Neuropharmacology 62 (2012) 209e216
Contents lists available at SciVerse ScienceDirect
Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm
N-desalkylquetiapine activates ERK1/2 to induce GDNF release in C6 glioma cells: A putative cellular mechanism for quetiapine as antidepressant Barbara Di Benedetto a, *, Ralf Kühn c, d, Caroline Nothdurfter a, b, Theo Rein a, Wolfgang Wurst c, d, e, Rainer Rupprecht a, b a
Max Planck Institute of Psychiatry, Munich, Germany Department of Psychiatry and Psychotherapy, University Regensburg, Germany Institute for Developmental Genetics, Helmholtz Center Munich, Munich, Germany d Technical University Munich, Munich, Germany e Deutsches Zentrum für Neurodegenerative Erkrankungen e.V., Munich, Germany b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 29 March 2011 Received in revised form 29 June 2011 Accepted 1 July 2011
Quetiapine is an atypical antipsychotic which has been suggested to possess also antidepressant efficacy in the treatment of bipolar and unipolar depression. Recently, a link between the activation of the ERK/ MAPK signalling pathway and the release of GDNF has been proposed as a specific feature of antidepressants. To obtain a first insight into the putative molecular mechanism of action of quetiapine, we examined its impact and that of its major metabolite norquetiapine on the activation of the ERK/MAPK signalling pathway in C6 glioma cells. Additionally, we investigated the induction of GDNF release as a possible physiological consequence of this activation. We found that norquetiapine, similarly to the antidepressant reboxetine, activated both ERK1 and ERK2 (pERK) with consequent enhanced release of GDNF; this release was dependent on pERK, as demonstrated by its reversibility after pre-treatment with a pharmacological pERK inhibitor. In contrast, quetiapine induced activation of ERK2 only. It also caused release of GDNF, but this release was independent of ERK activation. To test whether the simultaneous activation of ERK1 with ERK2 was critical for the observed pERK-dependent GDNF release, we specifically inactivated ERK1 mRNA via RNA interference. Our data show that indeed ERK1 plays an essential role, as GDNF release was hampered after Erk1 downregulation comparably to a pharmacological pERK inhibitor. Thus, activation of only ERK2 appears not to be sufficient for promoting GDNF release. Our results reveal the release of GDNF as a consequence of ERK/MAPK signalling activation by norquetiapine, which may contribute to the putative antidepressant properties of quetiapine. This article is part of a Special Issue entitled ‘Anxiety and Depression’. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Quetiapine Norquetiapine Antidepressant ERK/MAP kinase GDNF C6 cells
1. Introduction Quetiapine is an atypical antipsychotic which has recently been suggested to exert also antidepressant activity. Therefore, its efficacy as monotherapy for the acute treatment of depressive episodes associated with bipolar disorder (BD) has been studied in several clinical trials (Gajwani et al., 2007) and it is approved by the US
Abbreviations: ERK/MAPK, extracellular signal-regulated kinase/mitogen-activated protein kinase; GDNF, glia cell line-derived neurotrophic factor; QTP, quetiapine; NORQTP, norquetiapine; RBX, reboxetine. * Corresponding author. Max Planck Institute of Psychiatry, Molecular Psychopharmacology, Kraepelinstrasse 2, D-80804 Munich, Germany. Tel.: þ49 89 30622506; fax: þ49 89 30622402. E-mail address:
[email protected] (B. Di Benedetto). 0028-3908/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2011.07.001
Food and Drug Administration (FDA) for the respective indication (Suppes et al., 2010). Moreover, its efficacy for the treatment also of unipolar depression has recently been investigated in several clinical trials (Bortnick et al., 2010; Cutler et al., 2009; El-Khalili et al., 2010; Weisler et al., 2009). Although the clinical efficacy of quetiapine as a monotherapy for unipolar depression is still a matter of discussion, in a rat model of chronic stress quetiapine exerts antidepressant-like effects (Orsetti et al., 2007). Nevertheless, the precise molecular mechanisms of action of quetiapine remain to be elucidated. Recently, the role of N-desalkylquetiapine (norquetiapine), a major metabolite of quetiapine, as a potent norepinephrine reuptake inhibitor (NRI) was discovered and its antidepressant-like activity has been shown in mice by the tail suspension test (Jensen et al., 2008). As certain antidepressants inhibit norepinephrine
210
B. Di Benedetto et al. / Neuropharmacology 62 (2012) 209e216
reuptake (e.g. venlafaxine, duloxetine and reboxetine), it has been suggested that the antidepressant activity of quetiapine could be mediated by norquetiapine (Jensen et al., 2008). The Extracellular signal-Regulated Kinase/Mitogen-Activated Protein Kinase (ERK/ MAPK) signalling pathway has been shown to be involved in various cellular functions (Kyriakis and Avruch, 1996; Pages et al., 1993; Weber et al., 1997). Moreover, recent findings emphasized its role in the pathophysiology of major depressive disorder (Duric et al., 2010). In particular, an altered MAPK activity has been observed in the prefrontal cortex of postmortem brain tissue from depressive patients (Yuan et al., 2010). Furthermore, it has been shown that the antidepressant fluoxetine can reverse the reduced MAPK activity in an animal model of depression (Qi et al., 2008). With regard to cellular specificity involved in depression, not only neuronal, but also non-neuronal