Neuroprotective effects of Paeonia Lactiflora extract against cell death of dopaminergic SH-SY5Y cells is mediated by epigenetic modulation

Background The Paeonia lactiflora extract (PLE) has been reported to have neuroprotective effect against neurodegeneration that are induced by cellular stress such as oxidative stress. Its underlying mechanisms, however, remain unclear. In latest decades, emerging evidence has suggested that epigenetic mechanisms play a key role in gene regulation in response to the cellular stress. We investigated whether epigenetic modulation was involved in neuronal cell death by the neurotoxicant, 1-Methyl-4-phenylpyridinium (MPP+), and the neuroprotective effect of PLE. Methods Differentiated SH-SY5Y, which is a well-established dopaminergic cell line model, was treated with 0 ~ 200 μg/ml PLE for 4 h prior to MPP+ treatment. The effect of PLE on cell viability was determined by MTT assays. Gene expression levels of oxidative stress responsive genes, such as Heme oxygenase 1 (HMOX1), and histone modifiers, such as histone acetyltransferases (HATs) and deacetylases (HDACs) were measured by quantitative RT PCR. In order to investigate the changes in epigenetic modifications, the acetylated lysine 9 (H3K9ac) and lysine 27 (H3K27ac) of Histone H3 were measured by western blot using histones extracted from the cells. Results MPP+-induced cell death in SH-SY5Y cells was significantly reduced by PLE pretreatment in a dose-dependent manner, indicating the potent neuroprotective effects of PLE. It was accompanied by induced expression of HMOX1. MPP+ treatment increased the expression of HATs and consistently increased H3K9ac and H3K27ac of Histone H3. PLE pretreatment impeded the changes in H3K9ac and H3K27ac, coincided with increased expression of HDAC5 without changes in HAT expression. Conclusions The results suggested that MPP+-induced cell death in the dopaminergic SH-SY5Y cells was related with transcriptional induction of HATs and increased histone H3 acetylation and that PLE might prevent the cells from MPP+-induced cell death via tempering histone H3 acetylation.


Background
Parkinson's disease, the second most common neurodegenerative disorder following Alzheimer's disease, is mainly caused by progressive degeneration of dopaminergic neurons [1]. One major risk factor leading to the degeneration is cellular oxidative stress [2]. Oxidative stress is a cellular oxidative imbalanced condition due to impaired mitochondrial function and elevated reactive oxygen species (ROS) in the cell. 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and its metabolite, MPP + , has been used as oxidative stressors that induce cell death in dopaminergic neurons in vivo and in vitro, which mimic Parkinson's disease [2]. MPP + enters dopaminergic neuronal cells via dopamine transporters and induces mitochondria dysfunction. It produces ROS [3] and releases cytochrome C (Cyt-C), which activates caspase-3 and eventually, leads dopaminergic neuronal cells to apoptosis [4]. MPP + -induced cell system serves as a useful model to study the molecular mechanism underlying loss of dopaminergic neurons and therapeutic potentials.
Paeonia lactiflora Pall is one of the most commonly used medicinal herbs across different traditional Chinese medicine treatment for its anti-inflammatory and antioxidant effects [5]. The herbal extract of this plant, PLE, has been reported to have neuroprotective effects against oxidative stress both in vivo [6] and in vitro [7,8], in addition to the main compounds such as albiflorin [9], paeoniflorin [10][11][12], and paeonol [13,14] that have been extracted from this plant. It indicates its potentials to protect dopaminergic neurons from cell death and to attenuate the process of Parkinson's disease [6,15,16]. The underlying molecular mechanism of PLE, however, remains elusive.
Recently, the emerging evidence on disease-preventive and therapeutic effects of phytochemicals and herb extracts suggests that they mainly exert their functions, at least in part, via epigenetic modulation [17]. Epigenetic mechanisms such as DNA methylation and histone modifications regulate gene expression via changes in chromatin accessibility in response to environmental stimuli such as oxidative stress [18]. Disturbed patterns of histone acetylation as well as altered activity of HATs and HDACs have been connected to chronic diseases, including cancer and neurodegenerative diseases [18]. Cellular stress alters histone acetylation at specific residues such as histone H3 lysine9 and histone H4 lysine12 [19][20][21] and changes the activities of HDACs [22,23], suggesting the potential development of therapeutic strategies for neurodegenerative conditions [24][25][26][27]. However, the epigenetic mechanism has been investigated on limited phytochemicals, for examples, curcumin enriched in turmeric and sulforaphane enriched in broccoli, which have effects on cell death and proliferation by inhibiting HATs and HDACs, respectively [28][29][30]. Consequently, to identify potential modulators of histone acetylation in neuronal cells may provide novel therapeutic avenues in neurodegenerative disorders. Therefore, we investigated whether the neuroprotective effects of PLE against oxidative stress in dopaminergic neuron cells are mediated by epigenetic modulation. Using the MPP + -induced SY5Y cell model, gene expression involving the antioxidant pathway as well as histone acetylation histone acetylation levels at histone H3 lysine 9 (H3K9) and lysine 27 (H3K27) were analyzed with and without PLE treatment.

