Abstract

   Valproic acid (VPA) is a multifunctional medication used for the
   treatment of epilepsy, mania associated with bipolar disorder, and
   migraine. The pharmacological effects of VPA involve a variety of
   neurotransmitter and cell signaling systems, but the molecular
   mechanisms underlying its clinical efficacy is to date largely unknown.
   In this study, we used the isobaric tags for relative and absolute
   quantitation shotgun proteomic analysis to screen differentially
   expressed proteins in VPA-treated SH-SY5Y cells. We identified changes
   in the expression levels of multiple proteins involved in Alzheimer’s
   disease, Parkinson’s disease, chromatin remodeling, controlling gene
   expression via the vitamin D receptor, ribosome biogenesis,
   ubiquitin-mediated proteolysis, and the mitochondrial oxidative
   phosphorylation and electron transport chain. Our data indicate that
   VPA may modulate the differential expression of proteins involved in
   mitochondrial function and vitamin D receptor-mediated chromatin
   transcriptional regulation and proteins implicated in the pathogenesis
   of neurodegenerative diseases.

   Keywords: valproic acid, proteomics, epigenetics, pathway analysis

1. Introduction

   Valproic acid (VPA), a branched-chain fatty acid, is widely used as an
   antiepileptic drug (AED) and an antimanic agent [[36]1]. VPA is
   frequently used as an antimanic agent compared to lithium with a narrow
   therapeutic window and the strict limitations of using other
   medications that alter renal function, such as angiotensin-converting
   enzyme inhibitors or nonsteroidal anti-inflammatory drugs [[37]2].
   However, hepatic function and the blood level of VPA are monitored from
   starting use due to its side effects of liver or pancreas problems,
   such as loss of appetite, upper stomach pain, ongoing nausea or
   vomiting, dark urine, swelling in the face, or jaundice [[38]3]. VPA
   also acts as a mood stabilizer in bipolar disorder and as a
   prophylactic for the prevention of migraine [[39]4]. Besides, VPA is
   associated with anti-cancer activity [[40]5,[41]6]. Although VPA is
   generally safe to use and is broadly applicable, its mechanisms of
   action remain controversial. VPA can potentiate the inhibitory effect
   of γ-aminobutyric acid (GABA) transmission, possibly by enhancing GABA
   synthesis and blocking metabolism [[42]7]. In addition, VPA may reduce
   voltage-gated sodium channel activity and modulate N-methyl-d-aspartate
   (NMDA) receptor function, effects that could explain its efficacy as
   both an antiepileptic drug and an antimanic agent [[43]8].

   There is growing evidence that the therapeutic effects of VPA involve
   the modulation of multiple protein signaling pathways. For example,
   microarray analysis of rat brain genes identified a variety of protein
   pathways modulated by VPA, including synaptic transmission, ion
   channels and transport, G-protein signaling, lipid, glucose, and amino
   acid metabolism, transcriptional and translational regulation, the
   phosphoinositol cycle, protein kinases and phosphatases, and apoptosis
   [[44]9]. Du et al. reported in their study that the downregulation of
   transient receptor potential channel 1 (TRPC1) gene expression and
   function in neurons may be involved in the mood-stabilizing action of
   VPA [[45]10], while Jiang et al. suggested that the downregulation of
   Homer1b/c is a potential mechanism for the neuroprotective effects of
   VPA and lithium [[46]11].

   Nevertheless, many of the detailed molecular mechanisms of VPA action
   are to date incompletely described or unknown. Comparative proteomics
   has emerged as a powerful tool in identifying differentially expressed
   proteins in biological systems under “experimental vs. control” or
   “disease vs. normal” conditions [[47]12]. The identification of
   differentially expressed proteins can provide new insights into the
   mechanisms of biological processes such as development, cancer
   formation, and drug treatment effect. Isobaric tags for relative and
   absolute quantitation (iTRAQ) shotgun proteomic analysis is an isobaric
   labeling method used in quantitative proteomics by tandem mass
   spectrometry in order to identify multiple proteins from different
   functional classes in a single experiment [[48]13]. The iTRAQ proteomic
   analysis is non-targeted and so allow for the unbiased investigation of
   numerous novel candidates across multiple biological pathways
   [[49]14,[50]15].

   In the present study, we employed iTRAQ shotgun proteomic analysis,
   real-time quantitative PCR (RT-qPCR), and immunoblotting analysis to
   identify novel differentially expressed proteins in SH-SY5Y
   neuroblastoma cells treated with VPA. The SH-SY5Y neuroblastoma cells
   were derived from SK-N-SH cells, which were initially established from
   a bone marrow biopsy of a neuroblastoma patient [[51]16]. The
   undifferentiated SH-SY5Y cells have been widely used as a model to
   study metabolism and neuroblastoma tumorigenesis [[52]17,[53]18].
   Moreover, the undifferentiated SH-SY5Y cells have been used to study
   the effects of mood stabilizers and antipsychotic drugs, and
   Parkinson’s disease [[54]19,[55]20,[56]21].

