Figure S2A, B in Supplement 1). This selective www.sobp.org/journal

530 BIOL PSYCHIATRY 2012;72:528 –536

L. Breuillaud et al. Crtc1–/– Mice Exhibit an Inhibited Locomotion in Novel Environments and Enhanced Emotional Response to a Footshock Stress Mood-related disorders often coexist with disturbed emotional regulation and anxiety. Therefore, we monitored the behavior of Crtc1–/– and WT littermate mice in the open field, elevated plus maze, and light/dark box tests. The Crtc1 mutant and WT control

Figure 1. Increased aggressive behavior in Crtc1–/– males and lactating females. In the resident-intruder assay, Crtc1–/– male mice (n ⫽ 18) displayed enhanced aggressiveness compared with wild-type (WT) littermates (n ⫽ 14) as measured by a decreased latency to first attack (A) and an increased number (B) and duration (C) of attacks. (D–F) Aberrant aggressive behavior of Crtc1–/– lactating females (n ⫽ 5) toward a female intruder in a maternal aggression test 4 –5 days after parturition. None of the WT females (n ⫽ 7) attacked the female intruder. Data are means ⫾ SEM. *p ⬍ .05, **p ⬍ .01, ***p ⬍ .001.

lack of preference for female-derived olfactory cues suggested that Crtc1–/– males have a deficit in female-directed behavior. Indeed, Crtc1 null males spent less time close to a partition that separated them from a receptive female (Figure 2F), and consistent with a reduced sexual motivation, none of them exhibited mounting behaviors when the receptive female was introduced in their home cage for 15 min (Figure 2G). Remarkably, among the eight resident Crtc1–/– males tested, three behaved aberrantly and attacked the female intruder. In keeping with their altered social abilities, CRTC1-deficient mice had an impaired nesting behavior (Figure S2C in Supplement 1), which is considered as a social behavior in mice and an important component of fitness. Altered social interactions, anger/aggression, social withdrawal, and decreased sexual motivation are symptoms associated with mood disorders. Hence, we examined depressive-like behaviors in Crtc1–/– mice by measuring immobility in the forced swim test (FST), a model of depression-like behavioral despair. During two consecutive daily sessions, Crtc1 null mice displayed significantly increased immobility as compared with their WT littermates (Figure 2H). Accordingly, Crtc1–/– mice had a reduced preference for a sucrose solution, revealing anhedonia-like symptoms (Figure 2I). www.sobp.org/journal

Figure 2. Altered social-related behaviors and increased depressive-like symptoms in Crtc1⫺/⫺ mice. (A) Crtc1⫺/⫺ male mice (n ⫽ 10) displayed reduced social investigation during the resident-intruder test compared with wild-type (WT) littermate control subjects (n ⫽ 8). (B) Crtc1⫺/⫺ male mice (n ⫽ 10) spent less time than WT (n ⫽ 9) exploring an unfamiliar male mouse stimulus in a homecage partition test. The perforated transparent partition prevents agonistic interactions, whereas it allows social investigation through its holes. (C–E) In a soiled bedding preference test, Crtc1⫺/⫺ and WT male mice (n ⫽ 9 –10) had a normal preference for male-soiled bedding over clean bedding (C), whereas Crtc1⫺/⫺ males exhibited an altered chemoinvestigation of female chemosignals as shown by their lack of preference for female-soiled bedding over malesoiled bedding (D) or clean bedding (E). (F) Crtc1⫺/⫺ male mice spent significantly less time than WT males close to a perforated transparent partition that separate them from an estrous female, reflecting a decreased sexual motivation (n ⫽ 8 for both genotypes). (G) Crtc1⫺/⫺ male mice did not exhibit sexual mounting behavior when the receptive female was introduced in their home cage for 15 min (Crtc1⫺/⫺: 0/8, WT: 5/8). (H, I) Increased depressive-like behaviors in CRTC1-deficient mice. Crtc1⫺/⫺ male mice (n ⫽ 20) were significantly more immobile than WT control mice (n ⫽ 19) in the two consecutive daily sessions of the forced swim test (FST) (H). In a two-bottle choice test (I), Crtc1⫺/⫺ male mice (n ⫽ 11) had a reduced preference for a 2.5% sucrose solution over water compared with WT control mice (n ⫽ 7). Data are means ⫾ SEM. *p ⬍ .05, **p ⬍ .01, ***p ⬍ .001.