cells such as astrocytes might play a major role in the pathophysiology of this disease (reviewed in Rajkowska and Miguel-Hidalgo, 2007; Banasr and Duman, 2008). Recently, several antidepressants, but not other psychotropic drugs such as diazepam or lithium, have been shown to activate the ERK/MAPK signalling pathway in rat C6 glioma cells, a cell line commonly used as a model for astrocytes, which further underscores the relevance of this cell type for the clinical response to antidepressant treatment. The ERK/MAPK activation correlated with the release of glial cell-derived neurotrophic factor (GDNF), an established protective factor for dopaminergic neurons (Hisaoka et al., 2007; Lin et al., 1993), thereby suggesting a functional link between the activation of this specific signalling pathway and the respective physiological cellular response after antidepressant treatment. Moreover, also several atypical antipsychotics such as quetiapine, but also clozapine and risperidone, exerted protective effects on rat differentiated PC12 cells (Bai et al., 2002). Furthermore, both quetiapine and clozapine, but also the typical antipsychotic haloperidol, can induce GDNF release in C6 glioma cells (Shao et al., 2006). Therefore, we studied whether a putative molecular mechanism shared specifically by quetiapine, its metabolite norquetiapine and antidepressants involves activation of the ERK/MAPK signalling pathway with subsequent release of GDNF. We could indeed show that norquetiapine, similarly to the NRI reboxetine, activated both ERK isoforms via phosphorylation (pERK) with subsequent induction of GDNF release in a MAPK dependent fashion. In contrast, quetiapine activated only ERK2; it also induced GDNF release similarly to other antipsychotics, but in a pERK-independent manner. Using short hairpins designed to target specifically ERK1 mRNA via RNA interference, we demonstrated that the activation of both ERK isoforms induced by norquetiapine and reboxetine is a necessary condition to trigger GDNF release. Our results support the hypothesis that the intracellular mechanisms of action of quetiapine might be conferred by its metabolite norquetiapine, involving a direct link between ERK/MAPK pathway activation and induction of GDNF release.
to serum-free DMEM before drug treatment 24 h later. In all assays, drugs were delivered in serum-free DMEM at concentrations specified for each experiment. 2.3. Drug treatment and western blot procedure for ERK and pERK detection For the time course experiments, following FCS deprivation for 24 h, cells were treated with 25 mM of each drug of interest. At the required time points, medium was removed and plates were washed with cold PBS; cells were lysed thereafter (as in Rubovitch et al., 2004), collected and cleared by centrifugation (10 min (min), 12,000 g). Total protein content was quantified with Lowry standard assay (Bio-Rad, Hercules, CA, USA) and equivalent amounts of protein lysates (15 mg) were denatured in Laemmli buffer (95 C for 2 min) before separating them on 12% SDS-PAGE and transblotting them to nitrocellulose membranes (Whatman, Kent, UK). Membranes were blocked with 5% bovine serum albumin in Tris-buffered saline with Tween-20 (TBS-T, 10 mM Tris base (pH 8.0), 150 mM NaCl, and 0.1% Tween-20) overnight at 4 C. Thereafter, they were incubated with anti-phosphorylated ERK1/2 or with antiERK1/2 (Cell Signalling, Boston, MA, USA, dilution 1:1000) together with antiHPRT, as standard internal control to normalize samples for semi-quantitative analysis (Santa Cruz Biotechnology, Santa Cruz, CA, USA, dilution 1:400), for 2 h at room temperature (RT). Blots were washed in TBS-T and incubated with HRP-goat anti-rabbit (Jackson ImmunoResearch, West Grove, PA, USA, dilution 1:5000) for 1 h at RT. After washing, they were visualized with ECL (Amersham Biosciences, GE Healthcare, Freiburg, Germany) and films were scanned and processed by densitometric analysis using NIH ImageJ software (http://rsb.info.nih.gov/ij/). 2.4. GDNF release enzyme-linked immunosorbent assay (ELISA) and lactate dehydrogenase (LDH) release assay For both assays, C6 cells were cultured as described before. To analyze the dosedependent release of GDNF, cells were treated with 5, 10 or 25 mM of each drug. To examine the pMAPK dependence of GDNF release, cells were treated for 30 min with 20 mM U0126, a pharmacological inhibitor of MEK1/2 kinases, the upstream activators of ERK1/2, and thereafter with the drug of interest. After 48 h of treatment, conditioned media were collected and GDNF protein levels in media were determined using a GDNF ELISA (Promega) according to the manufacturer’s instructions. To determine the cytotoxicity of the used drugs, lactate dehydrogenase (LDH) levels were measured in the same media using a cytotoxicity colorimetric assay kit (Oxford Biomedical Research, Rochester Hills, MI, USA) according to the manufacturer’s instructions. 2.5. Cloning and transfection of shRNA vectors to perform RNA interference experiments A pBluescript plasmid containing the human U6 promoter was opened with BseRI/BamHI to allow cloning of the shRNAs oligonucleotide pairs. Afterwards, a XhoI/ClaI fragment was cut and transcloned into a vector containing a PGK-eGFP to measure transfection efficiencies. Oligonucleotide sequences were as follows (underlined is the sequence of the hairpin loop): Erk1.1: GCGATTCCGCCATGAGAATGTTATACAAGCTTCTATAACATTCTCATGGCGGAATCGC; Erk1.2-GCAATGACCACATCTGCTATTTCAAGCTTCAAATAGCAGATGTGGTCATTGC; LacZGGCGTTACCCAACTTAATCGCCTTGCAAGCTTCCAAGGCGATTAAGTTGGGTAACGCC.
