Desipramine Protects Neuronal Cell Death and Induces Heme Oxygenase-1 Expression in Mes23.5 Dopaminergic Neurons

Background Desipramine is known principally as a tricyclic antidepressant drug used to promote recovery of depressed patients. It has also been used in a number of other psychiatric and medical conditions. The present study is the first to investigate the neuroprotective effect of desipramine. Methodology/Principal Findings Mes23.5 dopaminergic cells were used to examine neuroprotective effect of desipramine. Western blot, reverse transcription-PCR, MTT assay, siRNA transfection and electrophoretic mobility shift assay (EMSA) were carried out to assess the effects of desipramine. Desipramine induces endogenous anti-oxidative enzyme, heme oxygenase-1 (HO-1) protein and mRNA expression in concentration- and time-dependent manners. A different type of antidepressant SSRI (selective serotonin reuptake inhibitor), fluoxetine also shows similar effects of desipramine on HO-1 expression. Moreover, desipramine induces HO-1 expression through activation of ERK and JNK signaling pathways. Desipramine also increases NF-E2-related factor-2 (Nrf2) accumulation in the nucleus and enhances Nrf2-DNA binding activity. Moreover, desipramine-mediated increase of HO-1 expression is reduced by transfection with siRNA against Nrf2. On the other hand, pretreatment of desipramine protects neuronal cells against rotenone- and 6-hydroxydopamine (6-OHDA)-induced neuronal death. Furthermore, inhibition of HO-1 activity by a HO-1 pharmacological inhibitor, ZnPP IX, attenuates the neuroprotective effect of desipramine. Otherwise, activation of HO-1 activity by HO-1 activator and inducer protect 6-OHDA-induced neuronal death. Conclusions/Significance These findings suggest that desipramine-increased HO-1 expression is mediated by Nrf2 activation through the ERK and JNK signaling pathways. Our results also suggest that desipramine provides a novel effect of neuroprotection, and neurodegenerative process might play an important role in depression disorder.


Introduction
Parkinson's disease (PD) is a neurodegenerative disorder characterized by the progressive degeneration of dopaminergic neurons of the substantia nigra (SN), giving rise to dopamine depletion in the striatum [1]. The resulting loss of dopaminergic neurons leads to debilitating motor dysfunction including rigidity, resting tremor, mask face and bradykinesia. Although PD is well characterized by motor symptoms, clinical depression is the most common neuropsychiatric disorder in PD patients [2][3][4]. More than 40% of PD patients are observed in depression [5,6]. Importantly, numerous studies have shown that there is an increased incidence of depression before the onset of PD, and motor fluctuations may greatly affect the occurrence [7,8]. Depression has been classified as a disorder of the brain and CNS, and is manifested by a combination of symptoms that interferes with the ability to work, study, sleep, eat, and enjoy once pleasurable activities [9][10][11][12].
Desipramine is a tricyclic antidepressant (TCA), one of an antidepressant drug used to promote recovery of depressed patients. Desipramine does not affect mood or arousal but may cause sedation in non-depressed individuals. However, desipramine exerts a positive effect on mood in depressed individuals. TCAs are potent inhibitors of serotonin and norepinephrine reuptake [13,14]. It has been reported that desipramine significantly increases anti-apoptotic protein Bcl-2 expression, repairs serotonin and noradrenaline production [15], and prevents stressinduced depressive-like behavioral changes [16]. Interestingly, it has been reported that desipramine is able to reduce MPP +induced cell toxicity in SH-SY5Y, however, despite many studies of those links, the detail molecular mechanisms of antidepressants on neuroprotective effect remain unknown.

Cell Cultures
Mes23.5 cell line is a dopaminergic cell line hybridized from murine neuroblastoma-glioma N18TG2 cells with rat mesencephalic neurons, which shows several properties similar to those of primary neurons originated in the substantia nigra [31]. The culture protocol of Mes23.5 was followed our previous report [32]. Briefly, cells were cultured in DMEM/F12 containing Sato's components growth medium supplemented with 5% FBS at 37uC in a humidified incubator in an atmosphere of 5% CO 2 . Confluent cultures were passaged by trypsinization.

