International Journal of Neuropsychopharmacology (2014), 17, 1307–1313. doi:10.1017/S1461145714000212
© CINP 2014
B R I E F R E PO R T
Acute but not sustained aromatase inhibition displays antidepressant properties Nikolaos Kokras1,2, Nikolaos Pastromas1, Tatiany H. Porto1,3, Vasilios Kafetzopoulos1, Theodore Mavridis1 and Christina Dalla1 1
Department of Pharmacology, Medical School, University of Athens, Greece First Department of Psychiatry, Eginition Hospital, Medical School, University of Athens, Greece 3 Psychology Institute, University of São Paulo, Brazil 2
Abstract
Received 22 July 2013; Reviewed 3 September 2013; Revised 3 February 2014; Accepted 5 February 2014; First published online 27 March 2014 Key words: Oestrogens, females, letrozole, progesterone, testosterone.
Introduction Aromatase inhibitors block the conversion of androgens to oestrogens and are used in post-menopausal women for the treatment of hormone-responsive breast cancer. An expanded use of them is under consideration, for treating hormone-responsive tumors in premenopausal women (Winer et al., 2002; Dowsett and Haynes, 2003). However, in post-menopausal women, aromatase inhibitors associate with psychiatric and cognitive side-effects (Phillips et al., 2011), while their safety in premenopausal women needs clarification. Experimental evidence suggests that deprivation of ovarian hormones is a risk factor for developing depression, especially after exposure to stressful experiences (Lagunas et al., 2010). Furthermore, response to antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), appears modulated by the female hormonal milieu (Estrada-Camarena et al., 2010; Keating et al., 2011). Therefore, we asked whether decreased oestrogens following aromatase inhibition would lead to a depressivelike behavioural response in cycling female rats, which
Address for correspondence: Assistant Professor C. Dalla, Department of Pharmacology, Medical School, University of Athens, Mikras Asias 75, Goudi, 11527, Greece. Tel.: +30 2107462577 Fax: +30 2107462554 Email: [email protected]
simulate premenopausal women. Furthermore, we investigated if gonadal hormonal changes due to aromatase inhibition would interfere with the SSRI antidepressant response, and we asked whether short-term vs. sustained aromatase inhibition would result in a similar behavioural response in the forced swim test (FST).
Method We used female Sprague–Dawley rats, group-housed under controlled 12:12 light/dark cycles (lights on at 07:00 hours) and temperature (22 ± 2 °C), with free access to food and water, aged 12 wk at the beginning of the experiment (Experiment 1 n = 40, Experiment 2 n = 46). Efforts were made to minimize the number of animals used and their suffering, in accordance with institutional regulations and the EU directive 2010/63. In experiment 1, 40 females received saline (Vehicle A) or 5 mg/kg fluoxetine (Eli Lilly Hellas S.A., Greece) for 28 d (n = 20 per group). At the 27th day of fluoxetine treatment all rats were subjected to the modified FST. They were individually placed in a cylindrical tank measuring 50 cm height × 20 cm width, filled with water (24 ± 1 °C) to a depth of 40 cm (Kokras et al., 2012). On the first day, rats were forced to swim for 15 min. After carefully drying, and before returning to their home cages, rats were injected i.p. with 1 mg/kg letrozole (Novartis Pharma
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
Aromatase inhibitors block the conversion of androgens to oestrogens and are used for the treatment of hormone-responsive breast cancer in menopause and recently also in premenopausal women. We investigate whether decreased oestrogen synthesis following aromatase inhibition leads to a depressive-like behavioural response in cycling female rats. Using the forced swim test (FST) we estimate the response of acute (three injections in 24 h) and sustained (7 d) letrozole and fluoxetine administration. Acute aromatase inhibition decreases immobility duration in the FST, indicating its antidepressant potential. Instead, sustained aromatase inhibition did not show such antidepressant potential. Testosterone elevation associates with the decreased depressive behaviour in the FST following acute letrozole treatment, but interestingly progesterone explains the increased swimming behaviour. Present findings may have potential implications for women treated with aromatase inhibitors, especially before menopause, as well as for the role of gonadal hormones in the expression of depressive symptoms and antidepressant response.
1308 N. Kokras et al.
Results Experiment 1: acute aromatase inhibition Behavioural measurements A two-way ANOVA for immobility duration showed a significant letrozole × fluoxetine interaction (F1,36 = 4.350 p = 0.044). Subsequent post-hoc testing showed that in rats pretreated with saline and those pretreated with fluoxetine letrozole administration resulted in reduced immobility (p < 0.001; p = 0.035, respectively) (Fig. 1a).
(a)
Immobility duration #
100
Vehicle B Letrozole
* Duration (s)
80 60 40 20 0
Vehicle A pretreatment
(b)
Fluoxetine pretreatment
Swimming duration 250
# Vehicle B Letrozole
*
Duration (s)
200 150 100 50 0
(c)
Head twitching frequency 40
Frequency (counts)
Vehicle A pretreatment Fluoxetine pretreatment
*
*
Vehicle B Letrozole
30
20
10
0
Vehicle A pretreatment Fluoxetine pretreatment
Fig. 1. Duration of immobility (a), swimming (b) and frequency of head twitches (c) during the second 5 min forced swim test following pretreatment for 28 d of saline vehicle or fluoxetine (5 mg/kg) and three injections of either vehicle B or letrozole. Note the reduction of immobility and increase in swimming duration after three injections of letrozole, which is similar to the antidepressant behavioural response after 28 d of fluoxetine treatment. An asterisk (*) denotes a significant letrozole effect and a hash sign (#) a fluoxetine effect, as indicated by the appropriate post-hoc test. n = 10/group, means ± S.E.M.
Acute letrozole administration did not alter the fluoxetine-induced reduction in immobility levels. Similarly, with regards to swimming duration, there was a significant letrozole × fluoxetine interaction (F1,36 = 5.659 p = 0.23). Post-hoc testing indicated that acute letrozole significantly increased swimming in vehicle-pretreated rats (p = 0.002) (Fig. 1b). Chronic treatment with fluoxetine also enhanced swimming (p = 0.045). There were no significant differences in climbing (data not shown). However, there was a weak letrozole main effect in increasing head twitching frequency (F1,36 = 4.069 p = 0.051) (Fig. 1c).
