Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Frequent Promoter Hypermethylation of the APC and RASSF1A Tumour Suppressors in Parathyroid Tumours

  • C. Christofer Juhlin ,

    christofer.juhlin@ki.se

    Affiliations Department of Molecular Medicine and Surgery, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden, Department of Oncology-Pathology, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden

  • Nimrod B. Kiss,

    Affiliation Department of Molecular Medicine and Surgery, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden

  • Andrea Villablanca,

    Affiliation Department of Molecular Medicine and Surgery, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden

  • Felix Haglund,

    Affiliation Department of Molecular Medicine and Surgery, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden

  • Jörgen Nordenström,

    Affiliation Department of Molecular Medicine and Surgery, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden

  • Anders Höög,

    Affiliation Department of Oncology-Pathology, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden

  • Catharina Larsson

    Affiliation Department of Molecular Medicine and Surgery, Center for Molecular Medicine, Karolinska Institutet, Karolinska University Hospital Solna, Stockholm, Sweden

Frequent Promoter Hypermethylation of the APC and RASSF1A Tumour Suppressors in Parathyroid Tumours

  • C. Christofer Juhlin, 
  • Nimrod B. Kiss, 
  • Andrea Villablanca, 
  • Felix Haglund, 
  • Jörgen Nordenström, 
  • Anders Höög, 
  • Catharina Larsson
PLOS
x

Abstract

Background

Parathyroid adenomas constitute the most common entity in primary hyperparathyroidism, and although recent advances have been made regarding the underlying genetic cause of these lesions, very little data on epigenetic alterations in this tumour type exists. In this study, we have determined the levels of promoter methylation regarding the four tumour suppressor genes APC, RASSF1A, p16INK4A and RAR-β in parathyroid adenomas. In addition, the levels of global methylation were assessed by analyzing LINE-1 repeats.

Methodology/Principal Findings

The sample collection consisted of 55 parathyroid tumours with known HRPT2 and/or MEN1 genotypes. Using Pyrosequencing analysis, we demonstrate APC promoter 1A and RASSF1A promoter hypermethylation in the majority of parathyroid tumours (71% and 98%, respectively). Using TaqMan qRT-PCR, all tumours analyzed displayed lower RASSF1A mRNA expression and higher levels of total APC mRNA than normal parathyroid, the latter of which was largely conferred by augmented APC 1B transcription levels. Hypermethylation of p16INK4A was demonstrated in a single adenoma, whereas RAR-β hypermethylation was not observed in any sample. Moreover, based on LINE-1 analyses, parathyroid tumours exhibited global methylation levels within the range of non-neoplastic parathyroid tissues.

Conclusions/Significance

The results demonstrate that APC and RASSF1A promoter hypermethylation are common events in parathyroid tumours. While RASSF1A mRNA levels were found downregulated in all tumours investigated, APC gene expression was retained through APC 1B mRNA levels. These findings suggest the involvement of the Ras signaling pathway in parathyroid tumorigenesis. Additionally, in contrast to most other human cancers, parathyroid tumours were not characterized by global hypomethylation, as parathyroid tumours exhibited LINE-1 methylation levels similar to that of normal parathyroid tissues.

Introduction

Primary hyperparathyroidism (PHPT) is a common endocrine disorder which denotes the tumorous enlargement of one or more of the parathyroid glands. While the genetic background of PHPT has been partly elucidated, the involvement of epigenetic modifications remains unresolved. Mutations in the HRPT2 (also known as CDC73) and MEN1 tumour suppressor genes are two important genetic events propelling parathyroid tumorigenesis [1][3]. The protein product of HRPT2 is parafibromin, which is functionally linked to the canonical part of the Wingless type (Wnt) pathway [4]. Parafibromin is a part of the human polymerase associated factor complex (hPAF) which is required for facilitating transcription elongation and histone modifications [5][6]. Furthermore, parafibromin has been shown to associate with the Wnt oncoprotein β-catenin and to down-regulate c-Myc oncogene transcription through direct binding to the c-Myc promoter region, supporting the role of parafibromin as a tumor suppressor protein that inhibits Wnt signaling [7]. Interestingly, the MEN1 product menin was also coupled to the Wnt pathway, demonstrated by its epigenetic regulation of Axin2 and by expression profiling of MEN1 knock-out mice [8][9].

