Valproate-induced alterations in human theca cell gene expression: clues to the association between valproate use and metabolic side effects

Jennifer R. Wood1, Velen L. Nelson-Degrave2, Erik Jansen3, Jan M. McAllister2, Sietse Mosselman4 and Jerome F. Strauss, III1

1 Center for Research on Reproduction and Women’s Health, University of Pennsylvania, Philadelphia
2 Department of Cellular and Molecular Physiology, Pennsylvania State University College of Medicine, Hershey, Pennsylvania
3 Global Business Intelligence Center
4 Department of Pharmacology, NV Organon, Oss, The Netherlands


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Valproic acid (VPA) is an anti-epileptic drug that has been associated with polycystic ovary syndrome (PCOS)-like symptoms, including increased ovarian androgen production. The hyperandrogenemia likely reflects the stimulatory action of VPA on theca cell androgen synthesis and has been correlated to its activity as a histone deacteylase inhibitor in these cells. To determine whether VPA induces a PCOS-like genomic phenotype, we compared the gene expression profiles of untreated (UNT) normal, VPA-treated normal, and UNT PCOS theca cells. Hierarchal cluster analysis demonstrated similarities in the gene expression profiles of VPA-treated normal and PCOS theca cells. Statistical analysis identified 1,050 transcripts that have significantly altered mRNA abundance in both VPA-treated normal and UNT PCOS theca cells compared with normal UNT theca cells. Among these 1,050 transcripts were cAMP-GEFII and TRB3, which have increased and decreased mRNA abundance, respectively. The altered abundance of these two mRNAs was correlated to increased basal and insulin-induced phosphorylation of protein kinase B (Akt/PKB). Thus these studies indicate that VPA- and PCOS-induced changes in gene expression enhance Akt/PKB signal transduction in human theca cells. Furthermore, common changes in gene expression in PCOS and VPA-treated normal theca cells suggest a possible mechanism for the development of PCOS-like symptoms, including increased steroid synthesis and arrested follicle development in women receiving chronic VPA therapy.

microarray; polycystic ovary syndrome; antiepileptic drug; histone deacetylase inhibitor


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
POLYCYSTIC OVARY SYNDROME (PCOS) is a common endocrine and metabolic disease that is characterized by increased circulating testosterone levels and insulin resistance with associated hyperinsulinemia, resulting in anovulatory infertility (24). Interestingly, women with seizure disease exhibit reduced fertility, including the development of PCOS-like symptoms (3, 30). Although these reproductive defects may be attributed in part to the underlying neuropathy of seizure disease, there are a number of reports in the literature suggesting an association between antiepileptic drug therapy and the development of PCOS-like symptoms (2, 19, 25, 31, 35).

Valproic acid (VPA) is a short-chain fatty acid that acts as an anti-convulsant and mood stabilizer, and thus is widely used to treat several neuropathologies including epilepsy, bipolar disorder, migraine headaches, and chronic neuropathic pain (18). The acute actions of VPA are predominately mediated through its regulation of {gamma}-aminobutyric acid levels and the activity of voltage-dependent Na+ channels, both of which inhibit high-frequency firing of neurons (20). In addition, there are several proposed effects of chronic VPA treatment on cell function. Specifically, VPA modulates intracellular signaling through specific signal transduction cascades, including the mitogen-activated protein kinase (MAPK), protein kinase C (PKC), and protein kinase B (Akt/PKB) pathways (10, 42, 47). Chronic treatment of cells with VPA has also been shown to modulate the activity of transcription factors. For example, VPA enhances c-Fos and c-Jun binding to activator protein (AP)-1-binding sites and increases AP-1-mediated gene expression (1, 5). Likewise, VPA increases peroxisome proliferator-activated receptor (PPAR){delta}-mediated transactivation in F9 teratocarcinoma cells via a ligand-independent mechanism (17, 23, 43). Recent in vitro studies have also identified VPA as a histone deacetylase (HDAC) inhibitor (17, 36). HDACs, which interact with nucleosomes and remove acetyl groups from their core histones, induce a closed conformation in DNA, resulting in repression of transcription and decreased gene expression (9, 40). VPA reversibly inhibits the catalytic activity of class I HDACs, resulting in increased acetylation of histones and leading to altered expression of target genes (17, 36). These direct effects on gene expression may subsequently cause increased or decreased transcription of downstream genes and result in global changes in transcript abundance in VPA-treated cells.

Given the relatively high incidence of seizure disease, which is estimated to occur in 1–2% of the population, and the widespread use of VPA as a therapeutic drug, understanding the molecular mechanisms of VPA-induced alterations on ovarian function represents an important unmet need. Increased ovarian androgen synthesis is a hallmark phenotype of PCOS. Thus we had previously examined the effect of VPA on theca cell steroidogenesis. These studies demonstrated that a pharmacologically relevant dose of VPA (500 µM) increases dehydroepiandrosterone (DHEA) and androstendione ({Delta}4A) synthesis in human theca cells (33). Increased VPA-induced androgen synthesis was correlated to the HDAC-inhibitor activity of VPA and was associated with increased expression of the P450 side-chain cleavage (CYP11A) and 17{alpha}-hydroxylase,17–20 lyase (CYP17) but not steroidogenic acute regulatory protein (StAR) mRNAs (33). Together, these studies indicated that VPA induces a PCOS-like biochemical phenotype in human theca cells. In this study, we have demonstrated that VPA induces changes in the gene expression profile of normal theca cells such that they more closely resemble the gene expression profile of PCOS theca cells. Genes that exhibit common changes in expression in the VPA-treated and PCOS theca cells suggest a possible mechanism for VPA-induced PCOS-like symptoms, including increased androgen synthesis and/or arrested follicle development.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Theca cell culturing and RNA isolation.
Theca cells were isolated from 3- to 5-mm follicles from the ovaries of four normal women and five PCOS patients, and independent cultures were established using the cells isolated from each woman as previously described (32, 44). The diagnosis of PCOS and the steroidogenic capacity of each sample were determined as previously described (32, 44, 48). For microarray hybridizations, fourth-passage cells from the four normal theca cell samples were cultured for 48 h in serum-free medium with no treatment or with 500 µM VPA treatment (Sigma, St. Louis, MO). After treatment, the medium was removed, the cells were washed with phosphate-buffered saline (PBS), and RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA).