Western Blot Analysis
The rat Mes23.5 cell line was treated with desipramine for indicated time periods, and whole cell extracts were lysed on ice with radioimmunoprecipitation assay buffer. The nuclear extracts were prepared as described previously [33]. Cells were rinsed with PBS and suspended in hypotonic buffer A for 10 min on ice, and vortexed for 10 s. The lysates were separated into cytosolic and nuclear fractions by centrifugation at 12,000 g for 10 min. The supernatants containing cytosolic proteins were collected. Pellet containing nuclear fraction was re-suspended in buffer C for 30 min on ice.
The supernatants containing nuclear proteins were collected by centrifugation at 13,000 g for 20 min and stored at 280uC. Protein samples were separated by SDS-PAGE (sodium dodecyl sulfate-polyacrylamide) and transferred to PVDF membranes which were then blocked with 5% nonfat milk, and then probed overnight with primary antibody at 4uC. After undergoing PBS washes, the membranes were incubated with secondary antibodies. The blots were visualized by enhanced chemiluminescence using Kodak XOMAT LS film (Eastman Kodak, Rochester, NY). The blots were subsequently stripped through incubation in stripping buffer [34] and reprobed for b-actin as a loading control. Quantitative data were obtained using a computing densitometer and a shareware Image J.

Reverse Transcription-PCR (RT-PCR)
Total RNA was extracted from Mes23.5 cell line using a TRIzol kit (MDBio Inc., Taipei, Taiwan). The reverse transcription reaction was performed using 2 mg of total RNA that was reverse transcribed into cDNA with the oligo(dT) primer. After preincubation at 50uC for 2 min and 95uC for 10 min, the PCR was performed as 30 cycles of 95uC for 10 s and 60uC for 1 min. The threshold was set above the non-template control background and within the linear phase of target gene amplification to calculate the cycle number at which the transcript was detected (denoted as CT). The oligonucleotide primers were HO-1:

Measurement of Cell Viability
Cell viability was assessed by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Cells cultured in 96-well plates were pre-treated with various inhibitors before treated with desipramine for 8 h followed by treatment with rotenone or 6-OHDA. After treatment, MTT (0.5 mg/ml) was added for 60 min, the culture medium was then removed, cells were dissolved in dimethyl sulfoxide and shaken for 10 min. OD values at 550 nm were measured by microplate reader. The absorbance indicates the enzymatic activity of mitochondria and provides information of cell viability.

Sulphorhodamine B Assay
Cell viability was assessed by the sulphorhodamine B (SRB) assay. Cells cultured in 96-well plates were treated with various inhibitors before treated with desipramine for 8 h followed by stimulation with rotenone or 6-OHDA. After treatment, the culture medium was then removed and cells were fixed in situ by gentle addition of 50 ml per well of 10% TCA and plates were incubated for 1 h at 4uC. After discarding supernatant, plates were then washed for 5 times with PBS, and then added 50 ml of SRB solution to each well, and plates were incubated for 15 min at room temperature. After staining, cells were washed 5 times with 1% acetic acid and dissolved in tris-base. After 10 mins' shaking, OD values at 515 nm were measured by microplate reader. The absorbance provides information of cell viability.

Trypan Blue Exclusion Assay
The trypan blue exclusion assay was used to assess cell viability after 6-OHDA or rotenone exposure. Culture medium was removed and 0.2% trypan blue in TBS was added to the cells for 10 min. After the removal of the trypan blue solution, cells were fixed with 4% paraformaldehyde in PBS for 5 min, then replaced by TBS. Cells were then observed under bright field microscope.

Immunocytofluorescent Staining
Cells were seeded onto glass coverslips and exposed to desipramine for 2 h, then washed with PBS and fixed with 4% paraformaldehyde for 15 min, after which they were permeabilized with Triton X-100 for 30 min. After blocking with 5% nonfat milk in PBS buffer, cells were incubated with rabbit anti-Nrf2 antibody for 1 h at room temperature. After a brief wash, cells were incubated with a secondary antibody conjugated with Alexa 488-Flour (1:200; Invitrogen Life Technologies, Carlsbad, CA). Finally, cells were washed again, mounted, and visualized with fluorescence microscope. Quantitative data were obtained using a computing densitometer and Image J.

Transfection
Mes23.5 cells were grown to 80% confluent in 6-well plates and transfected with control siRNA or Nrf2 siRNA on the following day by Lipofectamine TM 2000 (LF2000; Invitrogen). Control siRNA or Nrf2 siRNA and LF2000 were premixed in OPTImedium for 20 min and applied to the cells (0.8 ml/well). Medium containing 5% FBS (0.8 ml) was added 4,6 h later. After 24 hrs' transfection, LF2000-containing medium was replaced with DMEM/F12 medium containing 2% FBS and treated with desipramine for another 24 h.

