- Top of page
- Materials and methods
- Supporting Information
Recent evidence underlines the crucial role of neuronal cytoskeleton in the pathophysiology of psychiatric diseases. In this line, the deletion of STOP/MAP6 (Stable Tubule Only Polypeptide), a microtubule-stabilizing protein, triggers various neurotransmission and behavioral defects, suggesting that STOP knockout (KO) mice could be a relevant experimental model for schizoaffective symptoms. To establish the predictive validity of such a mouse line, in which the brain serotonergic tone is dramatically imbalanced, the effects of a chronic fluoxetine treatment on the mood status of STOP KO mice were characterized. Moreover, we determined the impact, on mood, of a chronic treatment by epothilone D, a taxol-like microtubule-stabilizing compound that has previously been shown to improve the synaptic plasticity deficits of STOP KO mice. We demonstrated that chronic fluoxetine was either antidepressive and anxiolytic, or pro-depressive and anxiogenic, depending on the paradigm used to test treated mutant mice. Furthermore, control-treated STOP KO mice exhibited paradoxical behaviors, compared with their clear-cut basal mood status. Paradoxical fluoxetine effects and control-treated STOP KO behaviors could be because of their hyper-reactivity to acute and chronic stress. Interestingly, both epothilone D and fluoxetine chronic treatments improved the short-term memory of STOP KO mice. Such treatments did not affect the serotonin and norepinephrine transporter densities in cerebral areas of mice. Altogether, these data demonstrated that STOP KO mice could represent a useful model to study the relationship between cytoskeleton, mood, and stress, and to test innovative mood treatments, such as microtubule-stabilizing compounds.
Schizophrenia and mood disorders are common, chronic, and debilitating psychiatric illnesses, which have a high prevalence, regardless of countries and cultures, and have a considerable socioeconomic cost (Eaton et al. 2008). For example, unipolar major depression, bipolar disorder, and schizophrenia are ranked first, sixth, and ninth, respectively, in the World Health Organization estimates for disease-related lifetime disabilities, and 2% of humans are affected by schizophrenia or bipolar disorder (Lopez et al. 2006; Mathers and Loncar 2006). Although, the etiology of schizophrenia and mood disorders is yet poorly understood, converging evidence support the view that they can arise from a deficit in cerebral connectivity, synaptic plasticity, and/or neuronal architecture (Mirnics et al. 2001, Frankle et al. 2003, Owen et al. 2005, Schloesser et al. 2008).
Microtubules and microtubule effectors are of fundamental importance to neuronal differentiation and functions. Dysfunctions of the microtubule network have been shown to lead to neurodegenerative diseases and to psychiatric disorders (Gardiner et al. 2011). Recently, it was found that microtubule deregulation and alterations were related to modifications of integrated brain functions both in animal models and in psychiatric diseases. The first evidence for such a role of cytoskeleton disorganization in psychiatric-like characteristics arises from the deletion in mice of the microtubule-stabilizing protein STOP (Stable Tubule Only Polypeptide, Andrieux et al. 2002). Indeed, STOP knockout (KO) mice exhibit abnormalities of glutamatergic, dopaminergic, acetylcholinergic/nicotinic, serotonergic, and noradrenergic neurotransmissions, deficits of neuronal and synaptic plasticity, sensorimotor gating impairment, associated with profound and widespread behavioral defects (Andrieux et al. 2002; Brun et al. 2005; Fradley et al. 2005; Bouvrais-Veret et al. 2007, 2008; Powell et al. 2007; Delotterie et al. 2010; Fournet et al. 2010, 2012; Kajitani et al. 2010). The overall phenotype of STOP KO mice suggests that they represent a relevant experimental model for schizoaffective-like characteristics. Other studies, based on human genetics, also indicate relationship between microtubule-regulatory proteins and mental functions. For example, dysbindin-1 gene mutations have been reported in both schizophrenic (Straub et al. 2002; Benson et al. 2004; Norton et al. 2006) and bipolar patients (Maier 2008; Domschke et al. 2011), and this protein interacts with and regulates microtubules (Talbot et al. 2006). Similarly, mutations in Disrupted-In-Schizophrenia 1 (DISC1) gene are associated with several psychiatric diseases [schizophrenia, bipolar disorders, depression, and autism (Millar et al. 2000; Ishizuka et al. 2006; Blackwood et al. 2007; Chubb et al. 2008; Kilpinen et al. 2008)] and its product is a multifunctional protein acting on microtubules and microtubule-regulatory proteins (Morris et al. 2003; Kamiya et al. 2006; Taya et al. 2007).