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
AG, Switzerland) or with vehicle B (saline with 5% Tween 80 and 5% ethanol, n = 10 per each group of vehicle or fluoxetine pre-treated rats). A second and a third letrozole or vehicle B injection were administered 19 and 23 h after the first injection, respectively. One hour after the third injection and 24 h after the first FST session, rats were subjected to a second 5 min FST session, during which the total duration of immobility, swimming, climbing and frequency of head twitching were manually measured (Mikail et al., 2012). Immediately after the second FST, rats were sacrificed and trunk blood was used for serum extraction. Estradiol, progesterone and testosterone assays were performed using commercially available RIA kits (Siemens Coat-A-Count for total testosterone and progesterone and double antibody estradiol). The detection limits were 0.4 ng/ml, 0.02 ng/ml and 1.4 pg/ml, respectively. In experiment 2, 46 females received vehicle A (saline, n = 24) or fluoxetine 5 mg/kg (n = 22) for 28 d, as in experiment 1. During the last week of treatment, rats received, along with vehicle A or fluoxetine, seven daily injections of vehicle B (saline with 5% Tween 80 and 5% ethanol; n = 12 for vehicle A pre-treated rats and n = 10 for fluoxetine pre-treated rats) or letrozole (n = 12 for vehicle A for pre-treated rats, n = 12 for fluoxetine). At the 27th day of fluoxetine and 6th day of letrozole treatment, rats were subjected to the first FST session and the following day (28th day of fluoxetine and 7th of letrozole) underwent the second FST test session. The FST and hormones assay procedure was identical to experiment 1, with the exception of treatments as described. During both experiments, vaginal smears were monitored as before (Dalla et al., 2009). Results were analysed with SPSS v.21 (IBM SPSS Inc, USA) using a two-way analysis of variance (ANOVA) with independent factors fluoxetine treatment (vehicle A or 5 mg/kg) and letrozole treatment (Vehicle B or 1 mg/kg). Significant interactions and main effects were explored with post-hoc tests using Bonferonni’s type I error correction. For each experiment associations between behaviour and hormones were explored initially using univariate linear regression. This was followed by a complete multivariate model, which included both treatments and all hormones. Statistical significance was set at p < 0.05. Results are reported as means ± S.E.M.
Letrozole in FST (a)
Testosterone
(d) *
250
1309
Progesterone 60
#
Serum levels (ng/ml)
Serum levels (ng/ml)
200 150
* 100 50 0
20
0 Vehicle A pretreatment
Fluoxetine pretreatment
Vehicle B
Letrozole
Vehicle A pretreatment Vehicle B
(e)
Testosterone–immobility regression
Letrozole
Progesterone–immobility regression 150
Immobility duration
150
100
50
100
50
0
0 0
100
200
300
400
0
500
20
Testosterone (ng/ml)
(c)
40
60
80
100
Progesterone (ng/ml)
(f )
Testosterone–swimming regression
Progesterone–swimming regression 300
Swimming duration
300
Swimming duration
Fluoxetine pretreatment
200
100
200
100
0
0 0
100
200
300
400
500
Testosterone (ng/ml)
0
20
40
60
80
100
Progesterone (ng/ml)
Fig. 2. Testosterone serum levels (a) and univariate linear regression of testosterone serum levels with immobility (b) and swimming (c) duration. Progesterone (d) and univariate linear regression of progesterone serum levels with immobility (e) and swimming (f) following 28 d of saline or fluoxetine (5 mg/kg) pretreatment and three injections of either vehicle B or letrozole. Note that testosterone was significantly associated with immobility duration but not with swimming duration. Inversely, progesterone was not associated with immobility duration in the univariate regression model, but with swimming duration. For multivariate regression models refer to the text. An asterisk (*) denotes a significant letrozole effect and a hash sign (#) a fluoxetine effect, as indicated by the appropriate post-hoc test. N = 10/group, means ± S.E.M.
Hormonal assays As expected, contrary to females receiving vehicle, female rats that received three injections of letrozole in 24 h had estradiol serum levels near or below the detection limit (data not shown), although vaginal smears were normal
in all females. As a result of the aromatase inhibition, testosterone levels were significantly enhanced (main effect: F1,35 = 11.939 p = 0.01) in all letrozole-treated rats previously treated with saline (vehicle A) or fluoxetine (p = 0.033; p = 0.001, respectively) (Fig. 2a). Interestingly, a marginally non-significant fluoxetine main effect
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
(b)
Immobility duration
40
1310 N. Kokras et al. was also observed in testosterone levels (p = 0.078). Specifically, fluoxetine increased testosterone in vehicle B and letrozole-treated rats, as indicated by post-hoc testing (p = 0.033; p = 0.001, respectively). With regards to progesterone, a significant letrozole × fluoxetine interaction was revealed (F1,35 = 5.126 p = 0.030). Further post-hoc testing showed that in saline-pretreated rats acute letrozole treatment had no effect in progesterone levels. In fluoxetine-pretreated rats letrozole only marginally reduced progesterone (p = 0.076). Instead, in salinepretreated rats, fluoxetine increased progesterone (p = 0.030), whereas this was not observed in rats receiving letrozole (Fig. 2d).
Univariate linear regressions showed that testosterone levels correlated with immobility (F1,38 = 4.250 p = 0.046) (Fig. 2b), whereas progesterone correlated with swimming (F1,38 = 7.457 p = 0.010) (Fig. 2f). A multivariate regression model including as predictors all treatments (fluoxetine, letrozole) and hormones (estradiol, testosterone, progesterone) significantly explained 48% of variance in immobility (F5,34 = 6.379 p < 0.001) and 31% of variance in swimming (F5,34 = 3.042 p = 0.022). Letrozole and estradiol were the strongest predictors in the regression model for immobility (p = 0.003 and p = 0.035, respectively), whereas progesterone weakly contributed to the model (p = 0.098). Instead, with regards to swimming, the strongest regression predictors were letrozole and progesterone (p = 0.016 and p = 0.014, respectively). The same multivariate regression model did not explain a significant percentage of variance of climbing and head twitching. Interestingly, head twitching showed marginal non-significant associations in the multivariate model with fluoxetine treatment and estradiol (p = 0.081 and p = 0.070, respectively).
Experiment 2: sustained aromatase inhibition Behavioural measurements A two-way ANOVA for immobility, swimming and climbing behaviour did not show any significant fluoxetine × letrozole interactions. The analysis showed a marginally non-significant fluoxetine main effect (F1,42 = 3.658 p = 0.063), given that fluoxetine tended to shorten the immobility duration. As expected, fluoxetine significantly increased swimming duration (F1,42 = 6.634 p = 0.014), but post-hoc pairwise comparisons with Bonferonni correction showed that this effect reached significance only in vehicle B treated rats (p = 0.015) and not in letrozole-treated rats. Sustained aromatase inhibition had no effects on behaviour except a significant main effect in head twitching behaviour (F1,42 = 7.038 p = 0.011). Post-hoc testing showed that only vehicle A pretreated and not fluoxetine-pretreated rats had more head swings
(p = 0.023)
Hormonal assays All rats receiving repeated letrozole injections for 1 wk had disrupted oestrous cycle and their oestrogen levels were near or below detection limit, which was not the case for rats receiving vehicle B (data not shown). Testosterone levels were markedly increased following repeated letrozole treatment (F1,42 = 58.363 p < 0.001). Post-hoc testing showed that letrozole alone and in combination with fluoxetine pretreatment significantly increased testosterone (p < 0.001 in both treatments). Interestingly, fluoxetine also had a significant effect on testosterone levels (F1,42 = 6.283 p = 0.016) and post-hoc pairwise comparisons showed that fluoxetine either alone or in combination with letrozole tended to increase progesterone. With regards to progesterone, sustained aromatase inhibition caused a significant decrease (F1,42 = 8.092 p = 0.007). Post-hoc testing showed that letrozole, either alone or in combination with fluoxetine, decreased progesterone (p = 0.043 and p = 0.058, respectively) (Supplementary materials Figure S2). Association of behavioural and hormonal indices None of the monitored behavioural indices correlated with hormones as indicated by correlation and linear regression models, with the exception of head twitching frequency, which was negatively correlated with progesterone serum levels (r=−0.298 p = 0.045 n = 46) and positively with testosterone levels (r = 0.339 p = 0.021 n = 46). Discussion Present findings show that acute aromatase inhibition induces an antidepressant behavioural response. Instead, sustained aromatase inhibition for 1 wk results in no effect. These findings may have potential implications for women treated with aromatase inhibitors, especially before menopause, as well as for the role of gonadal hormones in the expression of depressive symptoms and antidepressant response. Specifically, acute letrozole treatment affects swimming and not climbing in the FST, like fluoxetine and all SSRI (Detke et al., 1995). Given that swimming duration in the FST has been linked with serotonergic changes induced by SSRIs, it can be assumed that acute letrozole influences mainly the serotonergic system and its behavioural response. In line with the antidepressant properties of acute letrozole treatment, case reports evidence induction of mania and mood swings in early stages of treatment with aromatase inhibitors (Goodwin, 2006; Rocha-Cadman et al., 2012). On the other hand, letrozole treatment for 1 wk showed no antidepressant potential, as it had no significant effect in the FST.