To broaden the knowledge of epigenetic events in parathyroid tumour development, we sought to determine the levels of promoter methylation of four tumour suppressor genes; APC (adenomatous polyposis coli), p16INK4A, RASSF1A (Ras association domain family protein 1) and RAR-β (Retinoic acid receptor-beta), all known to be widely hypermethylated in various human cancers as well as exhibiting potential as parathyroid tumour suppressor genes. APC mutations are responsible for the autosomal dominant familial adenomatous polyposis (FAP) syndrome [10][11], and are also found in sporadic colorectal cancer as well as in other tumour types, such as cancer of the thyroid and mammary glands [12][13]. Hypermethylation of the APC promoter 1A has been demonstrated in e.g. cancer of the colorectum, breast and lung [14][15]. Furthermore, loss of APC immunoreactivity has recently been implicated in parathyroid malignant tumours [16]. The tumour suppressor protein p16ink4A regulates the G1 to S phase transition of the cell cycle through its inhibition of the interaction between Cdk 4/6 and cyclin D1. Parathyroid tumours often demonstrate over-expression of cyclin D1, thus making the p16ink4A gene an interesting candidate for promoter methylation assays since p16ink4A inactivation might produce parallel oncogenic effects analogous to cyclin D1 up-regulation [17]. p16ink4A gene mutations have been sought in parathyroid tumours due to the presence of 9p21-pter deletions in small subsets of parathyroid adenomas; however, no mutations have been demonstrated [18].

The RASSF1A gene localized to 3p21 is epigenetically inactivated with high regularity in human cancers as well as in endocrine tumours such as pheochromocytomas and follicular thyroid cancer [19][21]. Interestingly, RASSF1A has been shown to down-regulate cyclin D1 expression in addition to its role as a Ras-binding protein, hypothetically making the RASSF1A gene a potential candidate as a parathyroid tumour suppressor gene [22]. RAR-β is a member of the nuclear receptor superfamily with the main function to convey retinoid signaling into target gene transcription in the nucleus. The RAR-β promoter exhibit frequent hypermethylation in various tumours, as for instance in breast- and prostate cancer as well as in malignant mesothelioma [23][25]. Recently, it was shown that up-regulation of the oncogene c-myc increased the levels of RAR-β promoter methylation in a prostate cell line, suggesting that these events might be directly or indirectly coupled [26]. These finding are of particular interest as c-myc over-expression has been demonstrated in parathyroid tumours [27].

In addition to gene-specific hypermethylation, global hypomethylation is a common characteristic in human cancers [28][29]. LINE-1 (long interspersed nuclear elements-1) retrotransposons denote replicating repetitive elements which constitute about 15% of the human genome. CpG sites within the LINE-1 promoter region are normally heavily methylated to protect from retrotransposon activation, and LINE-1 hypomethylation is a noted epigenetic event in human tumourigenesis. For example, LINE-1 promoter hypomethylation is common in colon cancer, and the retrotransposal insertion of a LINE-1 sequence has been directly shown to disrupt the APC gene in a case of colon cancer [29][30]. Furthermore, quantifying LINE-1 promoter methylation densities have been shown to be a reliable substitute for global methylation assays [31].

Towards these ends, we have quantified promoter methylation of APC, RASSF1A, p16INK4A, RAR-β and LINE-1 repeats in a series of parathyroid tumours characterized for HRPT2 and MEN1 genotypes.

Results

The APC and RASSF1A Promoters Are Hypermethylated in Parathyroid Tumours

In total, 55 parathyroid tumors were assessed for APC, RASSF1A, p16INK4A and RAR-β regional promoter methylation using Pyrosequencing analysis. Examples of Pyrograms are illustrated in Figure 1 and the results are summarized in Table 1. The density of promoter methylation in each tumour was calculated as a mean for the individual CpG residues, and the results are also presented as the range from minimum to maximum methylation density for the individual sites assessed (Table 1). The results indicate that CpG-specific methylation in parathyroid tumours is a common event at the APC and RASSF1A promoters but rarely occur at the p16INK4A and RAR-β promoter loci (Figure 2). For APC, the mean CpG methylation ranged from 1.5% (case RAD1) to 77.5% (case H2) in our panel (Table 1). Normal parathyroid tissues (Normal PT 1–3) displayed very low levels of APC promoter methylation (2.2%, 4.0% and 3.3%; Table 1). In addition, SHPT1, thymus tissue from case H1 as well as leukocyte DNA from patient RAD2 also exhibited low APC methylation (2.3% and 2.5%), in agreement with previous report of low APC methylation density in other non-neoplastic tissues such as colonic mucosa (≤5%) [32]. In our study, a total of 39 out of 55 tumours (71%) were considered to exhibit a hypermethylated APC promoter 1A (above 10% density) as compared to the 16 tumours (29%) with very low methylation (below 10% density). Twenty-four out of the 34 cystic parathyroid adenomas (71%) exhibited hypermethylation, as well as 14 out of 19 (74%) of the regular parathyroid adenomas. For the two cases with germ-line HRPT2 gene mutations (H1-2), the atypical adenoma case H1 presented with virtually no methylation (2.3%), whereas the parathyroid adenoma H2 was endowed with the highest density of APC methylation in the entire series (77.5%).