A large portion of psychiatric patients are refractory to therapeutic drugs and, during drug treatments, some symptoms are moderately improved or resistant to the current therapy. For example, antipsychotics do not improve negative symptoms and cognitive deficits (Keefe et al. 2007), in spite of a therapeutic benefit for positive schizophrenia symptoms (Seeman et al. 2006). In addition, some drugs need a delay for their therapeutic action, as in the case of antidepressants that necessitate 3–6 weeks to be active (Blier and Montigny 1994). Finally, most of psychiatric drugs elicit a broad range of undesirable side effects, which often lead patients to cease their treatment. Based on such evidence, there is the need to find innovative targets and develop novel therapeutic drugs. In addition, an essential prerequisite for the suitability of an experimental rodent line to model psychiatric-like symptoms is that some deficits will be improved by current therapy (pharmacological or predictive validity).
In the case of STOP KO mice, chronic treatments by both typical and atypical antipsychotics improve some defects, such as the reduced number of hippocampal synaptic vesicles, the post-tetanic potentiation, and/or the long-term potentiation (PTP and LTP, respectively) deficits, the nursing behavior of STOP KO females, the locomotor hyperactivity, the fragmentous activity, and the social interaction (Andrieux et al. 2002; Brun et al. 2005; Fradley et al. 2005; Delotterie et al. 2010; Merenlender-Wagner et al. 2010). Interestingly, a chronic treatment by epothilone D, a taxol microtubule-stabilizing compound (Kolman 2004; Nettles et al. 2004), also improved some deficits of STOP KO mice. In fact, it reduces the decrease of the hippocampal synaptic number, improves PTP and LTP, and alleviates their disorganized spontaneous activity and maternal care deficit (Andrieux et al. 2006).
We recently show that the deletion of the STOP protein triggers a high imbalance of serotonin (5-HT) neurotransmission, with dramatic consequences (Fournet et al. 2010, 2012). Indeed, STOP KO mice are highly depressed and very less anxious than their WT littermates and exhibit impaired short- and long-term memories and spatial learning. Therefore, we characterized the effects of chronic treatment by fluoxetine, a widely used antidepressant selective for 5-HT reuptake, as well as that of chronic treatment by epothilone D. Both chronic treatments were tested on mood status and cognitive memory of WT and STOP KO mice. Moreover, because of paradoxical responses of chronic control-treated STOP KO mice in some behavioral tasks, we tested their reactivity toward an acute stress. Finally, we measured the effects of fluoxetine and epothilone D chronic treatments on the density of serotonin (SERT) and norepinephrine (NET) transporters in brain areas of mice of both genotypes.
- Top of page
- Materials and methods
- Supporting Information
The effects on STOP KO mice of a chronic treatment with fluoxetine, a selective SERT inhibitor, could not be foreseeable, because of the dramatic decrease of SERT density in all brain projection areas of these mice (Fournet et al. 2010, 2012). However, our present study indicated that fluoxetine treatment exerted some effects on the mood of mutant mice. Indeed, chronic treatment by fluoxetine either improved or worsened the depression- and anxiety status of mutant mice. Control-treated STOP KO mice also exhibited paradoxical behaviors, compared with their basal status (Fournet et al. 2010). We hypothesized that the peculiar behavior of control-treated STOP KO mice, as well as the aggravating effects of chronic fluoxetine treatment, were triggered by an altered sensitivity of mutants to stress. Indeed, stress is believed to be a causal factor in the pathogenesis of psychiatric diseases, especially in mood disorders (McEwen 2003). Accordingly, we showed that acutely stressed STOP KO mice displayed a less depressed and more anxious status in some tests, in disagreement with their basal status. Mutant mice also exhibited enhanced plasma corticosterone level, but decreased stress-induced corticosterone stimulation. Worthy of note, our data demonstrated that both epothilone D and fluoxetine chronic treatments improved the short-term memory of STOP KO mice in the novel object recognition task. Finally, neither fluoxetine, nor epothilone D effects were because of variations of SERT and NET densities in the various brain areas tested.