Microarray hybridization.
The Affymetrix GeneChip Human Genome U133A and U133B microarray chips (Affymetrix, Santa Clara, CA) were hybridized at the University of Pennsylvania Microarray Core Facility, as previously described (45). Briefly, cRNA, which was generated from four culture-matched untreated and VPA-treated normal theca cell samples, was fragmented and sequentially hybridized to individual Affymetrix U133A and U133B gene array chips. The Affymetrix Microarray Suite (MAS) 5.0 software was used to normalize each chip to a trimmed mean signal of 150 to account for chip-to-chip differences in hybridization efficiency. cRNA from five untreated PCOS theca cell samples was previously hybridized and normalized on the U133A and U133B chips (45).

Gene expression analysis.
Each transcript on the U133A and U133B chips was determined to be present or absent in each theca cell sample using the statistical expression algorithm of the MAS 5.0 software. Each transcript was subsequently identified as expressed in the untreated (UNT) normal (NL), VPA-treated NL, or UNT PCOS theca cell samples if it was called present in at least three of the samples in each group. GeneSpring 7 (Silicon Genetics, Redwood City, CA) was used to normalize and compare the microarray data from each experimental group. Specifically, the fluorescence intensity of each transcript in UNT NL, VPA-treated NL, and UNT PCOS cells was normalized to the median fluorescence intensity of each transcript in UNT NL cells. The mean normalized fluorescence intensity of each transcript in each experimental group and the standard error associated with each transcript’s mean normalized fluorescence intensity were also determined. The profile of the expressed genes in the four UNT NL, four VPA-treated NL, and five UNT PCOS theca cell samples was subjected to hierarchal cluster analysis. Subsequently, statistically significant differences (P < 0.05) in the mean normalized fluorescence intensity of each transcript between the UNT NL and VPA-treated NL samples or UNT PCOS and UNT NL samples were determined by parametric testing, which incorporated all available error estimates. The gene associated with each transcript that was identified as differentially expressed in both the VPA-treated NL and UNT PCOS theca cells was determined. The microarray data described in this study have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) as series GSE1615.

Reverse transcription, PCR, and quantitative RT-PCR.
Reverse transcription and quantitative PCR (QPCR) were carried out as previously described (45). Briefly, total RNA (5 µg) from the culture-matched, UNT, and VPA-treated NL theca cells and UNT PCOS theca cells was pretreated with DNase I (Promega, Madison, WI) and reverse transcribed with Moloney Murine Leukemia Virus (Promega) in the presence of random primers (Promega). The cDNA was subsequently subjected to QPCR amplification for 10 transcripts that were identified as differentially expressed in both VPA-treated NL and UNT PCOS theca cell samples by microarray analysis. Primers were designed using the Primer Express 2.0 software (PE Applied Biosystems, Foster City, CA) for the following genes: Akt1 (5'-TGGACGATAGCTTGGAGGGA-3' and 5'-GAGGACAGCGTGGCTTCTCT-3'), CCAAT enhancer-binding protein-{gamma} (C/EBP{gamma}; 5'-GTCATTTTTGGCCACATTGCT-3' and 5'-AACCGGAGGGTGCAACTTG-3'), FoxO1A (5'-GGGCCCTAATTCGGTCATG-3' and 5'-GGTTCATACCCGAGGTGTGG-3'), prothymosin-{alpha} (PTMA; 5'-CAAGCGGGCAGCTGAAGA-3' and 5'-GGTCACGGCGGCCTTT-3'), sin 3-associated polypeptide, 18 kDa (SAP18; 5'-AGCTGCGGCAAGTTGAAGA-3' and 5'-TTTTATGTCATGGTACCCTGATCAAG-3'), transducer of ErbB2 1 (TOB1; 5'-AGCTCTCACCCAATGCCAAG-3' and 5'-GCCGCATCAAAGAAGAGGC-3'), and tribble 3 homolog (TRB3; 5'-AGCTGCCAACAGTGGATTGAG-3' and 5'-GCATGTGTGTGGAACAAAGCA-3'). Primer sets for cAMP-regulated guanine nucleotide exchange factor II (cAMP-GEFII), coxsackie and adenovirus receptor (CAR), and retinol dehydrogenase 2 (RoDH2) have been previously reported (45). Each primer set was tested empirically, and QPCR reactions were carried out using equivalent dilutions of each cDNA sample as previously described (45). To account for differences in starting material, QPCR was also carried out for each cDNA sample using the Applied Biosystems human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) 20x primer and probe reagent (PE Applied Biosystems). To define the relative abundance of each transcript in each experimental group, analysis of the resulting QPCR reactions was carried out as previously described (45). Briefly, the threshold cycle for each gene target and GAPDH in each cDNA sample was converted to an arbitrary value using a standard curve generated from serial dilutions of theca cell cDNA. The relative abundance of the target was divided by the relative abundance of GAPDH in each sample to generate a normalized abundance for each of the 10 transcripts tested. Analysis of variance was then used to determine the mean and standard error of the normalized abundance of each target in UNT and VPA-treated NL and UNT PCOS theca cells. The paired t-test was carried out to determine whether differences in the normalized mRNA abundance of each target were statistically significant between UNT and VPA-treated NL theca cell samples (P value <0.05). Likewise, the unpaired t-test was carried out to determine statistically significant differences in mRNA abundance of each target between UNT NL and PCOS theca cell samples.

