Aberrant Epigenetic Alteration of PAX1 Expression Contributes to Parathyroid Tumorigenesis

Abstract

Context

Primary hyperparathyroidism (PHPT) results from the hypersecretion of parathyroid hormone from parathyroid tumors. A transcription factor, namely Paired box1 (PAX1), is active in parathyroid gland development.

Objective

We aimed to study potential epigenetic-mediated mechanism of PAX1 gene in sporadic parathyroid adenomas.

Methods

In parathyroid adenomas tissues, we analyzed the DNA methylation via bisulfite-specific polymerase chain reaction (BSP) and histone modifications via chromatin immunoprecipitation in regulating the differential expression of PAX1.

Results

The results showed that mRNA and protein expression of PAX1 was significantly reduced in parathyroid adenomas. Bisulfite sequencing demonstrated hypermethylation in the promoter region of PAX1 (35%; 14/40) and lower levels of histone 3 lysine 9 acetylation (H3K9ac) were observed on the promoter region of PAX1 (6-fold; P < .004) in parathyroid adenomas. Furthermore, upon treatment with a pharmacologic inhibitor, namely 5′aza-2 deoxycytidine, in rat parathyroid continuous cells, we found re-expression of PAX1 gene.

Conclusion

Our study not only reveals expression of PAX1 is epigenetically deregulated but also paves a way for clinical and therapeutic implications in patients with PHPT.

Parathyroid glands play a central role in regulation of calcium and phosphate homeostasis. Parathyroid hormone (PTH) maintains serum calcium within a narrow range by its actions on bone, kidney, and intestine (1). The parathyroid gland has predominantly chief cells, which are able to perceive small changes in circulating calcium levels, and modify PTH synthesis and secretion mediated by the calcium-sensing receptor (CaSR) in the parathyroid cell membrane (2). Parathyroid adenomas are the most common endocrine tumors, but the underlying molecular pathogenic mechanism(s) is poorly understood. The embryonic development transcription factors such as glial cells missing homolog 2 (GCM2), homeobox A3 (HOXA3), and GATA3 are involved in parathyroid embryogenesis and have been found to be expressed in human adult parathyroid tissues (3). Another embryonic transcription factor T-box 1 (TBX1), involved in parathyroid embryogenesis, has been demonstrated to be expressed and deregulated in parathyroid adenomas (4).

The paired box 1 (PAX1) gene has a paired box DNA binding domain with transcriptional activator activity. Its role is identified in the development of the skeleton, thymus gland, and parathyroid glands (5). In silico analysis of the GCM2 nucleotide sequence upstream of transcription start site (TSS) revealed a binding site for the transcription factor PAX1 (+8 to +17 from TSS) (6). GCM2 expression is reduced in mouse embryos in which PAX1 has been genetically inactivated, which explains the link between these 2 transcription factors (6, 7). In our literature search, the presence of PAX1 in the GCM2 promoter strongly suggests that the action of the embryonic transcription factor on parathyroid development is mediated through the transactivation of GCM2. However, there is no literature available about the expression and role of the PAX1 gene in human parathyroid adenoma formation.

Epigenetic modifications are known to play a role in parathyroid adenoma formation with well-described examples of silencing of tumor suppressor genes by promoter DNA methylation, or by histone modifications such as repressive histone methylation (8, 9). Histone acetylation/deacetylation also alters the status of open chromatin domains and thus affects gene transcription. This process is modulated by histone acetyltransferases and histone deacetylases (HDACs). Loss of histone 3 and 4 acetylation attributes to the imbalanced recruitment of HDACs and results in transcriptional repression of tumor suppressor genes in cancers (10).

We hypothesized that epigenetic alterations such as DNA methylation and histone modifications of PAX1 would be associated with sporadic parathyroid tumorigenesis and may serve as potential biomarker for parathyroid adenoma detection. Accordingly, we systematically investigated PAX1 gene methylation and histone acetylation in parathyroid tumor tissues using bisulfite-specific PCR (BSP) and chromatin immunoprecipitation (ChIP) qPCR and compared with the normal parathyroid tissues. Since parathyroid tumor tissues revealed silencing of PAX1 gene expression and correlated with DNA methylation, we selected rat parathyroid cell line as a disease model to study the in vitro effects of inhibitors of DNA methylation and histone deacetylation.