thumbnail
Figure 1. Examples of Pyrograms for quantification of methylation densities.

10 CpG sites of the APC promoter 1A, 5 CpG sites of the RASSF1A promoter and 7 CpG sites of the p16INK4A promoter are shown for case RAD 13, a tumour carrying a somatic MEN1 mutation. The tumour displays methylation densities of 33.9% at APC (range 16–41%), 58.2% at RASSF1A (range 49–68%) and 1.6% at p16INK4A (range 1–3%), respectively. Nucleotides at position #16 of the APC promoter region 1A and at position #32 of the p16INK4A promoter (yellow) are examples of CpG free cytosine residues which serve as internal controls for satisfactory bisulphite-driven C to T conversion for each DNA sample.

http://dx.doi.org/10.1371/journal.pone.0009472.g001

thumbnail
Figure 2. Global and regional hypermethylation in parathyroid tumours.

Individual value plot illustrating the mean densities of CpG methylation of the APC, RASSF1A, p16INK4A, RAR-β and LINE-1 promoters for the 55 parathyroid tumours (T) and 3 normal parathyroid controls (N). The Y axis represents the mean density of methylation calculated as an average of the individual CpG residues assayed.

http://dx.doi.org/10.1371/journal.pone.0009472.g002

thumbnail
Table 1. Results from the gene promoter methylation analyses in parathyroid tumours and control tissues.

http://dx.doi.org/10.1371/journal.pone.0009472.t001

For RASSF1A, the methylation ranged from 8% (CAD13) to 80% (RAD4) and 54 out of 55 (98%) tumours displayed methylation densities above 10% (Table 1, Figure 2). The three normal parathyroid samples presented with almost no RASSF1A promoter methylation and additional non-tumorous controls consisting of normal thymus and leukocyte DNA also displayed low levels of RASSF1A methylation. Case SHPT1 however exhibited an average methylation density of 30.2% (Table 1, Figure 2). There was a significant but weak positive correlation between APC and RASSF1A methylation levels, suggesting that tumours with high levels of APC methylation density frequently demonstrate high RASSF1A methylation levels (Pearson's correlation 0.301, p = 0.022, Table 2).

thumbnail
Table 2. Correlations between hypermethylation levels and tumour characteristics.

http://dx.doi.org/10.1371/journal.pone.0009472.t002

Regarding p16INK4A promoter methylation, one case was found to be hypermethylated as compared to non-tumorous controls (RAD2, 14.9%). The remaining 54 cases (98%) displayed very low mean levels of promoter methylation densities (<10%) (Table 1). However, nine of these cases were found to exhibit methylation levels above 10% at one or more individual CpG sites, with preferentially involvement of CpG site 7. For RAR-β, all tumours and normal controls displayed low levels of promoter methylation (<10%) (Table 1).

Assessment of Global Methylation in Parathyroid Tumours

Regarding LINE-1 promoter methylation, normal parathyroid tissues ranged from 59.7% to 72.3%, thereby encompassing the vast majority (45/55; 82%) of all parathyroid tumours investigated.

Comparison of Promoter Methylation Density with Clinical and Genetic Parameters

Possible correlations were evaluated between methylation density of the individual hypermethylated APC and RASSF1A promoters with clinical and genetic characteristics (Table 1, Table 2, Table 3 and Table S1). No significant correlations were observed between APC promoter hypermethylation and serum levels of calcium, PTH or glandular weight (Table 2). For RASSF1A, no correlation between methylation levels and serum calcium, PTH or glandular weight was demonstrated; however, increased levels of RASSF1A methylation correlated negatively with serum calcium (Table 2).