2. Materials and Methods

2.1. Cell Culture and VPA Treatment

   The human SH-SY5Y neuroblastoma cell line (ATCC: CRL-2266) was cultured
   in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10%
   fetal bovine serum (FBS), penicillin (100 units/mL), streptomycin (100
   μg/mL), and L-glutamine (2 mM). The culture was maintained in a
   humidified atmosphere of 5% CO[2] at 37 °C, and the medium was changed
   every two to three days. VPA sodium salt was purchased from
   Sigma-Aldrich (P4543). A stock solution was made in phosphate-buffered
   saline (PBS) and diluted in the medium to the final concentration
   indicated. An equal volume of PBS was added to control cultures.

2.2. Cell Viability Assay

   The SH-SY5Y cells were plated in 96-well plates in a 10,000/well
   incubated for 24 h and then treated with VPA at the indicated
   concentrations for 24 h in serum-free DMEM. After VPA treatment, the
   cells were washed twice with PBS and cultured in DMEM with 10% FBS for
   two days. Subsequently, the cells were incubated with 0.5 mg/mL of the
   3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT)
   (Sigma Chemical Co., St. Louis, MO, USA) in DMEM for 4 h. Viable cells
   converted MTT to formazan, which is blue-purple when dissolved in
   dimethyl sulfoxide. Therefore, the color intensity (absorbance) is
   proportional to cell number. Absorbance at 545 nm was measured using a
   microplate reader (Varioskan Flash, Thermo Fisher Scientific, Vantaa,
   Finland) and % survival was calculated by dividing the absorbance of
   VPA-treated samples by absorbance of corresponding PBS-treated
   controls.

2.3. Protein Sample Preparation and iTRAQ Peptide Labeling

   Cells treated as indicated were washed twice with cold PBS,
   re-suspended in lysis buffer containing 20 mM HEPES (pH = 7.6), 7.5 mM
   NaCl, 2.5 mM MgCl[2], 0.1 mM EDTA, 0.1% TritonX-100, 0.1 mM Na[3]VO[4],
   50 mM NaF, and protease inhibitor cocktail (one mini tablet/10 mL,
   Roche Diagnostics GmbH, Mannheim, Germany). The homogenates were
   centrifuged at 13,000 rpm for 30 min at 4 °C, and the supernatants were
   stored at −80 °C until use.

   Each specimen was taken out from −80 °C freezer, and the protein amount
   of cell lysate was determined using a Pierce BCA protein assay kit
   (Thermo Fisher Scientific, Rockford, IL, USA). The 4 protein samples of
   each group with an equal amount (20 μg per sample) were subjected to
   reduction (5 mM tris-(2-carboxyethyl)-phosphine, Sigma-Aldrich, Saint
   Louis, MO, USA), cysteine-blocking (10 mM methyl methanethiosulfonate,
   Sigma-Aldrich), and trypsin (0.8 μg, Promega) digestion at 37 °C for 16
   h in a solution containing 200 mM triethylammonium bicarbonate buffer.
   The peptides were then labeled with iTRAQ reagent (Applied Biosystems,
   Foster City, CA, USA) according to the manufacturer’s protocol. After
   incubation at room temperature for 1 h, the 4 labeled peptide mixtures
   were pooled, dried by vacuum centrifugation, and stored at −80 °C until
   use.

2.4. LC-MS/MS Analysis

   The dried peptide mixtures were reconstituted and desalted using a
   homemade C18 microcolumn. The resulted peptides were reconstituted and
   loaded onto a homemade column (Luna SCX 5 μm, 0.5 × 120 mm) at a flow
   rate of 3 μL/min for 40 min. The peptides were then eluted fractionated
   to 22 fractions by eluting with 0 to 100% HPLC buffer B (0.5 M ammonium
   chloride/30% acetonitrile/0.1% formic acid) using on-line 2D-HPLC
   (Dionex Ultimate 3000, Thermo Fisher, San Jose, CA, USA). Each SCX
   fraction was diluted in-line before trap onto a reverse-phase column
   (Zorbax 300SB-C18, 0.3 × 5 mm; Agilent Technologies, Santa Clara, CA,
   USA). The peptides were then separated on a homemade column (HydroRP
   2.5 μm, 75 μm I.D. × 24 cm with a 15 μm tip) using a multi-step
   gradient of HPLC buffer C (99.9% acetonitrile/0.1% formic acid) for 65
   min with a flow rate of 0.23 μL/min. The LC apparatus was coupled with
   a 2D linear ion trap mass spectrometer (LTQ-Orbitrap ELITE; Thermo
   Fisher, San Jose, CA, USA) operated using Xcalibur 2.2 software (Thermo
   Fisher, San Jose, CA, USA). The full-scan MS was performed in the
   Orbitrap over a range of 400 to 1600 Da and a resolution of 120,000 at
   m/z 400. For proteome analysis, the 12 data-dependent MS/MS scan events
   (HCD) were followed by one MS scan for the 12 most abundant precursor
   ions in the preview MS scan. The m/z values selected for MS/MS were
   dynamically excluded for 80 s with a relative mass window of 20 ppm.
   The electrospray voltage was set to 2.0 kV, and the temperature of the
   capillary was set to 200 °C. MS and MS^n automatic gain control were
   set to 1000 ms (full scan), 300 ms (MS^2 for HCD), or 3 × 10^6 ions
   (full scan), and 3 × 10^4 ions (MS^2 for HCD) for maximum accumulated
   time or ions, respectively.