Materials and Methods

Patients and Parathyroid Tissue Samples

Forty sporadic parathyroid adenomas after parathyroidectomy and 10 normal parathyroid tissues that were removed inadvertently during clinically indicated thyroid surgeries for multinodular goiter were collected for this study. We excluded patients with family history suggestive of genetic forms of PHPT such as multiple endocrine neoplasia type 1, multiple endocrine neoplasia type 2A, multiple endocrine neoplasia type 4, familial hypocalciuric hypercalcemia syndrome, neonatal severe hyperparathyroidism, HPT-jaw tumor, and familial isolated primary hyperparathyroidism. The study was approved by the institutional ethics committee of the Post Graduate Institute of Medical Education and Research, Chandigarh, India. Written informed consent was obtained from each participant. Each parathyroid tissue sample was examined by an experienced endocrine pathologist (U.N.S.) who confirmed the status as adenomatous or normal parathyroid tissue. All samples were stored in a deep freezer at –80°C until final molecular analyses.

Biochemical Measurements

Serum calcium (reference range [RR]: 8.2-10.2 mg/dL), serum phosphorus (RR: 2.7-4.5 mg/dL), serum creatinine (RR, 0.50-1.20 mg/dL), serum albumin (RR, 3.5 to –5.5 g/dL), and alkaline phosphatase (RR, 40-129 U/L) were measured by standard methods using an Olympus auto-analyzer, and serum calcium was adjusted for serum albumin level. Plasma PTH (RR, 15-65 pg/mL) and 25-hydroxyvitamin D (RR: 11-43 ng/mL) were measured using immunochemiluminiscence (ELECSYS- 2010: Roche Diagnostics, Germany) (8).

mRNA Isolation, cDNA Synthesis, and Quantitative Real-time PCR

Total mRNA (control and parathyroid adenoma tissues) was isolated using trizol reagent and reverse-transcribed as described previously (9). PAX1 mRNA levels were quantified on a StepOne Plus Real-Time PCR System (Applied Biosystems, USA) using 18s rRNA as housekeeping gene and the specific set of primers (for PAX1, forward primer (FP) 5′-GACCCCTTACATAGATTACACGC-3′ and reverse primer (RP) 5′-CTCCACAGGATTTTCTCCTC-3′) were used. Quantitative RT-PCR (qRT-PCR) reaction conditions were used as described previously (9). Experiments were performed in duplicate with 2 nontemplate controls as a negative control. For all experimental samples, the relative value was normalized to the 18s rRNA of the same sample. The 2^–(ΔΔCt) method was used to quantify the relative expression of PAX1 genes as a fold change in parathyroid adenoma from the control parathyroid.

Immunohistochemical Staining

For immunohistochemistry (IHC) of PAX1, 4- to 5-μm-thick sections of formalin-fixed paraffin-embedded blocks of normal and parathyroid adenoma for IHC staining as reported previously (8, 11). PAX1 was detected using a polyclonal antibody for PAX1 nuclear protein (Thermo Fisher Scientific, catalog no PA5-75433, RRID:AB_2719161 (12), dilution, 1:100). The IHC slides were examined semi-quantitatively based on the percentage of positive tumor cells and staining intensity by an endocrine pathologist. A minimum of 4 different regions covering 200 cells was taken from each IHC stained slide at 40× magnification was analyzed. The staining intensity for the tumors and normal parathyroid tissue sections was graded as weak positive (1+; <40%), moderate positive (2+; 40-60%), and strong positive (3+; >60%) cells.

