Accepted Manuscript Pre-treatment with amitriptyline causes epigenetic up-regulation of neuroprotection-associated genes and has anti-apoptotic effects in mouse neuronal cells

Nguyen Quoc Vuong Tran, An Nghia Nguyen, Kyoko Takabe, Zentaro Yamagata, Kunio Miyake



PII:
DOI: Reference:

To appear in:

Received date: Revised date: Accepted date:

S0892-0362(16)30114-3
doi: 10.1016/j.ntt.2017.05.002 NTT 6700

Neurotoxicology and Teratology

17 October 2016 9 May 2017
11 May 2017

Please cite this article as: Nguyen Quoc Vuong Tran, An Nghia Nguyen, Kyoko Takabe, Zentaro Yamagata, Kunio Miyake , Pre-treatment with amitriptyline causes epigenetic up- regulation of neuroprotection-associated genes and has anti-apoptotic effects in mouse neuronal cells, Neurotoxicology and Teratology (2017), doi: 10.1016/j.ntt.2017.05.002

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Pre-treatment with amitriptyline causes epigenetic up-regulation of neuroprotection-associated genes and has anti-apoptotic effects in mouse neuronal cells

Nguyen Quoc Vuong Tran1, An Nghia Nguyen1, Kyoko Takabe, Zentaro Yamagata, Kunio Miyake*

Department of Health Sciences, Graduate School of Interdisciplinary Research, University of Yamanashi, 1110, Shimokato, Chuo, Yamanashi 409-3898, Japan

1These authors contributed equally to this work. *Corresponding author:
Kunio Miyake
Tel: +81-55-273-9557

Fax: +81-550273-9561
E-mail: kmiyake@yamanashi.ac.jp (K. Miyake)

1

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

ABSTRACT
Antidepressants, such as imipramine and fluoxetine, are known to alter gene expression patterns by inducing changes in the epigenetic status of neuronal cells. There is also some evidence for the anti-apoptotic effect of various groups of antidepressants; however, this effect is complicated and cell-type dependent. Antidepressants of the tricyclic group, in particular amitriptyline, have been suggested to be beneficial in the treatment of neurodegenerative disorders. We examined whether amitriptyline exerts an anti-apoptotic effect via epigenetic mechanisms. Using DNA microarray, we analyzed global gene expression in mouse primary cultured neocortical neurons after treatment with amitriptyline and imipramine. The neuroprotection-associated genes, activating transcription factor 3 (Atf3) and heme oxygenase 1 (Hmox1), were up-regulated at both mRNA and protein levels by treatment with amitriptyline. Quantitative chromatin immunoprecipitation assay revealed that amitriptyline increased enrichments of trimethylation of histone H3 lysine 4 in the promoter regions of Atf3 and Hmox1 and acetylation of histone H3 lysine 9 in the promoter regions of Atf3, which indicate an active epigenetic status. Amitriptyline pre-treatment attenuated 1-methyl-4-phenylpyridinium ion (MPP+)- or amyloid β peptide 1-42 (Aβ1- 42)-induced neuronal cell death and inhibited the activation of extracellular signal-regulated kinase 1 and 2 (ERK1/2). We found that Atf3 and Hmox1 were also up-regulated after Aβ1-42 treatment, and were further increased when pre-treated with amitriptyline. Interestingly, the highest up-regulation of Atf3 and Hmox1, at least at mRNA level, was observed after co-treatment with Aβ1-42 and amitriptyline, together with the loss of the neuroprotective effect. These findings suggest preconditioning and neuroprotective effects of amitriptyline; however, further investigations are needed for clarifying the contribution of epigenetic up-regulation of Atf3 and Hmox1 genes.

Abbreviations: AD, Alzheimer’s disease; Aβ, amyloid β; BDNF, brain-derived neurotrophic factor; CaMKII, calmodulin-dependent protein kinase II alpha; CBP, CREB-binding protein; CREB, cAMP response element- binding protein; DNMT, DNA methyl transferase; ERK1/2, extracellular signal-regulated kinase 1 and 2; GEO, Gene Expression Omnibus; HDAC, histone deacetylase; JNK, c-Jun N-terminal kinase; LDH, lactate dehydrogenase; MAPK, mitogen-activated protein kinase; MPP+, 1-methyl-4-phenylpyridinium ion; NAc, nucleus accumbens; PD, Parkinson’s disease; p-ERK1/2, phosphorylated ERK1/2; qChIP, quantitative chromatin immunoprecipitation; qRT-PCR, quantitative reverse transcription PCR; SNRI, serotonin-norepinephrine reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor; TCA, tricyclic antidepressant.

Keywords: Antidepressant; Amitriptyline; Neuroprotection; Neurodegenerative disorder; Epigenetic modification.

2

ACCEPTED MANUSCRIPT

1. Introduction

ACCEPTED MANUSCRIPT

Antidepressants have been used to treat depression for more than 60 years, and were used originally because they increased serotonin and norepinephrine. Thus, the most frequently used antidepressants belong to groups of selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors (SNRIs), and tricyclic antidepressants (TCAs). Depressive disorders are commonly seen in patients with neurodegenerative diseases; hence, antidepressants have been widely used and have improved the quality of life of patients with Alzheimer’s disease (AD) and Parkinson’s disease (PD) (McDonald et al., 2003; Moretti et al., 2002). In addition, advanced studies of the pharmacology of antidepressants suggest that antidepressants may have the capability to modify epigenetic status. Imipramine, a TCA, restores the impaired epigenetic state in depression by inhibiting HDAC5, which leads to an increase in histone acetylation and the subsequent expression of the brain-derived neurotrophic factor (Bdnf) gene in the hippocampal region in the mouse brain (Tsankova et al., 2006). Fluoxetine, an SSRI, is also known to have an epigenetic effect through a decrease in histone acetylation and an increase in histone H3 lysine 9 dimethylation in the calmodulin-dependent protein kinase II alpha (CaMKII) gene in the nucleus accumbens (NAc), a brain reward region in the human and mouse brain (Robison et al., 2014). Epigenetic mechanisms are an important aspect of gene regulation that maintains normal brain function. Particularly, there are evidences that the pathology of AD and PD may relate to epigenetic mechanisms by alternating DNA methylation and histone acetylation (Chen et al., 2009; Kontopoulos et al., 2006). More recent studies have identified an abnormal epigenetic status in the post-mortem brains of patients with mental and neurodegenerative diseases, including AD and PD (Feng et al., 2015; Mastroeni et al., 2015). In addition to these findings, HDAC inhibitors have reduced cell death and improved outcomes in mouse models of AD and PD (Coppede, 2014), suggesting that epigenetic drugs are a potential treatment for these diseases. Therefore, the epigenetic function of antidepressants requires exploration to determine whether it is a general characteristic of antidepressants. Moreover, inhibition of apoptosis increases neural degeneration outcomes (Ghavami et al., 2014), an indication for the use of drugs with neuroprotective effects in the treatment of neurodegenerative disorders. Antidepressants have been shown to protect against hippocampal volume loss in patients with depression, suggesting their neuroprotective effect (Sheline et al., 2003). Antidepressants, particularly fluoxetine, tianeptine, moclobemide, and imipramine, belong to different groups that have been proven to have anti-apoptotic effects in neuronal culture or neural stem cells (Drzyzga et al., 2009; Jantas et al., 2014). Nevertheless, the effect of antidepressants on apoptosis is multifarious and cell-type dependent (Jantas et al., 2014; Xia et al., 1999).

Amitriptyline, one of the earliest TCAs that has been in use, continues to be used in therapy for a number of mental disorders, as well as for neuropathic pain (Leucht et al., 2012; Moore et al., 2015); it appears to be highly effective compared with newer SSRIs (Anderson, 2000). In addition, in patients with depression, serum BDNF

3

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

concentration was increased by 13% after treatment with amitriptyline, but decreased by 12% after treatment with paroxetine, an SSRI (Hellweg et al., 2008). One study reported that amitriptyline induced global CCpGG hypomethylation in rat primary astrocytes, although no changes in specific gene expressions were observed (Perisic et al., 2010); the study suggested further investigation into epigenetic alterations by amitriptyline. Amitriptyline also appears to have an anti-apoptotic effect by preventing PC12 cells from cell death induced by hydrogen peroxide (Kolla et al., 2005) and by attenuating DNA damage induced by neurotoxins, N-(2- chloroethyl)-N-ehtyl-2-bromobenzylamine (DSP4) and camptothecin (CPT), in SH-SY5Y cells (Wang et al., 2015). Jang et al. (2009) demonstrated that amitriptyline, but not other TCAs (imipramine and clomipramine), protected primary cultured hippocampal neurons and in vivo hippocampal neurons from oxygen-glucose deprivation- and kainic acid-induced apoptosis, respectively; and that amitriptyline could independently activate tropomyosin receptor kinase A (TrKA) and TrKB and provoke their dimerization, which could not be seen with other TCAs (desipramine, imipramine, and trimipramine) or fluoxetine. Moreover, TCAs, particularly amitriptyline, have been shown to play a role in the treatment of neurodegenerative disorders. In a mouse model of AD, amitriptyline treatment significantly improved long- and short-term memory and increased neurogenesis and neurosynaptic marker proteins (Chadwick et al., 2011). In PD patients, nortriptyline was shown to be superior in effectiveness when compared to paroxetine or placebo in a randomized controlled trial (Menza et al., 2009). Furthermore, a retrospective cohort study pointed out that amitriptyline delayed the initiation of dopaminergic therapy in early PD, although this may possibly result from the increase in quality of life of patients with depression, or from the neuroprotective effect of the drug itself (Paumier et al., 2012). Later, the same group showed that in 6- hydroxydopamine model rats, pre-treatment with amitriptyline prevented tyrosine hydroxylase positive and NeuN positive cells from cell loss (Paumier et al., 2015). Amitriptyline and tranylcypromine (a monoamine oxidase inhibitor) protected PC12 from cell death induced by 1-methyl-4-phenylpyridinum (MPP+); in contrast, fluoxetine increased the toxicity of MPP+ (Han and Lee, 2009). Recently, a study in rats treated with rotenone as a model of PD showed that pre-treatment with amitriptyline and imipramine improved motor performance and coordination and attenuated the noxious effects of rotenone (Kandil et al., 2016). Taken together, these data support that amitriptyline may have neuroprotective effects besides its well-known anti-depressive effects. However, as mentioned above, the neuroprotective effects of amitriptyline in neuronal cells need to be verified. Moreover, it is unclear whether amitriptyline has epigenetic effects and whether its neuroprotective effects occur via epigenetic mechanisms.

In the present study, we sought to identify amitriptyline’s epigenetic mechanism in gene regulation and to support its neuroprotective effects. First, we examined changes in gene expression after TCA exposure (imipramine and amitriptyline). Interestingly, we found that genes associated with anti-apoptotic action are commonly up-regulated by TCAs; therefore, we selected them and screened them for epigenetic changes. Lastly,

4

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

we used MPP+ and amyloid β peptide 1-42 (Aβ1-42) as a model of neurotoxicity in PD and AD and confirmed the anti-apoptotic effect of amitriptyline. These analyses indicated that amitriptyline might have an epigenetic-related anti-apoptotic action in neuronal cells.