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
Association of behavioural and hormonal indices
after repeated letrozole treatment (Supplementary materials Figure S1).
Letrozole in FST
be involved, as implied by the fact that the multifactorial regression in experiment 2 did not explain a significant part of the variance. Although our hypothesis provides a plausible explanation, more studies are warranted to further explore such interactions. Moreover, whereas the FST is widely used as a screening tool for monitoring antidepressant potential (Porsolt, 1979; Cryan et al., 2005), chronic stress models would probably shed more light. In our study, letrozole dosage can hardly explain the observed difference between acute and sustained aromatase inhibition because we used a high dose that totally blocks the aromatase enzyme (Kafali et al., 2004; Aydin et al., 2008). This is further confirmed by an ongoing study (preliminary results communicated at SfN (Dalla et al., 2012)), where aromatase activity was found completely blocked with this dose. Regarding fluoxetine, previous studies in female rats have shown that chronic administration of a suboptimal 5 mg/kg fluoxetine dose may or may not result in an antidepressant-like behavioural response or have a small effect size (Mitic et al., 2013). Beyond this known issue of suboptimal SSRI dosing in chronic treatment, in our chronic experiment letrozole does not show antidepressant potential and also does not augment the low fluoxetine dose in producing an antidepressant effect when letrozole and fluoxetine are combined together. Finally, acute letrozole treatment enhances head twitching frequency, while fluoxetine has no effect. Also, in the second experiment, head twitching frequency correlates with progesterone and testosterone. These findings further associate head twitching in the FST with gonadal hormones. Previous findings from our group showed that females twitch their heads less than males, and that head twitching is higher in oestrous/proestrous (Drossopoulou et al., 2004; Kokras et al., 2009; Kokras et al., 2012). Head twitching is linked to 5-HT2 receptor activity in the open field (Wieland et al., 1993) and perhaps this FST behaviour might be used as representative of the effect of gonadal hormones on the serotonergic system. Present findings indicate that in considering treatment with aromatase inhibitors in premenopausal women, attention should be considered regarding possible psychotropic effects of aromatase inhibitors. Seemingly, treatment duration with aromatase inhibitors plays an important role in aromatase-induced mood effects. The unexpected acute antidepressant potential of an aromatase inhibitor raises the intriguing question of whether such molecules could be investigated in combination with known antidepressants and in short-term dosing regimens in order to kick-start an antidepressant response, or as a transient short-lived augmentation strategy in treatment-resistant depression. Supplementary material For supplementary material accompanying this paper, visit http://dx.doi.org/10.1017/S1461145714000212
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
A lack of effect in the FST was also reported after 21 d of letrozole treatment in ovariectomized mice (Meng et al., 2011). Such differentiated effects of acute and sustained aromatase inhibition are surprising, because oestrogens have been associated with antidepressant properties, so a decrease in oestrogen levels by aromatase inhibition should hypothetically result in a ‘depressivelike’ phenotype (Estrada-Camarena et al., 2003, 2010). Although treatment duration in rats does not equally translate to humans, our data support a time-dependent biphasic effect following aromatase inhibition. In support of this concept, previous studies on female mice knocked out since birth for CYP19A1 (aromatase gene) displayed in adulthood depressive symptomatology in the FST (Dalla et al., 2004). In addition, aromatase inhibition had detrimental effects on long-term potentiation and spine density in the hippocampus, and those effects were dependent on the duration of letrozole treatment (Zhou et al., 2010; Vierk et al., 2012). Thus, we support that aromatase inhibition initially results in mood elation, subsequently this phenomenon vanishes and, in line with the aforementioned studies at a later stage (beyond the 1–3 wk time interval), a pro-depressive effect might appear. Although this hypothesis needs further investigation, based on data reported herein there is now strong evidence that duration of treatment plays a crucial role in letrozole’s effects in brain and behaviour. A possible explanation of the remarkable antidepressant effect observed in this study might lay in hormonal adjustments following aromatase inhibition. Exogenous testosterone administration is known to exert an antidepressant effect in the FST (Frye and Walf, 2009), and letrozole substantially increases testosterone, as seen in both experiments. However, sustained aromatase inhibition causes a more pronounced testosterone increase in comparison to subacute inhibition, but no antidepressant effect was observed. In addition, although testosterone correlates with immobility in the acute experiment, the multivariate regression does not confirm a key role for testosterone. In fact, the antidepressant behavioural response is explained by letrozole and, interestingly, by oestrogen depletion with regards to immobility, and by progesterone increase with regards to swimming. Thus, we speculate that a rapid oestrogen decline, when combined with an equally rapid testosterone and progesterone increase, may result in antidepressant response. When this hormonal adjustment is sustained, as in experiment 2, other compensatory mechanisms kick in and, for example, progesterone stabilizes to lower levels. As progesterone withdrawal results to a ‘depressive-like’ symptomatology in the FST (Li et al., 2012), it can be suggested that, following sustained letrozole treatment, the prolonged progesterone decrease counteracts for the oestrogen decline and the testosterone enhancement, and thus the antidepressant effect of letrozole is no longer evident. However, our study cannot prove causal relationships between hormones and behaviour, and other factors might
1311
1312 N. Kokras et al. Acknowledgments The authors acknowledge Novartis Pharma AG for providing letrozole, Eli Lilly Hellas S.A. for fluoxetine, Mrs Despina Papassava and Dr Mikael Hudu Garba for excellent technical assistance. CD and NK are supported by a ‘Large Scale Cooperative Project’ (09SYN-21–1003) and VK by ‘Education and Lifelong Learning, Supporting Postdoctoral Researchers’. Both projects are co-financed by the European Social Fund (ESF) and the General Secretariat for Research and Technology, Greece. THP had a CAPES Foundation fellowship from the Brazilian Ministry of Education. None of these entities had any influence over this study.