Western blot analysis.
Two independent fourth-passage NL theca cell samples were cultured in serum-free, insulin-free media for 48 h in the absence (UNT) or presence of 500 µM VPA (Sigma-Aldrich). Likewise, fourth-passage PCOS theca cells were cultured for 48 h in serum-free, insulin-free medium. Increasing concentrations of insulin (0, 0.001, 0.01, 0.1, 1.0, or 10 ng/ml) were added to each culture group for the final 20 min of culture time. Whole cell lysates were collected from each culture in phosphoprotein cell lysate buffer (10 mM Tris, pH 7.4, 100 mM NaCl, 1 mM EDTA, 1 mM NaF, 20 mM Na4P2O7, 2 mM Na3VO4, 1% Triton X, 10% glycerol, 0.1% SDS, 4 mM DTT, and 0.5% sodium deoxycholate) containing protease inhibitors (PMSF, leupeptin, aprotinin, and pepstatin). To detect phosphorylated Akt/PKB in these samples, 35 µg of total protein were separated by SDS-PAGE, transferred to polyvinylidene difluoride membrane, and probed with the phospho-Akt (Ser473) antibody (Cell Signaling Technology, Beverly, MA). The immunoreactive band was detected with SuperSignal West Femto Sensitivity Substrate (Pierce, Rockford, IL). Membranes were stripped and subsequently probed for total Akt/PKB protein (Cell Signaling).


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
VPA-treated NL theca cells exhibit a gene expression profile that is more similar to PCOS theca cells than to UNT NL theca cells.
Both VPA-treated NL theca cells and UNT PCOS theca cells exhibit increased DHEA and {Delta}4A synthesis compared with UNT NL theca cells, demonstrating that VPA-treated and PCOS theca cells have a similar biochemical profile (32, 33). To determine whether VPA-treated NL theca cells have a PCOS-like genomic profile, we compared the gene expression profiles of four UNT NL, four VPA-treated NL, and five UNT PCOS theca cell samples using the Affymetrix U133 oligonucleotide chip set. We have previously shown that VPA treatment (500 µM) significantly increases androgen synthesis by the same cell stocks. Of the 45,000 transcripts interrogated on the U133 chip set, 16,197 transcripts were expressed in the UNT NL, VPA-treated NL, and/or UNT PCOS theca cells. To assess global gene expression in each of the theca cell samples, the fluorescence intensity of the 16,197 expressed transcripts in each sample was compared. This analysis demonstrated distinct differences in the gene expression patterns of the VPA-treated NL and UNT PCOS theca cells (Fig. 1). To determine whether the gene expression profiles of the VPA-treated NL and UNT PCOS theca cells were similar, hierarchal cluster analysis was carried out. As expected, the gene expression profiles clustered most tightly within experimental groups. However, the gene expression profiles of the four VPA-treated NL theca cell samples clustered more tightly with the five UNT PCOS theca cell samples than to the four UNT NL theca cell samples (Fig. 2). This finding was particularly intriguing, since the UNT NL and VPA-treated NL theca cells were culture-matched samples and thus each treatment pair had the same genetic background.



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 1. Gene expression profile of untreated (UNT) normal (NL), valproic acid (VPA)-treated NL, and untreated polycystic ovary syndrome (PCOS) theca cells. Four normal theca cell cultures were maintained in serum-free medium for 48 h in the absence or presence of 500 µM VPA (UNT NL and VPA NL, respectively). Likewise 5 PCOS theca cell cultures were maintained in serum-free medium for 48 h (UNT PCOS). RNA was collected from the cells, and the gene expression profile of each sample was assessed using GeneSpring 7. No change in gene expression is indicated in yellow, increased mRNA abundance is indicated in red, and decreased mRNA abundance is indicated in blue.

 


View larger version (49K):
[in this window]
[in a new window]
 
Fig. 2. Hierarchal cluster analysis of the UNT NL, VPA-treated NL, and UNT PCOS theca cell gene expression profiles. The gene expression profiles of the 4 UNT NL, 4 VPA NL, and 5 UNT PCOS theca cell samples were compared using hierarchal clustering. This analysis demonstrated tight association of the samples in each experimental group. Furthermore, gene expression in the VPA NL samples was more closely associated to gene expression in the 5 UNT PCOS samples than in the 4 UNT NL samples.

 
Microarray analysis of the individual samples indicated similarities in the gene expression profiles of the VPA-treated NL and UNT PCOS theca cells.
To determine whether there were genes that commonly exhibit altered mRNA abundance in both VPA-treated and PCOS cells, the mean fluorescence intensity of the 16,197 expressed transcripts in the UNT NL cells was compared with the mean fluorescence intensity of each transcript in the VPA-treated or PCOS cells. Statistical analysis of these comparisons demonstrated that 1,050 transcripts had statistically significant changes in mRNA levels in both the VPA-treated and PCOS cells. Upregulation and downregulation of gene expression in the VPA-treated and PCOS theca cells matched for the majority (99%) of the 1,050 transcripts. In fact, when the VPA-to-UNT ratio was compared with the PCOS-to-NL ratio of these 1,050 transcripts, there was a statistically significant correlation in the magnitudes of altered gene expression for this subset of genes (Fig. 3). Among the 1,050 transcripts were mRNAs for 38 genes involved in cell cycle regulation and apoptosis (Table 1), 51 genes involved in transcriptional regulation (Table 2), 59 genes involved in signal transduction cascades (Table 3), and 51 genes involved in cellular metabolism (Table 4). Taken together, these microarray analyses demonstrated similarities in the molecular signatures of VPA-treated NL and UNT PCOS theca cells and suggest possible mechanisms for the development of PCOS-like symptoms in VPA-treated NL theca cells.