2. Materials and Methods

2.1. Cell culture and treatment

Primary cultures of neocortical neuronal cells were prepared from ICR (CD-1) mouse fetuses on embryonic day 15. Briefly, neocortices were carefully dissected, minced, and digested with 0.25% trypsin. Isolated cells were suspended in Neurobasal medium (Life Technologies Ltd., Carlsbad, CA, USA) supplemented with 1% L- glutamine, penicillin/streptomycin, and B27 serum-free supplement (Life Technologies), and plated at a density of 2.5 x 105 cells/cm2 on poly-L-ornithine-coated plastic dishes. This method provided more than 95% of cells positive for microtubule-associated protein 2 (MAP2), a general marker for neurons, as confirmed by immunofluorescence (Supplementary Fig. 1). These cultures were maintained at 37° C in a humidified 5% CO2 atmosphere. Our experimental protocols were approved by the Institute Animal Care and Use Committee of the University of Yamanashi.

2.2. Drugs and treatments

The antidepressant drugs used in this study, imipramine and amitriptyline, were obtained from Sigma Aldrich, St. Louis, MO, USA and prepared at a concentration of 100 mM in distilled water for stocking. On the day of treatment, drugs were diluted to the desired concentration in culture medium. Drug concentrations were selected following a preliminary 3-[4,5-dimethyl-thiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay to identify the maximum dose that did not significantly reduce cell survival rates after 48 h (Supplementary Fig. 2). First, we evaluated the effects of 48 h exposure to antidepressant drugs in cultured cells. Briefly, cultured neuronal cells at in vitro day 3 (DIV3) were treated with 5 μM imipramine or 5 μM amitriptyline. The cultures were harvested after 48 h (DIV5) for further investigation. We used Aβ1-42 (Abcam, Tokyo, Japan) and MPP+ (Sigma Aldrich) to induce apoptosis in neuronal cell cultures. Stocking solution was prepared in line with product protocols. Treatment solution was prepared before each experiment by diluting in Neurobasal medium supplemented with 1% L-glutamine, penicillin/streptomycin, and B27 serum-free supplement minus AO (Life Technologies). In co-treatment experiments, 5 μM amitriptyline was added to Aβ1-42 treatment solution and treated to cultured neuronal cells at DIV5 for 24 h.

5

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

HDAC inhibitor SAHA was obtained from Histone Deacetylase (HDAC) Inhibitor Set II (Sigma-Aldrich). 10 mM stocking solution was prepared in DMSO and stored at -20oC. The drug was diluted to the desired concentration just before treatment. Cells grown in 96-well plates were treated with 0.5 or 1 μM SAHA, or a combination of 5 μM amitriptyline and 0.25, 0.5, or 1 μM SAHA for 48 h. After removal of the drug, cells were treated with 5 or 10 μM Aβ1-42 for 24 h. MTT assay was performed as described.

2.3. Microarray analysis

RNA was isolated from amitriptyline- and imipramine-treated cells and from untreated cells using High Pure RNA Isolation Kit (Roche Applied Science, Manheim, Germany) following the manufacturer’s protocol. Total RNA (100 ng) was amplified, converted to double-stranded cDNA, and labeled with biotin according to the manufacturer’s protocol (Affymetrix, Santa Clara, CA, USA). Biotinylated cDNA was hybridized to Mouse Gene 2.0 ST chips (Affymetrix). Gene chips were scanned using an Affymetrix Scanner 3000. All probe sets were normalized by using the RMA algorithm on the Affymetrix Expression Console. The data analysis was performed using Subio Platform version 1.19 (Subio Inc., Kagoshima, Japan). The statistical significance of the differences between the subjects and the controls was evaluated using the cross-gene-error model in combination with one- way ANOVA (P < 0.05) and Bonferroni’s multiple testing correction. The data have been deposited in the Gene Expression Omnibus (GEO) database under accession number GSE63005.

2.4. Quantitative reverse transcription PCR

Quantitative reverse transcription PCR (qRT-PCR) was performed using a StepOnePlus real time PCR system (Applied Biosystems, CA, USA). cDNA was prepared using a Reverse Transcription Kit (Applied Biosystems) according to the manufacturer’s instructions. Real-time PCR was then performed with Fast SYBR Green Master Mix (Applied Biosystems). Primer sequences are described in Supplementary Table 1. Gene expression levels were calculated using the comparative CT method after normalization against β-actin.

2.5. Western blotting

Primary cultures of neocortical neuronal cells were lysed in buffer containing 50 mM Tris, 25 mM KCl, 5 mM MgCl2, 250 mM sucrose, 5% Triton X-100, and a protease inhibitor cocktail or phosphatase inhibitor cocktail (both from Nacalai Tesque Inc., Kyoto, Japan). Ten-microgram protein extracts were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride membranes (EMD Millipore, Billerica, MA, USA). Membranes were blocked using Blocking One P solution (Nacalai Tesque Inc.,

6

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Kyoto, Japan) for 40 min in order to be used with phospho-specific antibodies. Membranes were incubated overnight at 4° C with one of the following primary antibodies: rabbit anti-Hmox1 (diluted 1:200, sc-10789, Santa Cruz Biotechnology, Dallas, TX, USA); rabbit anti-Atf3 (diluted 1:500, ab180842, Abcam); goat anti-Gdf15 (diluted 1:1000, ab39999, Abcam); rabbit anti-cleaved-caspase 3 (diluted 1:1000. #9664S); rabbit anti-ERK1/2 (diluted 1:1000; #4695S); rabbit anti-p-ERK1/2 (diluted 1:2000, #4370S); rabbit anti-CREB (diluted 1:1000, #4820S); rabbit anti-p-CREB (diluted 1:1000, #9198S); or mouse anti-β-actin (diluted 1:1000, #3700S) (all from Cell Signaling Technology Inc., Danvers, MA, USA). We then incubated the membranes with the appropriate secondary horseradish peroxidase-linked antibodies. Immunoreactivity was visualized by a chemiluminescence detection system (Nacalai Tesque) and the images were analyzed using LAS-4000 software (Fujifilm Inc., Tokyo, Japan). Staining intensities were normalized as a ratio relative to β-actin. For relative ratio of p-ERK/ERK, the quantification of ERK1/2 and pERK1/2 was first normalized against beta-actin, after which the normalized values were used for quantification of p-ERK/ERK. Representative bands for each protein were selected from one single membrane per experiment; however, for simplicity, we cropped the unused lanes, as indicated by black vertical lines in the figures.

2.6. Quantitative chromatin immunoprecipitation assay

Quantitative chromatin immunoprecipitation (qChIP) assays were performed using ChIP-IT Express Enzymatic Kits (Active Motif, Carlsbad, CA, USA) according to the manufacturer’s instructions. In brief, primary cultured neocortical neuronal cells were fixed with 1% formaldehyde for 10 min at room temperature to form DNA-protein cross-links. Fixation was stopped by the addition of glycine. Cells were collected and re-suspended in lysis buffer, then homogenized on ice using a dounce homogenizer. After centrifugation for 10 min at 5000 rpm, the pellets were re-suspended in digestion buffer for 5 min at 37° C, and the chromatin was sheared to an average length of 200−500 base pairs. Sheared chromatin was incubated with rotation at 4° C overnight in the presence of 5 μg mouse anti-acetyl-H3K9 (Cat#. 61251, Active Motif), 10 μg rabbit anti-acetyl-H3K27 (#4353S, Cell Signaling Technology), or 5 μg rabbit anti-tri-methyl-H3K4 (#9727S, Cell Signaling Technology), or with IgG (Jackson ImmunoResearch, West Grove, PA, USA) and protein G beads. The beads were then washed and eluted. Cross-links were reversed by heating at 95° C for 15 min. After treatment with proteinase K for 1 h at 37° C, the reaction was stopped using a Proteinase K Stop Solution. Precipitated DNA was analyzed by real-time PCR with primer pairs corresponding to the upstream and downstream regions of mouse Atf3 and Hmox1. Primer sequences are described in Supplementary Table 1.

7

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

2.7. MTT and lactate dehydrogenase (LDH)-release assays

Apoptosis was induced in neuronal cell cultures using either MPP+ or Aβ1-42. Cultured neuronal cells at DIV3 were treated with 5 μM amitriptyline for 48 h. After removal of the drug, the cultures were maintained in medium without antioxidants with either MPP+ (50 or 200 μM) for 48 h or with Aβ1-42 (5 or 10 μM) for 24 h. Control group was maintained in medium without antioxidants. Cell viability in all treatment groups was examined by MTT reduction assays as described previously (Miyake and Nagai, 2009). Cell death rates were determined by LDH- release assays using Cytotoxicity Detection KitPLUS (Roche Ltd, Basel, Switzerland) in accordance with the manufacturer’s instructions. For the MTT assay, the untreated control was set as 100% survival. For the LDH- release assay, the untreated control was set as 0% dead and the positive control was set as 100% dead.

2.8. Immunocytochemistry

Cells used for immunostaining were washed 2 times with PBS, and then fixed in 4% paraformaldehyde. After fixing, cells were washed 2 times with PBS and permeabilized using 0.1% Triton X-100 in PBS. 1% skim milk in PBS was used to block unspecific binding. The primary antibodies were diluted in PBS, added to cells after removing the blocking solution, and incubated overnight at 4oC. The following primary antibodies were used: rabbit anti-MAP2 (diluted 1:200, #8707S, CST) and mouse anti-glial fibrillary acidic protein (anti-GFAP) (diluted 1:200, MAB3402, Millipore). The next day, cells were rinsed 5 times with PBS, and proper fluorophore-conjugated secondary antibodies were diluted 1:1000 in blocking solution and added to cells. Cells were incubated with secondary antibodies for 1 h at room temperature, and then washed again 5 times with PBS. After the last washing, Hoechst 33342 (Dojindo) was diluted in distilled water, added to cells, and incubated at room temperature for 15 min. After incubating, Hoechst 33342 was replaced by PBS. Images were acquired using EVOS® FL Cell Imaging System (Life Technologies).

2.9. Statistical analysis

All values are expressed as mean ± standard error of the mean (SEM) for n independent observations. Comparisons of multiple groups were performed with one-way analysis of variance (ANOVA), followed by Tukey’s post hoc test or Dunnett’s T3 test. Differences were considered significant at P ≤ 0.05.

3. Results

3.1. Identification of genes whose expression is commonly altered by antidepressants

8

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Our first goal was to identify genes whose expression was altered by the tested antidepressants, namely, imipramine and amitriptyline (TCAs), because such genes potentially could be associated with the therapeutic mechanisms of the antidepressants. For this purpose, we considered that 1 μM and 10 μM concentrations of amitriptyline are at plasma therapeutic level and cumulative level in the brain, respectively (Perisic et al., 2010) and are sufficient to induce changes in the mRNA level of neuroprotective enzyme superoxide dismutase (Li et al., 2000). We treated mouse primary neocortices neuron cultures with 5 μM amitriptyline and confirmed with MTT assay that this concentration did not induce cell death (Supplementary Fig. 2). Next, we performed a DNA microarray assay and a based-cluster analysis. The expression profiles of the two TCAs showed some similarity (Fig. 1A).

From the comparative analysis, we selected genes showing at least 1.5-fold higher or lower fold changes in expression between the treated and non-treated cells. In total, 60 and 43 genes were up-regulated by amitriptyline and imipramine, respectively (Fig. 1B). Of these, 17 genes were up-regulated by both antidepressants (Supplementary Table 2). We identified 74 and 45 genes that were down-regulated by amitriptyline and imipramine, respectively (Fig. 1C). Of these, 27 genes were down-regulated by both antidepressants (Supplementary Table 3).