2.5. Protein Identification

   Liquid chromatography-tandem mass spectrometry (LC-MS/MS) data were
   analyzed using Proteome Discoverer software (version 1.4, Thermo Fisher
   Scientific), including the reporter ions quantifier node for iTRAQ
   quantification. The MS/MS spectra were searched against the UniProt
   sequence database (released on 16 March 2016, extracted for Homo
   sapiens, 20,199 entries) using the Mascot search engine (Matrix
   Science, London, UK; version 2.5). For peptide identification, 10 ppm
   of mass tolerance was permitted for intact peptide masses and 0.05 Da
   for HCD fragment ions with allowance for two missed cleavages made by
   trypsin digestion: oxidized methionine; acetyl (protein N-terminal);
   iTRAQ4plex (N-terminal); and iTRAQ4plex (lysine), as variable
   modifications and methylthio (cysteine) as the fixed modification. The
   protein ratio was calculated using the intensity of iTRAQ reporter ions
   (114, 115, 116, and 117) in the spectrum, which matched to the specific
   peptide in the sequence database. If one of the 4 channels was missing
   and the missing value was not replaced (e.g., replaced to 0 or minimal
   intensity of reporter ion in the LC-MS/MS experiment), the protein
   ratio would be left to blank. The false discovery rate of
   peptide-spectrum matching was calculated using Percolator (q value <
   0.01) [[57]22]. Protein ratio with at least two peptide identifications
   was normalized based on the median ratio of all proteins in each group.
   After ratio normalization, the mean ± 1.5 × standard deviation of the
   log2 transformed ratio was used as thresholds for differential protein
   selection.

2.6. Pathway Analysis

   DAVID (the database for annotation, visualization and integrated
   discovery) is a web-based bioinformatics resource intended for
   functional interpretation of large lists of genes or proteins [[58]23].
   In total, 1195 differentially expressed proteins (mean ± 1.5 × standard
   deviation (SD) of the log2) were subjected to the DAVID database using
   the official gene symbol method (DAVID v6.8) [[59]24]. Differentially
   expressed proteins associated with the Biological Biochemical Image
   Database (BBID), Biocarta, and Kyoto Encyclopedia of Genes and Genomes
   (KEGG) pathways were highlighted, annotated, and viewed in pathway maps
   [[60]25,[61]26,[62]27].

2.7. Immunoblotting Analysis

   Cellular protein was mixed with sample buffer (62.5 mM Tris-HCl pH =
   6.8, 2% SDS, 25% glycerol, 0.01% Bromophenol Blue, 5%
   β-mercaptoethanol) and denatured by heating at 95 °C for 10 min.
   Immunoblotting analysis was performed using standard protocols with the
   following primary antibodies: rabbit anti-ATP5D (A9929, ABclonal,
   Woburn, MA, USA); rabbit anti-ATP5J (tcea21587, Taiclone, Taipei,
   Taiwan); rabbit anti-DPH6 (23993-1-AP, Proteintech, Rosemont, IL, USA);
   rabbit anti-FBXO4 (tcea12933, Taiclone); rabbit anti-FSTL1 (A15789,
   ABclonal); mouse anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
   (G8795, Sigma-Aldrich); mouse anti-HDAC1 (GTX100513, GeneTex, Hsinchu
   City, Taiwan); rabbit anti-HDAC2 (GTX112957, GeneTex); rabbit
   anti-HIST1H2B (MAB15119, Abnova, Taipei, Taiwan); rabbit anti-KCNAB2
   (17890-1-AP, Proteintech); rabbit anti-NPTX2 (A12031, ABclonal); rabbit
   anti-SCG3 (A7799, ABclonal); rabbit anti-SMARCA4 (tcea559, Taiclone);
   rabbit anti-UBE1L (tcea10708, Taiclone); rabbit anti-UBE2D1 (tcea7105,
   Taiclone); rabbit anti-UQCRB (tcea18400, Taiclone); rabbit anti-VDAC1
   (tcea521, Taiclone); rabbit anti-VSNL1 (A6999, ABclonal). Horseradish
   peroxidase-conjugated donkey anti-rabbit IgG (NA934, Amersham, Little
   Chalfont, UK) or goat anti-mouse IgG (074-186, KPL) were used as the
   secondary antibodies. Immunolabeled proteins were visualized using the
   ECL detection system (GE Health Care Bio-Sciences AB, Uppsala, Sweden).
   The membrane was then stained by submersion in Amido Black (0.1% [w/v]
   Naphthol blue-black in 10% methanol, 2% acetic acid, Sigma-Aldrich,
   Saint Louis, MO, USA) for the control of protein loading and transfer
   efficiency. Immunoblot intensity was assessed using NIH ImageJ software
   (NIH image) [[63]28]. Treatment group means were compared by analysis
   of variance (ANOVA). A p < 0.05 was considered significant for all
   tests.