All MEN1 mutated tumours in the current study displayed APC promoter 1A hypermethylation, and a weak but significant correlation (Pearson's correlation 0.322, p = 0.027) was revealed between high APC methylation density and presence of an MEN1 mutation (Figure 3). However, no association was observed between RASSF1A methylation levels and MEN1 genotype (Table 2). Furthermore, there was no correlation between APC or RASSF1A promoter methylation densities and HRPT2 genotype (Table 2).

thumbnail
Figure 3. APC promoter 1A hypermethylation in parathyroid tumours with a coupling to MEN1 mutations.

Boxplot displaying APC promoter methylation in comparison to MEN1 genotype for all cases where genotype data was available. The endpoints of the boxes represent the first and third quartiles respectively, and the horizontal line corresponds to the median. High APC promoter methylation density is statistically correlated to MEN1 mutational status.

http://dx.doi.org/10.1371/journal.pone.0009472.g003

APC and RASSF1A mRNA Expression Analyses by qRT-PCR

For APC, we assessed both total levels of APC mRNA and an APC 1B specific transcript regulated by the APC promoter 1B. No TaqMan Gene Expression Assay exists exclusively for the APC 1A transcript. For the total levels of APC mRNA, all 16 parathyroid adenomas available for qRT-PCR exhibited greater relative expression levels compared to the normal parathyroid mean, ranging from 1.3–9 (normal parathyroid mean  = 1) and a mean relative expression of 4.5. For the APC 1B transcript, all tumours displayed greater relative expression levels compared to the normal parathyroid mean ranging from 2–20 (normal parathyroid mean  = 1) and a mean relative expression of 8.7 (Figure 4). For RASSF1A, all 16 adenomas investigated have been shown to exhibit RASSF1A hypermethylation ranging from 37.2–71.2%. Using qRT-PCR, all tumours were found to exhibit lower relative RASSF1A mRNA expression levels compared to the normal parathyroid mean value, ranging from 0.05–0,8 (normal parathyroid mean  = 1). The mean relative expression was 0.26, suggesting that parathyroid adenomas display a noteworthy reduction in RASSF1A expression compared to normal parathyroid (Figure 4).

thumbnail
Figure 4. Interval plots demonstrating relative RASSF1A and APC gene expression in parathyroid tumours.

16 adenomas were compared to the normal parathyroid mean regarding (A) RASSF1A and (B) total APC and APC 1B gene expression. The endpoints of the intervals represent the 95% confidence interval for the mean, and the diamond corresponds to the mean value.

http://dx.doi.org/10.1371/journal.pone.0009472.g004

Discussion

In this study we demonstrate frequent hypermethylation of the APC and RASSF1A tumour suppressor genes in the majority of parathyroid adenomas. RASSF1A promoter hypermethylation led to diminished RASSF1A mRNA levels in parathyroid adenomas as compared to normal parathyroid tissues, whereas APC gene expression is retained through augmented expression of APC 1B mRNA regulated by APC promoter 1B. In addition, based on LINE-1 promoter methylation results parathyroid benign tumours were found to exhibit global methylation levels within range of the normal parathyroid tissues examined.

APC gene expression is regulated by at least two separate promoter regions, named 1A and 1B. Hypermethylation of promoter 1A has been readily demonstrated in various human tumours; however to our knowledge there is no or little evidence suggesting epigenetic inactivation of the APC promoter 1B through hypermethylation [33]. Hence, for this purpose we focused solely on the APC promoter 1A regarding Pyrosequencing analyses. In our tumour panel, hypermethylation of the APC promoter 1A was found in the majority of benign parathyroid tumours, as well as in one familial HRPT2 related case. These results are supported by research in colorectal tumorigenesis, where promoter 1A methylation is regarded as an early event and is readily detected in small colorectal adenomas [33]. Furthermore, in this study all 16 adenomas analyzed by qRT-PCR displayed augmented APC 1B mRNA levels as compared to normal parathyroid tissues, strongly suggesting that the APC promoter 1B is not epigenetically silenced in these tumours.