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 3. One thousand fifty transcripts exhibit altered mRNA abundance in VPA-treated NL theca cells and PCOS theca cells. The PCOS-to-NL ratio of mRNA abundance was plotted against the VPA-treated-to-UNT ratio of mRNA abundance for the 1,050 transcripts that exhibited altered abundance in both experimental groups. The slope and y-intercept of the line are indicated. Regression analysis of the data demonstrated statistically significant correlation in altered gene expression between the two groups (F value is indicated).

 

View this table:
[in this window]
[in a new window]
 
Table 1. Differentially expressed cell cycle regulation and apoptosis genes

 

View this table:
[in this window]
[in a new window]
 
Table 2. Differentially expressed transcriptional regulation genes

 

View this table:
[in this window]
[in a new window]
 
Table 3. Differentially expressed intracellular signaling genes

 

View this table:
[in this window]
[in a new window]
 
Table 4. Differentially expressed cellular metabolism genes

 
Quantitative real-time RT-PCR validates altered mRNA abundance of cAMP-GEFII and TRB3 in the VPA-treated and UNT PCOS theca cells.
The true power of the microarray analysis is the ability to identify groups of differentially expressed genes whose protein functions impact a specific cell function. In this array analysis, several genes that may contribute to the PCOS phenotypes of increased androgen synthesis and/or arrested follicle development were identified as differentially expressed by the microarray analysis. For example, the mRNA levels of RoDH2, which metabolizes vitamin A to retinaldehyde (45), were increased in the VPA-treated NL and UNT PCOS theca cells (Table 4). Likewise, TRB3 and cAMP-GEFII, which impact Akt/PKB signal transduction, showed decreased and increased mRNA abundance, respectively, in the VPA-treated and PCOS theca cells (Table 3). To assess the reproducibility of the microarray analysis, the independent methodology of quantitative real-time RT-PCR (QPCR) was carried out using primers toward RoDH2, cAMP-GEFII, TRB3, and seven other genes that are involved in intracellular signaling and transcriptional regulation of gene expression (Fig. 4). QPCR demonstrated that six genes exhibit reproducible changes in mRNA levels in both VPA-treated and PCOS theca cells, including cAMP-GEFII, TRB3, and FoxO1A, which are all components of the Akt/PKB signal transduction pathway (12, 28, 34).



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 4. Quantitative PCR (QPCR) validates increased and decreased mRNA abundance of 6 genes identified as differentially expressed in both VPA-treated and PCOS theca cells by microarray analysis. QPCR was carried out using cDNA from UNT NL, VPA-treated NL, and UNT PCOS theca cells and primers specific for Akt1, cAMP-GEFII, FoxO1A, TRB3, coxsackie and adenovirus receptor (CAR), retinol dehydrogenase 2 (RoDH2), sin 3-associated polypeptide, 18 kDa (SAP18), CCAAT enhancer-binding protein-{gamma} (C/EBP{gamma}), prothymosin-{alpha} (PTMA), and transducer of ErbB2 1 (TOB1). The normalized mean relative abundance and SE of each target gene were plotted. Statistically significant differences in mRNA abundance between UNT and VPA-treated cells or between NL and PCOS cells were determined using parametric testing (*P < 0.05).

 
Altered TRB3 and cAMP-GEFII mRNA levels are correlated to increased insulin-induced Akt/PKB phosphorylation in VPA-treated and PCOS theca cells.
Akt/PKB is a protein kinase that mediates the mitogenic and metabolic actions of insulin in target cells (39). TRB3, which was downregulated in VPA-treated and PCOS theca cells, is the human ortholog of the Drosophila tribbles gene and is an inactive kinase that binds to Akt/PKB and directly inhibits the transmission of the Akt/PKB signal (Fig. 5A; Ref. 12). cAMP-GEFII, which is upregulated in VPA-treated and PCOS theca cells, activates the small GTPase Rap 1, which subsequently augments phosphoinositide triphosphate kinase (PI3K) activity and indirectly increases Akt/PKB signal transduction (Fig. 5A; Ref. 28).



View larger version (58K):
[in this window]
[in a new window]
 
Fig. 5. Insulin-induced Akt/PKB phosphorylation was increased in VPA-treated and PCOS theca cells. A: schematic representation of cAMP-GEFII and TRB3 regulation of Akt/PKB (Akt) phosphorylation. cAMP-GEFII expression is upregulated (+) and TRB3 expression is downregulated (–) according to the QPCR and microarray data. B: NL theca cells were cultured in the absence (UNT NL) or presence (VPA NL) of VPA for 48 h in serum-free, insulin-free medium. Likewise, PCOS cells (UNT PCOS) were cultured for 48 h in serum-free, insulin-free medium. Cells were then stimulated with increasing concentrations of insulin (0, 0.001, 0.01, 0.1, 1.0, or 10 ng/ml), and the protein extracts were separated and transferred to polyvinylidene difluoride membranes. The membranes were probed using antibody against Akt, which is phosphorylated at Ser473 (Akt-P). The blots were stripped and probed using an antibody against total Akt (Akt).