Paradoxical effects of fluoxetine on the mood status of STOP KO mice
We recently demonstrated that STOP KO mice exhibited high variations in SERT density, which increase in 5-HT somas and highly decrease in all the projection areas, triggering dramatic consequences on mood (Fournet et al. 2010, 2012). Actually, STOP KO mice displayed a clear-cut mood in basal conditions, i.e., a depressed and less anxious status. Our present data demonstrated that fluoxetine treatment triggered effects on the mood status of STOP KO mice, in spite of the high disequilibrium of their 5-HT tone.
However, whereas chronic fluoxetine treatment clearly improved the grooming behavior of STOP KO mice, tested by the coat state and the splash test, it elicited a paradoxical response of mutant mice in the forced swimming test, another standardized paradigm for the assessment of despair behavior. Indeed, chronic fluoxetine treatment of STOP KO mice worsened their depressed status, by increasing their immobility time and decreasing (although not significantly) their climbing attempts and latency. The same paradoxical effect of fluoxetine was also found in the forced swimming test after an acute treatment of STOP KO mice (Fournet and Martres, unpublished observations). In the same manner, chronic fluoxetine treatment elicited an anxiolytic effect on STOP KO mice in the marble burying test, but an anxiogenic effect in the light/dark box test, by decreasing the time spent and number of visits in the light box of mutants. These paradoxical effects of chronic fluoxetine could unlikely be because of the fluoxetine dosage selected for chronic treatment. The relatively low dose of fluoxetine was chosen according to its acute effect on the tail suspension test (Fournet et al. 2012). At the dose of 10 mg/kg, fluoxetine had no effect on the immobility of WT mice, whereas it significantly decreased the immobility of STOP KO mice. Also, the aggravating effects of fluoxetine were not because of opposite effects on WT mice, as fluoxetine parallely affected mood of WT and STOP KO mice in these tests.
Interestingly, the two tests upon which chronic fluoxetine exerted a paradoxical effect, i.e., pro-depressant in the forced swimming test and anxiogenic in the light/dark box test, were also those in which control-treated STOP KO mice responded in a paradoxical manner.
Mutant mice were hyper-reactive to acute stress and not tolerant to chronic stress
Although STOP KO mice clearly exhibited a highly depressed status and decreased anxiety status on a series of different tests (Fournet et al. 2012), they exhibited paradoxical responses to some despair and anxiety tests after chronic treatment with vehicle (control-treated). For example, they displayed a depressed-like behavior in the splash test, but they were less depressed than control-treated WT mice in the forced swimming test. In the same manner, whereas control-treated STOP KO were lesser anxious in the marble burying test than WT mice, they were equally anxious in the light/dark box test. However, such an inverted behavior of STOP KO mice was not because of changes in the mood status of control-treated WT mice. Indeed, control treatment of WT mice had variable effects in the forced swimming and no effect in the light/dark box. Accordingly, these opposite behaviors of mutant mice prompted us to test the effects of an acute stress on their mood.
We showed that STOP KO mice were hyper-reactive to acute stress, contrasting with WT mice. In fact, an acute mild stress, induced by a peritoneal administration of saline 30 min before testing, could reverse both the depressed and the less anxious phenotype of STOP KO mice in selected tests. Acute stress exerted an antidepressant effect in mutant mice in the forced swimming and in the tail suspension tests, compared with basal (non-injected) conditions. In the same manner, acute stress had an anxiogenic effect on STOP KO mice in the light/dark box test. This hyper-reactivity of STOP KO mice to acute mild stress has already been reported on their locomotor activity (Brun et al. 2005; Fradley et al. 2005; Begou et al. 2007). In addition, our data suggested that STOP KO mice were not tolerant to chronic stress, as acute and chronic vehicle administration induced the same inverted effects on their mood (see Tables 1 and S5).