The following drugs were used: quetiapine and norquetiapine (a generous gift from AstraZeneca, Södertälje, Sweden), clozapine and haloperidol (RBI, Köln, Germany), reboxetine (a generous gift from Pfizer, New York City, NY, US) and U0126 (Promega, Madison, WI, USA). For stock solutions, all drugs were dissolved in 100% DMSO, except reboxetine that was dissolved in water.
Before transfection, cells were seeded into twelve-well plates at a density of 105/well in 1 ml DMEM with FCS. The day after, they were transfected using the Lipofectamine 2000 Reagent (Invitrogen), following manufacturer’s instructions. They were grown for 48 h and then medium was changed to serum-free medium before drug treatment 24 h later (72 h after transfection). After 48 h of treatment, conditioned media were collected and GDNF protein levels were analyzed as described before. In parallel, cells were lysed and total proteins were collected as described above (in Section 2.3) to analyze the downregulation efficiency for each replicate in each experiment. Three independent experiments were performed, with three replicates for each treatment condition. Values were included in the statistical analysis only when it was verified by Western blot a reduction of basal ERK1 in respect to controls; and at the same time the respective pERK levels showed inhibition efficiencies comparable with the pharmacological inhibitor U0126 (as example, see representative Western blots in Figs. 4b and 5b).
2.2. Cell culture
2.6. Statistical analysis
For all experiments, rat C6 glioma cells were cultured in DMEM containing 4.5 g glucose/L, 10% foetal calf serum (FCS), 1% penicillin and streptomycin, 1 mM sodium pyruvate (all from GIBCO) and maintained in 5% CO2/95% air at 37 C. For each experiment, confluent cells were collected and seeded into twelve-well plates at a density of 105/well in 1 ml DMEM with FCS. After 24 h (hrs), medium was changed
All data are expressed as mean SD, and n refers to the number of replicates. Statistical differences were calculated with one-way ANOVA, followed by Dunnett’s (Figs. 1, 2 and 3 and 4c) or Tukey HSD (Figs. 4a and 5) post hoc tests using the SPSS 16 programme (SPSS Inc., Chicago, IL, USA). A value of p < 0.05 was considered statistically significant.
2. Material and methods 2.1. Chemicals and drugs
B. Di Benedetto et al. / Neuropharmacology 62 (2012) 209e216
211
Fig. 1. Time courses of activation of ERK/MAPK upon quetiapine or norquetiapine treatment. Both drugs activated ERK/MAPK but with differential specific profiles. Quetiapine can activate only ERK2, whereas norquetiapine affects both ERK isoforms (n ¼ 6e8; ANOVA, aap < 0.01, aaap < 0.001 for ERK2; *** p< 0.001 for ERK1 as compared to control, C). b) A representative Western blot of the different time points shows the differential activation of the two ERK isoforms (pERK1/2) together with the internal control HPRT that was used for normalization and semi-quantitative analysis.
3. Results 3.1. Time course of pERK activation upon drug treatment Because the antidepressant amitriptyline activates ERK/MAPK pathway with a peak within the first minutes after drug treatment that goes down to basal levels afterwards (Hisaoka et al., 2007), we examined the time course of ERK1/2 activation at 5, 10, 15, 30, 60 and 120 min as well as 24 and 48 h after treatment of cells with 25 mM quetiapine or norquetiapine, respectively (Fig. 1a). A oneway ANOVA revealed a significant effect of time on ERK activation for both quetiapine (F(1,8) ¼ 5.42; ***p < 0.001 for ERK1; F(1,8) ¼ 4.45; ***p < 0.001 for ERK2) and norquetiapine (F(1,8) ¼ 4.05; ***p < 0.001 for ERK1; F(1,8) ¼ 3.39; **p < 0.01 for ERK2). A Dunnett post hoc test revealed differential effects between both drugs: quetiapine caused only an ERK2 activation peak after 5 min (154.3 14.2%, aap < 0.01) and an inactivation of ERK1 (48.9 7.9%, ***p < 0.001) 30 min after treatment. On the contrary, norquetiapine induced both ERK1 and ERK2 activation 10 min after treatment (183.1 20.8%, ***p < 0.001 for ERK1 and 187.1 15.8%, aaap < 0.001 for ERK2). For both drugs, ERK activity reached basal levels again after 60 min and remained unchanged afterwards. 3.2. Effect of quetiapine and norquetiapine on GDNF release In view of the peak of ERK activity observed with both drugs, we studied whether this effect was related to a release of GDNF as previously shown for other antidepressants (Hisaoka et al., 2007). Both quetiapine and norquetiapine induced GDNF release in a dosedependent fashion (Fig. 2a and Suppl Fig. S1). A one-way ANOVA showed a dose-dependent significant effect on the amount of GDNF released for both drugs (F(1,3) ¼ 30.6; *** p< 0.001 for quetiapine;
Fig. 2. Effects of quetiapine and norquetiapine on the release of GDNF and LDH enzyme. a) Both quetiapine and norquetiapine induced a release of GDNF in a dosedependent fashion (n ¼ 6; ANOVA, * p< 0.05, ** p< 0.01, *** p< 0.001 as compared with control; specifically, for quetiapine: 5 mM, 146.8 15.4%; 10 mM, 155.7 13.6%; 25 mM, 243.1 46.5%; for norquetiapine: 5 mM, 308.9 54.5%; 10 mM, 388.8 50.8%; 25 mM, 608.0 129.7%); b) Only norquetiapine showed a toxic effect at a 25 mM concentration as measured with LDH release (n ¼ 6; 136.3 23.7%; ANOVA, ** p< 0.01).