 
To determine whether changes in the mRNA levels of TRB3 and cAMP-GEFII were correlated with functional changes in Akt/PKB activity, the phosphorylation of Akt/PKB in UNT NL, VPA-treated NL, and UNT PCOS theca cells stimulated with increasing concentrations of insulin was determined by Western blot analysis. In the absence of insulin, there were basal levels of phosphorylated Akt/PKB detected in the UNT NL cells. When increasing concentrations of insulin were added to the culture medium, Akt/PKB phosphorylation increased and exhibited maximal phosphorylation at a concentration of 1.0 ng/ml insulin (Fig. 5B). However, in the VPA-treated NL theca cells, Akt/PKB phosphorylation was maximal in the absence of insulin, and this maximal Akt/PKB phosphorylation was maintained at all concentrations of insulin tested (Fig. 5B). Likewise, Akt/PKB phosphorylation was increased in the PCOS theca cells compared with NL theca cells in the absence of insulin. This high level of Akt/PKB phosphorylation was maintained when up to 10 ng/ml insulin were added to the cells (Fig. 5B). Taken together, these data suggest that altered expression of TRB3 and cAMP-GEFII contributes to the enhanced Akt/PKB phosphorylation detected in VPA-treated NL theca cells and in PCOS theca cells. Furthermore, insulin signaling through the Akt/PKB pathway may be potentiated in the VPA-treated NL and UNT PCOS theca cells, providing a possible mechanism for enhanced steroid synthesis and/or altered cell proliferation in these cells.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
HDACs, which are a family of 11 proteins categorized into two classes, interact with nucleosomes and remove acetyl groups from their core histones (9). This activity of the HDAC induces a closed conformation of DNA, resulting in repression of transcription and decreased gene expression (9, 40). The specificity of HDAC-induced inhibition of gene expression is conferred by its interaction with transcription factors and transcriptional corepressors that localize HDACs to specific DNA sequences in the promoter region of target genes (40). VPA reversibly inhibits the catalytic activity of class I HDACs, resulting in increased acetylation of histones and leading to altered expression of target genes (17, 36). These direct effects on gene expression may subsequently cause increased or decreased transcription of downstream genes and result in global changes in transcript abundance in VPA-treated cells. Indeed, several studies have demonstrated that VPA increases histone H3 and H4 acetylation, which has been correlated to increased expression of candidate genes in diverse cell culture systems (17, 27, 33, 36, 46). Furthermore, Wang et al. (41) described 28 genes that are differentially expressed in VPA-treated compared with untreated rat cerebral cortical cells using microarray analysis. Our gene expression profiling experiments demonstrated that VPA alters global gene expression in the human theca cell such that it exhibits a PCOS-like molecular phenotype.

In the human theca cells we studied, VPA induces increased acetylation of histone H3, increased expression of CYP17 and CYP11A, and increased synthesis of DHEA and {Delta}4A (33). Furthermore, these effects of VPA on theca cell steroidogenesis are augmented when the cells are also exposed to the adenylate cyclase activator forskolin. However, the microarray analysis of VPA-treated compared with untreated normal theca cells did not identify statistically significant changes in CYP17 or CYP11A mRNA levels. One explanation for these data is that the theca cells were not stimulated with forskolin, resulting in levels of CYP17 and CYP11A mRNA that are below the sensitivity of the Affymetrix oligonucleotide probe set used in the analysis. This observation is consistent with previous array analysis of unstimulated normal and PCOS theca cells (45). Despite these results, there is evidence from this study that VPA stimulates theca cell steroidogenesis. For example, the mRNA abundance of cytochrome b5, which serves as an allosteric effector activating the lyase activity of CYP17 (15), was increased in VPA-treated compared with untreated normal theca cells, providing evidence that VPA stimulates theca cell steroidogenesis. In addition, VPA-dependent alterations in gene expression were correlated to increased sensitivity of normal theca cells to insulin-dependent Akt/PKB phosphorylation, which has been shown to increase steroid synthesis in ovarian cells (11, 37, 38, 49). Thus the HDAC-inhibitory property of VPA affects the steroidogenic capacity of the theca cell, which may lead to the PCOS-like phenotype of increased androgen synthesis.

Among the genes with altered mRNA abundance in both VPA-treated normal theca cells and PCOS theca cells are TRB3 and cAMP-GEFII. TRB3 regulates glucose metabolism in hepatic cells, and cAMP-GEFII is a component of the insulin secretory pathway in pancreatic ß-cells (12, 21, 22). Furthermore, TRB3 negatively regulates and cAMP-GEFII positively regulates phosphorylation of Akt/PKB, suggesting that these two genes modulate the transmission of the insulin signal in target cells (12, 16). This altered gene expression is associated with increased Akt/PKB phosphorylation upon insulin stimulation, suggesting that VPA-treated and PCOS theca cells have increased sensitivity to insulin actions. Interestingly, women with PCOS exhibit insulin resistance in muscle and fat tissues and insulin sensitivity in ovarian tissues (13, 14). The microarray data suggest that altered expression of TRB3 and cAMP-GEFII in VPA-treated and PCOS theca cells sensitizes ovarian cells to insulin actions. Furthermore, VPA may be differentially regulating gene expression in muscle and fat compared with ovarian tissues to produce a phenotype of insulin resistance. Taken together, these observations suggest a plausible mechanism for the dichotomy of insulin signaling in PCOS women.