Parafibromin, encoded by HRPT2, is a vital part of the hPAF complex involved in epigenetic modifications for facilitating gene transcription, and has also been shown to specifically regulate c-Myc oncogene expression through binding directly to the c-Myc promoter region [7]. Our results do not point towards a specific correlation between APC promoter methylation and HRPT2 gene mutations, suggesting that parafibromin may not play an apparent role in the regulation of APC promoter 1A methylation. However, APC promoter 1A methylation displayed a weak but significant correlation to the tumour suppressor MEN1 genotype, and all parathyroid tumours with MEN1 mutations in our panel displayed APC hypermethylation ranging from 11.6–65.9% (Figure 2). Interestingly, the MEN1 protein product menin has recently been implicated as an epigenetic modifier of Axin2, a component within the Wnt pathway as well as an APC-associated protein [8]. Our study vaguely implies that menin might regulate the levels of APC promoter methylation at the DNA level in parathyroid cells, since cases with MEN1 mutations in general displayed higher levels of methylation than wild-type cases. Further molecular studies are needed to investigate the potential role of menin as a DNA regulator of APC promoter methylation status in addition to its known role as an Axin2 regulator through histone methylation.

The RASSF1A promoter was found to be hypermethylated in nearly all parathyroid adenomas examined and RASSF1A gene expression levels were considerably reduced in adenomas compared to normal parathyroid tissue. These data strongly support the involvement of this tumour suppressor gene in parathyroid tumorigenesis. In unrelated tumour tissues, RASSF1A promoter methylation has been shown to greatly diminish the levels of RASSF1A mRNA transcription, and subsequent usage of demethylating agents such as 5-Aza-CdR reconstitutes the RASSF1A transcription levels [34]. The role of RASSF1A as a cyclin D1 posttranslational modifier motivated us to correlate our methylation data to previous cyclin D1 expression results for the parathyroid cystic adenomas published elsewhere [35]. However, no statistical correlation between increased RASSF1A promoter methylation density and over-expressed cyclin D1 levels were seen (data not shown), suggesting that RASSF1A hypermethylation at least in part might contribute to excessive parathyroid tumour growth by other means than cyclin D1 accumulation, for example through its role as a regulator of apoptosis and mitotic arrest [19].

Regarding APC, we demonstrate three cases of normal parathyroid tissue (Normal PT1-3) as well as a case of secondary hyperparathyroidism with hyperplasia (case SHPT1) displaying virtually no CpG promoter methylation, suggesting that methylation of this locus might be conserved to true tumorous lesions of the parathyroids, such as adenomas and carcinomas. Moreover, studies have shown that normal adrenal as well as normal colonic mucosa in general shows APC promoter 1A methylation levels below 5%, as compared to the higher levels observed in the corresponding tumours [20], [32]. Also, the vast majority of cases of normal breast and lung tissue display an un-methylated APC promoter 1A as compared to the corresponding tumours [15], [36]; consequently these are examples which illustrate the tumour-specific phenotype for APC promoter 1A methylation in some tissue types.

For RASSF1A, the three normal parathyroid samples were endowed with virtually no methylation (2.0%, 3.6% and 3.0% respectively) whereas case SHPT1 was endowed with a methylation density of approximately 30%, suggesting that RASSF1A methylation also can be found in parathyroid tissue with an disproportionate growth pattern, albeit not a tumour per definition. Interestingly, cyclin D1 over-expression is also found in a fraction of cases with secondary hyperparathyroidism, suggesting that a subset of these lesions are propelled in parts by processes similar to those present in parathyroid adenomas and carcinomas [37]. Hence, the finding of RASSF1A hypermethylation in case SHPT1 could imply that this process could contribute to pathological growth in secondary hyperparathyroidism in addition to adenomas. In other non-neoplastic endocrine tissues such as the normal adrenal medulla and normal thyroid tissue, the RASSF1A promoter habitually exhibit low levels (<10%) of methylation densities with few exceptions [20][21]. In our study, normal thymus and leukocyte DNA was used as additional controls and displayed low levels of RASSF1A methylation, validating the specificity of our findings. Furthermore, a tumour case in our study displayed virtually no RASSF1A promoter methylation (CAD13, 8%), suggesting that parathyroid tissue not routinely exhibit high levels of methylation at this locus.