2.8. Total RNA Preparation and RT-qPCR

   Total RNA of cells was purified using TRIzol reagent (Invitrogen)
   according to the manufacturer’s standard protocol and stored at −80 °C
   until use. cDNA was prepared using SuperScript^TM III Reverse
   Transcriptase (Invitrogen) according to the manufacturer’s standard
   protocol. The RT-qPCR was performed using an Applied Biosystems
   PRISM7900 Sequence Detection System. We used a comparative ^ΔΔCt method
   to validate the differential gene expression. The ^ΔΔCt value of each
   sample was obtained by subtracting the Ct of the target gene from the
   Ct of the GAPDH gene of the sample as the endogenous gene for
   normalization and then from the ^ΔCt of the first sample. Differences
   of fold change (2^^−ΔΔCT) for selected genes among four groups were
   assessed by ANOVA test. All data were expressed as mean ± SD.
   Statistically significant differences were defined as those with a
   p-value <0.05. All experiments were performed in at least triplicate.

3. Results

3.1. Effects of VPA on SH-SY5Y Cell Viability

   The effects of VPA (0.2, 2, and 20 mM for 24 h) on SH-SY5Y cell
   viability were examined using the MTT assay ([64]Figure 1). The %
   survival relative to PBS-treated controls was 75% after exposure to 20
   mM VPA for 24 h (p = 0.015) and 113% after exposure to 0.2 mM VPA (p =
   0.038). Therefore, high-dose VPA had modest inhibitory effects on cell
   survival and/or proliferation, while low-dose VPA appeared to promote
   survival and/or proliferation.

Figure 1.

   [65]Figure 1
   [66]Open in a new tab

   Valproic acid (VPA) has reciprocal concentration-dependent effects on
   SH-SY5Y cell viability. The 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyl
   tetrazolium bromide (MTT) assay was used to estimate cell number
   following 24 h treatment with the indicated VPA concentration. The
   results were calculated as the 545 nm optical density (OD545) of
   VPA-treated cultures relative to phosphate-buffered saline
   (PBS)-treated control cultures and expressed as means ± SD. VPA
   enhanced survival or proliferation at the low concentration and reduced
   survival or proliferation at the high concentration. A statistically
   significant difference between VPA- and PBS-treated cultures was
   detected by analysis of variance (ANOVA), and a post hoc test using the
   least significant difference (LSD) setting was performed (* p < 0.05, n
   = 24).

3.2. Identification of Differentially Expressed Proteins by the iTRAQ
Analysis

   The iTRAQ shotgun proteomic analysis method was used to identify
   proteins differentially expressed by SH-SY5Y cells exposed to different
   VPA concentrations for 24 h. Among 5325 protein identifications in the
   iTRAQ-labeled samples, 5288 proteins were able to be quantified in all
   groups ([67]Table 1). Global normalization [[68]29] was used to correct
   the starting protein amount in different treatments. By assuming the
   most proteins in different treatments remained unchanged, all protein
   ratios were normalized using the median ratio in each group. Protein
   ratio with at least two peptide identifications was normalized based on
   the median ratio of all proteins in each group ([69]Tables S1 and S2).
   After ratio normalization, mean ± 1.5 × SD of log2 transformed ratio
   was used as thresholds for differential protein selection. After global
   normalization, a total of 1195 differentially expressed proteins (mean
   ± 1.5 × SD) were subjected to pathway enrichment analysis. These
   differentially expressed proteins are listed in the [70]Supplementary
   Table S3.

Table 1.

   Summary of quantification results.
                      # Proteins       0.2 mM VPA/PBS 2 mM VPA/PBS 20 mM VPA/PBS
                 Total Quantification       5293          5295         5294
   ≥mean + 1.5 × s.d.
   (fold change) 213
   (≥1.35) 261
   (≥1.28) 302
   (≥1.39)
   ≤mean − 1.5 × s.d.
   (fold change) 229
   (≤0.73) 332
   (≤0.76) 280
   (≤0.72)
                   ≥1.5 fold change          83            49           176
                   ≤1.5 fold change         122           117           153
   [71]Open in a new tab