NK has received speaker’s fees, consultancy honoraria and travel support from Janssen, Lundbeck, SanofiAventis, Medochemie Generics and Elpen S.A. None of those are relevant to this study.
References Aydin M, Yilmaz B, Alcin E, Nedzvetsky VS, Sahin Z, Tuzcu M (2008) Effects of letrozole on hippocampal and cortical catecholaminergic neurotransmitter levels, neural cell adhesion molecule expression and spatial learning and memory in female rats. Neuroscience 151:186–194. Cryan JF, Valentino RJ, Lucki I (2005) Assessing substrates underlying the behavioural effects of antidepressants using the modified rat forced swimming test. Neurosci Biobehav Rev 29:547–569. Dalla C, Antoniou K, Papadopoulou-Daifoti Z, Balthazart J, Bakker J (2004) Oestrogen-deficient female aromatase knockout (ArKO) mice exhibit depressive-like symptomatology. Eur J Neurosci 20:217–228. Dalla C, Papachristos EB, Whetstone AS, Shors TJ (2009) Female rats learn trace memories better than male rats and consequently retain a greater proportion of new neurons in their hippocampi. Proc Natl Acad Sci U S A 106:2927–2932. Dalla C, Kokras N, Pastromas N, Kafetzopoulos V, Tzouveka E, Balthazart J, Cornil CA, Papadopoulou-Daifoti Z (2012) Sex differences in the forced swim test of antidepressant activity: effects of gonadal- and brain- derived estrogens. Program No. 12.04.2012 Neuroscience Meeting Planner. New Orleans, LA: Society for Neuroscience, 2012. Online. Detke MJ, Rickels M, Lucki I (1995) Active behaviours in the rat forced swimming test differentially produced by serotonergic and noradrenergic antidepressants. Psychopharmacology (Berl) 121:66–72. Dowsett M, Haynes B (2003) Hormonal effects of aromatase inhibitors: focus on premenopausal effects and interaction with tamoxifen. J Steroid Biochem Mol Biol 86:255–263. Drossopoulou G, Antoniou K, Kitraki E, Papathanasiou G, Papalexi E, Dalla C, Papadopoulou-Daifoti Z (2004) Sex differences in behavioural, neurochemical and neuroendocrine effects induced by the forced swim test in rats. Neuroscience 126:849–857.
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
Statement of Interest
Estrada-Camarena E, Fernandez-Guasti A, Lopez-Rubalcava C (2003) Antidepressant-like effect of different estrogenic compounds in the forced swimming test. Neuropsychopharmacol 28:830–838. Estrada-Camarena E, Lopez-Rubalcava C, Vega-Rivera N, Recamier-Carballo S, Fernandez-Guasti A (2010) Antidepressant effects of estrogens: a basic approximation. Behav Pharmacol 21:451–464. Frye CA, Walf AA (2009) Depression-like behaviour of aged male and female mice is ameliorated with administration of testosterone or its metabolites. Physiol Behav 97:266–269. Goodwin GM (2006) Aromatase inhibitors and bipolar mood disorder: a case report. Bipolar disorders 8:516–518. Kafali H, Iriadam M, Ozardali I, Demir N (2004) Letrozole-induced polycystic ovaries in the rat: a new model for cystic ovarian disease. Arch Med Res 35:103–108. Keating C, Tilbrook A, Kulkarni J (2011) Oestrogen: an overlooked mediator in the neuropsychopharmacology of treatment response? Int J Neuropsychopharmacol 14:553–566. Kokras N, Antoniou K, Dalla C, Bekris S, Xagoraris M, Ovestreet DH, Papadopoulou-Daifoti Z (2009) Sex-related differential response to clomipramine treatment in a rat model of depression. J Psychopharmacol 23:945–956. Kokras N, Dalla C, Sideris AC, Dendi A, Mikail HG, Antoniou K, Papadopoulou-Daifoti Z (2012) Behavioural sexual dimorphism in models of anxiety and depression due to changes in HPA axis activity. Neuropharmacol 62:436–445. Lagunas N, Calmarza-Font I, Diz-Chaves Y, Garcia-Segura LM (2010) Long-term ovariectomy enhances anxiety and depressive-like behaviours in mice submitted to chronic unpredictable stress. Horm Behav 58:786–791. Li Y, Pehrson AL, Budac DP, Sanchez C, Gulinello M (2012) A rodent model of premenstrual dysphoria: progesterone withdrawal induces depression-like behaviour that is differentially sensitive to classes of antidepressants. Behav Brain Res 234:238–247. Meng F-T, Ni R-J, Zhang Z, Zhao J, Liu Y-J, Zhou J-N (2011) Inhibition of oestrogen biosynthesis induces mild anxiety in C57BL/6J ovariectomized female mice. Neuroscience Bulletin 27:241–250. Mikail HG, Dalla C, Kokras N, Kafetzopoulos V, Papadopoulou-Daifoti Z (2012) Sertraline behavioural response associates closer and dose-dependently with cortical rather than hippocampal serotonergic activity in the rat forced swim stress. Physiol Behav 107:201–206. Mitic M, Simic I, Djordjevic J, Radojcic MB, Adzic M (2013) Gender-specific effects of fluoxetine on hippocampal glucocorticoid receptor phosphorylation and behaviour in chronically stressed rats. Neuropharmacol 70:100–111. Phillips KA, Ribi K, Fisher R (2011) Do aromatase inhibitors have adverse effects on cognitive function? Breast Cancer Res 13:203. Porsolt RD (1979) Animal model of depression. Biomedicine 30:139–140. Rocha-Cadman X, Massie MJ, Du Hamel K (2012) Aromatase inhibitors and mood disturbances. Palliat Support Care 10:225. Vierk R, Glassmeier G, Zhou L, Brandt N, Fester L, Dudzinski D, Wilkars W, Bender RA, Lewerenz M, Gloger S (2012) Aromatase inhibition abolishes LTP generation in female but not in male mice. J Neurosci 32:8116–8126.
Letrozole in FST Wieland S, Fischette CT, Lucki I (1993) Effect of chronic treatments with tandospirone and imipramine on serotonin-mediated behavioral responses and monoamine receptors. Neuropharmacol 32:561–573. Winer EP, Hudis C, Burstein HJ, Chlebowski RT, Ingle JN, Edge SB, Mamounas EP, Gralow J, Goldstein LJ, Pritchard KI, Braun S, Cobleigh MA, Langer AS, Perotti J, Powles TJ, Whelan TJ, Browman GP (2002) American society of clinical oncology technology assessment
1313
on the use of aromatase inhibitors as adjuvant therapy for women with hormone receptor–positive breast cancer: status report 2002. J Clin Oncol 20:3317–3327. Zhou L, Fester L, von Blittersdorff B, Hassu B, Nogens H, Prange-Kiel J, Jarry H, Wegscheider K, Rune GM (2010) Aromatase inhibitors induce spine synapse loss in the hippocampus of ovariectomized mice. Endocrinol 151:1153–1160.