It is not known whether the hypermethylation of APC promoter 1A bears any pathological consequences to parathyroid growth, since APC protein expression has been previously demonstrated in parathyroid adenomas [16], [27]. However, a recent study has shown reduced APC promoter 1A mRNA levels in colorectal tumours with APC promoter 1A hypermethylation although retained APC protein expression through promoter 1B activity, and the authors advocate that even a slight decrease of APC gene transcription through the epigenetic silencing of promoter 1A could have an impact on Wnt signalling, although via unknown mechanisms [33]. Furthermore, another study regarding colorectal cancer found no correlations between APC methylation levels and APC protein expression [38]. In this study, we found notably augmented levels of total APC mRNA levels (from promoters 1A and 1B) as well as APC 1B mRNA in the adenomas investigated by qRT-PCR, of which 13 out of 16 samples displayed APC promoter 1A hypermethylation (Figure 4). These data imply that APC 1B mRNA is highly expressed in parathyroid adenomas, suggesting that the APC promoter 1B is not epigenetically inactivated in these cases. This presents a credible explanation to the fact that parathyroid adenomas display retained APC protein expression as demonstrated in previous publications although possibly endowed with APC promoter 1A hypermethylation [16], [27]. Furthermore, the finding of elevated levels of total APC mRNA is most likely a result of elevated levels of APC 1B mRNA expression, and does not argue directly against the theoretical, specific down-regulation of the APC 1A mRNA, the latter which is an observed phenomenon in colon cancer research [33]. In studies of unrelated tumour tissues, APC promoter 1A methylation is coupled to down-regulation of APC 1A mRNA, and treatment with demethylating agents results in re-expression of the specific transcript, demonstrating the sensible correlation between APC promoter methylation and APC mRNA transcript levels in vivo [39].

LINE-1 bisulphite Pyrosequencing analysis has been shown to be an efficient substitute for global methylation assays [31], and our data thus suggests that parathyroid tumours display comparable methylation levels globally as normal parathyroid cells. This finding was unexpected, considering the current notion that human cancers generally exhibit global hypomethylation. In support of our findings, a current study recently demonstrated global hypermethylation in patients with myelodysplastic syndrome (MDS) as compared to normal controls using LINE-1 Pyrosequencing [40]. The authors furthermore observed noteworthy clinical improvement among MDS patients when treated with hypomethylation-inducing agents, suggesting that the global hypermethylation was of clinical significance in this tumour type. Hence, these data suggest that not all tumours display relative hypomethylation globally compared to their corresponding normal tissues. In addition, we found that male cases displayed significantly higher levels of LINE-1 methylation than female cases (Pearson's correlation −0.466, p<0.001), and cases with higher age at surgery displayed lower levels of LINE-1 methylation than younger cases (Pearson's correlation 0.320, p = 0.018). These observations may suggest an influence of age or gender upon global levels of methylation in parathyroid tissues.

In summary, this study is the first to identify APC and RASSF1A promoter hypermethylation in the majority of parathyroid tumours investigated. The frequently observed hypermethylation of the RASSF1A promoter and downregulation of RASSF1A mRNA levels in parathyroid adenomas point toward an association between the inactivation of this tumour suppressor gene and parathyroid aberrant growth, whereas APC gene expression is retained in parathyroid adenomas due to increased transcription from APC promoter 1B. Moreover, LINE-1 analyses indicate that parathyroid tumours do not exhibit global hypomethylation as compared to normal parathyroid tissues.

Materials and Methods

Ethics Statement

Fifty-five samples of parathyroid adenoma were collected with informed verbal consent. Informed verbal consent constitutes standard procedure at the Karolinska University Hospital, and the consent is subsequently documented in the patient's medical journal as according to Swedish Biobank law. The procedure has been specifically approved by the Karolinska University Hospital Ethics Committee and by the Karolinska Institute Research Ethics Committee, and these approvals include collection of normal parathyroid biopsies as well as normal tissues (including blood and thymus tissue) whenever motivated from patients undergoing surgery for primary and secondary hyperparathyroidism.

Patients and Tissue Samples

Fifty-five samples of parathyroid adenoma were collected for the study (Tables 1 and 3, Table S1). Histopathological classifications were according to the guidelines of the World Health Organization (WHO) [41]. Thirty-four cases were parathyroid adenomas with cystic features (CAD1-34), which have previously been reported in detail including screening for HRPT2 and MEN1 mutations [3], [35]. Thirty-three of these cases were sporadic and one single case (CAD11) was derived from a 1q linked FIHP family. In addition one case (CAD25) was classified as an atypical adenoma. Two additional HRPT2 related adenomas were derived from patients with germ-line HRPT2 mutations (H1-2). Tumor H1 exhibited atypical histological findings that are sometimes seen in malignant parathyroid tumours, but was without definite criteria of malignancy in agreement with the diagnosis atypical adenoma or equivocal carcinomas as previously applied [42]. Nineteen cases were regular parathyroid adenomas (RAD1-19) that have been previously characterized for MEN1 genotype [43][44]. Moreover, a case of parathyroid hyperplasia due to secondary hyperparathyroidism (SHPT1) was also included in the study. As non-tumorous normal controls, three samples of normal parathyroid tissue derived from surgical biopsies were included together with human leukocyte DNA from patient RAD2 as well as thymic tissue from patient H1.