Fig. 1. Global gene expression patterns in primary cultures of mouse neuronal cells treated with amitriptyline (Ami) or imipramine (Imi). (A) Hierarchical clustergram of transcription profiles. Red: higher gene expression. Blue: lower gene expression. Venn diagram of the genes that were (B) up-regulated or (C) down-regulated by antidepressant treatment. Cultured neuronal cells at DIV3 were treated with 5 μM imipramine or 5 μM amitriptyline for 48 h.

3.2. Up-regulation of neuroprotection-associated genes

Of the 17 genes that were up-regulated by both imipramine and amitriptyline, we selected three genes, activating transcription factor 3 (Atf3), heme oxygenase 1 (Hmox1), and growth differentiation factor 15 (Gdf15), which are known to have neuroprotective actions (Hettiarachchi et al., 2014; Liang et al., 2009; Strelau et al.,



9

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

2000). We performed qRT-PCR and western blot to confirm the up-regulation of these three genes by both antidepressants at both mRNA and protein levels. Result analyses demonstrated that, at both mRNA and protein levels, Atf3 was significantly up-regulated solely by amitriptyline (Figs. 2A and 2C), while Hmox1 was significantly up-regulated by both imipramine and amitriptyline (Figs. 2B and 2D).

However, for Gdf15, although both antidepressants up-regulated at mRNA level, none was significantly up- regulated at protein level (Supplementary Fig. 3). Therefore, we excluded Gdf15 from further investigation.

Fig. 2. qRT-PCR assay and western blots of Atf3 and Hmox1 in mouse neuronal cells treated with 5 μM imipramine or 5 μM amitriptyline for 48 h. (A and B) qRT-PCR analysis of Atf3 and Hmox1. (C and D) Western blot of Atf3 and Hmox1. Bars represent the relative ratios of treated to untreated controls. Data are based on three independent experiments. *P < 0.05; **P < 0.01 versus control.

3.3. Amitriptyline-induced active histone modifications in the Atf3 and Hmox1 promoter regions

As imipramine is known to induce epigenetic changes by increasing histone acetylation, we asked whether the up-regulation of Atf3 and Hmox1 by amitriptyline was related to changes in histone modification. To screen for changes to active histone modifications in the promoter regions of the up-regulated genes, we performed ChIP against histone H3 lysine 4 trimethylation (H3K4me3), H3K9 acetylation (H3K9ac), or H3K27 acetylation (H3K27ac), followed by qPCR. These analyses showed that amitriptyline significantly increased the level of histone H3K4me3 and H3K9ac within the Atf3 promoter region encompassing the MEF2 and CRE binding sites (Zhang et al., 2011) (Figs. 3A and 3B). ChIP qPCR for Hmox1 promoter showed that amitriptyline significantly increased H3K4me3, but not H3K9ac, in the region encompassing the AP-1 and HSF binding sites (Alam and



10

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Cook, 2007) (Figs. 3C and 3D). H3K27ac was not enriched in the promoter region of either gene (Figs. 3B and 3D).

Given that H3K4me3 and H3K9ac are well-defined as histone codes for active promoters and H3K27ac is enriched at active promoters and enhancers, collectively, our results suggest that amitriptyline up-regulated Atf3 and Hmox1 genes by inducing active histone modifications in the promoter regions of Atf3 and Hmox1.

Fig. 3. Changes in histone modifications in the promoter regions of Atf3 and Hmox1. (A) Schematic representation of the upstream region of Atf3 encompassing the MEF2 and CRE binding sites and the TATA box. Numbers (1 to 7) represent the regions assayed by ChIP qPCR. (B) Changes in histone H3K9ac, H3K27ac, and H3K4me3 modifications after treatment with 5 μM amitriptyline for 48 h. (C) Schematic representation of the upstream region of Hmox1 encompassing the AP-1 and HSF binding sites. Numbers (1 to 8) represent the regions assayed by ChIP qPCR. (D) Changes in histone H3K9ac, H3K27ac, and H3K4me3 modifications after treatment with 5 μM amitriptyline for 48 h. Data were normalized over input and based on three independent experiments. *P < 0.05, **P < 0.01 versus control.



11

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

3.4. Neuroprotective action of amitriptyline in primary cultures of mouse neuronal cells

Since amitriptyline induced up-regulation of the neuroprotection-associated Atf3 and Hmox1 genes at both the mRNA and protein levels in mouse primary neocortical neuron culture, we theorized that amitriptyline might have a neuroprotective effect on neuronal cells. We induced cell damage by treating cells with either MPP+ (50 or 200 μM) or Aβ1-42 (5 or 10 μM) for 48 h or 24 h, respectively. Previous studies have shown that at these concentrations, MPP+ had a neurotoxic effect and Aβ1-42 significantly induced apoptosis and increased caspase-3 level in rat neurons (Marini and Nowak, 2000; Wei et al., 2008). Amitriptyline was found to increase cell survival rates slightly but significantly in an MTT assay of cultures treated with 200 μM MPP+ (Fig. 4A) and with 5 or 10 μM Aβ1-42 (Fig. 4C).

Amitriptyline significantly decreased the proportion of dead cells in an LDH-release assay in cultures treated with 50 or 200 μM MPP+ (Fig. 4B), and in cultures treated with 5 or 10 μM Aβ1-42 (Fig. 4E). Furthermore, amitriptyline significantly reduced cleaved caspase 3, a critical marker for neuronal cell death (Ferrer and Planas, 2003), in cultures treated with 200 μM MPP+ (Fig. 4C) or 5 μM Aβ1-42 (Fig. 4F). When results are considered in toto, amitriptyline appeared to exercise epigenetic regulation of neuroprotection-associated genes and exhibited neuroprotective effects at a concentration as low as 5 μM.

12

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT



Fig. 4. Assessment of the neuroprotective effects of amitriptyline in mouse neuronal cells. Cells at DIV3 were treated with 5 μM amitriptyline for 48 h, after removal of the drug, cells were treated with either MPP+ (50 or 200 μM) for 48 h or Aβ1-42 (5 or 10 μM) for 24 h. (A) MTT assay for MPP+-induced cell damage. (B) LDH-release assay for MPP+-induced cell damage. (C) Western blot for MPP+-induced changes in cleaved caspase 3 expression. (D) MTT assay for Aβ1-42 -induced cell damage. (E) LDH-release assay for Aβ1-42-induced cell damage. (F) Western blot for Aβ1-42 -induced changes in cleaved caspase 3 expression. For cell survival and cell death assay, each value represents the mean ± SEM of 12 wells in three independent experiments. For western blot of cleaved caspase 3, each value represents the mean ± SEM in three independent experiments. Bars represent the relative ratios of treated to untreated controls. *P < 0.05; **P < 0.01; ***P < 0.001.

3.5. Amitriptyline inhibited the activation of mitogen-activated protein kinase (MAPK) family members extracellular signal-regulated kinase 1 and 2 (ERK1/2)

MAPK families contain c-Jun N-terminal kinase (JNK), p38-MAP kinase, and ERK1 and 2, which control many intracellular processes, including gene expression, cell growth, cell differentiation, cell survival, and cell death (Chen et al., 2001). MAPK sub-families ERK1/2 have both anti-apoptotic and pro-apoptotic functions. Activation of ERK1/2 may result either in the activation of anti-apoptotic proteins such as cAMP response element-binding protein (CREB) and CREB-binding protein (CBP) and the inhibition of pro-apoptotic proteins

13

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

such as caspase 8 and 9 or lead to the enhancement of pro-apoptotic molecules such as caspase 3 and p53 under some specific statuses (Lu and Xu, 2006). Primary cortical neuron cultures and SH-SY5Y cells exposed to oxidative stress induced by H2O2, ethanol, or Aβ1-42 were reported to have increased phosphorylated ERK1/2 (p- ERK1/2) in cells (Bartov et al., 2006; Liu et al., 2010; Yu et al., 2016). Therefore, we presumed that ERK1/2 and p-ERK1/2 were altered in our experiments. To investigate this presumption, we performed western blot on control and amitriptyline-treated primary neocortices neuron cultures with and without MPP+ or Aβ1-42.

As expected, p-ERK1/2 increased in cells exposed to 200 μM MPP+ (Fig. 5B) or 5 μM Aβ1-42 (Fig. 5E), while total ERK1/2 remained unchanged (Figs. 5A and 5D). Amitriptyline neither altered total ERK1/2 in control and cell damage-induced groups (Figs. 5A and 5D), nor changed p-ERK1/2 in control groups (Figs. 5B and 5E). However, amitriptyline significantly reduced the elevated p-ERK1/2 caused by MPP+ or Aβ1-42 (Figs. 5B and 5D). Analyses of the ratio of p-ERK1/2 to total ERK1/2 also revealed that amitriptyline successfully reduced the p-ERK1/2 to ERK1/2 ratio compared with the control group treated with MPP+ (Fig. 5C) or showing a trend toward significance compared with the control group treated with Aβ1-42 (Fig. 5F). To exclude the possibility that activation of ERK1/2 gave rise to the MAPK anti-apoptotic pathway, we examined changes in total CREB and p-CREB. As predicted, we found no alteration in total CREB or p-CREB between control and cell damage-induced groups (Supplementary Fig. 4).

Taken together, our results demonstrated that MPP+ and Aβ1-42 activated ERK1/2 followed the pro-apoptotic pathway, and consequently increased activation of caspase 3. Further, amitriptyline exercised a neuroprotective effect partly involved in inhibiting ERK1/2 activation and reducing cleaved-caspase 3.

14

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT



Fig. 5. Assessment of the changes in ERK1/2 and p-ERK1/2 in mouse neuronal cells. Cells at DIV3 were treated with 5 μM amitriptyline for 48 h, after removal of the drug, cells were treated with either 200 μM MPP+ for 48 h or 10 μM Aβ1-42 for 24 h. (A and B) Western blot for MPP+-induced changes in ERK1/2 and p-ERK1/2 expressions. (C) Ratio of p-ERK1/2 to ERK1/2 in MPP+-induced groups. (D and E) Western blot for Aβ1-42-induced changes in ERK1/2 and p-ERK1/2 expressions. (F) Ratio of p-ERK1/2 to ERK1/2 in Aβ1-42-induced groups. Each value represents the mean ± SEM in three independent experiments. Bars represent the relative ratios of treated to untreated controls; *P < 0.05; **P < 0.01.

3.6. Co-treatment also up-regulated Atf3 and Hmox1 but not retained neuroprotection

We further asked whether amitriptyline also possessed neuroprotective effect when treated at the same time with Aβ1-42. In MTT assay of co-treatment cultures, amitriptyline failed to increase cell survival rates compared to control (Fig. 6A), suggesting that the effect on cell survival of amitriptyline is related to the changes induced by pre-treatment. For this reason, we wondered how Atf3 and Hmox1 changed followed treatment with Aβ1-42 and co-treatment. To this end, we performed RT-PCR and western blot to further examine the alterations of Atf3 and Hmox1 after Aβ1-42 treatment and after co-treatment with Aβ1-42 and amitriptyline. Amitriptyline slightly but significantly up-regulated Atf3 and Hmox1 at both mRNA and protein levels after 24 h of treatment, which were

15

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

further increased after 48 h (Figs. 6B, 6C, 6D, and 6E). Interestingly, treatment with Aβ1-42 only for 24 h up- regulated Atf3 and Hmox1 similar to the effect of treatment with amitriptyline for 48 h (Figs. 6B, 6C, 6D, and 6E).