F(1,3) ¼ 20.32; *** p< 0.001 for norquetiapine), which was detectable already at the lowest drug concentration (5 mM; 146.8 15.4%, * p< 0.05 for quetiapine; 308.9 54.5%, *** p< 0.001 for norquetiapine) (Dunnett’s post hoc test). In order to determine whether this release of GDNF was due to a leakage of damaged cells as a consequence of putative drug toxicity, we measured the content of LDH enzyme in the same conditioned media previously tested for GDNF content. Only norquetiapine showed a toxic effect at 25 mM dosage (ANOVA, F(1,3) ¼ 5.96, 136.3 23.7%, ** p< 0.01) (Fig. 2b). This was the concentration used in the first experiment to unravel
212
B. Di Benedetto et al. / Neuropharmacology 62 (2012) 209e216
Fig. 3. The impact of quetiapine and norquetiapine versus classical antipsychotics or antidepressants on ERK/MAPK activation profile. Only norquetiapine could activate both ERK isoforms after 10 min of pharmacological treatment, like the antidepressant reboxetine; on the contrary, quetiapine could activate only ERK2, similar to the classical antipsychotics; specifically, quetiapine was inducing ERK activation already by 5 min after treatment (Fig. 1a), whereas at 10 min after treatment there was only a tendency towards activation (25 mM for all drugs; n ¼ 6e8; ANOVA, * p< 0.05, ** p< 0.01, *** p< 0.001).
differences in ERK1 or ERK2 activation by Western blot. Unfortunately, due to the low sensitivity of this method, it was not possible to examine drug dose-dependent ERK activity alterations with Western blot to determine whether lower drug concentrations might also induce ERK activation. 3.3. Comparison of pMAPK activation by quetiapine, norquetiapine and other antipsychotics or antidepressants We next compared the effects of quetiapine and norquetiapine with those of other antipsychotics, clozapine and haloperidol, and of the NRI antidepressant reboxetine with regard to pMAPK activation profiles 10 min after treatment with 25 mM of each compound. As is evident from a comparison between Figs. 1a and 3, quetiapine showed an activation profile at 5 min after treatment that is similar to that of the antipsychotics clozapine and haloperidol at 10 min, whereas norquetiapine displayed an activation pattern similar to that of the antidepressant reboxetine. Moreover, a one-way ANOVA revealed a significant effect of drug treatment on both ERK isoforms (F(1,2) ¼ 13.5 for ERK1, *** p< 0.001 and F(1,2) ¼ 8.8 for ERK2, *** p< 0.001). In particular, only norquetiapine and reboxetine significantly activated both ERK1 and ERK2 (Dunnett post hoc test, *** p< 0.001 for both isoforms; specifically, for norquetiapine: 183.1 20.8% for ERK1 and 187.1 15.8% for ERK2; for reboxetine: 148.2 14.8% for ERK1 and 158.2 8.0% for ERK2), whereas clozapine and haloperidol predominantly activated only ERK2 at the same time point (141.8 9.4%, ** p< 0.01 for clozapine and 137.6 14.8%, * p< 0.05 for haloperidol). 3.4. Dependence of GDNF release on pMAPK activation To examine whether the previously observed ERK/MAPK activation was directly responsible for the GDNF release, we performed
two complementary experiments. In the first, we blocked ERK phosphorylation by a 30 min pre-treatment of cells with the pMAPK inhibitor U0126. This pre-treatment followed by 5 mM norquetiapine administration significantly reduced GDNF release compared to norquetiapine alone (Fig. 4a and Suppl Fig. S2), thereby strongly suggesting a causal link between ERK activation and GDNF release for this compound (ANOVA, F(1,2) ¼ 44.5, *** p< 0.001; 273.0 37.9% for NORQTP alone and 209.5 42.6% for NORQTP þ U0126; Tukey post hoc, ** p< 0.01). This was in contrast to results obtained after 5 mM quetiapine application and U0126 co-treatment (ANOVA, F(1,2) ¼ 10.8, *** p< 0.001; 151.6 39.1% for QTP alone and 164.9 22.8% for QTP þ U0126; Tukey post hoc, n.s., not significant) and pointed to different mechanisms underlying the induction of GDNF release between the two compounds. Following the hypothesis that the antidepressant efficacy of quetiapine might be mediated by the NRI properties of its metabolite, and in view of the similarity in pMAPK activation profiles between norquetiapine and reboxetine (Fig. 3), we further tested the ability of reboxetine to evoke GDNF release, in the presence or absence of U0126. As shown in Fig. 4a, 10 mM reboxetine (minimal dosage inducing GDNF release, Di Benedetto et al, unpublished results) induced an increase in GDNF release that could also be only partially reversed by pre-treatment of the cells with U0126 (ANOVA, F(1,2) ¼ 42.9, *** p< 0.001; 168.2 19.7% for RBX alone and 134.0 10.4% for RBX þ U0126; Tukey post hoc, *** p< 0.001), similarly to what we observed for norquetiapine. The lack of reversibility of GDNF release after U0126 and quetiapine coadministration was not attributable to a leakiness of pharmacological inhibition of ERK phosphorylation, because for coadministration experiments we used the concentration of U0126 that demonstrated the higher efficiency in phospho-ERK inhibition after 48 h of treatment (20 mM, Fig. 4b). In the second