   Because we were not able to validate all the differentially expressed
   proteins identified from iTRAQ assay, 2 down-expressed proteins (KRT6A
   and KRT14) and 7 up-expressed proteins (ATP5D, DPH6, FSTL1, KCNAB2,
   NPTX2, SCG3, and VSNL1) in all treated groups were selected for
   verification with biological replicated SH-SY5Y cells using the
   immunoblotting assay. However, KRT6A and KRT14 protein were not well
   detected in the SH-SY5Y cells (data not shown). We compared the
   expression levels of seven up-expressed proteins in biological
   replicated SH-SY5Y cells treated with either PBS or VPA (0.2, 2, and 20
   mM) using immunoblotting analysis. The fold differences in the
   expression of these proteins between VPA-treated groups and PBS control
   were calculated ([72]Figure 2A,B). We were able to verify 6 proteins
   (ATP5D, DPH6, FSTL1, KCNAB2, NPTX2, and SCG3) that were significantly
   up expressed by the treatment of VPA.

Figure 2.

   [73]Figure 2
   [74]Open in a new tab

   Immunoblotting analysis to validate the differential expression for 7
   proteins in VPA-treated and PBS-treated biological replicated SH-SY5Y
   cells. (A) Immunoblotting showed the expression of ATP5D, DPH6, FSTL1,
   KCNAB2, NPTX2, SCG3, and VSNL1 in VPA-treated and PBS-treated SH-SY5Y
   cells. (B) Quantification showing the fold differences in the
   expression of ATP5D, DPH6, FSTL1, KCNAB2, NPTX2, SCG3, and VSNL1
   between VPA-treated groups and PBS control. Glyceraldehyde 3-phosphate
   dehydrogenase (GAPDH) was a loading control. Data are expressed as fold
   change to PBS ± SD (* p < 0.05, n = 3).

3.3. Pathway Enrichment Analysis

   Next, the proteins with altered expression in each group were submitted
   to a pathway analysis using the functional annotation of DAVID 6.8. The
   BBID, Biocarta, and KEGG pathway analyses showed that differentially
   expressed proteins in the different VPA concentrations associated with
   several pathways and diseases ([75]Supplementary Table S4). After
   Benjamini multiple correlations (p < 0.05), many differentially
   expressed proteins were found to be involved in the control of gene
   expression via the vitamin D receptor, oxidative phosphorylation,
   electron transport in mitochondria, ribosome biogenesis in eukaryotes,
   Parkinson’s disease, ubiquitin-mediated proteolysis, and Alzheimer’s
   disease ([76]Table 2). VPA altered the expression of 11 proteins
   (ATP5A1, ATP5B, ATP5D, ATP5F1, ATP5J, SDHA, SDHB, UQCRB, UQCRC1,
   UQCRC2, and UQCRH) involved in Alzheimer’s disease ([77]Figure S1A); 16
   proteins (ATP5A1, ATP5B, ATP5D, ATP5F1, ATP5J, SDHA, SDHB, SLC25A6,
   UBA7, UQCRB, UQCRC1, UQCRC2, UQCRFS1, UQCRH, VDAC1, and VDAC3) involved
   in Parkinson’s disease ([78]Figure S1B); 13 proteins (ATP5A1, ATP5B,
   ATP5D, ATP5F1, ATP5J, ATP5L, ATP6V1G1, SDHA, SDHB, UQCRB, UQCRC1,
   UQCRC2, and UQCRH) involved in oxidative phosphorylation ([79]Figure
   S1C); 5 proteins (ATP5A1, SDHA, SDHB, SLC25A6, and UQCRC1) involved in
   electron transport reaction in mitochondria (Electron Transprot Chain)
   [[80]30]; 9 proteins (BAZ1B, HDAC1, NCOA2, SMARCA4, SMARCC1, SMARCD1,
   SMARCE1, SUPT16H, and TOP2B) involved in the control of gene expression
   by vitamin D receptor [[81]31]; 10 proteins (ANAPC4, CDC23, FBXO2,
   FBXO4, NEDD4, TRAF6, UBE2D1, UBE2E1, UBE2E3, and UBE2G2) involved in
   ubiquitin mediated proteolysis ([82]Figure S1D); 10 proteins (BMS1,
   DKC1, GNL2, GTPBP4, LSG1, NOP10, RIOK2, TCOF1, WDR36, and WDR75)
   involved in ribosome biogenesis in eukaryotes ([83]Figure S1E).

Table 2.