When comparing between treatment with Aβ1-42 only, pre-treatment, and co-treatment with 5 μM amitriptyline, we found that, at mRNA level, Atf3 and Hmox1 were significantly up-regulated after pre-treatment with amitriptyline compared to treatment with Aβ1-42 alone. Unexpectedly, the highest increased mRNA level of Atf3 and Hmox1 was observed after co-treatment, which significantly higher than pre-treatment (Figs. 6B and 6C). However, protein level of ATF3 only showed a trend of increasing similarly with mRNA level without significance (Fig. 6D). Western blot assay for HMOX1 revealed that pre-treatment with amitriptyline significantly up-regulated HMOX1 compared to treatment with Aβ1-42 alone. In co-treatment group, HMOX1 was also up-regulated but without significance compared to treatment with Aβ1-42 only (Fig. 6E).



16

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Fig. 6. (A) The neuroprotective effect of amitriptyline was lost when treated at the same time with Aβ1-42 in mouse neuronal cells as assessment by MTT assay. Cells at DIV5 were treated with 5 μM amitriptyline and Aβ1-42 (5 or 10 μM) at the same time (Co-treatment) for 24 h. Each value represents the mean ± SEM of 12 wells in three independent experiments. (B and C) qRT-PCR analysis of Atf3 and Hmox1 before and after Aβ1-42 -induced cell damage. (D and E) Western blot of Atf3 and Hmox1 before and after Aβ1-42 -induced cell damage. Cells were treated with only 5 μM amitriptyline for 24 h or 48 h (Ami 24h and Ami 48h, respectively); or with 5 μM Aβ1-42 with or without pre-treatment with 5 μM amitriptyline; or co-treated with 5 μM amitriptyline and 5 μM Aβ1-42 for 24 h. Each value represents the mean ± SEM in three independent experiments. Bars represent the relative ratios of treated to untreated controls. *P < 0.05, **P < 0.01, ***P < 0.001 versus control; #P < 0.05, ##P < 0.01 versus treated with 5 μM Aβ1-42 only; $P < 0.05, $$PP < 0.01 between pre-treatment and co-treatment with 5 μM amitriptyline.

4. Discussion

In this study, we analyzed changes in the expression of a number of genes, investigated the underlying epigenetic mechanisms for altered gene expression levels following treatment with tricyclic antidepressants, and verified the neuroprotective action of amitriptyline. First, by screening for genes whose expression was altered by imipramine and amitriptyline, we identified 17 up-regulated genes and 27 down-regulated genes that responded to both tested drugs. Next, we confirmed the up-regulation at both mRNA and protein levels of two neuroprotection-related genes, Atf3 and Hmox1, after treatment with amitriptyline, and of Hmox1 after treatment with imipramine. In addition, we found that the changes in Atf3 and Hmox1 expression induced by amitriptyline were associated with epigenetic alterations to their promoter regions. Finally, we demonstrated neuroprotective action of amitriptyline in mouse primary cultured neocortical neuronal cells potentially by lessening ERK1/2 activation.

Hmox1 is known to have neuroprotective effect against various toxic stimuli (Reviewed by Schipper et al., 2009), including Aβ1-42 and MPP+ (Hettiarachchi et al., 2014; Huang and Chuang, 2010; Hung et al., 2008). Overexpression of HMOX1 in rat protected Dopaminergic neurons from MPP+-induced cell death and attenuated asymmetric behavior (Hung et al., 2008). Moreover, HMOX1 overexpressed 24 h before and at the same time with adding MPP+ had higher effectiveness than overexpressed after adding MPP+ (Hung et al., 2008). In primary cortical neurons, up-regulated Hmox1 was also proven to have protective effect possibly by reducing H2O2 level (Huang and Chuang, 2010). Previous study showed that antidepressants, desipramine (a TCA) and fluoxetine, up-regulated Hmox1 by activating nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and MAPK pathways, although the activation of MAPK pathway was only observed up to 2 h after treatment (Lin et al., 2012), which is discussed

17

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

later in this paper; the authors also pointed out that up-regulation of Hmox1 by desipramine helped to protect from neurotoxicity induced by 6-OHDA and rotenone. Another gene we selected for further investigation was Atf3, of which neuroprotection is still controversial. Indeed, overexpression of ATF3 in CA3 hippocampal neurons of rat prior to kainic acid administration significantly reduced apoptotic cell death (Francis et al., 2004). In a stable PC12 cell model of Huntington disease induced by mutant huntingtin (mHTT), overexpression of ATF3 exerted a neuroprotective effect against mHTT while Atf3 knockout enhanced the effects of mHTT (Liang et al., 2009). The evidences above support our hypothesis that amitriptyline up-regulated ATF3 before induced cell death with Aβ1- 42 or MPP+ may aid in amitriptyline-induced neuroprotection. On the contrary, a recent study showed that Atf3 knockdown prevented SH-SY5Y cells from MPP+-induced apoptosis (Zhao et al., 2016), which was also contrast to an earlier study showing that primary neuron cultures from Atf3 knockout mice did not exhibit increased survival compared to cultures from wild-type mice (Bernstein and O’Malley, 2013). We also looked at Gdf15, which was reported to rescue phenotype and have neuroprotection effect on dopaminergic neurons from 6OHDA-induced neurotoxicity (Strelau et al., 2000). However, we observed a discrepancy between Gdf15 mRNA and protein level. While treatment with imipramine and amitriptyline increased Gdf15 mRNA significantly compared to control, our results showed only a trend toward increasing GDF15 protein after treatment with these two antidepressants. The translation from mRNAs to proteins is under the control of complicated post- transcriptional and -translational mechanisms. In fact, a previous study showed that only 44% of changes in protein level can be expected from mRNA level (Schwanhausser et al., 2011). Moreover, it is suggested that, the correlation between mRNA and protein levels remarkably change over time when cells are under stimulation (Cheng et al., 2016). Thus, in our study, one possible explanation is that we only examined genes’ expression for only one timepoint, 48 h after treatment with antidepressants, which may not be enough for the increase of GDF15 protein level.

We found that Atf3 and Hmox1 were up-regulated by treatment with 5 μM Aβ1-42 alone in a same manner with pre-treatment with amitriptyline for 48 h, further increased by pre-treatment with amitriptyline, but unexpectedly, even further increased, at least at mRNA level, in co-treated cells. A possible explanation is that co-treatment suddenly up-regulated Atf3 and Hmox1 up to a very high level, which made cells unstable and led to the loss of neuroprotective effect. In addition, earlier studies have highlighted the neuroprotection of preconditioning, mainly in the context of ischemia and injury (recently reviewed by Li et al., 2017). Noticeably, Atf3 and Hmox1 were implicated to participate in the effect of preconditioning. Although studies on ischemic preconditioning of the heart have pointed out changes in Atf3 expression, location, and activation in cardioprotective effect (Brooks et al., 2014; Liu et al., 2009), as far as we can tell, there are not many studies focus on Atf3 in brain preconditioning. Atf3 was found to be one of the early response genes that up-regulated after one hour of hypoxia preconditioning (Tang et al., 2006). Thus, together with the controversial findings in

18

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

neuroprotective effect of Atf3 mentioned above, further studies on gene knockout as well as overexpression are needed to elucidate the mechanisms and confirm the roles of Atf3 in preconditioning and neuroprotection. In contrast, many studies have subjected the up-regulation of Hmox1 as an important and common event in preconditioning by different factors, for instance, hypoxia (Shu et al., 2016), sublethal ischemia (Lee et al., 2015a), ginkgo biloba (Nada and Shah, 2012), or helium (Li et al., 2016). Furthermore, our observation was similar to studies that investigated Hmox1 after cell damage initiation followed preconditioning, where the results showed a higher increase compared to preconditioning or cell damage initiation alone (Lee et al., 2015a; Nada and Shah, 2012; Shu et al., 2016). Taken with the loss of neuroprotective effect in co-treatment, our results suggested that pre-treatment with amitriptyline may bring about preconditioning effect that increased tolerance of primary neuron cultures to Aβ1-42-induced neurotoxicity.

According to our results, amitriptyline induced active epigenetic changes, including H3K4me3 and H3K9ac but not H3K27ac, in promoters of Atf3 and Hmox1 of neurons. Previous study has provided evidence that amitriptyline induced minor cytosine demethylation by reducing DNA methyl transferase (DMNT) activities but did not cause any changes in histone acetylation and astrocyte-specific genes in rat primary astrocytes (Perisic et al., 2010). To our knowledge, this is the first report that amitriptyline alters histone modifications and consequently changes gene expression in neurons. However, the mechanism is still unclear; future studies may usefully focus on histone methyltransferases and histone demethylases after antidepressant exposure. Moreover, an earlier study of myeloma cells has shown that amitriptyline inhibited several HDACs and increased histone H3 acetylation in tumor suppressor genes p21 and p27 (Mao et al., 2011); thus, amitriptyline acted as an HDAC inhibitor in myeloma cells. In line with this earlier study, our results suggest that amitriptyline may have HDAC inhibitor properties. However, this property of amitriptyline may be cell-type dependent and warrants further investigation. Amitriptyline is used as an antidepressant as well as being the first line of treatment for neuropathic pain, a common problem in clinical practice that results from many peripheral and central nervous system diseases including diabetic neuropathy and cancer (Moore et al., 2015). Therefore, further studies are needed to compare the alterations in HDAC activities and HDAC expressions in several cell types to understand more about this ability of amitriptyline.

Amitriptyline is known to inhibit neurotransmitter reuptake, and thereby reverse serotonin and norepinephrine accumulation in synapses in patients with depression (Guaiana et al., 2007). It is also reported that amitriptyline has ability as a sodium channel blockade, which may involve in the effect of a painkiller and migraine prophylaxis of amitriptyline (Yan et al., 2010). Additionally, amitriptyline has a neuroprotective effect by countering mental stress-induced down-regulation of anti-apoptotic genes through the up-regulation of nerve growth factor 1A, mineralocorticoid receptors, and glucocorticoid receptors (Johansson et al., 1998; Okugawa et al., 1999). Together with these findings, our study suggests that epigenetic up-regulation of anti-apoptotic genes,

19

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

such as Atf3 and Hmox1, may contribute to the neuroprotective effect of amitriptyline. It is worth mentioning that HDAC inhibitors have been proven to improve cell viability and inhibit the activation of caspase 3 in SK-N-SH and MES 23.5 cells treated with MPP+ (Kidd and Schneider, 2010). We also showed that suberoylanilide hydroxamic acid (SAHA), also known as vorinostat, improved cell viability in primary neurons treated with Aβ1-42 (Supplementary Fig. 5A). In our result, SAHA showed a higher neuroprotective effect compared to amitriptyline. Moreover, in control cells, treated with 1 μM SAHA showed a significant increase in MTT assay compare to untreated cells (Supplementary Fig. 5A). These observations could be explained by the fact that, in addition to its innate neuroprotection against oxidative stress (Ryu et al., 2003), SAHA was also reported to protect neurons against the gradual shrinkage in primary cultures (Chen et al., 2012). The neuroprotective effect of SAHA may rely on glial cells (Chen et al., 2012), which were lacked in our experiments, thus, the neuroprotection of SAHA may not be at full potential, resulted in similar effect between SAHA 0.5 and 1 μM. However, combination of amitriptyline and SAHA exhibited neurotoxicity in a dose dependent manner of SAHA (Supplementary Fig. 5B). SAHA is known as a pan-HDAC inhibitor that targets a range of HDACs class I and II. At high concentration, HDAC inhibitors, particularly SAHA at 2.5 μM, may resulted in aberrant gene expression and induced cell death (Peart et al., 2005; Rouaux et al., 2004). Furthermore, neuronal cells were suggested to be more susceptibility to changes in acetylation level and both hyper- and hypo-methylation may induced cell death (Rouaux et al., 2004). Therefore, we supposed that SAHA and amitriptyline, when combined together, may exceedingly perturb gene expression and induced cell damage. In the present study, amitriptyline showed neuroprotective effects at the concentration examined for epigenetic changes. This concentration is lower than that reported in a previous study of PC12 cells, which used 50 μM and 100 μM (Kolla et al., 2005). As mentioned earlier in this discussion, the neuroprotective effect of antidepressants, particularly amitriptyline, is presumed to depend on cell types and doses. Recently, amitriptyline has been shown to diminish lidocaine-induced rat dorsal root ganglion neurons in a dose-dependent manner up to 10 μM, but exerted toxicity at higher concentrations (Zheng et al., 2016). Previous studies showed that amitriptyline at 50 μM reduced cell viability in colon carcinoma cells (Arimochi and Morita, 2006), and more recently, concentrations at 60 μM and 120 μM caused mitochondria-mediated neurotoxicity and oxidative stress in SH-SY5Y cells (Lee et al., 2015b). Collectively, we hypothesized that amitriptyline when treated in low dose, induced preconditioning of neuronal cells without causing any harm to cells viability and, subsequently, had to neuroprotective effect. On the other hand, when treated with a high dose, amitriptyline resulted in neurotoxicity.