Cell differentiation and treatment
To obtain dopaminergic neuronal cells, SY5Y cells were differentiated using DMEM with 10 μM retinoic acid (Sigma Aldrich, St. Louis, MO, USA), 1 % FBS and100 units/ml penicillin, and 100 μg/ml streptomycin for 4 days. The medium was changed after 2 days during differentiation. The differentiation was confirmed by the increase in mRNA level of Tyrosine hydroxylase gene (TH

MTT assays
3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazoliumbromide (MTT) (Sigma Aldrich, St. Louis, MO, USA) was dissolved in PBS at the concentration 5 mg/ml. The MTT solution was added to one-tenth the original culture volume and the cells were incubated at 37°C for 2 h. Then, DMSO was added in cells to convert soluble MTT to insoluble purple formazan in mitochondria of living cells. Viable cells forming formazan were measured at 562 nm using a microplate reader (Lab tech international, Uckfield, UK).

RNA isolation and reverse transcription
Total RNA was isolated from cells by Trizol reagent (Tel-test Inc, Friendswood, TX, USA). Reverse transcription of 500 ng or 1 μg total RNA was performed for cDNA synthesis. In brief, RNA samples were treated with DNase I (Thermo Fisher Scientific, Waltham, MA, USA) to remove genomic DNA, after which RNA samples were incubated with dNTP (Thermo Fisher Scientific, Waltham, MA, USA), oligo dT (Thermo Fisher Scientific, Waltham, MA, USA), and reverse transcriptase (Thermo Fisher Scientific, Waltham, MA, USA) to synthesize cDNA, following the manufacturer's protocol.

Histone extraction
Histones were extracted from SH-SY5Y cells by acid extraction method in accordance with Shechter, et al.'s protocol [31]. Cells were washed with 1X PBS and pelleted by centrifugation at 1000 g for 3 min. Cell pellets were incubated in hypotonic lysis buffer (10 mM Tris, 1 mM KCl, 1 mM DTT, 1.5 mM MgCl 2 , 10 mM Sodium butyrate, 1 mM PMSF, and 7X protease inhibitor cocktail (Roche, Robert Rohrberg, Germany)) on a rotator at 4°C for 30 min. The nuclei were collected by centrifugation at 10,000 x g at 4°C for 10 min, and the pelleted nuclei were suspended in 0.4 N H 2 SO 4 and placed on a rotator at 4°C overnight. Next day, the nuclear lysate was collected from the supernatants after centrifugation at 14,620 x g at 4°C for 10 min to remove nuclear debris. Histones were precipitated by TCA and washed with acetone-HCl and acetone. Precipitated histones were collected by centrifugation at 14,620 x g at 4°C for 5 min, air-dried at room temperature for 15 min and then dissolved in distilled water. Histones were stored at -80°C. All steps were carried out by keeping samples on ice whenever possible.