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
© CINP 2014
B R I E F R E PO R T
Acute but not sustained aromatase inhibition displays antidepressant properties Nikolaos Kokras1,2, Nikolaos Pastromas1, Tatiany H. Porto1,3, Vasilios Kafetzopoulos1, Theodore Mavridis1 and Christina Dalla1 1
Department of Pharmacology, Medical School, University of Athens, Greece First Department of Psychiatry, Eginition Hospital, Medical School, University of Athens, Greece 3 Psychology Institute, University of São Paulo, Brazil 2
Abstract
Received 22 July 2013; Reviewed 3 September 2013; Revised 3 February 2014; Accepted 5 February 2014; First published online 27 March 2014 Key words: Oestrogens, females, letrozole, progesterone, testosterone.
Introduction Aromatase inhibitors block the conversion of androgens to oestrogens and are used in post-menopausal women for the treatment of hormone-responsive breast cancer. An expanded use of them is under consideration, for treating hormone-responsive tumors in premenopausal women (Winer et al., 2002; Dowsett and Haynes, 2003). However, in post-menopausal women, aromatase inhibitors associate with psychiatric and cognitive side-effects (Phillips et al., 2011), while their safety in premenopausal women needs clarification. Experimental evidence suggests that deprivation of ovarian hormones is a risk factor for developing depression, especially after exposure to stressful experiences (Lagunas et al., 2010). Furthermore, response to antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), appears modulated by the female hormonal milieu (Estrada-Camarena et al., 2010; Keating et al., 2011). Therefore, we asked whether decreased oestrogens following aromatase inhibition would lead to a depressivelike behavioural response in cycling female rats, which
Address for correspondence: Assistant Professor C. Dalla, Department of Pharmacology, Medical School, University of Athens, Mikras Asias 75, Goudi, 11527, Greece. Tel.: +30 2107462577 Fax: +30 2107462554 Email: [email protected]
simulate premenopausal women. Furthermore, we investigated if gonadal hormonal changes due to aromatase inhibition would interfere with the SSRI antidepressant response, and we asked whether short-term vs. sustained aromatase inhibition would result in a similar behavioural response in the forced swim test (FST).
Method We used female Sprague–Dawley rats, group-housed under controlled 12:12 light/dark cycles (lights on at 07:00 hours) and temperature (22 ± 2 °C), with free access to food and water, aged 12 wk at the beginning of the experiment (Experiment 1 n = 40, Experiment 2 n = 46). Efforts were made to minimize the number of animals used and their suffering, in accordance with institutional regulations and the EU directive 2010/63. In experiment 1, 40 females received saline (Vehicle A) or 5 mg/kg fluoxetine (Eli Lilly Hellas S.A., Greece) for 28 d (n = 20 per group). At the 27th day of fluoxetine treatment all rats were subjected to the modified FST. They were individually placed in a cylindrical tank measuring 50 cm height × 20 cm width, filled with water (24 ± 1 °C) to a depth of 40 cm (Kokras et al., 2012). On the first day, rats were forced to swim for 15 min. After carefully drying, and before returning to their home cages, rats were injected i.p. with 1 mg/kg letrozole (Novartis Pharma
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
Aromatase inhibitors block the conversion of androgens to oestrogens and are used for the treatment of hormone-responsive breast cancer in menopause and recently also in premenopausal women. We investigate whether decreased oestrogen synthesis following aromatase inhibition leads to a depressive-like behavioural response in cycling female rats. Using the forced swim test (FST) we estimate the response of acute (three injections in 24 h) and sustained (7 d) letrozole and fluoxetine administration. Acute aromatase inhibition decreases immobility duration in the FST, indicating its antidepressant potential. Instead, sustained aromatase inhibition did not show such antidepressant potential. Testosterone elevation associates with the decreased depressive behaviour in the FST following acute letrozole treatment, but interestingly progesterone explains the increased swimming behaviour. Present findings may have potential implications for women treated with aromatase inhibitors, especially before menopause, as well as for the role of gonadal hormones in the expression of depressive symptoms and antidepressant response.
1308 N. Kokras et al.
Results Experiment 1: acute aromatase inhibition Behavioural measurements A two-way ANOVA for immobility duration showed a significant letrozole × fluoxetine interaction (F1,36 = 4.350 p = 0.044). Subsequent post-hoc testing showed that in rats pretreated with saline and those pretreated with fluoxetine letrozole administration resulted in reduced immobility (p < 0.001; p = 0.035, respectively) (Fig. 1a).
(a)
Immobility duration #
100
Vehicle B Letrozole
* Duration (s)
80 60 40 20 0
Vehicle A pretreatment
(b)
Fluoxetine pretreatment
Swimming duration 250
# Vehicle B Letrozole
*
Duration (s)
200 150 100 50 0
(c)
Head twitching frequency 40
Frequency (counts)
Vehicle A pretreatment Fluoxetine pretreatment
*
*
Vehicle B Letrozole
30
20
10
0
Vehicle A pretreatment Fluoxetine pretreatment
Fig. 1. Duration of immobility (a), swimming (b) and frequency of head twitches (c) during the second 5 min forced swim test following pretreatment for 28 d of saline vehicle or fluoxetine (5 mg/kg) and three injections of either vehicle B or letrozole. Note the reduction of immobility and increase in swimming duration after three injections of letrozole, which is similar to the antidepressant behavioural response after 28 d of fluoxetine treatment. An asterisk (*) denotes a significant letrozole effect and a hash sign (#) a fluoxetine effect, as indicated by the appropriate post-hoc test. n = 10/group, means ± S.E.M.
Acute letrozole administration did not alter the fluoxetine-induced reduction in immobility levels. Similarly, with regards to swimming duration, there was a significant letrozole × fluoxetine interaction (F1,36 = 5.659 p = 0.23). Post-hoc testing indicated that acute letrozole significantly increased swimming in vehicle-pretreated rats (p = 0.002) (Fig. 1b). Chronic treatment with fluoxetine also enhanced swimming (p = 0.045). There were no significant differences in climbing (data not shown). However, there was a weak letrozole main effect in increasing head twitching frequency (F1,36 = 4.069 p = 0.051) (Fig. 1c).