DNA Extraction, Bisulfite Conversion, and Bisulfite Sequencing PCR

DNA from tissues samples (normal parathyroid and parathyroid adenoma) were extracted using a DNeasy Tissue mini kit (Qiagen, USA), according to the manufacturer’s instructions. Bisulfite modification of the genomic DNA was achieved using the EZ DNA methylation kit (Zymo Research, USA) as per the manufacturer’s instructions. The promoter region of PAX1 was selected from the eukaryotic promoter database and bisulfite specific primers were designed using the online tool MethPrimer. Primer sequences (For PAX1 FP 5′-TTGAAAGGGGTTTAGAGTAGTGGAA-3′, RP 5′- CCCAAACCCAAAATAAACTTCATC-3′) were used. PCR was carried out using Takara EpiTaq hot-start Taq Polymerase (Takara Bio Inc, Country). The amplified PCR products were subjected to 2% agarose gel electrophoresis to confirm the product size. The resultant PCR products were subjected further for bisulfite sequencing. Sequencing data were analyzed with bisulfite sequencing DNA methylation analysis (BISMA) software to generate quantitative results for each CpG site.

Chromatin Immunoprecipitation Assay

The ChIP assay was adapted with some changes and carried out in 23 parathyroid adenoma samples according to the Weinmann and Farnham, 2002, protocol (8, 11, 13). The nucleoprotein complexes were sonicated to reduce the sizes of DNA fragments to 300 to 500 bp using a sonicator. Normal rabbit IgG (Abcam, catalog number ab171870, RRID:AB_2687657 (14)) was used as the negative control and anti- trimethyl histone H3K9 antibody (Abcam, catalog number ab8898, RRID:AB_306848 (15)), trimethyl H3K27 (Abcam, catalog number ab6002, RRID:AB_305237 (16)), and anti-acetyl histone H3K9 antibody (Cell Signaling Technology, catalog number RRID:AB_823528 (17)) were used for each immunoprecipitation. Immunoprecipitated DNA was amplified by PCR using a Sybr green method (Takara, Japan) on StepOne Plus real-time PCR system. The ChIP primers were designed to target the promoter region of PAX1, ChIP primers sequences (for PAX1 FP 5′-CGCCCCTTAGCAGCAGATCG-3′, RP 5′- TGGGGGTGATTTCGGTGATGC-3′) were used.

Effect of 5-aza2′ Deoxycytidine and Trichostatin A Treatment on PAX1 Expression in PTH-C1 Cell Line

To investigate the role of DNA hypermethylation and H3K9ac on PAX1 expression, we treated PTH-C1 cells (18) with a DNA methyltransferase inhibitor, 5-aza-2′ deoxycytidine (DAC) and a histone deacetylase inhibitor, trichostatin A (TSA) respectively. DAC, 100 mM (10 mg/438µL) solution (189826) in dimethyl sulfoxide and TSA (T1952), dissolved in dimethyl sulfoxide as a 5 mM stock solution, were purchased from Sigma Aldrich. PTH-C1 cells were derived from rat parathyroid continuous cell line, as there is no human parathyroid cell line available. The cells were cultured in Ham F12 (Sigma,) medium and was supplemented with 10% fetal bovine serum (Thermo Fisher Scientific) and 100 U/mL penicillin/100 Ag/mL streptomycin (Thermo Fisher Scientific). A total of 5000 PTH-C1 cells were seeded into 96-well plates. After culturing for 24 hours, cells were treated with DAC at various concentrations (0.625, 1.25, 2.5, 5, 10, and 20 μM) and incubated for the 72 hours (DAC was changed every 24 hours) while TSA was exposed to different concentrations (5, 10, 25, 50, 100 nM) prepared from a stock solution dissolved in DMSO for 24 hours, respectively. Cells treated with identical concentrations of DMSO were used as control. The cells were assayed by MTT (3-(4,5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide). All experiments were performed in triplicate independently.

We performed qRT-PCR analysis to understand whether the mRNA expression of PAX1 was re-activated after treatment with DAC for 72 hours (10 µM, fresh drug was changed every 24 hours), and TSA (25 nM) for 24 hours respectively. Total RNA was isolated from cells, which had or had not undergone DAC treatment and reverse transcribed, and qRT-PCR analysis was performed as described previously for parathyroid tissue samples. Rat-specific PAX1 primers for qRT-PCR (FP 5′- CTGCGCATCGTGGAATTAGC -3′ and RP 5′-CGCCAGGATTTTGCTCACAC-3′) were used during the experiments. Data were normalized to the expression of β-actin as a housekeeping gene for cell lines and relative expression was calculated using a 2-(ΔΔCt) method. In this analysis, data from 3 independent experiments were averaged.