Electrophoretic Mobility Shift Assay (EMSA)
The electrophoretic mobility shift assay gel shift kit (Panomics, Redwood City, CA) was used according to our previous report [35]. Nuclear extract (4 mg) of Mes23.5 was incubated with poly d(I-C) at room temperature for 5 min. The nuclear extract was then incubated with biotin-labeled probes at room temperature for 30 min. After electrophoresis on an 8% polyacrylamide gel, the samples on gel were transferred onto a presoaked Immobilon-Nyt membrane (Millipore, Billerica, MA). The membrane was crosslinked in an oven for 1 min and then developed with blocking buffer and streptavidin-horseradish peroxidase conjugate before being subjected to Western blot analysis.

Statistics
Statistical analysis was performed using software Graphpad Prism 4.01 (Graph Pad Software Inc., San Diego, CA). The values given are means 6 S.E.M. Statistical analysis between two samples was performed using Student's t-test. Statistical comparisons of more than two groups were performed using one-way ANOVA with Bonferroni's post hoc test. In all cases, a value of p less than 0.05 was considered significant.

Desipramine and Fluoxetine Induce HO-1 Expression in Mes23.5 Cells
We recently reported that Omega-3 polyunsaturated fatty acids [32] and berberine [15] increase HO-1 expression and exert antineuroinflammatory responses in glial cells. Firstly, we identified the effect of desipramine on the regulation of HO-1 in Mes23.5 dopaminergic neurons. Stimulation of desipramine with various concentrations (5, 10, or 20 mM for 24 h) and for indicated time periods (0, 4, 8 or 24 h) increased HO-1 protein expression ( Fig. 1A and 1B). Moreover, desipramine also increased HO-1 mRNA expression in time-dependent and concentration-dependent manners ( Fig. 1E and 1F). We further analyzed whether a different type of antidepressant SSRI fluoxetine regulates HO-1 expression. Stimulation of fluoxetine with various concentrations (5, 10, or 20 mM for 24 h) and for indicated time periods (0, 4, 8 or 24 h) also increased HO-1 protein expression ( Fig. 1C and 1D). In addition, fluoxetine also increased HO-1 mRNA up-regulation ( Fig. 1G and 1H). ERK, JNK Signaling Pathways are Involved in Desipramine-mediated HO-1 up-regulation MAP kinase pathways are the most important signaling pathways that participate in transducing a variety of biological responses. Numerous evidences reported that activation of MAP kinase pathways contribute to regulation of HO-1 expression [36][37][38]. Therefore, we examined the effect of desipramine on the activation of MAP kinases in Mes23.5 dopaminergic neurons. As shown in Figure 2A, desipramine-increased HO-1 expression could be inhibited by ERK (PD98059) and JNK (SP600125) inhibitors, but not inhibitor of p38 (SB203580). Desipramine also resulted in the JNK and ERK phosphorylation initiated at 10 min and sustained to 120 min ( Fig. 2B and 2C). Similarly, treatment of JNK inhibitor SP600125 also effectively reduced fluoxetineinduced HO-1 expression (Fig. 2D). Stimulation of cells with fluoxetine also increased JNK and ERK phosphorylation timedependently ( Fig. 2E and 2F).