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
AG, Switzerland) or with vehicle B (saline with 5% Tween 80 and 5% ethanol, n = 10 per each group of vehicle or fluoxetine pre-treated rats). A second and a third letrozole or vehicle B injection were administered 19 and 23 h after the first injection, respectively. One hour after the third injection and 24 h after the first FST session, rats were subjected to a second 5 min FST session, during which the total duration of immobility, swimming, climbing and frequency of head twitching were manually measured (Mikail et al., 2012). Immediately after the second FST, rats were sacrificed and trunk blood was used for serum extraction. Estradiol, progesterone and testosterone assays were performed using commercially available RIA kits (Siemens Coat-A-Count for total testosterone and progesterone and double antibody estradiol). The detection limits were 0.4 ng/ml, 0.02 ng/ml and 1.4 pg/ml, respectively. In experiment 2, 46 females received vehicle A (saline, n = 24) or fluoxetine 5 mg/kg (n = 22) for 28 d, as in experiment 1. During the last week of treatment, rats received, along with vehicle A or fluoxetine, seven daily injections of vehicle B (saline with 5% Tween 80 and 5% ethanol; n = 12 for vehicle A pre-treated rats and n = 10 for fluoxetine pre-treated rats) or letrozole (n = 12 for vehicle A for pre-treated rats, n = 12 for fluoxetine). At the 27th day of fluoxetine and 6th day of letrozole treatment, rats were subjected to the first FST session and the following day (28th day of fluoxetine and 7th of letrozole) underwent the second FST test session. The FST and hormones assay procedure was identical to experiment 1, with the exception of treatments as described. During both experiments, vaginal smears were monitored as before (Dalla et al., 2009). Results were analysed with SPSS v.21 (IBM SPSS Inc, USA) using a two-way analysis of variance (ANOVA) with independent factors fluoxetine treatment (vehicle A or 5 mg/kg) and letrozole treatment (Vehicle B or 1 mg/kg). Significant interactions and main effects were explored with post-hoc tests using Bonferonni’s type I error correction. For each experiment associations between behaviour and hormones were explored initially using univariate linear regression. This was followed by a complete multivariate model, which included both treatments and all hormones. Statistical significance was set at p < 0.05. Results are reported as means ± S.E.M.
Letrozole in FST (a)
Testosterone
(d) *
250
1309
Progesterone 60
#
Serum levels (ng/ml)
Serum levels (ng/ml)
200 150
* 100 50 0
20
0 Vehicle A pretreatment
Fluoxetine pretreatment
Vehicle B
Letrozole
Vehicle A pretreatment Vehicle B
(e)
Testosterone–immobility regression
Letrozole
Progesterone–immobility regression 150
Immobility duration
150
100
50
100
50
0
0 0
100
200
300
400
0
500
20
Testosterone (ng/ml)
(c)
40
60
80
100
Progesterone (ng/ml)
(f )
Testosterone–swimming regression
Progesterone–swimming regression 300
Swimming duration
300
Swimming duration
Fluoxetine pretreatment
200
100
200
100
0
0 0
100
200
300
400
500
Testosterone (ng/ml)
0
20
40
60
80
100
Progesterone (ng/ml)
Fig. 2. Testosterone serum levels (a) and univariate linear regression of testosterone serum levels with immobility (b) and swimming (c) duration. Progesterone (d) and univariate linear regression of progesterone serum levels with immobility (e) and swimming (f) following 28 d of saline or fluoxetine (5 mg/kg) pretreatment and three injections of either vehicle B or letrozole. Note that testosterone was significantly associated with immobility duration but not with swimming duration. Inversely, progesterone was not associated with immobility duration in the univariate regression model, but with swimming duration. For multivariate regression models refer to the text. An asterisk (*) denotes a significant letrozole effect and a hash sign (#) a fluoxetine effect, as indicated by the appropriate post-hoc test. N = 10/group, means ± S.E.M.
Hormonal assays As expected, contrary to females receiving vehicle, female rats that received three injections of letrozole in 24 h had estradiol serum levels near or below the detection limit (data not shown), although vaginal smears were normal
in all females. As a result of the aromatase inhibition, testosterone levels were significantly enhanced (main effect: F1,35 = 11.939 p = 0.01) in all letrozole-treated rats previously treated with saline (vehicle A) or fluoxetine (p = 0.033; p = 0.001, respectively) (Fig. 2a). Interestingly, a marginally non-significant fluoxetine main effect
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
(b)
Immobility duration
40
1310 N. Kokras et al. was also observed in testosterone levels (p = 0.078). Specifically, fluoxetine increased testosterone in vehicle B and letrozole-treated rats, as indicated by post-hoc testing (p = 0.033; p = 0.001, respectively). With regards to progesterone, a significant letrozole × fluoxetine interaction was revealed (F1,35 = 5.126 p = 0.030). Further post-hoc testing showed that in saline-pretreated rats acute letrozole treatment had no effect in progesterone levels. In fluoxetine-pretreated rats letrozole only marginally reduced progesterone (p = 0.076). Instead, in salinepretreated rats, fluoxetine increased progesterone (p = 0.030), whereas this was not observed in rats receiving letrozole (Fig. 2d).
Univariate linear regressions showed that testosterone levels correlated with immobility (F1,38 = 4.250 p = 0.046) (Fig. 2b), whereas progesterone correlated with swimming (F1,38 = 7.457 p = 0.010) (Fig. 2f). A multivariate regression model including as predictors all treatments (fluoxetine, letrozole) and hormones (estradiol, testosterone, progesterone) significantly explained 48% of variance in immobility (F5,34 = 6.379 p < 0.001) and 31% of variance in swimming (F5,34 = 3.042 p = 0.022). Letrozole and estradiol were the strongest predictors in the regression model for immobility (p = 0.003 and p = 0.035, respectively), whereas progesterone weakly contributed to the model (p = 0.098). Instead, with regards to swimming, the strongest regression predictors were letrozole and progesterone (p = 0.016 and p = 0.014, respectively). The same multivariate regression model did not explain a significant percentage of variance of climbing and head twitching. Interestingly, head twitching showed marginal non-significant associations in the multivariate model with fluoxetine treatment and estradiol (p = 0.081 and p = 0.070, respectively).
Experiment 2: sustained aromatase inhibition Behavioural measurements A two-way ANOVA for immobility, swimming and climbing behaviour did not show any significant fluoxetine × letrozole interactions. The analysis showed a marginally non-significant fluoxetine main effect (F1,42 = 3.658 p = 0.063), given that fluoxetine tended to shorten the immobility duration. As expected, fluoxetine significantly increased swimming duration (F1,42 = 6.634 p = 0.014), but post-hoc pairwise comparisons with Bonferonni correction showed that this effect reached significance only in vehicle B treated rats (p = 0.015) and not in letrozole-treated rats. Sustained aromatase inhibition had no effects on behaviour except a significant main effect in head twitching behaviour (F1,42 = 7.038 p = 0.011). Post-hoc testing showed that only vehicle A pretreated and not fluoxetine-pretreated rats had more head swings
(p = 0.023)
Hormonal assays All rats receiving repeated letrozole injections for 1 wk had disrupted oestrous cycle and their oestrogen levels were near or below detection limit, which was not the case for rats receiving vehicle B (data not shown). Testosterone levels were markedly increased following repeated letrozole treatment (F1,42 = 58.363 p < 0.001). Post-hoc testing showed that letrozole alone and in combination with fluoxetine pretreatment significantly increased testosterone (p < 0.001 in both treatments). Interestingly, fluoxetine also had a significant effect on testosterone levels (F1,42 = 6.283 p = 0.016) and post-hoc pairwise comparisons showed that fluoxetine either alone or in combination with letrozole tended to increase progesterone. With regards to progesterone, sustained aromatase inhibition caused a significant decrease (F1,42 = 8.092 p = 0.007). Post-hoc testing showed that letrozole, either alone or in combination with fluoxetine, decreased progesterone (p = 0.043 and p = 0.058, respectively) (Supplementary materials Figure S2). Association of behavioural and hormonal indices None of the monitored behavioural indices correlated with hormones as indicated by correlation and linear regression models, with the exception of head twitching frequency, which was negatively correlated with progesterone serum levels (r=−0.298 p = 0.045 n = 46) and positively with testosterone levels (r = 0.339 p = 0.021 n = 46). Discussion Present findings show that acute aromatase inhibition induces an antidepressant behavioural response. Instead, sustained aromatase inhibition for 1 wk results in no effect. These findings may have potential implications for women treated with aromatase inhibitors, especially before menopause, as well as for the role of gonadal hormones in the expression of depressive symptoms and antidepressant response. Specifically, acute letrozole treatment affects swimming and not climbing in the FST, like fluoxetine and all SSRI (Detke et al., 1995). Given that swimming duration in the FST has been linked with serotonergic changes induced by SSRIs, it can be assumed that acute letrozole influences mainly the serotonergic system and its behavioural response. In line with the antidepressant properties of acute letrozole treatment, case reports evidence induction of mania and mood swings in early stages of treatment with aromatase inhibitors (Goodwin, 2006; Rocha-Cadman et al., 2012). On the other hand, letrozole treatment for 1 wk showed no antidepressant potential, as it had no significant effect in the FST.