ChIP assay was performed as described previously for parathyroid tissue samples, to check whether H3K9ac levels increased in PTH-C1 cells, which had or had not undergone TSA treatment. Rat-specific PAX1 primers for ChIP-qPCR (FP 5′-TAACTTTCCCGCACGATCCC-3′ RP 5′-GAGCGGAGAAAGTCGCTAGG-3′) were used.

Statistical Analysis

Statistical analyses were performed using Graph Pad Prism 6. All values are presented as mean ± SD except the ChIP-qPCR data, which were expressed as mean ± standard error of the mean (SEM). An unpaired t-test was used for the comparison of data between the 2 groups (parathyroid adenoma vs control). The Spearman correlation test was used to analyze the association between expression data and DNA promoter methylation and ChIP-qPCR data respectively as well as all data sets with the disease parameters (calcium, phosphorus, alkaline phosphatase (ALP), 25(OH)D, PTH, and adenoma weight). P <.05 was considered statistically significant for all of our analyses in the study.

Results

Characteristics of PHPT Patients

A total of 40 PHPT patients with a mean age of 43 ± 13.5 years (range, 18-65 years; 31 women) and 10 control subjects with a normal parathyroid function who had undergone surgery for benign thyroid disease (mean age, 43 ± 13.3 years; range, 25-61 years; 7 women) were included in the study. All PHPT patients were symptomatic having complaints of bone pain, weakness, and fatigue, kidney stones, gall stone disease, fractures, and pancreatitis. Sixteen (40%) patients had vitamin D deficiency (25-hydroxyvitamin D level <20 ng/mL) at the time of presentation and the median parathyroid adenoma weight was 2 g (first interquartile-third interquartile range (1-4.5)) (8). Histopathological analysis showed that all 40 tissue sections had features of parathyroid adenoma.

PAX1 Expression in Sporadic Parathyroid Adenoma

The PAX1 mRNA expression was reduced in parathyroid adenomas with a mean fold decrease of 0.32 ± 0.33 (0.001-0.91) compared with control parathyroid tissue samples (P < .0001; Fig. 1A). The mean proportion of PAX1 positive cells were 49.8% in adenomatous vs 77.5% in normal parathyroid tissues. The staining intensity was 1+ in 18 (45%), 2+ in 13 (32.5%), and 3+ intensity in 9 (22.5%) adenomas. All normal parathyroid tissue cells showed 3+ nuclear positivity (Fig. 1B). The mean proportion of PAX1 positive cells was 49.8 ± 23% in parathyroid adenoma and 77.5 ± 4.1% in normal parathyroid tissue sections. Thus, the present findings showed a significant loss of PAX1 expression at the protein level in parathyroid adenoma compared with normal parathyroid tissue sections (P < .001). The decreased protein expression of PAX1 showed a significant inverse relationship with high serum calcium (r = –0.29, P < .03) (Fig. 1C) and high plasma PTH (r = –0.37, P < .01) (Fig. 1D) in PHPT patients.

Figure 1.

The expression of PAX1 in parathyroid adenoma. (A) mRNA levels of PAX1 in sporadic parathyroid adenoma (n = 40) and control parathyroid samples (n = 10) were assessed by qRT-PCR. Total RNA was isolated, reverse transcribed and the resulting cDNA was amplified using primers for the 18s rRNA and PAX1 gene. The PAX1 mRNA expression was significantly lower in tumor tissues than in corresponding control tissues (P < 0.0001). (B) Representative images of IHC by polyclonal anti-PAX1 antibody of formalin-fixed paraffin embedded sections of the parathyroid and normal endometrium for PAX1 was used as a positive control. (1) Normal showed a positive nuclear immunostaining for PAX1 (2,3 and 4) parathyroid adenoma show intensity in adenoma as 1+, 2+, and 3+, respectively. Images captured at magnification at 40X, scale bar = 400μm. (C) and (D) Scatter plot showing correlation analysis between decreased protein expression of PAX1 and high serum calcium (r= -0.29, P < .03) (D) and with plasma PTH levels (r= -0.37, P < .01) in PHPT patients.