L. Breuillaud et al.

BIOL PSYCHIATRY 2012;72:528 –536 531 Figure 3. Inhibited locomotor exploration in novel environments, and enhanced immediate fear response and cued fear conditioning in Crtc1⫺/⫺ mice. (A) Crtc1⫺/⫺ and wild-type (WT) male mice spent a similar time in the lit area of the light/dark box, the central zone of the open field, and the open arms of the elevated plus maze. (B–D) Crtc1⫺/⫺ males exhibited inhibited ambulation in the behavioral tests used to assess anxiety. In the open field (B), CRTC1-deficient mice (n ⫽ 18) had a significantly decreased locomotion compared with WT control mice (n ⫽ 18) during the first minute of the test. In the elevated plus maze (C), CRTC1-deficient mice (n ⫽ 18) exhibited a decreased distance traveled in 1-min time bins compared with WT control mice (n ⫽ 17). In the light/ dark box (D), Crtc1⫺/⫺ male mice (n ⫽ 19) made fewer transitions between the two compartments than WT males (n ⫽ 18). (E) Crtc1⫺/⫺ male mice (n ⫽ 18) froze significantly more than WT control mice (n ⫽ 17) after the first tone-shock pairing of the fear conditioning paradigm. The 3 horizontal lines indicate the auditory tones, which were coterminated with foot shocks (vertical bars). (F) Crtc1 null males reacted with a higher velocity than WT males during the 2-sec period after the foot shocks. (G) Crtc1–/– and WT male mice had a similar contextual fear conditioning response when placed in the same context 24 hours after the training session. (H) Crtc1⫺/⫺ male mice exhibited an increased auditory cued fear conditioning when exposed to the tone (horizontal line) in a novel context. Notice the emotional-like response (freezing) of CRTC1-deficient mice before the tone onset, suggesting a generalization of fear responses to unknown environments. Data are means ⫾ SEM. *p ⬍ .05, **p ⬍ .01.

mice did not spend a significantly different time in the anxiogenic area of the three tests, suggesting that anxiety-related behaviors are not drastically affected in Crtc1–/– mice (Figure 3A; Figure S3A–D in Supplement 1). However, like CREB␣␦ knockout mice (26), Crtc1 null mice persistently exhibited a neophobia-induced hypolocomotion. When placed in the novel environment of the open field, Crtc1–/– mice showed a significant inhibition of ambulatory activity restricted to the first minute of the test (Figure 3B). Nevertheless, the distance traveled in the open field after this initial period was similar for Crtc1–/– and WT mice, indicating that the decreased exploration at the beginning of the test was not due to a locomotor impairment. A general decrease in Crtc1 null mice locomotor activity was also observed in the elevated plus maze (Figure 3C), and mutant mice made fewer transitions between the two compartments of the light/dark box (Figure 3D). Together, these results show that CRTC1-deficient mice are inhibited and less active in novel, moderately stressful environments without showing major differences in other anxiety-like behavior parameters. The Crtc1–/– mice were subjected to the fear conditioning paradigm, to further evaluate their emotional behavior. Mostly used to study learning and memory of fear, this test is also useful to assess emotional response to a traumatic stress and is considered as a model of post-traumatic stress disorder (27). The Crtc1 null mice showed significantly higher shock reactivity, as revealed by an increased velocity immediately after the shocks and more freezing in re-

sponse to the first tone-shock pairing (Figure 3E, F). The Crtc1–/– and WT control mice displayed a similar contextual fear conditioning the day after the training session (Figure 3G), whereas Crtc1 mutant mice had an increased emotional-like response in the novel environment of the auditory-cued fear conditioning (Figure 3H). Their freezing behavior before the presentation of the auditory cue likely resulted from a generalization of fear responses to different contexts. Together, these results provide evidence that Crtc1–/– mice exhibit an increased emotional response to novel environments in the form of inhibited ambulation reminiscent of the psychomotor retardation occurring in human depression as well as increased immediate, generalized, and conditioned fear responses to stressful events. Crtc1–/– Mice Have a Reduced Dopamine and Serotonin Turnover in the PFC Altered monoamine levels in specific regions of the brain are associated with anxiety, impulsive aggression, feeding behavior, and depression (28,29). We thus monitored by high-performance liquid chromatography the levels of monoamines and metabolites in the PFC and hippocampus (Figure 4) as well as the hypothalamus and nucleus accumbens (Figure S4 in Supplement 1) of Crtc1–/– and WT male mice. We found significantly lower levels of the dopamine metabolites 3,4 dihydroxy-phenylacetic acid and homovanillic acid as well as of the serotonin metabolite 5-hydroxyindole acetic acid in www.sobp.org/journal

532 BIOL PSYCHIATRY 2012;72:528 –536

L. Breuillaud et al. nucleus accumbens (Figure S4B in Supplement 1) of Crtc1–/– mice. Reduced dopamine and serotonin release in the PFC has been associated with aggressive behaviors, impulsivity, and depressivelike behaviors, suggesting that these changes in monoamine turnover might account for the altered behavioral phenotype of CRTC1deficient mice.