The phosphorylation of Akt/PKB not only transmits the insulin signal but also regulates the proliferative and/or differentiative state of the cell (8). For example, Akt/PKB increases NF-{kappa}B transcriptional activity. This increased activity is accomplished through phosphorylation and subsequent degradation of the inhibitory protein I{kappa}B or through phosphorylation and increased association of the coactivator p300/CBP with DNA-bound NF-{kappa}B (26, 29, 34). Studies in non-small cell lung cancer cells demonstrate that HDAC inhibitors stimulate Akt/PKB-dependent NF-{kappa}B transcriptional activity, which is correlated to reduced apoptosis of these cells (26). Akt/PKB also phosphorylates forkhead receptors including FoxO1A and FoxO3A, resulting in the translocation of the transcription factors from the nucleus to the cytoplasm (7, 34). In the porcine granulosa cell, transcriptionally active FoxO1A increases the expression of the cell cycle regulator p27kip and inhibits cell cycle progression (6). Conversely, when FoxO1A is localized in the cytoplasm of porcine granulosa cells, p27kip expression is decreased and cell cycle progression is increased, resulting in proliferation of the granulosa cells. Interestingly, FoxO1A expression is increased in both VPA-treated and PCOS theca cells (Figs. 1 and 2), providing a possible feedback mechanism for FoxO1-mediated cell cycle arrest. Ablation of FoxO3A expression in female mice increases initiation of primordial follicle growth including granulosa cell proliferation (4), suggesting that inhibition of FoxO3A transcriptional activity also promotes cell proliferation. Taken together, these studies indicate that increased Akt/PKB phosphorylation enhances cell proliferation in the ovary and may provide a mechanism for theca cell hypertrophy, abnormal follicular development, and/or anovulation, which are characteristic phenotypes in women with PCOS.

The microarray analysis described in this manuscript indicates that VPA affects the expression of genes responsible for a variety of cell functions, including proliferation, differentiation, and metabolism. The identification of a subset of genes with altered mRNA abundance in both VPA-treated and PCOS theca cells also provides plausible mechanisms for the PCOS-like symptoms in women receiving VPA therapy. Furthermore, the cell specificity of HDAC-regulated gene expression emphasizes the importance of understanding the repertoire of factors that are impacted by HDAC inhibitors, including VPA, to better predict the clinical outcomes of these drugs.


    ACKNOWLEDGMENTS
 
We thank Lisa Salvador for helpful comments during the preparation of this manuscript.

This work was supported by funding from the National Institutes of Health (U54-HD-34449 and HD-07305) and the Mellon Foundation (J. R. Wood).


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: J. F. Strauss III, Center for Research on Reproduction and Women’s Health, 1349 BRB II/III, 421 Curie Blvd., Philadelphia, PA 19104 (E-mail: jfs3{at}mail.med.upenn.edu).