   Pathway enrichment.
   Group Trend Category Term Count Proteins
   0.2 mM/PBS Down KEGG_pathway: map04120 Ubiquitin mediated proteolysis
   10 UBE2E3, NEDD4, UBE2G2, ANAPC4, FBXO2, CDC23, FBXO4, TRAF6, UBE2D1,
   UBE2E1
   2 mM/PBS Up KEGG_pathway: map05012 Parkinson’s disease 10 UQCRC2,
   ATP5D, UQCRH, ATP5F1, UBA7, UQCRFS1, VDAC3, VDAC1, ATP5J, UQCRB
   2 mM/PBS Down BIOCARTA Control of Gene Expression by Vitamin D Receptor
   ^# 9 NCOA2, BAZ1B, SMARCE1, HDAC1, SMARCC1, SMARCD1, SUPT16H, TOP2B,
   SMARCA4
   2 mM/PBS Down KEGG_pathway: map03008 Ribosome biogenesis in eukaryotes
   10 WDR75, GTPBP4, WDR36, DKC1, TCOF1, LSG1, GNL2, RIOK2, BMS1, NOP10
   20 mM/PBS Up BIOCARTA Electron Transport Reaction in Mitochondria ^$ 5
   SDHA, SDHB, UQCRC1, SLC25A6, ATP5A1
   20 mM/PBS Up KEGG_pathway: map00190 Oxidative phosphorylation 13 ATP5D,
   UQCRC2, UQCRC1, ATP5B, ATP5F1, ATP6V1G1, SDHA, SDHB, UQCRH, ATP5L,
   ATP5A1, UQCRB, ATP5J
   20 mM/PBS Up KEGG_pathway: map05012 Parkinson’s disease 12 UQCRC2,
   SDHA, ATP5D, SDHB, UQCRC1, UQCRH, ATP5B, SLC25A6, ATP5F1, ATP5A1,
   ATP5J, UQCRB
   20 mM/PBS Up KEGG_pathway: map05010 Alzheimer’s disease 11 UQCRC2,
   SDHA, ATP5D, SDHB, UQCRC1, UQCRH, ATP5B, ATP5F1, ATP5A1, ATP5J, UQCRB
   [84]Open in a new tab

   KEGG website: [85]https://www.genome.jp/kegg/. ^#
   [86]https://www.gsea-msigdb.org/gsea/msigdb/cards/BIOCARTA_VDR_PATHWAY;
   ^$
   [87]https://www.gsea-msigdb.org/gsea/msigdb/cards/BIOCARTA_ETC_PATHWAY.

3.4. Validation of Three Differentially Expressed Epigenetic Proteins by
Immunoblotting

   Because VPA is an inhibitor of histone deacetylases (HDACs), we
   selectively performed immunoblotting analysis on three epigenetic
   proteins (HDAC1, HDAC2, and HIST1H2B) identified by iTRAQ assay.
   Immunoblotting analysis revealed that VPA reduced HDAC1 expression at 2
   mM ([88]Figure 3A) and HDAC2 at 2 and 20 mM ([89]Figure 3B), while VPA
   upregulated HIST1H2B expression at 2 and 20 mM ([90]Figure 3C).

Figure 3.

   [91]Figure 3
   [92]Open in a new tab

   VPA induces the differential expression of the epigenetic regulators
   HDAC1, HDAC2, and HIST1H2B in SH-SY5Y cells. (A–C) Representative
   immunoblots and quantification of target protein band intensities for
   lysates from SH-SY5Y cultures treated with the indicated VPA
   concentration or PBS (control). (A) HDAC1. (B) HDAC2. (C) HIST1H2B. The
   results are expressed as means ± SD. A statistically significant
   difference between VPA and PBS was detected by ANOVA (* p < 0.05, n = 5
   to 10).

3.5. Immunoblotting and RT-qPCR Analysis in Biological Replicated SH-SY5Y
Cells

   Nine differentially expressed proteins and 14 differentially expressed
   genes on enriched pathways were selected for verification with
   biological replicated SH-SY5Y cells. We compared the expression levels
   of nine proteins in biological replicated SH-SY5Y cells treated with
   either PBS or VPA (0.2, 2, and 20 mM) using immunoblotting analysis.
   The fold differences in the expression of these proteins between
   VPA-treated groups and PBS control were calculated ([93]Figure 4A,B).
   Seven proteins (ATP5J, FBXO4, HDAC1, HDAC2, SMARCA4, UQCRB, and VDAC1)
   were differentially expressed by the treatment of VPA ([94]Figure 4B).
   We compared the mRNA expression levels of 14 genes in biological
   replicated SH-SY5Y cells treated with either PBS or VPA (0.2, 2, and 20
   mM) using RT-qPCR assay. The fold differences between VPA-treated
   groups and PBS control of these 14 genes are shown in [95]Figure 5.
   Primer sequences and the size of each amplicon are listed in the
   [96]Supplementary Table S5. Among the 14 genes assayed, we found that
   eight genes were differentially expressed by the treatment of VPA,
   including BAZ1B, FBXO2, HDAC1, SMARCA4, UBA7, UBE2D1, UQCRB, and WDR36.
   In the HDAC1, UBE2D1, UQCRB, and WDR36 gene, change in mRNA levels and
   protein levels were opposite by the treatment of VPA.

Figure 4.