Table 1. Compared effects of acute stress, chronic stress, and chronic fluoxetine on the mood status of WT and STOP KO males
The corticosterone plasma level in basal conditions was elevated in STOP KO mice compared with WT, indicating that mutant mice were more stressed than their WT littermates. However, 30 min after saline administration, the increase in corticosterone level, expressed as percent of respective basal levels, was significantly lower in STOP KO than in WT mice. This suggests that the hypothalamic–pituitary–adrenal (HPA) axis in mutant mice may be desensitized, possibly as a consequence of a chronic state of stress. Moreover, the HPA axis being excitated by both noradrenergic and serotonergic neurotransmissions (Herman et al. 2003; Lanfumey et al. 2008), the decreased levels of 5-HT and NE found in projection areas of STOP KO (Fournet et al. 2012) could under-regulate the HPA axis.
Such a desensitization of the HPA axis in mutant mice was in disagreement with their behavioral hyper-reactivity to acute and chronic stress. An explanation of this discordance will be that the tests chosen to characterize the effect of stress, i.e., the forced swimming, tail suspension, and light/dark box tests, triggered a significantly higher additional stress and that the HPA axis in mutant mice, whereas desensitized to mild stress, was hyper-reactive to higher stress. Another explanation will be that the stress induced by these behavioral tests will imply different molecular pathways from those dependent of the HPA axis.
The parallelism between the paradoxical effects of fluoxetine and the paradoxical behaviors of control-treated STOP KO mice suggested that both chronic fluoxetine treatment and chronic stress acted by the same molecular mechanism(s). Finally, as only some tests were sensitive to stress, whereas other were not, it appears to be necessary to use a battery of tests to characterize the depression and anxiety status of mutant mice as STOP KO mice, to avoid stress artifacts.
Chronic epothilone D had little if any effect on the mood of STOP KO mice
The chronic treatment by epothilone D, a microtubule-stabilizing taxol analog—used in cancerology, which can cross the blood–brain barrier—only marginally affected the mood of STOP KO mice and had no effect on the mood of WT mice. It acted as a pro-depressant on the immobility time of mutant mice in the forced swimming test and as an anxiolytic compound on the time spent by STOP KO mice in the light/dark box. It had no effect on all other parameters and tests. The administered dose and the duration of the chronic treatment by epothilone D were selected according to previous study (Andrieux et al. 2006) and to Andrieux and Schweitzer (personal communication). Indeed, after 8-week treatment of STOP KO mice, 0.3–3 mg/kg/week epothilone D has been shown to be efficacious on some deficits and ineffective on others (Andrieux et al. 2006).
Nevertheless, the absence of notable effects of chronic epothilone D treatment on the mood of WT and STOP KO mice suggests that administration of this microtubule-stabilizing drug in adult mice could not have a direct impact on the 5-HT and the NE neurotransmissions and/or the HPA axis.
Both epothilone D and fluoxetine improved short-term memory of STOP KO mice
Very interestingly, we showed that chronic epothilone D and fluoxetine treatments improved the short-term memory of STOP KO mice in the novel object recognition task. We previously showed that STOP KO mice exhibit preserved very short-term memory in the spontaneous alternation test, but impaired short- and long-term memories in the novel object recognition task, as well as learning and memory in the Morris watermaze test (Bouvrais-Veret et al. 2007; Fournet et al. 2012). In this work, control-treated STOP KO mice did not distinguish between the familiar and the novel objects after a time interval of 10 min, as in basal conditions. Very interestingly, they were able to preferentially explore novel objects after chronic epothilone D and fluoxetine treatments.