Analysis of PAX1 Methylation in Sporadic Parathyroid Adenomas

A total of 20 CpG sites were present on PAX1 promoter from nucleotide –337 to –8 with respect to TSS (Fig. 2A). Genomic DNA isolated from parathyroid adenomas and control parathyroids, were bisulfite converted and amplified by (BSP. PCR products of 345 bp were confirmed on 2% agarose gel using AlphaImager gel documentation system (Protein Simple) (Fig. 2B). After sequencing analysis, we found that the promoter region of PAX1 was methylated in 35% of parathyroid adenomas (14 of 40) with a mean methylation density of 20.2 ± 10.3%. Five cases of parathyroid adenoma had a mean methylation density of 6.1 ± 0.5% and 21 cases did not show methylation at any CpG sites. Control parathyroid samples were not methylated except at 1 CpG site with a methylation density of 5.6%. Representative sequencing chromatograms are displayed in Fig. 2C. Moreover, the correlation between PAX1 mRNA expression showed significant negative association with DNA methylation of PAX1 in parathyroid adenomas (r = –0.47, P < .01; Fig. 2D). The results suggested that the PAX1 methylation status may be an important factor for the aberrant mRNA expression of PAX1 in parathyroid adenoma. We found 7 CpG sites that were maximally hypermethylated (methylation more than 10%) compared to other CpG sites in the PAX1 promoter region of sporadic parathyroid adenomas (Fig. 2E).

Figure 2.

PAX1 promoter methylation in sporadic parathyroid adenoma (n = 40). (A) Schematic representation of the promoter region of PAX1. There are 20 CpG sites in this region. (B) Representative image showing 345 bp PCR products of PAX1 promoter separated electrophoretically on 2% agarose gel. (C) Representative bisulfite sequencing chromatograms showing methylated site (blue arrow) in adenoma compared with control parathyroid samples. (D) Scatter plot showing the correlation analyses between PAX1 mRNA expression and DNA methylation in parathyroid adenoma and control samples (r = - 0.4; P < .01). (E) Tabular representation of 20 CpG sites at PAX1 promoter showing average percentage methylation at each CpG site.

Modification of Histone H3 Associated With PAX1 Promoter in Sporadic Parathyroid Adenoma

Histone modification was quantified as the relative amount of each histone changes in control and parathyroid adenoma groups. In the PAX1 promoter region, the ChIP-qPCR analysis indicated a significant increase of the H3K9me3 levels in parathyroid adenomas (mean ± SEM 20.5 ± 4.42 compared with control parathyroid samples (mean ± SEM 1.05 ± 0.18; P < .0001; Fig. 3A). Besides, the H3K27me3 level was increased by 4.5-fold in sporadic parathyroid adenomas (mean ± SEM 9.98 ± 5.76) compared with control parathyroids (mean ± SEM 2.18 ± 0.33; P = .47; Fig. 3B). However, H3K9ac was decreased by 6-fold in parathyroid adenomas (mean ± SEM 1.03 ± 0.27) compared with control parathyroids (mean ± SEM 6.23 ± 2.87; P < .004; Fig. 3C). H3K9ac levels at the PAX1 promoter were inversely correlated with tumor weight.

Figure 3.

Chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR) for PAX1 promoter region in sporadic parathyroid adenoma. Reactions were performed by qPCR using input DNA. (A) Relative fold enrichment in H3K9me3 (***P < .0001) (B) H3K27me3 (P = .47) (C) H3K9ac (**P < .004) in the PAX1 promoter region in parathyroid adenomas relative to negative control (nonspecific IgG) and normalized with input DNA. Bars are represented as mean fold enrichment (mean+ SEM; n = 23). The data were compared using the student’s unpaired t test. (D) Scatter plot showing the correlation analysis between PAX1 H3K9ac levels and tumor weight of the excised parathyroid adenomas (r= -0.5, P < .03).

In correlation analysis, a positive association between H3K9ac and mRNA expression of PAX1 (r = 0.63, P < .003) was noted (Fig. 3D). Therefore, we focused on the investigation of the potential role of low H3K9ac levels on the promoter region of PAX1 in sporadic parathyroid adenomas.