   [97]Figure 4
   [98]Open in a new tab

   Immunoblotting analysis to validate the differential expression for 9
   proteins in VPA-treated and PBS-treated biological replicated SH-SY5Y
   cells. (A) Immunoblotting showed the expression of ATP5J, FBXO4, HDAC1,
   HDAC2, SMARCA4, UBE1L, UBE2D1, UQCRB, and VDAC17 in VPA-treated and
   PBS-treated SH-SY5Y cells. (B) Quantification showing the fold
   differences in the expression of ATP5J, FBXO4, HDAC1, HDAC2, SMARCA4,
   UBE2D1, UQCRB, and VDAC1 between VPA-treated groups and PBS control.
   GAPDH was a loading control. The data are expressed as fold change to
   PBS ± SD (* p < 0.05, n = 3 to 6).

Figure 5.

   [99]Figure 5
   [100]Open in a new tab

   RT-qPCR assay showing the expression of 14 genes (ATP5F1A, ATP5PF,
   BAZ1B, BMS1, FBXO2, HDAC1, HDAC2, SDHA, SMARCA4, UBA7, UBE2D1, UQCRB,
   VDAC1, and WDR36) in VPA-treated and PBS-treated SH-SY5Y cells. The
   GAPDH gene was used as the endogenous gene for normalization. The data
   are expressed as fold change to PBS ± SD (* p < 0.05, n = 3 to 6).

4. Discussion

   VPA is an inhibitor of histone deacetylases (HDACs) and has pleiotropic
   effects on many biological processes such as cancer cell growth arrest,
   proliferation, differentiation, invasion, apoptosis, immune system, and
   neuroprotection [[101]32,[102]33,[103]34,[104]35,[105]36,[106]37].
   Several studies indicate that HDAC inhibitors may be promising
   therapeutic agents for neurodegenerative diseases [[107]38,[108]39].
   Recently, VPA emerges as a novel potent anti-cancer drug potentially
   applicable for clinical practice [[109]40]. However, the comprehensive
   mechanisms underlying VPA anti-cancer effects and neuroprotection
   remain to be elucidated due to its pleiotropic effects on multiple
   signaling pathways. In line with the findings from the present study
   and other research groups [[110]41,[111]42], we used proteomic shotgun
   analysis to identify multiple differentially expressed proteins in
   VPA-treated SH-SY5Y neuroblastoma cells. Among these differentially
   expressed proteins, we were able to confirm six proteins that are
   differentially up expressed under the administration of VPA. However,
   two down-expressed proteins (KRT6A and KRT14) were not well detected in
   the SH-SY5Y cells. The results may be due to the antibody specificity.
   According to subsequent pathway analyses, VPA appears to regulate a
   myriad of physiological and two neurodegenerative disease-associated
   signaling pathways, including Parkinson’s disease and Alzheimer’s
   disease. Immunoblotting and RT-qPCR confirmed that VPA regulated a
   subset of proteins and genes enriched in these pathways. However, in
   the HDAC1, UBE2D1, UQCRB, and WDR36 gene, change in mRNA levels and
   protein levels were opposite in SH-SY5Y cells treated with VPA. The
   lack of correlation between mRNA and protein abundances can be
   explained by transcriptional regulation, translation efficiency, and
   protein turnover rates [[112]43,[113]44].

   In the current study, VPA altered the expression of 11 proteins
   implicated in Alzheimer’s disease (ATP5A1, ATP5B, ATP5D, ATP5F1, ATP5J,
   SDHA, SDHB, UQCRB, UQCRC1, UQCRC2, and UQCRH), while 16 proteins
   implicated in Parkinson’s disease (ATP5A1, ATP5B, ATP5D, ATP5F1, ATP5J,
   SDHA, SDHB, SLC25A6, UBA7, UQCRB, UQCRC1, UQCRC2, UQCRFS1, UQCRH,
   VDAC1, and VDAC3). Of note, except UBA7, above all differentially
   expressed proteins are implicated in mitochondrial oxidative
   phosphorylation chain. Furthermore, our findings identified targeting
   mitochondrion proteins as an approach to alleviate Alzheimer’s disease
   and Parkinson’s disease. Alzheimer’s disease is a complex
   neurodegenerative disease with multifaceted pathogenesis that usually
   starts slowly and gradually worsens over time [[114]45]. Both in vitro
   and in vivo studies have shown strong modulatory actions of VPA on
   neurogenesis, neurite outgrowth, and neuronal survival
   [[115]19,[116]46]. VPA promoted neurogenesis and GSK-3β dependent
   neurite outgrowth in a mouse model of Alzheimer’s disease
   [[117]47,[118]48]. Therefore, VPA may be useful for the treatment of
   behavioral disturbance in dementia patients
   [[119]49,[120]50,[121]51,[122]52]. Parkinson’s disease is a progressive
   neurodegenerative disorder that primarily affects the motor control
   system. Animal model studies support the therapeutic potential of VPA
   may be Parkinson’s disease [[123]53]. Clinically, VPA is used to treat
   dopamine dysregulation syndrome in Parkinson’s disease [[124]54],
   although the molecular mechanisms underlying this therapeutic effect
   are to date unclear.