Our detailed results showed that amitriptyline exhibited an anti-apoptotic effect by suppressing the activation of ERK1/2 and its downstream caspase 3. The interplay between amitriptyline and the MAPK pathways seems to be more complex and relate to cell types and cell conditions. In brief, our data identified that amitriptyline inhibits ERK1/2 activation in neurons induced by MPP+ or Aβ1-42. On the other hand, others have

20

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

shown that amitriptyline up-regulates early growth response-1 (Egr-1) and Bdnf mRNA by promoting the activation of ERK1/2 in rat astrocytes and microglia (Chung et al., 2007; Hisaoka-Nakashima et al., 2016). The length of time of exposure to oxidative stress is another concern, as some studies have identified different consequences from transient and sustained ERK1/2 activation (Marshall, 1995; Rybakova et al., 2012). When transiently activated, ERK1/2 assists anti-apoptotic factors and cell survival; however, sustained activation boosts pro-apoptotic factors and leads to cell death. The same complexity in the MAPK pathway involving ERK1/2 was also observed in MPP+- or Aβ1-42-induced neurotoxicology models. Previous studies on PC12 cells showed that treatment with MPP+ for up to 120 min inhibited the activation of ERK1/2 and compounds with the effects of activating ERK1/2 successfully prevented MPP+-induced cell death, which is lost by ERK1/2 inhibition (Di Segni et al., 2006; Hashimoto et al., 2012). In contrast, p-ERK started increasing after 8h treatment with MPP+ and this effect increased overtime up to at least 32 h (Zhu et al., 2007). Moreover, adding MAPK/ERK Inhibitor UO126 before or up to 12 h after MPP+ initiation significantly prevented cell death induced by MPP+, however, if administered after 20 h, the protective effect of UO126 was lost (Zhu et al., 2007). In primary cortical neurons, treatment with MPP+ for 42 h was shown to increase caspase 3 activation, which could not be observed after 6 h treatment (Sun and Chang, 2003), and treatment with MPP+ for 24 h increased p-ERK in a dose dependent manner (Fragkouli and Doxakis, 2014). In our study, the activation of ERK and caspase 3 after treatment with Aβ1-42 was in line with other studies in SH-SY5Y cells, mouse primary cortical neurons and primary neural progenitor cells (Ha et al., 2012; Li and Qian, 2016; Mendell et al., 2016). Moreover, chemicals that can inhibit the activation of ERK1/2 were found to reduce the neurotoxicity of Aβ1-42 (Li and Qian, 2016; Mendell et al., 2016), supporting our finding in amitriptyline. Aβ1-42 was shown to induce phosphorylation of ERK in a dose dependent manner, however, this p-ERK level returned to normal after 60 min of treatment (Young et al., 2009). In contrast to our result, Thangnipon et al. (2013) showed that on rat primary cortical neuron cultures at DIV5, treatment with 10 μM Aβ1-42 did not affect p-ERK but increased activated caspase-3. Collectively, the activation of ERK1/2 and its downstream caspase 3 played important role in the neurotoxicity induced by MPP+ or Aβ1-42. Together with other studies, we exposed primary neuron cultures to MPP+ or Aβ1-42 for 48 h and 24 h, respectively, which would activate ERK1/2 and lead to apoptosis, and showed that pre-treatment with amitriptyline successfully prevented this effect.

Although we have adequately investigated the epigenetic and neuroprotective effects of amitriptyline, there are limitations remaining to be explored. Firstly, in our study, we treated primary neuron cultures at DIV3 with amitriptyline, and after 48 h (DIV5) with Aβ1-42 or MPP+. Therefore, the observation was limited to immature neurons. There are evidences that the neurotoxicity effects of staurosporine, doxorubicin, cyclosporine, N-methyl- D-aspartate, and glutamate are different in primary neuronal cultures at different DIVs. In fact, neurons at earlier stages were more prone to apoptosis induced by staurosporine, doxorubicin, and cyclosporine, while neurons at

21

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

later stages showed higher susceptibility to glutamate and N-methyl-D-aspartate (Jantas-Skotniczna et al., 2006; Jantas et al., 2016; Jantas and Lason, 2009; McDonald et al., 1997). Related to models used in our study, hippocampal neurons at DIV30 treated with Aβ caused neurite degeneration without affecting cell survival (Rapoport and Ferreira, 2000). Thus, the changes in gene expression and the neuroprotective effect of amitriptyline maybe different in mature primary cultures of neurons. Secondly, in our study, we examined the neuroprotection of amitriptyline with only one concentration, which corresponded to the concentration investigated for epigenetic effect. Moreover, as discussed above, the MPP+- or Aβ1-42-induced changes in ERK1/2 were different over time. To comprehensively understand the neuroprotection effects of amitriptyline, further study is needed to investigate in dose- and time-dependent changes in gene expression, neuroprotection, and ERK1/2 pathway activation.

5. Conclusion

In conclusion, our study has provided insights into the epigenetic regulation of the preconditioning and neuroprotective effect of the tricyclic antidepressant amitriptyline in mouse primary cultured neurons. However, the pathways that taken place in amitriptyline’s preconditioning and neuroprotection require further investigations. Moreover, it remains to be determined whether these neuroprotective effects occur in vivo (e.g., in the brains of mice under stress) or in the patients’ neurons. To address this question, we plan to investigate the neuroprotective mechanism of antidepressants using recently established neuronal cells derived from human induced pluripotent stem cells derived from patients with a mental disorder (Andoh-Noda et al., 2015).

Conflicts of interest

There are no conflicts of interest.

Acknowledgments

This study was supported by a Grant-in-Aid for Scientific Research (C) 15K09592 from the Ministry of Education, Culture, Sports, Science, and Technology in Japan. We thank Dr. Takeo Kubota for providing helpful comments. We would like to thank Editage (www.editage.jp) for English language editing.

22

ACCEPTED MANUSCRIPT

References

ACCEPTED MANUSCRIPT

Alam J, Cook JL. 2007. How many transcription factors does it take to turn on the heme oxygenase-1 gene? Am J Respir Cell Mol Biol;36:166-74. http://dx.doi.org/10.1165/rcmb.2006-0340TR

Anderson IM. 2000. Selective serotonin reuptake inhibitors versus tricyclic antidepressants: a meta-analysis of efficacy and tolerability. J Affect Disord;58:19-36. http://www.ncbi.nlm.nih.gov/pubmed/10760555

Andoh-Noda T, Akamatsu W, Miyake K, Matsumoto T, Yamaguchi R, Sanosaka T, et al. 2015. Differentiation of multipotent neural stem cells derived from Rett syndrome patients is biased toward the astrocytic lineage. Mol Brain;8:31. http://dx.doi.org/10.1186/s13041-015-0121-2

Arimochi H, Morita K. 2006. Characterization of cytotoxic actions of tricyclic antidepressants on human HT29 colon carcinoma cells. European journal of pharmacology;541:17-23. http://dx.doi.org/10.1016/j.ejphar.2006.04.053

Bartov O, Sultana R, Butterfield DA, Atlas D. 2006. Low molecular weight thiol amides attenuate MAPK activity and protect primary neurons from Abeta(1-42) toxicity. Brain Res;1069:198-206. http://dx.doi.org/10.1016/j.brainres.2005.10.079

Bernstein AI, O’Malley KL. 2013. MPP+-induces PUMA- and p53-dependent, but ATF3-independent cell death. Toxicology letters;219:93-8. http://dx.doi.org/10.1016/j.toxlet.2013.03.003

Brooks AC, Guo Y, Singh M, McCracken J, Xuan YT, Srivastava S, et al. 2014. Endoplasmic reticulum stress- dependent activation of ATF3 mediates the late phase of ischemic preconditioning. Journal of molecular and cellular cardiology;76:138-47. http://dx.doi.org/10.1016/j.yjmcc.2014.08.011

Chadwick W, Mitchell N, Caroll J, Zhou Y, Park SS, Wang L, et al. 2011. Amitriptyline-mediated cognitive enhancement in aged 3xTg Alzheimer’s disease mice is associated with neurogenesis and neurotrophic activity. PLoS One;6:e21660. http://dx.doi.org/10.1371/journal.pone.0021660

Chen KL, Wang SS, Yang YY, Yuan RY, Chen RM, Hu CJ. 2009. The epigenetic effects of amyloid-beta(1-40) on global DNA and neprilysin genes in murine cerebral endothelial cells. Biochemical and biophysical research communications;378:57-61. http://dx.doi.org/10.1016/j.bbrc.2008.10.173

Chen SH, Wu HM, Ossola B, Schendzielorz N, Wilson BC, Chu CH, et al. 2012. Suberoylanilide hydroxamic acid, a histone deacetylase inhibitor, protects dopaminergic neurons from neurotoxin-induced damage. British journal
of pharmacology;165:494-505. http://dx.doi.org/10.1111/j.1476-5381.2011.01575.x

Chen Z, Gibson TB, Robinson F, Silvestro L, Pearson G, Xu B, et al. 2001. MAP kinases. Chemical reviews;101:2449-76. http://www.ncbi.nlm.nih.gov/pubmed/11749383

Cheng Z, Teo G, Krueger S, Rock TM, Koh HW, Choi H, et al. 2016. Differential dynamics of the mammalian mRNA and protein expression response to misfolding stress. Molecular systems biology;12:855. http://dx.doi.org/10.15252/msb.20156423

Chung EY, Shin SY, Lee YH. 2007. Amitriptyline induces early growth response-1 gene expression via ERK and JNK mitogen-activated protein kinase pathways in rat C6 glial cells. Neuroscience letters;422:43-8. http://dx.doi.org/10.1016/j.neulet.2007.05.057

Coppede F. 2014. The potential of epigenetic therapies in neurodegenerative diseases. Frontiers in genetics;5:220. http://dx.doi.org/10.3389/fgene.2014.00220

23

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Di Segni A, Farin K, Pinkas-Kramarski R. 2006. ErbB4 activation inhibits MPP+-induced cell death in PC12- ErbB4 cells: involvement of PI3K and Erk signaling. Journal of molecular neuroscience : MN;29:257-67. http://dx.doi.org/10.1385/JMN:29:3:257