Figure 4. Reduced dopamine (DA) and serotonin (5-HT) turnover in the prefrontal cortex of Crtc1–/– mice. High-performance liquid chromatography (HPLC) analysis of DA, 3,4 dihydroxy-phenylacetic acid (DOPAC), homovanillic acid (HVA), norepinephrine (NE), 5-HT, and 5-hydroxyindole acetic acid (5-HIAA) in the prefrontal cortex (A) and hippocampus (B) of Crtc1–/– and wild-type (WT) male mice (n ⫽ 8 for both genotypes). The significant decrease of DA (DOPAC, HVA) and 5-HT metabolites (5-HIAA) in the prefrontal cortex of Crtc1–/– mice reflects a reduced DA and 5-HT turnover in this structure. Data are means ⫾ SEM. *p ⬍ .05, **p ⬍ .01, ***p ⬍ .001.

the PFC of Crtc1 null mice (Figure 4A), which indicated reduced dopaminergic and serotonergic activities in this structure. No significant differences in levels of monoamines and their metabolites were observed in any other brain regions tested, except for a slight decrease of serotonin in the hippocampus (Figure 4B) and the

Decreased CRE-Mediated Gene Expression in Relevant Prefrontocortical-Limbic Structures of Crtc1–/– Mice To unravel the impact of CRTC1 deletion on CRE-mediated transcription in the brain, we examined the expression of well-characterized CREB-regulated genes implicated in mood disorders. The PFC, hippocampus, hypothalamus, and nucleus accumbens, where Crtc1 is expressed (Figure S1 in Supplement 1), have been the main focuses of our investigations due to their implications in aggression, obesity, and depression. The Bdnf gene has multiple promoters, and CREB is mainly contributing to its regulation by binding to the CRE present in the activity-dependent promoter IV (17,30). We measured specific Bdnf transcripts from the most characterized five first promoters as well as total Bdnf expression with primers in the common coding exon IX. Of major interest to Bdnf regulation, we found a decreased expression of Bdnf exon IV in the PFC, hippocampus, and hypothalamus of Crtc1–/– mice (Figure 5A, B; Figure S5A in Supplement 1). As a result, total Bdnf expression was significantly decreased in the PFC and hippocampus, with a likely additive effect of exon V downregulation in the PFC. Hence, we found reduced BDNF protein levels in the PFC and hippocampus of Crtc1–/– mice (Figure 5E, F). In the hypothalamus, total Bdnf levels were not significantly reduced, despite exon I and IV decreased expression, whereas no significant alteration in Bdnf expression was found in the nucleus accumbens (Figure S5B in Supplement 1). The BDNF receptor TrkB has also been associated with mood disorders (31,32) and shown to be regulated by CREB (33,34). Interestingly, TrkB messenger RNA (mRNA) levels were decreased in all four structures tested, further contributing to a decrease in neurotrophin signaling (Figure 5C, D; Figure S5C, D in Supplement 1).

Figure 5. Reduced prefrontocortical-limbic expression of brain-derived neurotrophic factor (Bdnf) and tyrosine kinase B (TrkB) in Crtc1⫺/⫺ mice. Quantitative real-time reverse transcriptase-polymerase chain reaction analysis of total RNA samples extracted from the prefrontal cortex (PFC) (A, C), and hippocampus (Hipp) (B, D) of Crtc1–/– and wild-type (WT) male mice (n ⫽ 6 for both genotypes) under basal home cage housing conditions. (A, B) Bdnf exon I, II, III, IV, V transcripts as well as total Bdnf expression. (C, D) BDNF receptor TrkB messenger RNA (mRNA) levels. Expression levels were normalized to ␤-actin and are presented as fold change relative to WT levels, which were set to 1.0. (E, F) Brain-derived neurotrophic factor (BDNF) enzyme-linked immunosorbent assay analysis of total BDNF protein levels in the PFC (E) and the Hipp (F) of Crtc1⫺/⫺ and WT male mice (n ⫽ 4 for both genotypes) under basal home cage housing conditions. Data are means ⫾ SEM. *p ⬍ .05, **p ⬍ .01, ***p ⬍ .001.

www.sobp.org/journal

L. Breuillaud et al.

BIOL PSYCHIATRY 2012;72:528 –536 533

Figure 6. Reduced prefrontocortical-limbic expression of nuclear receptor 4A (Nr4a)1-3 and relevant cyclic adenosine monophosphate (cAMP) response element-regulated genes in Crtc1–/– mice. Quantitative polymerase chain reaction analysis of total RNA samples extracted from the prefrontal cortex (A, C) and hippocampus (B, D) of Crtc1–/– and wild-type (WT) male mice (n ⫽ 6 for both genotypes) under basal home cage housing conditions. (A, B) Orphan nuclear receptors Nr4a1-3 messenger RNA (mRNA) levels. (C, D) Expression of the cAMP response element binding protein target genes c-fos, FosB, Crem-Icer, somatostatin (Sst), Cartpt (Cart), and Pgc-1␣. Expression levels were normalized to ␤-actin and are presented as fold change relative to WT levels, which were set to 1.0. Data are means ⫾ SEM. *p ⬍ .05, **p ⬍ .01, ***p ⬍ .001.