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
Association of behavioural and hormonal indices
after repeated letrozole treatment (Supplementary materials Figure S1).
Letrozole in FST
be involved, as implied by the fact that the multifactorial regression in experiment 2 did not explain a significant part of the variance. Although our hypothesis provides a plausible explanation, more studies are warranted to further explore such interactions. Moreover, whereas the FST is widely used as a screening tool for monitoring antidepressant potential (Porsolt, 1979; Cryan et al., 2005), chronic stress models would probably shed more light. In our study, letrozole dosage can hardly explain the observed difference between acute and sustained aromatase inhibition because we used a high dose that totally blocks the aromatase enzyme (Kafali et al., 2004; Aydin et al., 2008). This is further confirmed by an ongoing study (preliminary results communicated at SfN (Dalla et al., 2012)), where aromatase activity was found completely blocked with this dose. Regarding fluoxetine, previous studies in female rats have shown that chronic administration of a suboptimal 5 mg/kg fluoxetine dose may or may not result in an antidepressant-like behavioural response or have a small effect size (Mitic et al., 2013). Beyond this known issue of suboptimal SSRI dosing in chronic treatment, in our chronic experiment letrozole does not show antidepressant potential and also does not augment the low fluoxetine dose in producing an antidepressant effect when letrozole and fluoxetine are combined together. Finally, acute letrozole treatment enhances head twitching frequency, while fluoxetine has no effect. Also, in the second experiment, head twitching frequency correlates with progesterone and testosterone. These findings further associate head twitching in the FST with gonadal hormones. Previous findings from our group showed that females twitch their heads less than males, and that head twitching is higher in oestrous/proestrous (Drossopoulou et al., 2004; Kokras et al., 2009; Kokras et al., 2012). Head twitching is linked to 5-HT2 receptor activity in the open field (Wieland et al., 1993) and perhaps this FST behaviour might be used as representative of the effect of gonadal hormones on the serotonergic system. Present findings indicate that in considering treatment with aromatase inhibitors in premenopausal women, attention should be considered regarding possible psychotropic effects of aromatase inhibitors. Seemingly, treatment duration with aromatase inhibitors plays an important role in aromatase-induced mood effects. The unexpected acute antidepressant potential of an aromatase inhibitor raises the intriguing question of whether such molecules could be investigated in combination with known antidepressants and in short-term dosing regimens in order to kick-start an antidepressant response, or as a transient short-lived augmentation strategy in treatment-resistant depression. Supplementary material For supplementary material accompanying this paper, visit http://dx.doi.org/10.1017/S1461145714000212
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
A lack of effect in the FST was also reported after 21 d of letrozole treatment in ovariectomized mice (Meng et al., 2011). Such differentiated effects of acute and sustained aromatase inhibition are surprising, because oestrogens have been associated with antidepressant properties, so a decrease in oestrogen levels by aromatase inhibition should hypothetically result in a ‘depressivelike’ phenotype (Estrada-Camarena et al., 2003, 2010). Although treatment duration in rats does not equally translate to humans, our data support a time-dependent biphasic effect following aromatase inhibition. In support of this concept, previous studies on female mice knocked out since birth for CYP19A1 (aromatase gene) displayed in adulthood depressive symptomatology in the FST (Dalla et al., 2004). In addition, aromatase inhibition had detrimental effects on long-term potentiation and spine density in the hippocampus, and those effects were dependent on the duration of letrozole treatment (Zhou et al., 2010; Vierk et al., 2012). Thus, we support that aromatase inhibition initially results in mood elation, subsequently this phenomenon vanishes and, in line with the aforementioned studies at a later stage (beyond the 1–3 wk time interval), a pro-depressive effect might appear. Although this hypothesis needs further investigation, based on data reported herein there is now strong evidence that duration of treatment plays a crucial role in letrozole’s effects in brain and behaviour. A possible explanation of the remarkable antidepressant effect observed in this study might lay in hormonal adjustments following aromatase inhibition. Exogenous testosterone administration is known to exert an antidepressant effect in the FST (Frye and Walf, 2009), and letrozole substantially increases testosterone, as seen in both experiments. However, sustained aromatase inhibition causes a more pronounced testosterone increase in comparison to subacute inhibition, but no antidepressant effect was observed. In addition, although testosterone correlates with immobility in the acute experiment, the multivariate regression does not confirm a key role for testosterone. In fact, the antidepressant behavioural response is explained by letrozole and, interestingly, by oestrogen depletion with regards to immobility, and by progesterone increase with regards to swimming. Thus, we speculate that a rapid oestrogen decline, when combined with an equally rapid testosterone and progesterone increase, may result in antidepressant response. When this hormonal adjustment is sustained, as in experiment 2, other compensatory mechanisms kick in and, for example, progesterone stabilizes to lower levels. As progesterone withdrawal results to a ‘depressive-like’ symptomatology in the FST (Li et al., 2012), it can be suggested that, following sustained letrozole treatment, the prolonged progesterone decrease counteracts for the oestrogen decline and the testosterone enhancement, and thus the antidepressant effect of letrozole is no longer evident. However, our study cannot prove causal relationships between hormones and behaviour, and other factors might
1311
1312 N. Kokras et al. Acknowledgments The authors acknowledge Novartis Pharma AG for providing letrozole, Eli Lilly Hellas S.A. for fluoxetine, Mrs Despina Papassava and Dr Mikael Hudu Garba for excellent technical assistance. CD and NK are supported by a ‘Large Scale Cooperative Project’ (09SYN-21–1003) and VK by ‘Education and Lifelong Learning, Supporting Postdoctoral Researchers’. Both projects are co-financed by the European Social Fund (ESF) and the General Secretariat for Research and Technology, Greece. THP had a CAPES Foundation fellowship from the Brazilian Ministry of Education. None of these entities had any influence over this study.