Results
MPP + -mediated changes of cell viability and SOD2 and HMOX1 expression in differentiated SH-SY5Y cells We analyzed cell viability of SH-SY5Y cells upon MPP + treatment using MTT assays. Treatment with MPP + concentrations ranging from 0 to 2 mM decreased cell viability in a dose-dependent manner, down to about 52 % (Fig. 1a). To investigate whether MPP + treatment changed the expression of genes related with the cellular anti-oxidative pathway in response to oxidative stress, mRNA expression levels of SOD2 and HMOX1 were analyzed ( Fig. 1b and c). Gene expression level of HMOX1 significantly increased at in 1 to 2 mM MPP + treatment and SOD2 significantly decreased at the 2 mM concentration. The results indicated that MPP + indeed acted as a neurotoxicant in dopaminergic neuronal cells by inducing oxidative stress-mediated cell death. While the dose at 1 mM MPP + is relatively low, it was sufficiently toxic to significantly elevate HMOX1 gene expression and reduce cell viability. Consequently, for subsequent experiment we used the 1 mM dose to analyze the neuroprotective effects provided by PLE.

Effects of PLE on MPP + -mediated cell death and HMOX1 expression
To investigate the effect of PLE against MPP + -induced cell death, we pretreated SH-SY5Y cells with PLE in the range of 0 to 200 μg/ml for 4 h prior to a 24 h 1 mM MPP + treatment. PLE pretreatment attenuated MPP + -induced cell death in a dose-dependent manner (Fig. 2a). Significantly, 200 μg/ml PLE recovered cell viability up to 93 % compared with MPP + nontreated group.
Interestingly, treatment with PLE did not alter SOD2 gene expression levels, a key component of the antioxidant pathway (Fig. 2b), while HMOX1 gene expression levels in 20-200 μg/ml PLE samples were significantly increased 1.5 to 1.9 fold, respectively, when compared to 0 μg/ml PLE group (Fig. 2c).

Changes in acetylated levels of histones in MPP + treated, differentiated SH-SY5Y cells
To determine whether MPP + -induced cell death was a result of epigenetic changes, we investigated alteration in acetylation levels of histones and HAT gene expression levels. The acetylation levels of H3K9 and H3K27 (relative to total histones H3 and H4) were analyzed using western blot analysis. When treating cells with MPP + ranging from 0.5 to 2 mM, we observed a linear correlation with increased acetylation levels of H3K9 and H3K27 (Fig. 3a). In addition, gene expression levels of the major HAT proteins, EP300 and CREBBP, were also increased in the MPP + treated samples (Fig. 3b). The results suggested that increased HAT expression under neurotoxic condition led to increased acetylation levels of H3K9 and H3K27.  Next, we determined whether pretreatment of PLE had neuroprotective effect against MPP + -induced cell death by means of epigenetic modulation. We investigated alteration of acetylation levels of histones and HATs and HDACs gene expression levels. We measured acetylation levels of H3K9 and H3K27 in MPP + treated cells with and without PLE using western blotting analysis. In the PLE pretreated group, we observed that acetylation level of H3K9 decreased, but not H3K27 acetylation (Fig. 4a). The decreased acetylated H3K9 seemed to be due to induction of the HDAC5 gene expression (Fig. 4b), rather than altered EP300 and CREBBP expression (Fig. 4c), as HDAC5 robustly increased with elevated PLE concentration. Our data suggest that HDAC5 decreased acetylation levels of H3K9 in the 200 μg/ml PLE group, overriding the effect from HATs, potentiation PLE as an antineurotoxicant by tempering histone acetylation.