We were also particularly interested in the prototypic CREB target genes coding for the nuclear receptor 4A (NR4A) family of orphan nuclear receptors, which were implicated in molecular mechanisms regulating dopaminoceptive and dopaminergic structures as well as the hypothalamic–pituitary–adrenal axis (35–38). We found a major decrease of the three NR4A members in the PFC as well as in the hippocampus of CRTC1-deficient mice (Figure 6A, B). The NR4A1 and NR4A3 were also significantly downregulated in the nucleus accumbens, whereas a significant decrease of NR4A3 was found in the hypothalamus (Figure S5E, F in Supplement 1). To gain deeper insight into the role of CRTC1 in the expression of other CREB-regulated genes, we analyzed the PFC and hippocampal mRNA levels of c-fos, FosB, Crem-Icer, Somatostatin, Cartpt, and Pgc-1␣, which contain one or more CRE motifs and have been implicated in several neural mechanisms, including neuroplasticity. The Crtc1–/– mice displayed a significantly decreased expression of all genes in both structures (Figure 6C, D), except for Pgc-1␣ that was decreased only in the PFC. The decreased expression of Crem, a member of the CREB family, prompted us to measure the levels of Creb, Crtc2, Crtc3, Cbp, and p300 mRNAs, which were actually not significantly different in the brain of Crtc1–/– mice and WT littermates (Figure S6A, B in Supplement 1). These findings provide further evidence of a major role of CRTC1 for the proper regulation of neuroplasticity genes involved in mood regulation. Differential Behavioral Response to Chronic Antidepressant Treatment in Crtc1–/– Mice To assess the role of CRTC1 in the antidepressant response, Crtc1⫺/⫺ males and WT littermates were treated orally with fluoxetine (18 mg/kg/day). This protocol of fluoxetine administration was based on prior studies showing that this dose resulted in plasma levels within the therapeutic range in humans (39). We first tested the effects of chronic fluoxetine on the aggressive behavior of Crtc1⫺/⫺ males in the RI paradigm (Figure 7A–C). The reduced number and duration of attacks of the Crtc1⫺/⫺ males treated with fluoxetine indicated that their aggressive behavior was diminished by this antidepressant.

We next wanted to explore the effects of chronic fluoxetine administration on the depressive-like behaviors of Crtc1–/–mice. Because several weeks of antidepressant drug administration are needed to treat depressed patients, animal models of depression responding only to chronic antidepressant exposure are preferred over those that react to acute antidepressant administration, such as for instance the FST. Moreover, a better face validity with respect to human depression is achieved if the animals treated by antidepressants display depressive-like symptoms that are usually induced by various chronic stress procedures (40). We chose to apply a recently developed chronic stress paradigm called the repeated open-space forced swim (OSFS) model of depression that induces chronic depressive-like symptoms in mice, which were reversed by chronic but not acute administration of antidepressants (41,42). This procedure is a modification of the FST in which mice are initially subjected to 15-min daily swims in rat tub cages for 4 consecutive days, during which they acquire a syndrome of passive, immobile water behavior that persists unchanged for at least 1 month as assessed by the maintenance swims performed at 3- or 4-day intervals. The Crtc1 null males exhibited, according to the depressivelike endophenotypes of Crtc1–/–mice, an increased depressive-like behavior in the OSFS model of depression as revealed by a significantly increased immobile water behavior compared with their WT littermates (Figure 7D). The WT mice chronically treated with fluoxetine displayed a gradual decreased immobility, which became statistically significant 19 days after the beginning of fluoxetine treatment. Interestingly, however, Crtc1–/–mice did not respond at all to fluoxetine in the repeated OSFS model of depression. These striking differences in behavioral response between WT and Crtc1–/– mice are likely not due to an altered hypothalamic–pituitary–adrenal axis in mutant mice, because their basal and stressinduced plasma corticosterone levels were similar to those of WT mice (Figure 7E, F). During the last 2 weeks of the repeated OSFS test, we investigated the depressive-like symptoms of the animals and their response to chronic fluoxetine in the tail suspension test (TST) and NIH test (Figure 7G, H). As in the OSFS test, chronic fluoxetine significantly reduced the immobility of WT but not Crtc1–/– www.sobp.org/journal