NK has received speaker’s fees, consultancy honoraria and travel support from Janssen, Lundbeck, SanofiAventis, Medochemie Generics and Elpen S.A. None of those are relevant to this study.
References Aydin M, Yilmaz B, Alcin E, Nedzvetsky VS, Sahin Z, Tuzcu M (2008) Effects of letrozole on hippocampal and cortical catecholaminergic neurotransmitter levels, neural cell adhesion molecule expression and spatial learning and memory in female rats. Neuroscience 151:186–194. Cryan JF, Valentino RJ, Lucki I (2005) Assessing substrates underlying the behavioural effects of antidepressants using the modified rat forced swimming test. Neurosci Biobehav Rev 29:547–569. Dalla C, Antoniou K, Papadopoulou-Daifoti Z, Balthazart J, Bakker J (2004) Oestrogen-deficient female aromatase knockout (ArKO) mice exhibit depressive-like symptomatology. Eur J Neurosci 20:217–228. Dalla C, Papachristos EB, Whetstone AS, Shors TJ (2009) Female rats learn trace memories better than male rats and consequently retain a greater proportion of new neurons in their hippocampi. Proc Natl Acad Sci U S A 106:2927–2932. Dalla C, Kokras N, Pastromas N, Kafetzopoulos V, Tzouveka E, Balthazart J, Cornil CA, Papadopoulou-Daifoti Z (2012) Sex differences in the forced swim test of antidepressant activity: effects of gonadal- and brain- derived estrogens. Program No. 12.04.2012 Neuroscience Meeting Planner. New Orleans, LA: Society for Neuroscience, 2012. Online. Detke MJ, Rickels M, Lucki I (1995) Active behaviours in the rat forced swimming test differentially produced by serotonergic and noradrenergic antidepressants. Psychopharmacology (Berl) 121:66–72. Dowsett M, Haynes B (2003) Hormonal effects of aromatase inhibitors: focus on premenopausal effects and interaction with tamoxifen. J Steroid Biochem Mol Biol 86:255–263. Drossopoulou G, Antoniou K, Kitraki E, Papathanasiou G, Papalexi E, Dalla C, Papadopoulou-Daifoti Z (2004) Sex differences in behavioural, neurochemical and neuroendocrine effects induced by the forced swim test in rats. Neuroscience 126:849–857.
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
Statement of Interest
Estrada-Camarena E, Fernandez-Guasti A, Lopez-Rubalcava C (2003) Antidepressant-like effect of different estrogenic compounds in the forced swimming test. Neuropsychopharmacol 28:830–838. Estrada-Camarena E, Lopez-Rubalcava C, Vega-Rivera N, Recamier-Carballo S, Fernandez-Guasti A (2010) Antidepressant effects of estrogens: a basic approximation. Behav Pharmacol 21:451–464. Frye CA, Walf AA (2009) Depression-like behaviour of aged male and female mice is ameliorated with administration of testosterone or its metabolites. Physiol Behav 97:266–269. Goodwin GM (2006) Aromatase inhibitors and bipolar mood disorder: a case report. Bipolar disorders 8:516–518. Kafali H, Iriadam M, Ozardali I, Demir N (2004) Letrozole-induced polycystic ovaries in the rat: a new model for cystic ovarian disease. Arch Med Res 35:103–108. Keating C, Tilbrook A, Kulkarni J (2011) Oestrogen: an overlooked mediator in the neuropsychopharmacology of treatment response? Int J Neuropsychopharmacol 14:553–566. Kokras N, Antoniou K, Dalla C, Bekris S, Xagoraris M, Ovestreet DH, Papadopoulou-Daifoti Z (2009) Sex-related differential response to clomipramine treatment in a rat model of depression. J Psychopharmacol 23:945–956. Kokras N, Dalla C, Sideris AC, Dendi A, Mikail HG, Antoniou K, Papadopoulou-Daifoti Z (2012) Behavioural sexual dimorphism in models of anxiety and depression due to changes in HPA axis activity. Neuropharmacol 62:436–445. Lagunas N, Calmarza-Font I, Diz-Chaves Y, Garcia-Segura LM (2010) Long-term ovariectomy enhances anxiety and depressive-like behaviours in mice submitted to chronic unpredictable stress. Horm Behav 58:786–791. Li Y, Pehrson AL, Budac DP, Sanchez C, Gulinello M (2012) A rodent model of premenstrual dysphoria: progesterone withdrawal induces depression-like behaviour that is differentially sensitive to classes of antidepressants. Behav Brain Res 234:238–247. Meng F-T, Ni R-J, Zhang Z, Zhao J, Liu Y-J, Zhou J-N (2011) Inhibition of oestrogen biosynthesis induces mild anxiety in C57BL/6J ovariectomized female mice. Neuroscience Bulletin 27:241–250. Mikail HG, Dalla C, Kokras N, Kafetzopoulos V, Papadopoulou-Daifoti Z (2012) Sertraline behavioural response associates closer and dose-dependently with cortical rather than hippocampal serotonergic activity in the rat forced swim stress. Physiol Behav 107:201–206. Mitic M, Simic I, Djordjevic J, Radojcic MB, Adzic M (2013) Gender-specific effects of fluoxetine on hippocampal glucocorticoid receptor phosphorylation and behaviour in chronically stressed rats. Neuropharmacol 70:100–111. Phillips KA, Ribi K, Fisher R (2011) Do aromatase inhibitors have adverse effects on cognitive function? Breast Cancer Res 13:203. Porsolt RD (1979) Animal model of depression. Biomedicine 30:139–140. Rocha-Cadman X, Massie MJ, Du Hamel K (2012) Aromatase inhibitors and mood disturbances. Palliat Support Care 10:225. Vierk R, Glassmeier G, Zhou L, Brandt N, Fester L, Dudzinski D, Wilkars W, Bender RA, Lewerenz M, Gloger S (2012) Aromatase inhibition abolishes LTP generation in female but not in male mice. J Neurosci 32:8116–8126.
Letrozole in FST Wieland S, Fischette CT, Lucki I (1993) Effect of chronic treatments with tandospirone and imipramine on serotonin-mediated behavioral responses and monoamine receptors. Neuropharmacol 32:561–573. Winer EP, Hudis C, Burstein HJ, Chlebowski RT, Ingle JN, Edge SB, Mamounas EP, Gralow J, Goldstein LJ, Pritchard KI, Braun S, Cobleigh MA, Langer AS, Perotti J, Powles TJ, Whelan TJ, Browman GP (2002) American society of clinical oncology technology assessment
1313
on the use of aromatase inhibitors as adjuvant therapy for women with hormone receptor–positive breast cancer: status report 2002. J Clin Oncol 20:3317–3327. Zhou L, Fester L, von Blittersdorff B, Hassu B, Nogens H, Prange-Kiel J, Jarry H, Wegscheider K, Rune GM (2010) Aromatase inhibitors induce spine synapse loss in the hippocampus of ovariectomized mice. Endocrinol 151:1153–1160.
Downloaded from http://ijnp.oxfordjournals.org/ at Bibliotheque de l'Universite Laval on November 16, 2015