Drzyzga LR, Marcinowska A, Obuchowicz E. 2009. Antiapoptotic and neurotrophic effects of antidepressants: a review of clinical and experimental studies. Brain research bulletin;79:248-57. http://dx.doi.org/10.1016/j.brainresbull.2009.03.009

Feng Y, Jankovic J, Wu YC. 2015. Epigenetic mechanisms in Parkinson’s disease. Journal of the neurological sciences;349:3-9. http://dx.doi.org/10.1016/j.jns.2014.12.017

Ferrer I, Planas AM. 2003. Signaling of cell death and cell survival following focal cerebral ischemia: life and death struggle in the penumbra. J Neuropathol Exp Neurol;62:329-39. http://www.ncbi.nlm.nih.gov/pubmed/12722825

Fragkouli A, Doxakis E. 2014. miR-7 and miR-153 protect neurons against MPP(+)-induced cell death via upregulation of mTOR pathway. Frontiers in cellular neuroscience;8:182. http://dx.doi.org/10.3389/fncel.2014.00182

Francis JS, Dragunow M, During MJ. 2004. Over expression of ATF-3 protects rat hippocampal neurons from in vivo injection of kainic acid. Brain Res Mol Brain Res;124:199-203. http://dx.doi.org/10.1016/j.molbrainres.2003.10.027

Ghavami S, Shojaei S, Yeganeh B, Ande SR, Jangamreddy JR, Mehrpour M, et al. 2014. Autophagy and apoptosis dysfunction in neurodegenerative disorders. Progress in neurobiology;112:24-49. http://dx.doi.org/10.1016/j.pneurobio.2013.10.004

Guaiana G, Barbui C, Hotopf M. 2007. Amitriptyline for depression. Cochrane Database Syst Rev;10.1002/14651858.CD004186.pub2:CD004186. http://dx.doi.org/10.1002/14651858.CD004186.pub2

Ha JS, Sung HY, Lim HM, Kwon KS, Park SS. 2012. PI3K-ERK1/2 activation contributes to extracellular H2O2 generation in amyloid beta toxicity. Neuroscience letters;526:112-7. http://dx.doi.org/10.1016/j.neulet.2012.08.023

Han YS, Lee CS. 2009. Antidepressants reveal differential effect against 1-methyl-4-phenylpyridinium toxicity in differentiated PC12 cells. European journal of pharmacology;604:36-44. http://dx.doi.org/10.1016/j.ejphar.2008.12.025

Hashimoto R, Yu J, Koizumi H, Ouchi Y, Okabe T. 2012. Ginsenoside Rb1 Prevents MPP(+)-Induced Apoptosis in PC12 Cells by Stimulating Estrogen Receptors with Consequent Activation of ERK1/2, Akt and Inhibition of SAPK/JNK, p38 MAPK. Evidence-based complementary and alternative medicine : eCAM;2012:693717. http://dx.doi.org/10.1155/2012/693717

Hellweg R, Ziegenhorn A, Heuser I, Deuschle M. 2008. Serum concentrations of nerve growth factor and brain- derived neurotrophic factor in depressed patients before and after antidepressant treatment. Pharmacopsychiatry;41:66-71. http://dx.doi.org/10.1055/s-2007-1004594

Hettiarachchi N, Dallas M, Al-Owais M, Griffiths H, Hooper N, Scragg J, et al. 2014. Heme oxygenase-1 protects against Alzheimer’s amyloid-beta(1-42)-induced toxicity via carbon monoxide production. Cell Death Dis;5:e1569. http://dx.doi.org/10.1038/cddis.2014.529

Hisaoka-Nakashima K, Kajitani N, Kaneko M, Shigetou T, Kasai M, Matsumoto C, et al. 2016. Amitriptyline induces brain-derived neurotrophic factor (BDNF) mRNA expression through ERK-dependent modulation of

24

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

multiple BDNF mRNA variants in primary cultured rat cortical astrocytes and microglia. Brain Res;1634:57-67. http://dx.doi.org/10.1016/j.brainres.2015.12.057

Huang JY, Chuang JI. 2010. Fibroblast growth factor 9 upregulates heme oxygenase-1 and gamma- glutamylcysteine synthetase expression to protect neurons from 1-methyl-4-phenylpyridinium toxicity. Free radical biology & medicine;49:1099-108. http://dx.doi.org/10.1016/j.freeradbiomed.2010.06.026

Hung SY, Liou HC, Kang KH, Wu RM, Wen CC, Fu WM. 2008. Overexpression of heme oxygenase-1 protects dopaminergic neurons against 1-methyl-4-phenylpyridinium-induced neurotoxicity. Mol Pharmacol;74:1564-75. http://dx.doi.org/10.1124/mol.108.048611

Jang SW, Liu X, Chan CB, Weinshenker D, Hall RA, Xiao G, et al. 2009. Amitriptyline is a TrkA and TrkB receptor agonist that promotes TrkA/TrkB heterodimerization and has potent neurotrophic activity. Chemistry & biology;16:644-56. http://dx.doi.org/10.1016/j.chembiol.2009.05.010

Jantas-Skotniczna D, Kajta M, Lason W. 2006. Memantine attenuates staurosporine-induced activation of caspase-3 and LDH release in mouse primary neuronal cultures. Brain Res;1069:145-53. http://dx.doi.org/10.1016/j.brainres.2005.11.055

Jantas D, Greda A, Golda S, Korostynski M, Lason W. 2016. The neuroprotective effects of orthosteric agonists of group II and III mGluRs in primary neuronal cell cultures are dependent on developmental stage. Neuropharmacology;111:195-211. http://dx.doi.org/10.1016/j.neuropharm.2016.09.003

Jantas D, Krawczyk S, Lason W. 2014. The predominant protective effect of tianeptine over other antidepressants in models of neuronal apoptosis: the effect blocked by inhibitors of MAPK/ERK1/2 and PI3-K/Akt pathways. Neurotoxicity research;25:208-25. http://dx.doi.org/10.1007/s12640-013-9430-3

Jantas D, Lason W. 2009. Protective effect of memantine against Doxorubicin toxicity in primary neuronal cell cultures: influence a development stage. Neurotoxicity research;15:24-37. http://dx.doi.org/10.1007/s12640-009- 9002-8

Johansson IM, Bjartmar L, Marcusson J, Ross SB, Seckl JR, Olsson T. 1998. Chronic amitriptyline treatment induces hippocampal NGFI-A, glucocorticoid receptor and mineralocorticoid receptor mRNA expression in rats. Brain Res Mol Brain Res;62:92-5. http://www.ncbi.nlm.nih.gov/pubmed/9795159

Kandil EA, Abdelkader NF, El-Sayeh BM, Saleh S. 2016. Imipramine and amitriptyline ameliorate the rotenone model of Parkinson’s disease in rats. Neuroscience;332:26-37. http://dx.doi.org/10.1016/j.neuroscience.2016.06.040

Kidd SK, Schneider JS. 2010. Protection of dopaminergic cells from MPP+-mediated toxicity by histone deacetylase inhibition. Brain Res;1354:172-8. http://dx.doi.org/10.1016/j.brainres.2010.07.041

Kolla N, Wei Z, Richardson JS, Li XM. 2005. Amitriptyline and fluoxetine protect PC12 cells from cell death induced by hydrogen peroxide. Journal of psychiatry & neuroscience : JPN;30:196-201. http://www.ncbi.nlm.nih.gov/pubmed/15944744

Kontopoulos E, Parvin JD, Feany MB. 2006. Alpha-synuclein acts in the nucleus to inhibit histone acetylation and promote neurotoxicity. Human molecular genetics;15:3012-23. http://dx.doi.org/10.1093/hmg/ddl243

Lee JC, Kim IH, Park JH, Ahn JH, Cho JH, Cho GS, et al. 2015a. Ischemic preconditioning protects hippocampal pyramidal neurons from transient ischemic injury via the attenuation of oxidative damage through upregulating heme oxygenase-1. Free radical biology & medicine;79:78-90. http://dx.doi.org/10.1016/j.freeradbiomed.2014.11.022

25

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Lee MY, Hong S, Kim N, Shin KS, Kang SJ. 2015b. Tricyclic Antidepressants Amitriptyline and Desipramine Induced Neurotoxicity Associated with Parkinson’s Disease. Molecules and cells;38:734-40. http://dx.doi.org/10.14348/molcells.2015.0131

Leucht C, Huhn M, Leucht S. 2012. Amitriptyline versus placebo for major depressive disorder. Cochrane Database Syst Rev;12:CD009138. http://dx.doi.org/10.1002/14651858.CD009138.pub2

Li M, Qian S. 2016. Gastrodin Protects Neural Progenitor Cells Against Amyloid beta (1-42)-Induced Neurotoxicity and Improves Hippocampal Neurogenesis in Amyloid beta (1-42)-Injected Mice. Journal of molecular neuroscience : MN;60:21-32. http://dx.doi.org/10.1007/s12031-016-0758-z

Li S, Hafeez A, Noorulla F, Geng X, Shao G, Ren C, et al. 2017. Preconditioning in neuroprotection: From hypoxia to ischemia. Progress in neurobiology;10.1016/j.pneurobio.2017.01.001. http://dx.doi.org/10.1016/j.pneurobio.2017.01.001

Li XM, Chlan-Fourney J, Juorio AV, Bennett VL, Shrikhande S, Bowen RC. 2000. Antidepressants upregulate messenger RNA levels of the neuroprotective enzyme superoxide dismutase (SOD1). Journal of psychiatry & neuroscience : JPN;25:43-7. http://www.ncbi.nlm.nih.gov/pubmed/10721683

Li Y, Liu K, Kang ZM, Sun XJ, Liu WW, Mao YF. 2016. Helium preconditioning protects against neonatal hypoxia- ischemia via nitric oxide mediated up-regulation of antioxidases in a rat model. Behavioural brain research;300:31-7. http://dx.doi.org/10.1016/j.bbr.2015.12.001

Liang Y, Jiang H, Ratovitski T, Jie C, Nakamura M, Hirschhorn RR, et al. 2009. ATF3 plays a protective role against toxicity by N-terminal fragment of mutant huntingtin in stable PC12 cell line. Brain Res;1286:221-9. http://dx.doi.org/10.1016/j.brainres.2009.06.049

Lin HY, Yeh WL, Huang BR, Lin C, Lai CH, Lin H, et al. 2012. Desipramine protects neuronal cell death and induces heme oxygenase-1 expression in Mes23.5 dopaminergic neurons. PLoS One;7:e50138. http://dx.doi.org/10.1371/journal.pone.0050138

Liu L, Cao JX, Sun B, Li HL, Xia Y, Wu Z, et al. 2010. Mesenchymal stem cells inhibition of chronic ethanol- induced oxidative damage via upregulation of phosphatidylinositol-3-kinase/Akt and modulation of extracellular signal-regulated kinase 1/2 activation in PC12 cells and neurons. Neuroscience;167:1115-24. http://dx.doi.org/10.1016/j.neuroscience.2010.01.057

Liu L, Zhu J, Glass PS, Brink PR, Rampil IJ, Rebecchi MJ. 2009. Age-associated changes in cardiac gene expression after preconditioning. Anesthesiology;111:1052-64. http://dx.doi.org/10.1097/ALN.0b013e3181bbcb2a