experiment, to prove that the peak of ERK activity observed 10 min after drug treatment was indeed responsible for the effects 48 h later, we treated cells with the different drugs for 10 min, then removed the drugs and let the cells further grow in drug-free medium for 48 h before measuring GDNF release. These data (Fig. 4c and Suppl Fig. S2) showed for both norquetiapine and reboxetine still a significantly higher amount of GDNF released after 48 h, but not for quetiapine. A one-way ANOVA indicated a significant effect of treatment on GDNF released (F(1,3) ¼ 5.79, ** p< 0.01); a post hoc Dunnett’s test showed that this effect was significant only for norquetiapine and reboxetine (135.3 16.7% and 129.5 15.2% respectively; * p< 0.05 for both in respect to control). 3.5. Effect of ERK1 downregulation on GDNF release As our previous experiments suggested ERK1 activation to be critical for GDNF release after norquetiapine and reboxetine treatment, we employed RNA interference to specifically inactivate the ERK1 isoform and examined the effects on GDNF release 48 h after drug administration. 5 mM norquetiapine and 10 mM reboxetine could significantly induce GDNF release (ANOVA; 133.3 6.5%, *** p< 0.001 for norquetiapine and 133.7 12.5%, *** p< 0.001 for reboxetine in comparison to control); the same could be observed after transfection of cells with a control plasmid bearing a shRNA against LacZ mRNA (“sh-nontarg”, nontargeting) (ANOVA, 151.2 10.4%, *** p< 0.001 for norquetiapine and 124.4 10.4%, * p< 0.05 for reboxetine), thus indicating that the plasmid backbone per se does not affect GDNF release after drug treatments. On the contrary, a Tukey post hoc analysis of the results after cell transfection with two different plasmids bearing shRNAs against ERK1 mRNA (shErk1.1 and shErk1.2) revealed an essential role of this ERK isoform on the amount of GDNF released. In fact, one-way ANOVA indicated a reduction of GDNF release both for norquetiapine
B. Di Benedetto et al. / Neuropharmacology 62 (2012) 209e216
213
Fig. 4. Dependence of GDNF release on ERK/MAPK activation after 48 h quetiapine, norquetiapine or reboxetine treatment. a) Quetiapine (5 mM) significantly induced GDNF release (QTP, n ¼ 6; ANOVA, ** p< 0.001 in comparison to control). The pMAPK inhibitor U0126 (20 mM) could not reverse this effect (QTP þ U0126, n ¼ 6; ANOVA, *** p< 0.001 in comparison to control and n.s., not significant, in comparison to QTP alone). Norquetiapine (5 mM) also significantly induced GDNF release (NORQTP, n ¼ 6; ANOVA, *** p< 0.001 in comparison to control). In contrast to quetiapine, this response could be at least partially reversed by U0126 (NORQTP þ U0126, n ¼ 6; ANOVA, ** p< 0.001 in comparison to NORQTP alone). However, NORQTP þ U0126 still showed a significant activation (ANOVA, *** p< 0.001 in comparison to control). Reboxetine (10 mM) significantly induced GDNF release (RBX, n ¼ 6; ANOVA, *** p< 0.001 in comparison to control). This effect was also only partially reversed by U0126 (RBX þ U0126, n ¼ 6; ANOVA, *** p< 0.01); however, RBX þ U0126 still caused a significant induction of GDNF release (ANOVA, *** p< 0.001 in comparison to control). b) Western blot of basal ERK and phospho-ERK after administration of U0126 demonstrated the higher efficacy of the 20 mM dosage of the pharmacological compound used in the co-administration experiments in inhibiting ERK activation after the 48 h course of the experiments. c) The effect on GDNF release after 4a was due to the acute ERK activation, as demonstrated by a “washout” experiment: cells were treated for 10 min with the different drugs and then further cultivated in drug-free medium until GDNF measurement 48 h later. Only norquetiapine and reboxetine still induced an increased GDNF release (ANOVA, n ¼ 6; n.s., not significant, * p< 0.05).
214
B. Di Benedetto et al. / Neuropharmacology 62 (2012) 209e216
Fig. 5. Effect of RNA interference-mediated specific downregulation of erk1 on GDNF release upon norquetiapine and reboxetine treatment. a) Norquetiapine (5 mM) and reboxetine (10 mM) significantly induced GDNF release (ANOVA, *** p< 0.001 in comparison to control); the same GDNF release was also observed after cell transfection with a control plasmid bearing a shRNA against LacZ mRNA (“sh-nontarg”; ANOVA, *** p< 0.001 for norquetiapine and * p< 0.05 for reboxetine), indicating that the plasmid alone does not affect cell responses to drug treatments. After transfection with plasmids containing shRNAs targeting two different regions of ERK1 mRNA (shErk1.1 and shErk1.2), the effect of the drug on GDNF release was reduced, suggesting a critical role for ERK1 in this response (ANOVA, ** p< 0.01 and * p< 0.05 in comparison to norquetiapine alone; and * p< 0.05 and ** p< 0.01 in comparison to reboxetine alone). n.s., not significant, *** p< 0.001. b) A representative Western blot of the different treatments shows the specificity of ERK1 mRNA downregulation, while ERK2 remains unaffected, as demonstrated by pERK2 activation that reaches the same levels in all samples and compared to controls.