534 BIOL PSYCHIATRY 2012;72:528 –536

L. Breuillaud et al.

Figure 7. Behavioral response of Crtc1⫺/⫺ mice to chronic antidepressant treatment. (A–C) Effect of long-term fluoxetine treatment (80 mg/L in drinking water corresponding to approximately 18 mg/kg/day, 5 weeks) on the aggressive behavior of Crtc1⫺/⫺ males in the resident-intruder assay. (A) The latencies to first attack of Crtc1⫺/⫺mice were significantly lower than those of the wild-type (WT) littermates (⫹⫹⫹p ⬍ .001), and fluoxetine had no effect on these latencies. (B) The increased number of attacks of Crtc1⫺/⫺ males (⫹⫹⫹p ⬍ .001 vs. WT) was reduced by the chronic fluoxetine treatment (**p ⬍ .01 vs. Crtc1⫺/⫺ water group). (C) The longer duration of attacks of Crtc1⫺/⫺ males (⫹p ⬍ .05 vs. WT) was also decreased by fluoxetine (*p ⬍ .05 vs. Crtc1⫺/⫺ water group). (D) Effect of chronic fluoxetine administration on the repeated open-space forced swim (OSFS) model of depression in Crtc1⫺/⫺ mice and WT littermates. Time of floating (immobility in seconds) during the 15-min sessions of the OSFS test increased for all groups during the four consecutive daily swims (Days ⫺4 to ⫺1). The two groups of Crtc1⫺/⫺mice were significantly more immobile than WT control mice on the third and fourth sessions (⫹p ⬍ .05 vs. WT). From Day 0 to Day 40, one group of WT mice and one group of Crtc1–/– mice received approximately 18 mg/kg/day of fluoxetine in the drinking water. Crtc1–/–mice spent much more time floating than WT control mice (⫹⫹⫹p ⬍ .001 vs. WT). Fluoxetine treatment significantly reduced immobility time in WT mice (*p ⬍ .05 vs. WT fluoxetine Day 2, #p ⬍ .05 vs. WT water group), whereas it had no effect on Crtc1⫺/⫺mice. (E) Similar plasma corticosterone levels in Crtc1–/–and WT mice under basal conditions. (F) Plasma corticosterone levels in the 4 groups of mice that were subjected to the repeated OSFS test before the first swim (basal), after the four consecutive daily swims, and after the 10th swim on Day 19. The stressful conditions of the repeated forced swim sessions significantly increased corticosterone levels (***p ⬍ .001 vs. basal), which were still high on Day 19 for all the groups with no significant effect ofthefluoxetinetreatment.(G)AbsenceofbehavioralresponseofCrtc1⫺/⫺micetochronicfluoxetineinthetailsuspensiontest.AfterDay23oftherepeatedOSFSprocedure, immobility of the four groups of mice was measured in the tail suspension test. Chronic fluoxetine significantly reduced the immobility of WT mice (***p ⬍ .001 vs. WT water group), whereas it had no effect on Crtc1⫺/⫺mice. (H) Response of Crtc1–/–mice to chronic fluoxetine in the novelty-induced hypophagia paradigm. Mean latencies to consume sweetened condensed milk in the home and novel environment are shown for the mice that underwent the repeated swim procedure. Of the 35 mice tested, 2 wereeliminatedfromanalysisforhavinglatencyscores⬎2SDfromthemean.Crtc1⫺/⫺miceexhibitedanincreasedlatencyrelativetoWTmice(⫹p⬍.05vs.WTwatergroup), and chronic fluoxetine significantly reduced the latency of Crtc1⫺/⫺mice (*p ⬍ .05 vs. Crtc1⫺/⫺ water group). Data are means ⫾ SEM.

www.sobp.org/journal

L. Breuillaud et al. mice in the TST. The NIH paradigm is a conflict test in which mice are trained to approach a reward (sweetened condensed milk) in their home cage and then placed in a novel brightly lit cage. The latency to approach and drink the sweetened milk is a measure of the anxiety-related behavior associated with this task, which has been reported to be sensitive to chronic but not acute antidepressant treatment (43). In keeping with their previously observed depressive-like behaviors, Crtc1–/– mice displayed a higher latency to drink the reward in the novel environment compared with WT mice (Figure 7H). In this test, however, Crtc1–/– mice responded to chronic fluoxetine as assessed by the significant reduction of the latency to drink sweetened milk in the anxiogenic environment. Taken together, these data further strengthen the depression-related behaviors of Crtc1–/– mice and show a differential response to fluoxetine in these mutant mice.