5-aza-2′ Deoxycytidine and Trichostatin A Cytotoxicity Detection Using MTT Assay

To detect the cytotoxicity of DAC and TSA, we performed MTT assay. As shown in Fig. 4A and 4B, there was no significant difference between the experimental group and control group (without DAC or TSA). We concluded that 10 μM DAC (Fig. 4C and 4D) and 25 nM TSA (Fig. 4E and 4F) produced no cytotoxicity in PTH-C1 cells respectively.

Figure 4.

Effects of DAC and TSA on morphological change and cell viability in PTH-C1 cell line. PTH-C1 cells treated with (A) DAC at indicated concentrations (0.625-20 μM) and (B) TSA at indicated concentrations (5–100 nM) for 72 hours and 24 hours respectively. Cell viability of PTH-C1 cells was determined by MTT assay. Results shown are the means SD of 3 independent experiments performed in triplicate. PTH-C1 cell were treated with control (C) or 10 μM DAC (D) for 72 hours and control (E) or 25 nM TSA (F) for 24 hours were examined with light microscopy (×200 magnification) for changes in cell morphology.

Effect of 5-aza-2′ Deoxycytidine and Trichostatin A on PAX1 Reactivation in PTH-C1 Cells

To confirm that DNA promoter methylation decreases the expression level of PAX1 mRNA, qRT-PCR was performed in PTH-C1 cells treated with DAC. It is known that DAC treatment alone suppresses DNA methylation in cancer cells (19). Treatment with 10 μM DAC for 72 hours caused a significant upregulation (P < .03) of PAX1 by 1.8-fold in PTH-C1 cells (Fig. 5A).

Figure 5.

Effect of DAC and TSA induces expression of PAX1. PTH-C1 cells were treated with 10 μM DAC for 72 hours. Total RNA was isolated, transcribed and resulting cDNA was amplified using PAX1 primers. (A) qRT-PCR showing expression levels of PAX1 control and treated with 10 μM DAC for 72 hours. (B) ChIP-qPCR analysis of H3K9ac on PAX1 promoter in PTH-C1 cells after treatment with TSA alone. Control: No treatment. Bars represent mean ± SEM calculated from 3 independent experiments. The data were compared using the student’s unpaired t test. DAC = 5-aza-2′ deoxycytidine; TSA = trichostatin A.

We detected an association between PAX1 silencing and decreased levels of H3K9ac in parathyroid control and adenoma tissue samples using ChIP assay followed by qPCR. To confirm this finding, ChIP-qPCR was performed to examine changes in histone modification after TSA treatment in PTH-C1 cells. We found that the levels of H3K9ac across the promoter regions of PAX1 (P < .05) were significantly increased by TSA (25 nM for 24 hours) treatment in PTH-C1 cells (Fig. 5B). However, the mRNA level of PAX1 was upregulated in TSA-treated PTH-C1 cells but the difference was insignificant.

Discussion

This study reports the epigenetic changes of the PAX1 gene in sporadic parathyroid adenomas. We found the reduced expression of PAX1 both at the gene and protein level and their gene expression was positively associated with the lower H3K9ac levels in sporadic parathyroid adenomas. In the present study, a subset of sporadic parathyroid adenomas had hypermethylation in the promoter region of PAX1, moreover, treatment with DAC in the PTH-C1 cell line was able to significantly induce PAX1 reactivation. These observations suggested that loss of histone acetylation at the H3K9 site and DNA hypermethylation might play role in altering PAX1 expression in sporadic parathyroid adenomas. Previous studies have revealed that there are numerous transcription factor binding sites in the GCM2 promoter region, including PAX1 (20). The GCM2 gene, which regulates CaSR, is also affected by DNA and histone methylation (H3K9me3) in sporadic parathyroid adenoma (8). Accordingly, it has been suggested that loss of PAX1 expression is further associated with GCM2 gene regulation in sporadic parathyroid adenoma. During embryogenesis, PAX1 is involved in the pouch patterning and is expressed in the endoderm of the third pharyngeal pouch (5). During murine embryogenesis, loss of PAX1 results in hypoplastic or absent parathyroid glands (3). Besides the role of PAX1 in embryonic development, evidence about its involvement in cancers is increasing, but the possible role of PAX1 expression in adult parathyroid tumors has not been investigated. In the present study, qRT-PCR and IHC analysis of parathyroid adenomas showed reduced expression of PAX1, where protein expression inversely correlated with serum calcium and plasma PTH of PHPT patients. This suggests that decreased expression of PAX1 could be associated with disease severity in symptomatic PHPT patients. PAX1 may function as a tumor suppressor gene and is downregulated in several cancers including cervical, oral, ovarian, and head, and neck squamous cell cancer (21-25). Moreover, the regulatory region of PAX1 has been found hypermethylated in human cervical and ovarian cancer (5). However, a previous study examined the DNA methylation status of PAX1 gene in parathyroid adenomas but was rarely observed (26). By contrast, PAX1 promoter was hypermethylated in 35% of the sporadic parathyroid adenomas in our study. Furthermore, the correlation analysis revealed an inverse relationship between PAX1 promoter methylation and its mRNA expression. This suggests a potential pathogenic role of PAX1 gene hypermethylation, at least in a subset of parathyroid adenomas.