Lu Z, Xu S. 2006. ERK1/2 MAP kinases in cell survival and apoptosis. IUBMB life;58:621-31. http://dx.doi.org/10.1080/15216540600957438

Mao X, Hou T, Cao B, Wang W, Li Z, Chen S, et al. 2011. The tricyclic antidepressant amitriptyline inhibits D- cyclin transactivation and induces myeloma cell apoptosis by inhibiting histone deacetylases: in vitro and in silico evidence. Mol Pharmacol;79:672-80. http://dx.doi.org/10.1124/mol.110.068122

Marini AM, Nowak TS, Jr. 2000. Metabolic effects of 1-methyl-4-phenylpyridinium (MPP(+)) in primary neuron cultures. Journal of neuroscience research;62:814-20. http://dx.doi.org/10.1002/1097- 4547(20001215)62:6<814::AID-JNR8>3.0.CO;2-F

Marshall CJ. 1995. Specificity of receptor tyrosine kinase signaling: transient versus sustained extracellular signal-regulated kinase activation. Cell;80:179-85. http://www.ncbi.nlm.nih.gov/pubmed/7834738

26

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Mastroeni D, Delvaux E, Nolz J, Tan Y, Grover A, Oddo S, et al. 2015. Aberrant intracellular localization of H3k4me3 demonstrates an early epigenetic phenomenon in Alzheimer’s disease. Neurobiology of aging;36:3121- 9. http://dx.doi.org/10.1016/j.neurobiolaging.2015.08.017

McDonald JW, Behrens MI, Chung C, Bhattacharyya T, Choi DW. 1997. Susceptibility to apoptosis is enhanced in immature cortical neurons. Brain Res;759:228-32. http://www.ncbi.nlm.nih.gov/pubmed/9221941

McDonald WM, Richard IH, DeLong MR. 2003. Prevalence, etiology, and treatment of depression in Parkinson’s disease. Biological psychiatry;54:363-75. http://www.ncbi.nlm.nih.gov/pubmed/12893111

Mendell AL, Creighton CE, Kalisch BE, MacLusky NJ. 2016. 5alpha-Androstane-3alpha,17beta-Diol Inhibits Neurotoxicity in SH-SY5Y Human Neuroblastoma Cells and Mouse Primary Cortical Neurons. Endocrinology;157:4570-8. http://dx.doi.org/10.1210/en.2016-1508

Menza M, Dobkin RD, Marin H, Mark MH, Gara M, Buyske S, et al. 2009. A controlled trial of antidepressants in patients with Parkinson disease and depression. Neurology;72:886-92. http://dx.doi.org/10.1212/01.wnl.0000336340.89821.b3

Miyake K, Nagai K. 2009. Inhibition of alpha-mannosidase attenuates endoplasmic reticulum stress-induced neuronal cell death. Neurotoxicology;30:144-50. http://dx.doi.org/10.1016/j.neuro.2008.10.010

Moore RA, Derry S, Aldington D, Cole P, Wiffen PJ. 2015. Amitriptyline for neuropathic pain in adults. Cochrane Database Syst Rev;10.1002/14651858.CD008242.pub3:CD008242. http://dx.doi.org/10.1002/14651858.CD008242.pub3

Moretti R, Torre P, Antonello RM, Cazzato G, Bava A. 2002. Depression and Alzheimer’s disease: symptom or comorbidity? American journal of Alzheimer’s disease and other dementias;17:338-44. http://www.ncbi.nlm.nih.gov/pubmed/12501480

Nada SE, Shah ZA. 2012. Preconditioning with Ginkgo biloba (EGb 761(R)) provides neuroprotection through HO1 and CRMP2. Neurobiology of disease;46:180-9. http://dx.doi.org/10.1016/j.nbd.2012.01.006

Okugawa G, Omori K, Suzukawa J, Fujiseki Y, Kinoshita T, Inagaki C. 1999. Long-term treatment with antidepressants increases glucocorticoid receptor binding and gene expression in cultured rat hippocampal neurones. Journal of neuroendocrinology;11:887-95. http://www.ncbi.nlm.nih.gov/pubmed/10520140

Paumier KL, Siderowf AD, Auinger P, Oakes D, Madhavan L, Espay AJ, et al. 2012. Tricyclic antidepressants delay the need for dopaminergic therapy in early Parkinson’s disease. Movement disorders : official journal of the Movement Disorder Society;27:880-7. http://dx.doi.org/10.1002/mds.24978

Paumier KL, Sortwell CE, Madhavan L, Terpstra B, Daley BF, Collier TJ. 2015. Tricyclic antidepressant treatment evokes regional changes in neurotrophic factors over time within the intact and degenerating nigrostriatal system. Experimental neurology;266:11-21. http://dx.doi.org/10.1016/j.expneurol.2015.02.005

Peart MJ, Smyth GK, van Laar RK, Bowtell DD, Richon VM, Marks PA, et al. 2005. Identification and functional significance of genes regulated by structurally different histone deacetylase inhibitors. Proc Natl Acad Sci U S A;102:3697-702. http://dx.doi.org/10.1073/pnas.0500369102

Perisic T, Zimmermann N, Kirmeier T, Asmus M, Tuorto F, Uhr M, et al. 2010. Valproate and amitriptyline exert common and divergent influences on global and gene promoter-specific chromatin modifications in rat primary astrocytes. Neuropsychopharmacology;35:792-805. http://dx.doi.org/10.1038/npp.2009.188

27

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Rapoport M, Ferreira A. 2000. PD98059 prevents neurite degeneration induced by fibrillar beta-amyloid in mature hippocampal neurons. Journal of neurochemistry;74:125-33. http://www.ncbi.nlm.nih.gov/pubmed/10617113

Robison AJ, Vialou V, Sun HS, Labonte B, Golden SA, Dias C, et al. 2014. Fluoxetine epigenetically alters the CaMKIIalpha promoter in nucleus accumbens to regulate DeltaFosB binding and antidepressant effects. Neuropsychopharmacology;39:1178-86. http://dx.doi.org/10.1038/npp.2013.319

Rouaux C, Loeffler JP, Boutillier AL. 2004. Targeting CREB-binding protein (CBP) loss of function as a therapeutic strategy in neurological disorders. Biochemical pharmacology;68:1157-64. http://dx.doi.org/10.1016/j.bcp.2004.05.035

Rybakova Y, Akkuratov E, Kulebyakin K, Brodskaya O, Dizhevskaya A, Boldyrev A. 2012. Receptor-mediated oxidative stress in murine cerebellar neurons is accompanied by phosphorylation of MAP (ERK 1/2) kinase. Current aging science;5:225-30. http://www.ncbi.nlm.nih.gov/pubmed/23387889

Ryu H, Lee J, Olofsson BA, Mwidau A, Dedeoglu A, Escudero M, et al. 2003. Histone deacetylase inhibitors prevent oxidative neuronal death independent of expanded polyglutamine repeats via an Sp1-dependent pathway. Proc Natl Acad Sci U S A;100:4281-6. http://dx.doi.org/10.1073/pnas.0737363100

Schipper HM, Song W, Zukor H, Hascalovici JR, Zeligman D. 2009. Heme oxygenase-1 and neurodegeneration: expanding frontiers of engagement. Journal of neurochemistry;110:469-85. http://dx.doi.org/10.1111/j.1471- 4159.2009.06160.x

Schwanhausser B, Busse D, Li N, Dittmar G, Schuchhardt J, Wolf J, et al. 2011. Global quantification of mammalian gene expression control. Nature;473:337-42. http://dx.doi.org/10.1038/nature10098

Sheline YI, Gado MH, Kraemer HC. 2003. Untreated depression and hippocampal volume loss. The American journal of psychiatry;160:1516-8. http://dx.doi.org/10.1176/appi.ajp.160.8.1516

Shu L, Wang C, Wang J, Zhang Y, Zhang X, Yang Y, et al. 2016. The neuroprotection of hypoxic preconditioning on rat brain against traumatic brain injury by up-regulated transcription factor Nrf2 and HO-1 expression. Neuroscience letters;611:74-80. http://dx.doi.org/10.1016/j.neulet.2015.11.012

Strelau J, Sullivan A, Bottner M, Lingor P, Falkenstein E, Suter-Crazzolara C, et al. 2000. Growth/differentiation factor-15/macrophage inhibitory cytokine-1 is a novel trophic factor for midbrain dopaminergic neurons in vivo. J Neurosci;20:8597-603. http://www.ncbi.nlm.nih.gov/pubmed/11102463

Sun DS, Chang HH. 2003. Differential regulation of JNK in caspase-3-mediated apoptosis of MPP(+)-treated primary cortical neurons. Cell biology international;27:769-77. http://www.ncbi.nlm.nih.gov/pubmed/12972283

Tang Y, Pacary E, Freret T, Divoux D, Petit E, Schumann-Bard P, et al. 2006. Effect of hypoxic preconditioning on brain genomic response before and following ischemia in the adult mouse: identification of potential neuroprotective candidates for stroke. Neurobiology of disease;21:18-28. http://dx.doi.org/10.1016/j.nbd.2005.06.002

Thangnipon W, Puangmalai N, Chinchalongporn V, Jantrachotechatchawan C, Kitiyanant N, Soi-Ampornkul R, et al. 2013. N-benzylcinnamide protects rat cultured cortical neurons from beta-amyloid peptide-induced neurotoxicity. Neuroscience letters;556:20-5. http://dx.doi.org/10.1016/j.neulet.2013.09.071

Tsankova NM, Berton O, Renthal W, Kumar A, Neve RL, Nestler EJ. 2006. Sustained hippocampal chromatin regulation in a mouse model of depression and antidepressant action. Nat Neurosci;9:519-25. http://dx.doi.org/10.1038/nn1659

28

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Wang Y, Hilton BA, Cui K, Zhu MY. 2015. Effects of Antidepressants on DSP4/CPT-Induced DNA Damage Response in Neuroblastoma SH-SY5Y Cells. Neurotoxicity research;28:154-70. http://dx.doi.org/10.1007/s12640-015-9534-z

Wei Z, Song MS, MacTavish D, Jhamandas JH, Kar S. 2008. Role of calpain and caspase in beta-amyloid- induced cell death in rat primary septal cultured neurons. Neuropharmacology;54:721-33. http://dx.doi.org/10.1016/j.neuropharm.2007.12.006

Xia Z, Bergstrand A, DePierre JW, Nassberger L. 1999. The antidepressants imipramine, clomipramine, and citalopram induce apoptosis in human acute myeloid leukemia HL-60 cells via caspase-3 activation. Journal of biochemical and molecular toxicology;13:338-47. http://www.ncbi.nlm.nih.gov/pubmed/10487422

Yan L, Wang Q, Fu Q, Ye Q, Xiao H, Wan Q. 2010. Amitriptyline inhibits currents and decreases the mRNA expression of voltage-gated sodium channels in cultured rat cortical neurons. Brain Res;1336:1-9. http://dx.doi.org/10.1016/j.brainres.2010.04.016

Young KF, Pasternak SH, Rylett RJ. 2009. Oligomeric aggregates of amyloid beta peptide 1-42 activate ERK/MAPK in SH-SY5Y cells via the alpha7 nicotinic receptor. Neurochemistry international;55:796-801. http://dx.doi.org/10.1016/j.neuint.2009.08.002

Yu J, Zheng J, Lin J, Jin L, Yu R, Mak S, et al. 2016. Indirubin-3-Oxime Prevents H2O2-Induced Neuronal Apoptosis via Concurrently Inhibiting GSK3beta and the ERK Pathway. Cellular and molecular neurobiology;10.1007/s10571-016-0402-z. http://dx.doi.org/10.1007/s10571-016-0402-z