Discussion The present reverse genetic study reveals a pivotal role of CRTC1 for adapted social and mood-related behavior, likely through the regulation of several susceptibility genes involved in neuroplasticity. Mice lacking CRTC1 combine major endophenotypes related to core mood disorder symptoms (27,44). In addition, we and others have previously reported that Crtc1–/– mice have increased feeding behavior and obesity, which are symptoms associated with depression (9). Together, these findings suggest that CRTC1 alteration might be a common molecular substrate for comorbid pathological aggression, obesity, and depression-related behaviors, supporting the view that CREB pathway hypofunction is primarily involved in mood disorder pathogenesis. A network of prefrontal cortical, limbic, and striatal neuronal circuits has been implicated in mood regulation and thus supports the behavioral manifestations of mood disorders (15,22,45). To explore the mechanisms underlying the behavioral phenotype of Crtc1–/– mice, we analyzed monoamines and metabolites as well as gene expression profiles in key structures of these affective circuits. We found decreased levels of dopamine and serotonin metabolites in the PFC of Crtc1–/– mice, which reflect a lower release of these monoamines that has been associated with impulsivity, aggressiveness/irritability, and depressive-like behaviors (28,29). Moreover, Crtc1–/– mice displayed altered expression of several genes associated with neuropsychiatric disorders in all the brain structures tested, with a greater extent in the PFC and the hippocampus, which are central in the pathophysiology of mood disorders. Interestingly, the expression of Bdnf exon IV, total Bdnf, as well as BDNF protein and its cognate receptor TrkB was significantly lower in both structures of CRTC1-deficient mice. According to the neurotrophic hypothesis of depression, the behavioral phenotype of Crtc1–/– mice might thus partly result from a decrease in BDNF function in these relevant brain structures. A comparison of the behavioral profile of Crtc1–/– mice with genetically modified Creb or Bdnf mouse models reveals both similarities and differences at several levels. In apparent contradiction with the postulated role of CREB in depression and antidepressant response, CREB-deficient mice showed a reduced immobility in the FST and TST and a normal response to antidepressants in these tests (46). In contrast, Crtc1–/– mice displayed an increased immobility in the FST and the OSFS model of depression and a blunted response to fluoxetine in this latter test and the TST. By contrast, CREB knockout mice behaved similarly in the NIH test compared with Crtc1–/– mice, with an increased anxiety in the mutant mice that was reduced by chronic antidepressant treatment (47). Of interest, up-

BIOL PSYCHIATRY 2012;72:528 –536 535 regulating CREB function in the basolateral amygdala or in the nucleus accumbens increased depressive-like behaviors, which might explain the antidepressant-like phenotype of CREB knockout mice (20). Like Crtc1–/– mice, Bdnf heterozygous mice and BDNF Valine66Methionine transgenic mice exhibit enhanced aggressive behavior and develop a hyperphagic, obese phenotype (48 –50). Specific genetic deletion of Bdnf in the dorsal forebrain results in a decreased activity in the open field without alteration in anxiety parameters and an increased fear response to foot shocks (51). Moreover, the genetic inactivation of Bdnf promoter IV-driven transcription resulted in depressive-like behaviors and reduced activity in the open field test (52). Genetically modified Bdnf mouse models show a blunted behavioral response to antidepressants, both in the FST and the NIH test (14,50,53), unlike Crtc1–/– mice that responded to fluoxetine in the NIH test. The characteristics of the phenotype of Crtc1⫺/⫺ mice might also be attributed to the lower expression levels of the transcription factors NR4A1-3 and the AP-1 family members c-fos and FosB as well as other CRE-regulated genes, which play a role in neuroactivity-dependent plasticity and are relevant to mood disorders (see Supplementary Discussion in Supplement 1). In addition to the role of CRTC1 in controlling the expression of the genes shown in the present study, several pieces of evidence support its involvement in neuronal plasticity processes (23,54 –56). We hypothesize, according to the network hypothesis of depression (21), that impairment of CRTC1 levels or activity might lead to a decrease of neuronal connectivity within specific neural networks and to the associated malfunction of information processing underlying mood disorders. In summary, CRTC1-deficient mice provide a new and valuable animal model that uniquely combines major behavioral, neurochemical, and molecular endophenotypes associated with mood disorders commonly mediated through the alteration of the CRTC1-CREB molecular pathway. This work was supported by grants from the Swiss National Science Foundation 3100A0-120699 and 31003A-135692, the Jérôme Lejeune Foundation, and the Société Académique Vaudoise. Author PJM is the recipient of the Asterion Foundation Chair. We thank A. Porret, A-M. Mérillat, and E. Hummler (Transgenic Animal Facility, University of Lausanne) for blastocyst injection, technical assistance, and suggestions; C. Centeno and E. Grouzmann (Division of Clinical Pharmacology and Toxicology of the Lausanne University Medical Center) for high-performance liquid chromatography measurement of monoamines and metabolites; F. Magara for critical discussions and suggestions. The authors report no biomedical financial interests or potential conflicts of interest. Supplementary material cited in this article is available online. 1. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE (2005): Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 62:593– 602. 2. Mann JJ (2003): Neurobiology of suicidal behaviour. Nat Rev Neurosci 4:819 – 828. 3. Mann JJ, Arango VA, Avenevoli S, Brent DA, Champagne FA, Clayton P, et al. (2009): Candidate endophenotypes for genetic studies of suicidal behavior. Biol Psychiatry 65:556 –563. 4. Fava M (1998): Depression with anger attacks. J Clin Psychiatry 59(suppl 18):18 –22. 5. Painuly N, Sharan P, Mattoo SK (2005): Relationship of anger and anger attacks with depression: A brief review. Eur Arch Psychiatry Clin Neurosci 255:215–222. 6. Benazzi F, Akiskal H (2005): Irritable-hostile depression: Further validation as a bipolar depressive mixed state. J Affect Disord 84:197–207.