Further, to explore another epigenetic regulation in the PAX1 gene, we investigated 2 repressive (H3K9me3 and H3K27me3) and 1 active (H3K9ac) histone modification in sporadic parathyroid adenomas. In correlation analysis, reduced expression was associated with decreased H3K9ac levels of PAX1 in parathyroid adenomas. HDACs and DNA methyltransferases interact with Methyl-CpG-binding protein 2 catalyzing methylation that results in gene silencing (10). Methyl binding proteins (MeCP2) are known as gene-specific transcriptional repressors and have the potential to bind methylated CpG through recruiting the HDAC complex to target promoters (27). To the best of our knowledge, we are the first to show reduced PAX1 expression is affected by DNA hypermethylation and loss of H3K9ac in sporadic parathyroid tumors. Thus, transcription factor, namely PAX1, has an important role in developmental processes and is implicated as a tumor suppressor according to our studies. DAC can effectively reverse DNA methylation and has been approved for the treatment of acute myeloid leukemia and colorectal cancer (8, 28-30). We found a significant re-expression of PAX1 after DAC treatment in PTH-C1 cells. Treatment with TSA revealed pronounced H3K9ac levels in PTH-C1 cells when ChIP-qPCR performed while PAX1 mRNA levels were upregulated but the increase was insignificant of PAX1 in PTH-C1 cells. Previous in vitro studies have reported that a downregulated PAX1 gene could be reactivated by a HDACs inhibitor independent of promoter methylation, suggesting a major role for chromatin remodeling in cervical cancer cells (31). Together with our previous work (8, 9, 32, 33), the present study provides further new insights into parathyroid tumorigenesis, although functional studies of PAX1 in parathyroid adenomas are also required and might help to improve our understanding of pathogenesis of parathyroid adenoma.

It should be noted that there are some limitations in the present study. First, the in vitro studies have performed in rat parathyroid cell line due to the nonavailability of the human parathyroid cell line. Second, protein expression of PAX1 in PTH-C1 cells after treatment of DAC was not performed in parathyroid adenoma cases. Third, we could not check the effect of inhibitors on DNA methylation levels in PTHC-1 cells.

Conclusion

Our study demonstrated that both promoter hypermethylation and deacetylation of histone H3 of lysine 9 (H3K9ac) are responsible for silencing of PAX1 in sporadic parathyroid adenomas. PAX1 expression can be restored by DAC treatment in PTHC-1 cells. Thus, the embryonic transcription factor PAX1 may provide promising biomarker for parathyroid adenoma detection. However, further detailed in-vitro studies with methylation levels and protein expression are necessary to confirm our findings.

Glossary

Abbreviations

BSP

bisulfite-specific polymerase chain reaction

ChIP

chromatin immunoprecipitation

DAC

5′-aza-2 deoxycytidine

FP

forward primer

IHC

immunohistochemistry

PAX 1

paired box 1

PBS

phosphate-buffered saline

PHPT

primary hyperparathyroidism

PTH

parathyroid hormone

RP

reverse primer

RR

reference range

TSA

trichostatin A

TSS

transcription start site

Data Availability

The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request. Some or all datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.