   Interestingly, we found that VPA altered the expression levels of 15
   mitochondrial proteins involved in oxidative phosphorylation (ATP5A1,
   ATP5B, ATP5D, ATP5F1, ATP5J, ATP5L, ATP6V1G1, SDHA, SDHB, UQCRB,
   UQCRC1, UQCRC2, and UQCRH) and five involved in mitochondrial electron
   transport chain (ATP5A1, SDHA, SDHB, SLC25A6, and UQCRC1). The
   mitochondrion is a double-membrane-bound organelle and produces ATP
   from products of the citric acid cycle, fatty acid oxidation, and amino
   acid oxidation. In eukaryotes, electron transport reaction is found in
   the inner mitochondrial membrane, where it serves as the site of
   oxidative phosphorylation. Mitochondrion dysfunction has been suggested
   as a contributory factor in the pathology of various neuropsychiatric
   disorders, including neurodegenerative disorders and neurodevelopmental
   disorders [[125]55,[126]56,[127]57]. Taken together, the clinical
   efficacy of VPA may be related to mitochondrial electron transport or
   oxidative phosphorylation chain in patients with neurodegenerative
   disease.

   In the current study, VPA altered the expression of nine proteins
   involved in the control of gene expression by vitamin D receptor
   (BAZ1B, HDAC1, NCOA2, SMARCA4, SMARCC1, SMARCD1, SMARCE1, SUPT16H, and
   TOP2B). Notably, SMARCA4, SMARCC1, SMARCD1, and SMARCE1 are members of
   the SWI/SNF family of proteins, whose members display helicase and
   ATPase activities and which are thought to regulate transcription of
   specific genes by altering the chromatin structure around those genes
   [[128]58]. VPA is an inhibitor of HDACs, which together with histone
   acetyltransferases control the state of histone acetylation, suggesting
   that the epigenetic regulation of gene expression may contribute to
   effects of VPA on neuropsychiatric diseases [[129]33,[130]59]. In the
   present study, the immunoblotting assay confirmed that VPA reduced
   HDAC1 and HDAC2 expression and induced HIST1H2B expression. The results
   suggest that the inhibition of HDACs and the concomitant enhancement of
   histone acetylation (which facilitates gene transcription) contribute
   to the clinical efficacy of VPA. Vitamin D is essential for bone
   health, and vitamin D deficiency is linked to osteoporosis and reduced
   mineral density [[131]60]. Clinical observations have shown that
   chronic VPA use has variable effects on bone metabolism. For example,
   the long-term use of AEDs, including VPA, has been associated with
   decreased bone mineral density [[132]61], including in young epilepsy
   patients [[133]62]. However, the mechanisms for this potential side
   effect are unclear. A previous iTRAQ proteomic study reported that VPA
   reduces collagen and osteonectin in a cellular model of spinal muscular
   atrophy [[134]63]. Vitamin D receptor regulates chromatin structure in
   cooperation with various histone modifiers and chromatin remodelers
   [[135]64,[136]65]. Taken together, we speculate that VPA actions on
   bone metabolism are associated with vitamin D receptor-mediated
   chromatin transcriptional regulation of target genes.

   In addition, we found that VPA can modulate the expression of proteins
   associated with ubiquitin-mediated proteolysis and ribosome biogenesis,
   which is in accordance with whole-genome microarray analysis of primary
   tumors from VPA-treated patients showing upregulation of several
   hundred genes, including genes that are related to ribosomal proteins
   and ubiquitin-mediated proteolysis [[137]66].

   In the present study, we observed that VPA could modulate the
   differential expression of several proteins in SH-SY5Y cells by iTRAQ
   shotgun proteomic analysis. However, this study has several
   limitations. Firstly, we did not perform a biological replicate in the
   iTRAQ experiment. Thus, we cannot account for biological variations
   among different cell passages. Secondly, the sample size is small and
   was conducted using only the SH-SY5Y cell line. Thirdly, differential
   protein expression was confirmed by immunoblotting and RT-qPCR assay
   for only a small subset of proteins. Hence, to obtain more insight into
   the molecular mechanism of VPA, we need to verify other genes and
   proteins in the future, and further replicated studies are needed to
   consolidate the findings in this study.

5. Conclusions

   In conclusion, by using iTRAQ analysis, immunoblotting, and RT-qPCR
   analysis, we observed that VPA could induce the differential expression
   of several proteins in SH-SY5Y cells. The identification of these
   differentially expressed proteins both enhances our knowledge of the
   molecular mechanisms underlying the clinical effects of VPA, such as
   the ubiquitin-mediated proteolysis, ribosome biogenesis in eukaryotes,
   chromatin-mediated transcriptional regulation, mitochondria electron
   transport reaction, and mitochondria oxidative phosphorylation. It
   provides potentially novel pathogenic mechanisms and treatment targets
   for various neurodegenerative diseases and cancers.

Acknowledgments