www.sobp.org/journal

536 BIOL PSYCHIATRY 2012;72:528 –536 7. Smith DJ, Muir WJ, Blackwood DH (2004): Is borderline personality disorder part of the bipolar spectrum? Harv Rev Psychiatry 12:133–139. 8. Garno JL, Gunawardane N, Goldberg JF (2008): Predictors of trait aggression in bipolar disorder. Bipolar Disord 10:285–292. 9. McElroy SL, Kotwal R, Malhotra S, Nelson EB, Keck PE, Nemeroff CB (2004): Are mood disorders and obesity related? A review for the mental health professional. J Clin Psychiatry 65:634 – 651, quiz 730. 10. Goldbacher EM, Matthews KA (2007): Are psychological characteristics related to risk of the metabolic syndrome? A review of the literature. Ann Behav Med 34:240 –252. 11. Carpenter KM, Hasin DS, Allison DB, Faith MS (2000): Relationships between obesity and DSM-IV major depressive disorder, suicide ideation, and suicide attempts: Results from a general population study. Am J Public Health 90:251–257. 12. Kato T (2007): Molecular genetics of bipolar disorder and depression. Psychiatry Clin Neurosci 61:3–19. 13. Duman RS, Monteggia LM (2006): A neurotrophic model for stressrelated mood disorders. Biol Psychiatry 59:1116 –1127. 14. Martinowich K, Manji H, Lu B (2007): New insights into BDNF function in depression and anxiety. Nat Neurosci 10:1089 –1093. 15. Krishnan V, Nestler EJ (2008): The molecular neurobiology of depression. Nature 455:894 –902. 16. Lonze BE, Ginty DD (2002): Function and regulation of CREB family transcription factors in the nervous system. Neuron 35:605– 623. 17. Greer PL, Greenberg ME (2008): From synapse to nucleus: Calciumdependent gene transcription in the control of synapse development and function. Neuron 59:846 – 860. 18. Martinowich K, Lu B (2008): Interaction between BDNF and serotonin: Role in mood disorders. Neuropsychopharmacology 33:73– 83. 19. Carlezon WA Jr, Duman RS, Nestler EJ (2005): The many faces of CREB. Trends Neurosci 28:436 – 445. 20. Blendy JA (2006): The role of CREB in depression and antidepressant treatment. Biol Psychiatry 59:1144 –1150. 21. Castren E (2005): Is mood chemistry? Nat Rev Neurosci 6:241–246. 22. Castren E, Rantamaki T (2010): The role of BDNF and its receptors in depression and antidepressant drug action: Reactivation of developmental plasticity. Dev Neurobiol 70:289 –297. 23. Kovacs KA, Steullet P, Steinmann M, Do KQ, Magistretti PJ, Halfon O, et al. (2007): TORC1 is a calcium- and cAMP-sensitive coincidence detector involved in hippocampal long-term synaptic plasticity. Proc Natl Acad Sci U S A 104:4700 – 4705. 24. Breuillaud L, Halfon O, Magistretti PJ, Pralong FP, Cardinaux JR (2009): Mouse fertility is not dependent on the CREB coactivator Crtc1. Nat Med 15:989 –990; author reply 991. 25. Altarejos JY, Goebel N, Conkright MD, Inoue H, Xie J, Arias CM, et al. (2008): The Creb1 coactivator Crtc1 is required for energy balance and fertility. Nat Med 14:1112–1117. 26. Hebda-Bauer EK, Watson SJ, Akil H (2004): CREB deficient mice show inhibition and low activity in novel environments without changes in stress reactivity. Eur J Neurosci 20:503–513. 27. Cryan JF, Holmes A (2005): The ascent of mouse: Advances in modelling human depression and anxiety. Nat Rev Drug Discov 4:775–790. 28. Nelson RJ, Trainor BC (2007): Neural mechanisms of aggression. Nat Rev Neurosci 8:536 –546. 29. Seo D, Patrick CJ, Kennealy PJ (2008): Role of serotonin and dopamine system interactions in the neurobiology of impulsive aggression and its comorbidity with other clinical disorders. Aggress Violent Behav 13:383– 395. 30. Hong EJ, McCord AE, Greenberg ME (2008): A biological function for the neuronal activity-dependent component of Bdnf transcription in the development of cortical inhibition. Neuron 60:610 – 624. 31. Dwivedi Y, Rizavi HS, Conley RR, Roberts RC, Tamminga CA, Pandey GN (2003): Altered gene expression of brain-derived neurotrophic factor and receptor tyrosine kinase B in postmortem brain of suicide subjects. Arch Gen Psychiatry 60:804 – 815. 32. Kohli MA, Salyakina D, Pfennig A, Lucae S, Horstmann S, Menke A, et al. (2010): Association of genetic variants in the neurotrophic receptorencoding gene NTRK2 and a lifetime history of suicide attempts in depressed patients. Arch Gen Psychiatry 67:348 –359.