(112.5 7.9% for shErk1.1 and 113.1 8.6% for shErk1.2, ** p< 0.01 and * p< 0.05, respectively, in comparison to norquetiapine alone) and for reboxetine (113.5 8.4% for shErk1.1 and 107.2 20.7% for shErk1.2, * p< 0.05 and ** p< 0.01, respectively, in comparison to reboxetine alone).
4. Discussion In this study we reveal a mechanism for the putative antidepressant activity of quetiapine via its major metabolite norquetiapine, ERK activation and the release of GDNF, a physiological cellular response conferred by other antidepressants (Hisaoka et al., 2007). Moreover, we suggest simultaneous activation of ERK1 and ERK2 as a hypothetical common feature of several antidepressants.
Recently, it has been shown that CHO cells transfected with the human 5HT1A receptor activate the MAPK signalling pathway with different intensities after administration of several antipsychotics, but specific differences between the two ERK isoforms in response to antipsychotic treatment have not been reported yet (Cussac et al., 2002). Moreover, downstream physiological consequences have not been investigated so far. Here we show a specific difference in the activation of the two ERK isoforms, entailing a differential GDNF release. Since also clozapine and haloperidol can induce GDNF release in the same C6 cell line (Shao et al., 2006), we also investigated to which extent a differential ERK activation might be responsible for this response. Indeed, we found that release of GDNF induced by norquetiapine is reversed by pretreatment of cells with a specific pharmacological inhibitor of ERK phosphorylation. Moreover, the early peak of ERK activation observed was sufficient to induce the long term response measured 48 h later only after norquetiapine and reboxetine treatment, but not with quetiapine. These results strongly suggested that the activation of only one ERK isoform as triggered by quetiapine, as well as by clozapine or haloperidol, might not be the mechanism for GDNF induction observed with antipsychotics. On the contrary, activation of both ERKs is necessary for GDNF induction, as observed with norquetiapine. This was additionally supported by the reduction of GDNF release increases after specific downregulation of the sole ERK1. Furthermore, our in vitro results are in line with recent findings showing that quetiapine might exert antidepressant effects. In particular, our study suggests that these effects are mediated, at least at the cellular level, by its metabolite norquetiapine, which has very recently been shown to possess antidepressant-like effects also in an animal model (Jensen et al., 2008). This might shed light on the apparent discrepancy between quetiapine plasma concentration with D2 or 5HTR2A receptor occupation and clinical effects (Mauri et al., 2007). So far, therapeutic drug monitoring of norquetiapine plasma concentration and its correlation with clinical antidepressant response is not available. Nevertheless, norquetiapine plasma levels might more likely reflect the clinical efficacy of quetiapine as an antidepressant than those of quetiapine itself. Moreover, therapeutic drug monitoring of norquetiapine might help avoiding adverse effects of quetiapine long term treatment in the case of poor metabolization of quetiapine into norquetiapine. The hypothetical antidepressant-like effects of quetiapine in humans would rather occur after chronic than after acute treatment (Orsetti et al., 2007), as shown for classical antidepressants. One might argue that this discourages undertaking a study to characterize acute effects of drugs with antidepressant properties. Nevertheless, the initiation of adaptive changes which might be responsible for clinical effects three to four weeks later may occur already at earlier phases after beginning of therapeutical intervention (for a review, see Andrade and Rao, 2010, and references therein). Therefore, it is of interest to understand which molecular events represent early responses to drug treatments. Moreover, our finding that the neurotrophin GDNF is released both after quetiapine and norquetiapine treatment is in line with the “neurotrophic theory” of depression, which postulates a lack of neurotrophins as one of the hallmarks of depression (reviewed in Nestler et al., 2002). Indeed, we could show that the acute (10 min) treatment of cells with both norquetiapine and reboxetine followed by a “washout” of the drugs was sufficient to induce the medium term consequences measured as an increased GDNF release 48 h later. This might represent the beginning of a cascade of events that leads to the clinically beneficial effects. Moreover, it is intriguing that norquetiapine’s effects resemble the ERK activation pattern of the selective norepinephrine reuptake inhibitor reboxetine. This similarity with NRIs is also sustained by the clinical study by Cutler et al.