Involvement of Nrf2 Activation in Desipramine-induced HO-1 Expression in Mes23.5 Dopaminergic Neurons
Numerous evidences reported that stress response element (StRE)/Nrf2 transcription factor pathway is an important transcriptional factor for HO-1 expression [39][40][41]. We therefore examined whether the Nrf2 signaling is involved in desipramineinduced HO-1 expression. Treatment of Mes23.5 dopaminergic neurons with desipramine enhanced Nrf2 expression time-dependently (Fig. 3A), and desipramine also resulted in an accumulation of Nrf2 in nucleus (Fig. 3B). Immunofluorescence staining of Nrf2 localization also showed that Nrf2 translocated from cytoplasm to nucleus in response to desipramine treatment for 2 h (Fig. 3C). Moreover, to evaluate desipramine-mediated Nrf2 transcription factor binding activity in Mes23.5 cells, EMSA were performed to test DNA binding activity. Stimulation of cells with desipramine for 1 or 2 h increased the DNA binding activity of Nrf2 in nuclear extracts, treatment with PD98059 or SP600125 reduced desipramine-increased DNA binding activity of Nrf2 (Fig. 4A). There was no detectable DNA binding complex without loading nuclear protein. To investigate whether the desipramineinduced HO-1 expression is mediated through Nrf2 activation, cells were transfected with Nrf2 siRNA for 24 h followed by stimulation with desipramine. Transfection with Nrf2 siRNA for 24 h in Mes23.5 dopaminergic neurons reduced Nrf2 expression. Moreover, tranfection with Nrf2 siRNA also significantly reduced desipramine-induced HO-1 expression in a concentration-dependent manner (Fig. 4B). These results suggest that desipramineincreased HO-1 expression was mediated through ERK, JNK and Nrf2 pathways in Mes23.5 dopaminergic neurons.
Desipramine-induced HO-1 Expression Protects Mes23.5 Cells from 6-OHDA or Rotenone-induced Neurotoxicity Rotenone and 6-OHDA are useful neurotoxins as resulting in dopaminergic neuron degeneration [42][43][44][45]. Studies using neurotoxins have provided insights into the molecular mechanisms of dopaminergic neuronal death. To ensure the actual cell loss was correlated with the cell death assay values, a cell count assay had been evaluated. The Mes23.5 dopaminergic neurons' cell death induced by rotenone and 6-OHDA were correlated with MTT assay value in a concentration-dependent manner ( Figure S1). Surprisingly, pretreatment of desipramine at 20 mM followed by rotenone ( Fig. 5A and 5C) and 6-OHDA ( Fig. 5B and 5D) dramatically protected cells from neurotoxicity. The neuroprotective effect of desipramine exerts only by pretreatment, but not cotreatment or posttreatment with the neurotoxins ( Figure S2).