www.sobp.org/journal

L. Breuillaud et al. 33. Kingsbury TJ, Murray PD, Bambrick LL, Krueger BK (2003): Ca(2⫹)-dependent regulation of TrkB expression in neurons. J Biol Chem 278: 40744 – 40748. 34. Deogracias R, Espliguero G, Iglesias T, Rodriguez-Pena A (2004): Expression of the neurotrophin receptor trkB is regulated by the cAMP/CREB pathway in neurons. Mol Cell Neurosci 26:470 – 480. 35. Murphy EP, Conneely OM (1997): Neuroendocrine regulation of the hypothalamic pituitary adrenal axis by the nurr1/nur77 subfamily of nuclear receptors. Mol Endocrinol 11:39 – 47. 36. Perlmann T, Wallen-Mackenzie A (2004): Nurr1, an orphan nuclear receptor with essential functions in developing dopamine cells. Cell Tissue Res 318:45–52. 37. Levesque D, Rouillard C (2007): Nur77 and retinoid X receptors: crucial factors in dopamine-related neuroadaptation. Trends Neurosci 30:22–30. 38. Hawk JD, Abel T (2011): The role of NR4A transcription factors in memory formation. Brain Res Bull 85:21–29. 39. Dulawa SC, Holick KA, Gundersen B, Hen R (2004): Effects of chronic fluoxetine in animal models of anxiety and depression. Neuropsychopharmacology 29:1321–1330. 40. Nestler EJ, Hyman SE (2010): Animal models of neuropsychiatric disorders. Nat Neurosci 13:1161–1169. 41. Stone EA, Lin Y, Quartermain D (2008): Evaluation of the repeated openspace swim model of depression in the mouse. Pharmacol Biochem Behav 91:190 –195. 42. Stone EA, Lin Y (2011): Open-space forced swim model of depression for mice. Curr Protoc Neurosci Chapter 9:Unit 9.36. 43. Dulawa SC, Hen R (2005): Recent advances in animal models of chronic antidepressant effects: The novelty-induced hypophagia test. Neurosci Biobehav Rev 29:771–783. 44. Nestler EJ, Barrot M, DiLeone RJ, Eisch AJ, Gold SJ, Monteggia LM (2002): Neurobiology of depression. Neuron 34:13–25. 45. Price JL, Drevets WC (2010): Neurocircuitry of mood disorders. Neuropsychopharmacology 35:192–216. 46. Conti AC, Cryan JF, Dalvi A, Lucki I, Blendy JA (2002): cAMP response element-binding protein is essential for the upregulation of brain-derived neurotrophic factor transcription, but not the behavioral or endocrine responses to antidepressant drugs. J Neurosci 22:3262–3268. 47. Gur TL, Conti AC, Holden J, Bechtholt AJ, Hill TE, Lucki I, et al. (2007): cAMP response element-binding protein deficiency allows for increased neurogenesis and a rapid onset of antidepressant response. J Neurosci 27:7860 –7868. 48. Lyons WE, Mamounas LA, Ricaurte GA, Coppola V, Reid SW, Bora SH, et al. (1999): Brain-derived neurotrophic factor-deficient mice develop aggressiveness and hyperphagia in conjunction with brain serotonergic abnormalities. Proc Natl Acad Sci U S A 96:15239 –15244. 49. Kernie SG, Liebl DJ, Parada LF (2000): BDNF regulates eating behavior and locomotor activity in mice. Embo J 19:1290 –1300. 50. Chen ZY, Jing D, Bath KG, Ieraci A, Khan T, Siao CJ, et al. (2006): Genetic variant BDNF (Val66Met) polymorphism alters anxiety-related behavior. Science 314:140 –143. 51. Gorski JA, Balogh SA, Wehner JM, Jones KR (2003): Learning deficits in forebrain-restricted brain-derived neurotrophic factor mutant mice. Neuroscience 121:341–354. 52. Sakata K, Jin L, Jha S (2010): Lack of promoter IV-driven BDNF transcription results in depression-like behavior. Genes Brain Behav 9:712–721. 53. Saarelainen T, Hendolin P, Lucas G, Koponen E, Sairanen M, MacDonald E, et al. (2003): Activation of the TrkB neurotrophin receptor is induced by antidepressant drugs and is required for antidepressant-induced behavioral effects. J Neurosci 23:349 –357. 54. Espana J, Valero J, Minano-Molina AJ, Masgrau R, Martin E, GuardiaLaguarta C, et al. (2010): beta-Amyloid disrupts activity-dependent gene transcription required for memory through the CREB coactivator CRTC1. J Neurosci 30:9402–9410. 55. Li S, Zhang C, Takemori H, Zhou Y, Xiong ZQ (2009): TORC1 regulates activity-dependent CREB-target gene transcription and dendritic growth of developing cortical neurons. J Neurosci 29:2334 –2343. 56. Finsterwald C, Fiumelli H, Cardinaux JR, Martin JL (2010): Regulation of dendritic development by brain-derived neurotrophic factor (BDNF) requires activation of the CREB-regulated transcription coactivator 1 (CRTC1) by glutamate. J Biol Chem 285:28587–28595.