Up to date, the only reports of a beneficial role of epothilone D or B on spatial learning and memory are on mouse models of tauopathy (Brunden et al. 2010; Barten et al. 2012; Zhang et al. 2012). In these studies, the cognitive improvement of the taxol derivatives is associated with increased microtubule density, axonal integrity, and decreased microtubule hyperdynamic. Such a relation between microtubule-targeting drugs and cognitive function is also found with the octapeptide NAP, a neuronal tubulin-preferring agent, in a mouse model of Alzheimer's disease (Matsuoka et al. 2008), or in heterozygous STOP mice (Merenlender-Wagner et al. 2010). In our case, the improvement of short-term memory of STOP KO mice by chronic epothilone D could be because of its beneficial effects on hippocampal synaptic number deficit, on post-tetanic and long-term potentiation defects and on their disorganized spontaneous activity (Andrieux et al. 2006).
Various neuropsychiatric disorders, including mood disorders, elicited impaired memory and cognitive functions (Levkovitz et al. 2002; Gallassi et al. 2006; Mostert et al. 2008). Thus, the effect of antidepressant therapy has been currently studied on a large scale of cognitive deficits, both in animal models and in human patients. Various studies reported the efficiency of chronic fluoxetine treatment on memory and learning deficits in several experimental mouse models: in two depressed models (learned helplessness and chronic mild stress, Song et al. 2006), in mice with ischemic stroke in hippocampus (Li et al. 2009) and in transgenic mice modeling the Down's syndrome (Bianchi et al. 2010). Interestingly, others showed that chronic fluoxetine treatment could decrease acetylated alpha-tubulin, indicating increased microtubule dynamics in rat hippocampus (Bianchi et al. 2009). Finally, fluoxetine therapy has positive effects regarding the cognitive impairments of depressed patients (Austin et al. 2001, Porter et al. 2003, Weiland-Fiedler et al. 2004, Gallassi et al. 2006), Alzheimer's patients (Mowla et al. 2007), or after traumatic brain injury (Horsfield et al. 2002).
Chronic treatments had no effect on SERT and NET densities
The various effects of chronic epothilone D and fluoxetine treatments were not associated with consequences on the density of SERT and NET, following a 7-day washout. However, we have not measured their uptake activity. Because of the delayed onset of clinical efficacy of antidepressant therapy in mood disorders, the adaptive processes in 5-HT neurotransmission to such treatments have been extensively studied. However, most works have focused on 5-HT receptor sensitivity. The consequences of prolonged antidepressant treatments on the SERT density are often controversial. For example, chronic administration of 2–10 mg/kg/day fluoxetine during 21 days induces either increase, or decrease, or has no effect on SERT brain density (Pineyro and Blier 1999; Benmansour et al. 2002; Hirano et al. 2005). Taken together, these data indicate that adaptive responses of SERT to chronic fluoxetine treatment are not correlated with antidepressant effects.
Interestingly, we found that the % variations of SERT and NET in both basal conditions and after control treatment were highly correlated in various brain areas of STOP KO mice. However, the slope of the linear regression in the case of SERT was significantly higher from 1, suggesting that the chronic mild stress induced by the control treatment exacerbated the 5-HT imbalance of STOP KO mice, whereas it was without consequence on the NE tone.
- Top of page
- Materials and methods
- Supporting Information
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
|jnc12027-sup-0001-FigureS1.tif||image/tif||12781K||Figure S1. (a) Schema of chronic treatments and of the test sequence. (b) Effects of chronic treatments on the body weight.|
|jnc12027-sup-0002-FigureS2.tif||image/tif||15193K||Figure S2. (a) Total fluid consumption. (b) Fluoxetine consumption.|
Table S1. Statistical analyses.
Table S2. Statistical analyses.
Table S3. Effects of chronic treatments on SERT densities in various areas of treated-mice.
Table S4. Effects of chronic treatments on NET densities in various areas of treated-mice.
Table S5. Performances of WT and STOP KO males after various treatments
Appendix S1. Supplementary data.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.