B. Di Benedetto et al. / Neuropharmacology 62 (2012) 209e216
(2009) that compared the effects of quetiapine as a monotherapy in major depressive disorder with duloxetine, another commonly used antidepressant with norepinephrine reuptake inhibitor properties. Although our in vitro study offers a first insight into one of the intracellular determinants that may contribute to the antidepressant activity of quetiapine and norquetiapine and is necessary to identify cell-autonomous responses to pharmacological stimuli, future complementary in vivo studies are demanded to further investigate the systemic mechanisms of action underlying the putative antidepressant properties of these drugs, as it has been recently shown for fluoxetine in an animal model of depression (Qi et al., 2008). With respect to the side effect profiles, the fact that not only quetiapine but also its metabolite show high antagonistic activity at histamine H1 receptors (Jensen et al., 2008) suggests that there might not be a specific advantage of the use of norquetiapine compared with quetiapine, as they may both induce sedation (Cohrs et al., 2004) or weight gain (Allison et al., 1999). However, data for norquetiapine with regard to clinical effects and side effects are lacking so far. The finding that a metabolite of quetiapine could be actually a major factor for the activation of molecular events that may contribute to the antidepressant efficacy of quetiapine is intriguing. Nevertheless, also other mechanisms of action, such as modulation of HPA axis activity by quetiapine (Cohrs et al., 2006) may contribute to its antidepressant activity. Further studies are needed to identify putative other signalling pathways that might also contribute to the GDNF release mediated by antipsychotics such as quetiapine. Moreover, additional downstream effectors, which may induce GDNF release synergistically together with ERK1/2 should be investigated, because GDNF release could not be completely counteracted by U0126 pre-treatment followed by norquetiapine or reboxetine co-administration. Among the MAPK signalling pathways, the p38MAPK or the JNK pathway would be of particular interest. Here, we specifically investigated the cellular responses of a glial cell line. However, it cannot be excluded that quetiapine and norquetiapine display a cell-type specific activity, which could not be detected in the present study. Therefore, it would be of interest to further examine a putative differential impact of these two compounds on various neuronal and non-neuronal cell types. 5. Conclusions In conclusion, in view of the differential ERK activation patterns of quetiapine and norquetiapine, it is conceivable that the putative antidepressant properties of this atypical antipsychotic are at least in part mediated through its major metabolite. Despite the obvious limitations of an in vitro study to draw conclusions about the effective mechanism of action of compounds used for human therapeutical purposes, we still believe that our results are intriguing, as several recent reports underscore the importance of the ERK/MAPK signalling pathway in the pathophysiology of major depressive disorder. Therefore, our findings not only support the importance of ERK/MAPK as a target for antidepressant activity, but also encourage future studies to identify further targets of quetiapine/norquetiapine in the treatment of depression. 6. Author’s contribution BDB and RR designed the research and wrote the manuscript; BDB performed the experiments and analyzed and interpreted data; RK and WW contributed essential reagents/tools to perform RNAi experiments; TR contributed valuable suggestions on the
215
performance of the LDH experiments; TR and CN contributed critical reading of the manuscript. RK, CN, TR and WW are listed in alphabetical order. All authors read and approved the final version of the manuscript. Acknowledgements The authors want to thank AstraZeneca for the generous gift of quetiapine and norquetiapine and Prof Dr B Hamprecht (Faculty of Chemistry and Pharmacy, Eberhard-Karls-University, Tübingen, Germany) for generously providing rat C6 glioma cells. BDB, TR, RK, WW and CN declare no conflict of interest. RR is on AstraZeneca advisory boards and is performing an investigator initiated study with quetiapine supported by AstraZeneca. This work was supported by a Fellow Group to RR from the Max Planck Society. Supplementary data Supplementary data related to this article can be found online at doi:10.1016/j.neuropharm.2011.07.001. References Allison, D.B., Mentore, J.L., Heo, M., Chandler, L.P., Cappelleri, J.C., Infante, M.C., Weiden, P.J., 1999. Antipsychotic-induced weight gain: a comprehensive research synthesis. Am. J. Psychiatry 156, 1686e1696. Andrade, C., Rao, N.S.K., 2010. How antidepressant drugs act: a primer on neuroplasticity as the eventual mediator of antidepressant efficacy. Indian J. Psychiatry 52, 378e386. Bai, O., Wei, Z., Lu, W., Bowen, R., Keegan, D., Li, X.M., 2002. Protective effects of atypical antipsychotic drugs on PC12 cells after serum withdrawal. J. Neurosci. Res. 69, 278e283. Banasr, M., Duman, R.S., 2008. Glial loss in the prefrontal cortex is sufficient to induce depressive-like behaviors. Biol. Psychiatry 64, 863e870. Bortnick, B., El-Khalili, N., Banov, M., Adson, D., Datto, C., Raines, S., Earley, W., Eriksson, H., 2010. Efficacy and tolerability of extended release quetiapine fumarate (quetiapine XR) monotherapy