10.1152/physiolgenomics.00193.2004.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Asghari V, Wang JF, Reiach JS, and Young LT. Differential effects of mood stabilizers on Fos/Jun proteins and AP-1 DNA binding activity in human neuroblastoma SH-SY5Y cells. Brain Res Mol Brain Res 58: 95–102, 1998.[ISI][Medline]
  2. Betts T, Yarrow H, Dutton N, Greenhill L, and Rolfe T. A study of anticonvulsant medication on ovarian function in a group of women with epilepsy who have only ever taken one anticonvulsant compared with a group of women without epilepsy. Seizure 12: 323–329, 2003.[CrossRef][ISI][Medline]
  3. Bilo L, Meo R, Valentino R, Di Carlo C, Striano S, and Nappi C. Characterization of reproductive endocrine disorders in women with epilepsy. J Clin Endocrinol Metab 86: 2950–2956, 2001.[Abstract/Free Full Text]
  4. Castrillon DH, Miao L, Kollipara R, Horner JW, and DePinho RA. Suppression of ovarian follicle activation in mice by the transcription factor Foxo3a. Science 301: 215–218, 2003.[Abstract/Free Full Text]
  5. Chen G, Yuan PX, Jiang YM, Huang LD, and Manji HK. Valproate robustly enhances AP-1 mediated gene expression. Brain Res Mol Brain Res 64: 52–58, 1999.[ISI][Medline]
  6. Cunningham MA, Zhu Q, and Hammond JM. FoxO1a can alter cell cycle progression by regulating the nuclear localization of p27kip in granulosa cells. Mol Endocrinol 18: 1756–1767, 2004.[Abstract/Free Full Text]
  7. Cunningham MA, Zhu Q, Unterman TG, and Hammond JM. Follicle-stimulating hormone promotes nuclear exclusion of the forkhead transcription factor FoxO1a via phosphatidylinositol 3-kinase in porcine granulosa cells. Endocrinology 144: 5585–5594, 2003.[Abstract/Free Full Text]
  8. Datta SR, Brunet A, and Greenberg ME. Cellular survival: a play in three Akts. Genes Dev 13: 2905–2927, 1999.[Free Full Text]
  9. de Ruijter AJ, van Gennip AH, Caron HN, Kemp S, and van Kuilenburg AB. Histone deacetylases (HDACs): characterization of the classical HDAC family. Biochem J 370: 737–749, 2003.[CrossRef][ISI][Medline]
  10. De Sarno P, Li X, and Jope RS. Regulation of Akt and glycogen synthase kinase-3 beta phosphorylation by sodium valproate and lithium. Neuropharmacology 43: 1158–1164, 2002.[CrossRef][ISI][Medline]
  11. Devoto L, Christenson LK, McAllister JM, Makrigiannakis A, and Strauss JF III. Insulin and insulin-like growth factor-I and -II modulate human granulosa-lutein cell steroidogenesis: enhancement of steroidogenic acute regulatory protein (StAR) expression. Mol Hum Reprod 5: 1003–1010, 1999.[Abstract/Free Full Text]
  12. Du K, Herzig S, Kulkarni RN, and Montminy M. TRB3: a tribbles homolog that inhibits Akt/PKB activation by insulin in liver. Science 300: 1574–1577, 2003.[Abstract/Free Full Text]
  13. Dunaif A. Insulin action in the polycystic ovary syndrome. Endocrinol Metab Clin North Am 28: 341–359, 1999.[ISI][Medline]
  14. Dunaif A. Insulin resistance and the polycystic ovary syndrome: mechanism and implications for pathogenesis. Endocr Rev 18: 774–800, 1997.[Abstract/Free Full Text]
  15. Geller DH, Auchus RJ, and Miller WL. P450c17 mutations R347H and R358Q selectively disrupt 17,20-lyase activity by disrupting interactions with P450 oxidoreductase and cytochrome b5. Mol Endocrinol 13: 167–175, 1999.[Abstract/Free Full Text]
  16. Gonzalez-Robayna IJ, Falender AE, Ochsner S, Firestone GL, and Richards JS. Follicle-stimulating hormone (FSH) stimulates phosphorylation and activation of protein kinase B (PKB/Akt) and serum and glucocorticoid-induced kinase (Sgk): evidence for a kinase-independent signaling by FSH in granulosa cells. Mol Endocrinol 14: 1283–1300, 2000.[Abstract/Free Full Text]
  17. Gottlicher M, Minucci S, Zhu P, Kramer OH, Schimpf A, Giavara S, Sleeman JP, Lo Coco F, Nervi C, Pelicci PG, and Heinzel T. Valproic acid defines a novel class of HDAC inhibitors inducing differentiation of transformed cells. EMBO J 20: 6969–6978, 2001.[Abstract/Free Full Text]
  18. Gurvich N and Klein PS. Lithium and valproic acid: parallels and contrasts in diverse signaling contexts. Pharmacol Ther 96: 45–66, 2002.[CrossRef][ISI][Medline]
  19. Isojarvi JI, Tauboll E, Pakarinen AJ, van Parys J, Rattya J, Harbo HF, Dale PO, Fauser BC, Gjerstad L, Koivunen R, Knip M, and Tapanainen JS. Altered ovarian function and cardiovascular risk factors in valproate-treated women. Am J Med 111: 290–296, 2001.[CrossRef][ISI][Medline]
  20. Johannessen CU. Mechanisms of action of valproate: a commentatory. Neurochem Int 37: 103–110, 2000.[CrossRef][ISI][Medline]
  21. Kashima Y, Miki T, Shibasaki T, Ozaki N, Miyazaki M, Yano H, and Seino S. Critical role of cAMP-GEFII—Rim2 complex in incretin-potentiated insulin secretion. J Biol Chem 276: 46046–46053, 2001.[Abstract/Free Full Text]
  22. Koo SH, Satoh H, Herzig S, Lee CH, Hedrick S, Kulkarni R, Evans RM, Olefsky J, and Montminy M. PGC-1 promotes insulin resistance in liver through PPAR-alpha-dependent induction of TRB-3. Nat Med 10: 530–534, 2004.[CrossRef][ISI][Medline]
  23. Lampen A, Carlberg C, and Nau H. Peroxisome proliferator-activated receptor delta is a specific sensor for teratogenic valproic acid derivatives. Eur J Pharmacol 431: 25–33, 2001.[CrossRef][ISI][Medline]
  24. Legro RS, Spielman R, Urbanek M, Driscoll D, Strauss JF III, and Dunaif A. Phenotype and genotype in polycystic ovary syndrome. Recent Prog Horm Res 53: 217–256, 1998.[Medline]
  25. Luef G, Abraham I, Trinka E, Alge A, Windisch J, Daxenbichler G, Unterberger I, Seppi K, Lechleitner M, Kramer G, and Bauer G. Hyperandrogenism, postprandial hyperinsulinism and the risk of PCOS in a cross sectional study of women with epilepsy treated with valproate. Epilepsy Res 48: 91–102, 2002.[CrossRef][ISI][Medline]
  26. Mayo MW, Denlinger CE, Broad RM, Yeung F, Reilly ET, Shi Y, and Jones DR. Ineffectiveness of histone deacetylase inhibitors to induce apoptosis involves the transcriptional activation of NF-kappa B through the Akt pathway. J Biol Chem 278: 18980–18989, 2003.[Abstract/Free Full Text]