Discussion
Clinical studies have suggested that depression and PD are closely related [46][47][48], and depression is a negative impact on the quality of life of PD patients and their families. Dopamine replacement therapy, especially with higher doses of levodopa, is sometimes accompanied by depression [49]. Moreover, the presence of levodopa-induced dyskinesia is correlated with an increased incidence of depression [50]. Furthermore, a number of PD patients who had subthalamic nucleus deep brain stimulation suffered from behavioral side effects including cognitive impairments and depression [51][52][53]. Treatment with TCAs is restricted by adverse effects, however, TCAs like desipramine are effective in treating depression in PD patients and may even reduce motor symptoms [54]. In this study, we used neurotoxins rotenone and 6- OHDA to induce dopaminergic neuron toxicity that mimics human neurodegenerative disease and neuropathological representative of PD. Previous report has reported the antioxidant effects of antidepressant agents [55]. Furthermore, it has also been reported that desipramine and fluoxetine protect neurons against microglial activation-mediated neurotoxicity [56]. However, the mechanism of antidepressants on neuroprotective effect remains unknown.
Desipramine is one of the most common tricyclic antidepressant (TCA) used to promote recovery of depressed patients. For adults, the therapeutic concentration of desipramine is approximately 1 mM [57]. Recently, it has been reported that a higher concentration of desipramine (up to 390 mM) exerts neuroprotective effect in primary culture [58]. Importantly, fluoxetine accumulates in the human brain relative to plasma, with brain concentrations ranging up to 30 mM [59]. Here, we showed that desipramine and fluoxetine at concentrations up to 20 mM increase HO-1 expression and protect neuronal cell death. Hence, Figure 3. Desipramine induces Nrf2 translocation from cytoplasm to nucleus in Mes23.5 dopaminergic neurons. Cells were incubated with desipramine (20 mM) for indicated time periods (4 or 8 h), and Nrf2 expression levels in whole cell lysates (A) and nuclear extracts (B) were determined by immunoblotting with Nrf2-specific antibody. PCNA was used as the loading control for nuclear fraction. The quantitative data are shown in below. The Nrf2 expression is significantly different between desipramine treatment group and control group in nuclear extract (one-way ANOVA followed by Bonferroni's post hoc test). Results are expressed as the means 6 S.E.M. from four independent experiments. *, p,0.05 as compared with the vehicle control group. Cells were treated with or without desipramine (20 mM) for 2 h, and the levels of Nrf2 were determined by immunoflourescence (C). Note that the Nrf2 translocates from cytoplasm to nucleus in response to desipramine stimulation. Results are the representatives of three independent experiments. doi:10.1371/journal.pone.0050138.g003 explore the therapeutic potential of antidepressants may help us to identify target molecules for drug development and therapy against neurodegeneration. To ascertain the significance of their neuroprotective effects in humans, further clinical trials with antidepressants are required as well as retrospective epidemiological studies assessing the prevalence of neurodegenerative diseases. The observation that antidepressants have neuroprotective properties might have important clinical implications since these drugs are heavily prescribed worldwide and chronic treatment often lasts several months.
It has been reported that HO-1 plays a significant role in antiapoptosis, anti-oxidant and neuroprotection [50,[60][61][62][63]. Numerous evidences have showed that neuronal and non-neuronal cells increase the synthesis of HO-1 under oxidative injury and inflammation conditions, which plays an important role in neuroprotective function [64,65]. We have also reported that increased HO-1 expression in glial cells may contribute to anti-neuroinflammatory responses and exert neuroprotection [15,32]. Importantly, the expression of anti-oxidative enzyme HO-1 exerts neuroprotective effect by protecting dopaminergic neurons [66][67][68][69] and might characterize the antidepressant mechanisms [32,49,[70][71][72]. Previous study also reveal that fluoxetine, an antidepressant of the SSRIs class, attenuates brain injury via enhancement of HO-1 expression [73]. ERK and JNK pathways are also associated with various plasticity processes including neurogenesis, axonal growth, and regulation of BDNF levels [74]. Alteration of the MAPK pathways such as ERK expression has been studied in post-mortem samples of depression patients and animal models [75][76][77]. Furthermore, previous report also shown that there is different potency of ERK activation between desipramine and fluoxetine [78]. In present study, the enhancement of HO-1 expression by desipramine could be regulated by ERK and JNK, but the effect of fluoxetine was only regulated by JNK. Moreover, it has also been reported that the difference in  potency between desipramine and fluoxetine exerts in many functions, such as inhibition of chemotaxis on polymorphonuclear leukocytes [79], acute hypnotic/sedative effect of ethanol [80] and development of ethanol-induced behavioral sensitization [81]. Our results reveal that desipramine-enhanced HO-1 expression sustains to 24 h, however, fluoxetine-increased HO-1 expression reaches a peak at 8 h. Our results and previous reports indicate that different types of antidepressant exert different molecular mechanisms which could provide a novel view for clinical drug choice.
It has been reported that Nrf2 may be a new drug target of treating depression [82]. Nrf2 is known as an important transcription factor involves in antioxidant response, binding to antioxidant response elements (ARE) encoding detoxification enzymes such as HO-1 [83]. Moreover, it has also been reported that activation of Nrf2 provides the insight of treatment of depression and neurodegenerative disease [84,85]. Previous reports have shown that ERK and JNK pathways are involved in the Nrf2 activation and HO-1 up-regulation in various cell types [86,87]. Here, our study also showed that desipramine increases HO-1 expression through ERK and JNK pathways, leading to Nrf2 activation. In line with our results, we also found that fluoxetine-induced HO-1 up-regulation is mediated by ERK and JNK pathways in Mes23.5 dopaminergic neurons. Our results demonstrated that pretreatment of desipramine protects neuronal cell death in dopaminergic neurons through HO-1 up-regulation. The neuroprotective effect of desipramine exerts only by pretreatment, but not co-treatment or posttreatment with the neurotoxin. It suggests that desipramine-increased HO-1 expression and protecting neuronal cell death might require a de novo synthesis pathway. Moreover, our study also showed that HO-1 inducer and activator effectively protect dopaminergic neurons from cell death. These findings indicate that when neuronal cells have adequate or enhanced levels of anti-oxidative agents, they can protect themselves from oxidative damage. Neuroprotection of dopaminergic neurons by the activation of Nrf2 in PD is also regarded as a promising therapeutic approach.
In conclusion, our study elucidates the neuroprotection effects of desipramine and the regulatory molecular mechanisms of desipramine-induced HO-1 expression through the ERK and JNK signaling pathways in Mes23.5 dopaminergic neurons. This is the first study reporting antidepressant against neuronal cell death. Connecting the evidence pointing to depression in PD and clinical studies will support exploring the novel therapeutic potential of desipramine and fluoxetine in treating neurodegeneration.