in major depressive disorder: a placebo-controlled, randomized study. J. Affect Disord. Cohrs, S., Rodenbeck, A., Guan, Z., Pohlmann, K., Jordan, W., Meier, A., Ruther, E., 2004. Sleep-promoting properties of quetiapine in healthy subjects. Psychopharmacology (Berl) 174, 421e429. Cohrs, S., Roher, C., Jordan, W., Meier, A., Huether, G., Wuttke, W., Ruther, E., Rodenbeck, A., 2006. The atypical antipsychotics olanzapine and quetiapine, but not haloperidol, reduce ACTH and cortisol secretion in healthy subjects. Psychopharmacology (Berl) 185, 11e18. Cussac, D., Duqueyroix, D., Newman-Tancredi, A., Millan, M.J., 2002. Stimulation by antipsychotic agents of mitogen-activated protein kinase (MAPK) coupled to cloned, human (h)serotonin (5-HT)(1A) receptors. Psychopharmacology (Berl) 162, 168e177. Cutler, A.J., Montgomery, S.A., Feifel, D., Lazarus, A., Astrom, M., Brecher, M., 2009. Extended release quetiapine fumarate monotherapy in major depressive disorder: a placebo- and duloxetine-controlled study. J. Clin. Psychiatry 70, 526e539. Duric, V., Banasr, M., Licznerski, P., Schmidt, H.D., Stockmeier, C.A., Simen, A.A., Newton, S.S., Duman, R.S., 2010. A negative regulator of MAP kinase causes depressive behavior. Nat. Med. 16, 1328e1332. El-Khalili, N., Joyce, M., Atkinson, S., Buynak, R.J., Datto, C., Lindgren, P., Eriksson, H., 2010. Extended-release quetiapine fumarate (quetiapine XR) as adjunctive therapy in major depressive disorder (MDD) in patients with an inadequate response to ongoing antidepressant treatment: a multicentre, randomized, double-blind, placebo-controlled study. Int. J. Neuropsychopharmacol. 13, 917e932. Gajwani, P., Muzina, D.J., Kemp, D.E., Gao, K., Calabrese, J.R., 2007. Update on quetiapine in the treatment of bipolar disorder: results from the BOLDER studies. Neuropsychiatr Dis. Treat. 3, 847e853. Hisaoka, K., Takebayashi, M., Tsuchioka, M., Maeda, N., Nakata, Y., Yamawaki, S., 2007. Antidepressants increase glial cell line-derived neurotrophic factor production through monoamine-independent activation of protein tyrosine kinase and extracellular signal-regulated kinase in glial cells. J. Pharmacol. Exp. Ther. 321, 148e157. Jensen, N.H., Rodriguiz, R.M., Caron, M.G., Wetsel, W.C., Rothman, R.B., Roth, B.L., 2008. N-desalkylquetiapine, a potent norepinephrine reuptake inhibitor and partial 5-HT1A agonist, as a putative mediator of quetiapine’s antidepressant activity. Neuropsychopharmacology 33, 2303e2312. Kyriakis, J.M., Avruch, J., 1996. Sounding the alarm: protein kinase cascades activated by stress and inflammation. J. Biol. Chem. 271, 24313e24316.
216
B. Di Benedetto et al. / Neuropharmacology 62 (2012) 209e216
Lin, L.F., Doherty, D.H., Lile, J.D., Bektesh, S., Collins, F., 1993. GDNF: a glial cell linederived neurotrophic factor for midbrain dopaminergic neurons. Science 260, 1130e1132. Mauri, M.C., Volonteri, L.S., Colasanti, A., Fiorentini, A., De Gaspari, I.F., Bareggi, S.R., 2007. Clinical pharmacokinetics of atypical antipsychotics: a critical review of the relationship between plasma concentrations and clinical response. Clin. Pharmacokinet. 46, 359e388. Nestler, E.J., Barrot, M., DiLeone, R.J., Eisch, A.J., Gold, S.J., Monteggia, L.M., 2002. Neurobiology of depression. Neuron 34, 13e25. Orsetti, M., Canonico, P.L., Dellarole, A., Colella, L., Di Brisco, F., Ghi, P., 2007. Quetiapine prevents anhedonia induced by acute or chronic stress. Neuropsychopharmacology 32, 1783e1790. Pages, G., Lenormand, P., L’Allemain, G., Chambard, J.C., Meloche, S., Pouyssegur, J., 1993. Mitogen-activated protein kinases p42mapk and p44mapk are required for fibroblast proliferation. Proc. Natl. Acad. Sci. U S A 90, 8319e8323. Qi, X., Lin, W., Li, J., Li, H., Wang, W., Wang, D., Sun, M., 2008. Fluoxetine increases the activity of the ERK-CREB signal system and alleviates the depressive-like behavior in rats exposed to chronic forced swim stress. Neurobiol. Dis. 31, 278e285. Rajkowska, G., Miguel-Hidalgo, J.J., 2007. Gliogenesis and glial pathology in depression. CNS Neurol. Disord. Drug Targets 6, 219e233.
Rubovitch, V., Gafni, M., Sarne, Y., 2004. The involvement of VEGF receptors and MAPK in the cannabinoid potentiation of Ca2þ flux into N18TG2 neuroblastoma cells. Brain Res. Mol. Brain Res. 120, 138e144. Shao, Z., Dyck, L.E., Wang, H., Li, X.M., 2006. Antipsychotic drugs cause glial cell linederived neurotrophic factor secretion from C6 glioma cells. J. Psychiatry Neurosci. 31, 32e37. Suppes, T., Datto, C., Minkwitz, M., Nordenhem, A., Walker, C., Darko, D., 2010. Effectiveness of the extended release formulation of quetiapine as monotherapy for the treatment of acute bipolar depression. J. Affect Disord. 121, 106e115. Weber, J.D., Raben, D.M., Phillips, P.J., Baldassare, J.J., 1997. Sustained activation of extracellular-signal-regulated kinase 1 (ERK1) is required for the continued expression of cyclin D1 in G1 phase. Biochem. J. 326 (Pt 1), 61e68. Weisler, R., Joyce, M., McGill, L., Lazarus, A., Szamosi, J., Eriksson, H., 2009. Extended release quetiapine fumarate monotherapy for major depressive disorder: results of a double-blind, randomized, placebo-controlled study. CNS Spectr. 14, 299e313. Yuan, P., Zhou, R., Wang, Y., Li, X., Li, J., Chen, G., Guitart, X., Manji, H.K., 2010. Altered levels of extracellular signal-regulated kinase signaling proteins in postmortem frontal cortex of individuals with mood disorders and schizophrenia. J. Affect Disord. 124, 164e169.