Discussion
In this study, we investigated whether MPP + induced cell death in the dopaminergic neuronal SH-SY5Y cells through epigenetic modulation, and whether PLE had a Fig. 3 Acetylation of H3K9 and H3K27 and mRNA expressions of HATs in MPP + treated cells. a MPP + treatment increased acetylation levels of H3K9 and H3K27. b MPP + treatment increased mRNA expressions of HATs such as EP300 and CREBBP. Figure 3 suggested HATs increase acetylation levels of H3K9 and H3K27.; The data was indicated as the mean ± S.E.M. P < 0.05 was defined statistical significance. One-way ANOVA followed by Duncan post hoc test  Figure 4 suggested HDAC5 rather than HATs decreases acetylation levels of H3K9. The data was indicated as the mean ± S.E.M. P < 0.05 was defined statistical significance. One-way ANOVA followed by Duncan post hoc test neuroprotective effect against cell death by reversing epigenetic changes. We demonstrated that MPP + -induced cell death decreased SOD2 gene expression and increased the HMOX1 gene expression levels. The results are consistent with previous reports showing downregulation of SOD2 and upregulation of HMOX1 under oxidative stress [15,32,33]. PLE was recently reported to have protective effect on neuronal cell death in studies both in vivo [6] and in vitro [7]. The positive influence of PLE on dopaminergic neuronal cells potentiates its use in attenuating the process of Parkinson's disease [15,16,34]. However, the molecular mechanism of this protective effect remains unclear.
One possibility is that the PLE effect involves the antioxidant pathway. Our results demonstrated that PLE upregulated HMOX1 gene expression, while it did not alter SOD2 gene expression. This indicates that the PLE effect may be restricted to the HMOX1 pathway. HMOX1 gene encodes heme oxygenase 1 (HMOX1), which catalyzes heme catabolism and plays a protective role as an antioxidant. HMOX1 gene is activated in response to oxidative stress, mainly via a major transcriptional factor, Nuclear factor erythroid 2 related factor 2 (NRF2), which exerts as a ROS scavenger [35]. A recent study showed that Bromodomain protein 4 (BRD4) also transcriptionally activates HMOX1 expression by targeting of SP1 binding sites in the HMOX1 promoter [36]. BRD4, an acetylated histone-binding protein, is a key factor for regulation of gene expression [37,38]. siRNA-mediated silencing of Bromodomain protein 4 (BRD4), in the absence of stress, resulted in downregulation of HMOX1 expression [36]. However, under oxidative stress, siRNAmediated knockdown of BRD4 upregulates HMOX1 expression further via downregulation of Kelch-like ECHassociated protein 1 (KEAP1), which inhibits the degradation of NRF2 [36]. The obvious connection between hypoacetylated histone residues, especially H3K9, is reflected by the silencing of BRD4 expression, since BRD4 binds to particularly acetylated histones.
The neurprotective effect of PLE on MPP + -induced cell death of the dopaminergic cells seems mediated by epigenetic modulation, supported by our observation of changes in acetylation of histone. Treatment of MPP + increased gene expression levels of HATs such as EP300 and CREBBP, resulting in hyperacetylation of H3K9 and H3K27. This is consistent with a previous report demonstrating that neurotoxicity induced hyperacetylation and transcriptional activation of pro-apoptotic genes, such as PKC gamma and IL-8 [39]. Conversely, the pretreatment of PLE increased gene expression of HDAC5, resulting in hypoacetylation of H3K9. HDAC5 has been reported to involve neuronal cell survival, proliferation, and differentiation possibly via regulation of Myocyte enhancer factor-2 (MEF2) [40][41][42]. The resulting hypoacetylation upon PLE treatment may mimic the effect of BRD4 silencing under stress, leading to upregulated HMOX1 expression.
Our study of the neuroprotective properties of PLE suggest that epigenetic modulation is a key component of neurotoxic stress, and that therapeutic intervention should consider this avenue of treatment in diseases such as Parkinson's disease.

Conclusions
MPP + -induced cell death in differentiated SH-SY5Y cells was significantly reduced by PLE pretreatment in a dosedependent manner, accompanied by induced expression of HMOX1. MPP + -treatment increased the expression of HATs and consistently increased acetylation of H3K9 and H3K27, while PLE pretreatment attenuated the hyperacetylation in H3K9, coincided with increased expression of HDAC5 without changes in HAT expression. Our study suggested that MPP + treatment induced cell death in dopaminergic SH-SY5Y cells via upregulation of HATs and hyperacetylation of H3K9 and H3K27 and PLE pretreatment may attenuate the cell death, in part, by upregulation of HMOX1 expression through hypoacetylation of H3K9, possibly mediated by HDAC5 upregulation. Further study is needed to identify direct targets for HDAC5-mediated hypoacetylation of H3K9 in the neuroprotective effects and to investigate whether it is recapitulated in vivo.

Availability of data and materials
The datasets supporting the conclusions of this article are included within the article.