Zhang SJ, Buchthal B, Lau D, Hayer S, Dick O, Schwaninger M, et al. 2011. A signaling cascade of nuclear calcium-CREB-ATF3 activated by synaptic NMDA receptors defines a gene repression module that protects against extrasynaptic NMDA receptor-induced neuronal cell death and ischemic brain damage. J Neurosci;31:4978-90. http://dx.doi.org/10.1523/JNEUROSCI.2672-10.2011

Zhao Q, Yang X, Cai D, Ye L, Hou Y, Zhang L, et al. 2016. Echinacoside Protects Against MPP(+)-Induced Neuronal Apoptosis via ROS/ATF3/CHOP Pathway Regulation. Neuroscience bulletin;32:349-62. http://dx.doi.org/10.1007/s12264-016-0047-4

Zheng X, Chen F, Zheng T, Huang F, Chen J, Tu W. 2016. Amitriptyline Activates TrkA to Aid Neuronal Growth
and Attenuate Anesthesia-Induced Neurodegeneration in Rat Dorsal Root Ganglion Neurons. Medicine;95:e3559. http://dx.doi.org/10.1097/MD.0000000000003559

Zhu JH, Horbinski C, Guo F, Watkins S, Uchiyama Y, Chu CT. 2007. Regulation of autophagy by extracellular signal-regulated protein kinases during 1-methyl-4-phenylpyridinium-induced cell death. The American journal of pathology;170:75-86. http://dx.doi.org/10.2353/ajpath.2007.060524

29

ACCEPTED MANUSCRIPT

Supplementary Table 1. List of primer sequences

ACCEPTED MANUSCRIPT

  

Experiment

qRT-PCR

ChIP assay

Name

Atf3

Gdf15

Hmox1

Hmox1 primer 1

Hmox1 primer 2

Hmox1 primer 3

Hmox1 primer 4

Hmox1 primer 5

Hmox1 primer 6

Hmox1 primer 7

Hmox1 primer 8

Atf3 primer 1

Atf3 primer 2

Atf3 primer 3

Atf3 primer 4

Atf3 primer 5

Sequence

F: CGGGAGTTTCATCAGA
R: TCCACCATGGCAGATC
F: GCTGAGCCGGCTGCAT
R: CAGGACTTGGTTCTGAGTTCGA F: CAGCCCCACCAAGTTCAAA

R: TCAGGTGTCATCTCCAGAGTGTTC

F: GCCTCCTGAGTGCTGAGATCA
R: TTAGAAGGGCTTAGTTGAAGATGACA F: CCACCCCCACCCAGGAT
R: CTGCGGGAAAAACAAAGTTCA
F: GTGTCCCCCCCATCTTTAGAG
R: ATGATGGGTTCAGACACAGTTGA
F: CCCTTGATGGGCAGTTGTTG
R: TGGTGGTTCTAAGTCCTCGACTCT
F: TCATGGTGTGGACCAGGTTCT
R: GTGCAGCACCCGAGTGTTT
F: GAGTTCCGGGAAGGTAGACTCA
R: GATGCTGGTGGGCAAACC
F: CATGTGGTTGCTGGGAATTG
R: CAGCGGTCAAGAGCACTGACT
F: GCTGAACTTCCTTCACTATGTAGACAGA R: CAGAGGGAGGGCAGATCTCTT
F: GGCAGGAGCAGATAGCATCAG
R: GCAGACTACCACCTCCAGTGTTT
F: TTGCCGCCGCCATTT
R: CCACTAATCAAACACTGGGTTTCC
F: TGGGTGTTGGGAGCTTTTCT
R: TGTGCGTGCGCATTCG
F: GCCCCGAGAGCCGTTT
R: GAAGCGCCAGGCTAATGTTT
F: CCGGCAGAGGAGGTCACA

     

30

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Atf3 primer 6

Atf3 primer 7

R: CAAAGTTACACTCCGATGAATCCA F: AGTTGCGCTCTCAGTTTGCA
R: ACCAGCACCCACCTTCCA
F: GCTCTCTCCCGTTGGATTGA

R: GAAATGGGAGTTGAAAATGAGCAT

  

31

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Supplementary Table 2. List of genes commonly up-regulated by amitriptyline and imipramine

Probe set ID

P-value (Imi)

Fold change (Imi)

P-value (Ami)

Fold change (Ami)

Gene symbol

mRNA accession

17218060

0.025803497

5.178868

0.02476389

4.2095003

Ptgs2

AK049923

17324546

0.03236918

2.895594

0.042722207

2.502995

Ostn

AY395730

17438995

0.022930728

2.7573016

0.030148624

2.135017

Cxcl2

AK137628

17502573

0.035132393

2.4298851

0.031818617

2.4150999

Hmox1

AK150934

17231033

0.039368454



2.046843



0.038839117



1.8478432





Atf3



AK133965

17445440

0.009769376

1.9874425

0.025825825

2.0059645

Steap1

AK010437

17329454

0.025165316

1.9531157

0.029150765

1.8893377

Leprel1



AK076434

17337228

0.005471213



1.9406939



0.002495439



1.7420701





Ier3



AK003466

17513287

0.001575976

1.8882313

0.013341134

1.8909816

—

ENSMUST00000083156

17383892

0.035890184



1.8527267

0.007604556



1.644332



Lcn2

AK149774

17420546

0.02574906

1.837759



0.025791548

1.7854931





Pla2g2e



AK148187

17398571

0.026061676

1.8332845



0.006602132

1.6405305



Tmem154



BC096432

17458439

0.005675832

1.7073649

0.019484038

1.5859574

Gpnmb

AK044764

17536364

0.02860776



1.6793344



0.027317973



1.6444608





—



n268784

17417844

0.039810725

1.6573004

0.029499989

1.6227481

Slc2a1

AK089353

17550428

0.042637687



1.6524079

0.022626651



1.8271943



—

NEAT1_2

17510114

0.030983923

1.6180384

0.018300436

1.6056564



Gdf15

AK144199

32

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Supplementary Table 3. List of genes commonly down-regulated by amitriptyline and imipramine

mRNA accession

AK050355 ENSMUST00000101956 BC034841
AK039777
AK020117
AK078170 XM_001005050 BC117110
AK042697
NM_029617
AK044924
AK046691
n293127
AK033321
n268974
AK040931
n268854
AK052772
AK141866
AK011090
AK013193
AK050821
AK031898
AK134045
BC023918
AK011276
n266831

Probe set ID

P-value (Imi)

Fold change (Imi)

P-value (Ami)

Fold change (Ami)

Gene symbol

17266851

0.038362

0.43585

0.005317

0.52552

Slfn9

17486023

0.022791

0.507985

0.044582

0.653756

—

17255058

0.02808

0.510159

0.020414

0.526663

Tmem100

17352862

0.00254

0.523658

0.010746

0.490587

Aqp4

17291005

0.04514



0.554479



0.019097



0.425298





Hist1h1b



17422443

0.010783

0.568145

0.012807

0.555916

Hes5

17490038

6.15E-04

0.588042

0.002825

0.625088

Gm5593

17285867

0.03964



0.594466



0.026461



0.456198





Hist1h2ab



17374833

0.007535

0.595006

0.003876

0.561358

Nusap1

17374692

0.026978



0.607735

0.023232



0.635185



Casc5

17404821

0.012694

0.61155





0.033563

0.551853



Ccna2



17545936

0.034863

0.613619



0.00609

0.595723



Car5b



17428803

0.013956

0.613939

0.001499

0.639792

—

17268909

0.0498



0.616667



0.022241



0.53151





Top2a



17480675

0.0086

0.620198

0.04916

0.609956

—

17276906

0.012773



0.620943

0.002202



0.581802



Smoc1

17535558

0.007519

0.621764



0.047871

0.655828

—

17497334

0.007005

0.628115

0.049489

0.642824

Mki67

17243839

0.043423

0.631999



0.037415



0.602959

Ascl1

17518636

0.006353

0.63404

0.041743



0.657931

2810417H13Rik

17339824

0.040438

0.642683

0.044259

0.591028

Vit

17523494

0.005378

0.650064

0.023545

0.529654

Kif15

17402787

0.017783

0.651767

0.011593

0.544107

Tacr3

17482739

0.04394

0.652155



0.025546

0.646653

Plk1

17267454

0.006387

0.652505



0.016977



0.611614

Prr11

17416234

0.005576

0.657695

0.033727

0.648857

Ppap2b

17449673

0.009737

0.664497



0.015083

0.603601

—

 

33

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT



Supplementary Fig. 1. (A) Immunofluorescence with MAP2 and GFAP showed that most cells are positive for MAP2 (marker for neurons) in primary cultures of mouse neuronal cells at in vitro day 3 and 5. (B) Bars represent the ratio of MAP2-positive cells and GFAP-positive cells to total cells. Each value represents the mean ± SEM of eight wells in two different experiments.

Supplementary Fig. 2. MTT assays for determination of the optimal concentrations of antidepressant drug for use in primary cultures of mouse neuronal cells. (A) Imipramine. (B) Amitriptyline. Each value represents the mean ± SEM in three independent experiments. Bars represent cell survival rate in the treated cultures compared to untreated (control) cultures.



34

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT



Supplementary Fig. 3. qRT-PCR assay and western blots of Gdf15/GDF15 in mouse neuronal cells treated with 5 μM imipramine or 5 μM amitriptyline for 48 h. (A) qRT-PCR analysis of Gdf15. (B) Western blot of GDF15. Each value represents the mean ± SEM in three independent experiments. Bars represent the relative ratios of treated to untreated controls. *P < 0.05; **P < 0.01 versus control.

Supplementary Fig. 4. Assessment of the changes in CREB and p-CREB in mouse neuronal cells. (A and B) Western blot for MPP+-induced changes in CREB and p-CREB expressions. (C and D) Western blot for Aβ1-42- induced changes in CREB and p-CREB expressions. Each value represents the mean ± SEM in three independent experiments. Bars represent the relative ratios of treated to untreated controls.



35

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT



Supplementary Fig. 5. (A) The neuroprotective effects of SAHA in mouse neuronal cells treated with Aβ1-42 as assessment by MTT assay. Cells at DIV3 were treated with 0.5 or 1 μM SAHA for 48 h, after removal of the drug, cells were treated with 5 or 10 μM Aβ1-42 for 24 h. (B) Combination of amitriptyline and SAHA induced toxicity in primary neuron cultures. Cells at DIV3 were treated with a combination of 5 μM amitriptyline and 0.25, 0.5, or 1 μM SAHA for 48 h, after removal of the drug, cells were treated with 5 or 10 μM Aβ1-42 for 24 h. Each value represents the mean ± SEM of 12 wells in three independent experiments. *P < 0.5, **P < 0.01, ***P < 0.001 compare to control in each group.

36

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Pre-treatment with amitriptyline causes epigenetic up-regulation of neuroprotection-associated genes and has anti-apoptotic effects in mouse neuronal cells

HIGHLIGHTS

  Antidepressants induce gene expression changes associated with neuronal cell death.

  Amitriptyline induces active histone modifications in the Atf3 and Hmox1 genes.

  Pre-treatment with amitriptyline exerts neuroprotective effect against neurotoxicity induced by MPP+ or Aβ1-
42.

  Amitriptyline inhibits the MPP+- or Aβ1-42-induced activation of ERK1/2.

37

ACCEPTED MANUSCRIPT