  27. McGarry LC, Winnie JN, and Ozanne BW. Invasion of v-Fos(FBR)-transformed cells is dependent upon histone deacetylase activity and suppression of histone deacetylase regulated genes. Oncogene 23: 5284–5292, 2004.[CrossRef][ISI][Medline]
  28. Mei FC, Qiao J, Tsygankova OM, Meinkoth JL, Quilliam LA, and Cheng X. Differential signaling of cyclic AMP: opposing effects of exchange protein directly activated by cyclic AMP and cAMP-dependent protein kinase on protein kinase B activation. J Biol Chem 277: 11497–11504, 2002.[Abstract/Free Full Text]
  29. Miao F, Gonzalo IG, Lanting L, and Natarajan R. In vivo chromatin remodeling events leading to inflammatory gene transcription under diabetic conditions. J Biol Chem 279: 18091–18097, 2004.[Abstract/Free Full Text]
  30. Morrell MJ. Reproductive and metabolic disorders in women with epilepsy. Epilepsia 44, Suppl 4: 11–20, 2003.
  31. Morrell MJ, Giudice L, Flynn KL, Seale CG, Paulson AJ, Done S, Flaster E, Ferin M, and Sauer MV. Predictors of ovulatory failure in women with epilepsy. Ann Neurol 52: 704–711, 2002.[CrossRef][ISI][Medline]
  32. Nelson VL, Legro RS, Strauss JF III, and McAllister JM. Augmented androgen production is a stable steroidogenic phenotype of propogated theca cells from polycystic ovaries. Mol Endocrinol 13: 946–957, 1999.[Abstract/Free Full Text]
  33. Nelson-DeGrave VL, Wickenheisser JK, Cockrell JE, Wood JR, Legro RS, Strauss JF III, and McAllister JM. Valproate potentiates androgen biosynthesis in human ovarian theca cells. Endocrinology 145: 799–808, 2004.[Abstract/Free Full Text]
  34. Nicholson KM and Anderson NG. The protein kinase B/Akt signalling pathway in human malignancy. Cell Signal 14: 381–395, 2002.[CrossRef][ISI][Medline]
  35. O’Donovan C, Kusumakar V, Graves GR, and Bird DC. Menstrual abnormalities and polycystic ovary syndrome in women taking valproate for bipolar mood disorder. J Clin Psychiatry 63: 322–330, 2002.[ISI][Medline]
  36. Phiel CJ, Zhang F, Huang EY, Guenther MG, Lazar MA, and Klein PS. Histone deacetylase is a direct target of valproic acid, a potent anticonvulsant, mood stabilizer, and teratogen. J Biol Chem 276: 36734–36741, 2001.[Abstract/Free Full Text]
  37. Poretsky L, Seto-Young D, Shrestha A, Dhillon S, Mirjany M, Liu HC, Yih MC, and Rosenwaks Z. Phosphatidyl-inositol-3 kinase-independent insulin action pathway(s) in the human ovary. J Clin Endocrinol Metab 86: 3115–3119, 2001.[Abstract/Free Full Text]
  38. Sekar N, Garmey JC, and Veldhuis JD. Mechanisms underlying the steroidogenic synergy of insulin and luteinizing hormone in porcine granulosa cells: joint amplification of pivotal sterol-regulatory genes encoding the low-density lipoprotein (LDL) receptor, steroidogenic acute regulatory (StAR) protein and cytochrome P450 side-chain cleavage (P450scc) enzyme. Mol Cell Endocrinol 159: 25–35, 2000.[CrossRef][ISI][Medline]
  39. Summers SA, Yin VP, Whiteman EL, Garza LA, Cho H, Tuttle RL, and Birnbaum MJ. Signaling pathways mediating insulin-stimulated glucose transport. Ann NY Acad Sci 892: 169–186, 1999.[Abstract/Free Full Text]
  40. Wade PA. Transcriptional control at regulatory checkpoints by histone deacetylases: molecular connections between cancer and chromatin. Hum Mol Genet 10: 693–698, 2001.[Abstract/Free Full Text]
  41. Wang JF, Shao L, Sun X, and Young LT. Glutathione S-transferase is a novel target for mood stabilizing drugs in primary cultured neurons. J Neurochem 88: 1477–1484, 2004.[CrossRef][ISI][Medline]
  42. Watterson JM, Watson DG, Meyer EM, and Lenox RH. A role for protein kinase C and its substrates in the action of valproic acid in the brain: implications for neural plasticity. Brain Res 934: 69–80, 2002.[CrossRef][ISI][Medline]
  43. Werling U, Siehler S, Litfin M, Nau H, and Gottlicher M. Induction of differentiation in F9 cells and activation of peroxisome proliferator-activated receptor delta by valproic acid and its teratogenic derivatives. Mol Pharmacol 59: 1269–1276, 2001.[Abstract/Free Full Text]
  44. Wickenheisser JK, Quinn PG, Nelson VL, Legro RS, Strauss JF III, and McAllister JM. Differential activity of the cytochrome P450 17{alpha}-hydroxylase and steroidogenic acute regulatory protein gene promoters in normal and polycystic ovary syndrome theca cells. J Clin Endocrinol Metab 85: 2304–2311, 2000.[Abstract/Free Full Text]
  45. Wood JR, Nelson VL, Ho C, Jansen E, Wang CY, Urbanek M, McAllister JM, Mosselman S, and Strauss JF III. The molecular phenotype of polycystic ovary syndrome (PCOS) theca cells and new candidate PCOS genes defined by microarray analysis. J Biol Chem 278: 26380–26390, 2003.[Abstract/Free Full Text]
  46. Yildirim E, Zhang Z, Uz T, Chen CQ, Manev R, and Manev H. Valproate administration to mice increases histone acetylation and 5-lipoxygenase content in the hippocampus. Neurosci Lett 345: 141–143, 2003.[CrossRef][ISI][Medline]
  47. Yuan PX, Huang LD, Jiang YM, Gutkind JS, Manji HK, and Chen G. The mood stabilizer valproic acid activates mitogen-activated protein kinases and promotes neurite growth. J Biol Chem 276: 31674–31683, 2001.[Abstract/Free Full Text]
  48. Zawadski J and Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Current Issues in Endocrinology and Metabolism. Polycystic Ovary Dyndrome, edited by Dunaif A, Givens J, Haseltine F, and Merriam G. Boston, MA: Blackwell Scientific, 1992, p. 377–384.
  49. Zhang G, Garmey JC, and Veldhuis JD. Interactive stimulation by luteinizing hormone and insulin of the steroidogenic acute regulatory (StAR) protein and 17alpha-hydroxylase/17,20-lyase (CYP17) genes in porcine theca cells. Endocrinology 141: 2735–2742, 2000.[Abstract/Free Full Text]




This Article
Abstract
Full Text (PDF)
All Versions of this Article:
20/3/233    most recent
00193.2004v1
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Wood, J. R.
Articles by Strauss, J. F.
Articles citing this Article
PubMed
PubMed Citation
Articles by Wood, J. R.
Articles by Strauss, J. F., III


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2005 by the American Physiological Society.