Bisulphite Treatment of DNA and Pyrosequencing Analyses

Sodium bisulphite modification of 500 ng DNA was carried out with the EZ DNA methylation kit (D5002, Zymo Research Corporation, CA, USA) following the manufacturer's protocol. The subsequent PCR and Pyrosequencing analysis to assess the level of methylation was carried out essentially as previously described for 10 CpG sites of the APC promoter 1A [20], 7 CpG sites of the p16INK4A promoter [45], 5 CpG sites of the RASSF1A promoter [46], 10 CpG sites of the RAR-β promoter [20] and 3 CpG sites of the LINE-1 promoter [20]. Primer sequences are available at the PyroMark Assay Database (Biotage, Sweden) for all assays with the exception of p16INK4A, and were designed to hybridize with CpG-free regions to secure methylation-independent amplification. HotStar Taq polymerase was used to amplify 1 µl of bisulphite–treated DNA. PCR conditions were 95°C 15 min followed by 45 cycles of: 95°C 20 s, 50°C (RAR-β and LINE-1) or 55°C (APC, RASSF1A, p16INK4A) 20 s, 72°C 20 s, followed by 72°C 5 min and finally 4° C. All samples and blank controls for each reaction (PCR water control and additional non-PCR water control) were then analyzed using Pyrosequencing analysis and the PyroMark Q24 system (Biotage). Subsequent quantification of methylation density was done using the PyroMark Q24 software. Cases of bisulphite-treated normal leukocyte and thymic DNA were assessed as additional negative controls. Subsets of cases were re-analyzed with repeated PCR and Pyrosequencing reactions to test the validity of the initial Pyrosequencing findings for all promoter regions included, with similar end results (data not shown). Subsets of cases (including both tumour and normal parathyroid samples) were also repeatedly bisulphite treated and subsequently analyzed using PCR and Pyrosequencing, with analogous results (data not shown).

We chose to define hypermethylation as an average CpG residue methylation above 10% for APC, RASSF1A, p16INK4A and RAR-beta since the normal parathyroid mean as well as other normal tissues in this study were below this value.

RNA Isolation, cDNA Synthesis and Quantitative RT-PCR (qRT-PCR)

Total RNA was available for 16 of the parathyroid adenomas and two of the normal parathyroid glands (normal PT1-2). All samples were extracted using the TRIzol Reagent (Invitrogen, CA, USA) and subsequently purified with the RNeasy kit (Qiagen, CA, USA). The RNA quality was verified using the NanoDrop ND-1000 (Thermo Scientific, DE, USA) for confirmation of acceptable 260/280 nm absorbance ratio as well as determining the sample concentrations. In addition, all RNA samples were analyzed in agarose gels where the integrity of the 28S and 18S bands was verified. Subsequent cDNA synthesis was performed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, CA, USA) following the company protocol. Subsequent assessment of APC and RASSF1A mRNA levels was carried out by quantitative RT-PCR (qRT-PCR) using the TaqMan Gene Expression Assay Technique and the 7900HT Fast Real-Time PCR System (Applied Biosystems, CA, USA). For APC, two assays were included; Hs00181051_m1 which covers total levels of APC mRNA (covering three transcript variants, one regulated from promoter 1A and two from 1B) and Hs01568282_m1 which covers the specific APC 1B transcript regulated by promoter 1B. There is no assay available exclusively for the analysis of the APC 1A transcript regulated by the APC promoter 1A using the TaqMan Gene Expression Assay technique. The RPLP0 (ribosomal protein, large, P0), also known as 36B4, has been used in previous publications by the authors and was selected as a housekeeping reference known to demonstrate little variation in parathyroid tissues (Hs99999902_m1) [35]. For the APC promoter 1A, 13/16 adenomas available for qRT-PCR displayed hypermethylation ranging from 10.3–64.1%. For RASSF1A, all 16 adenomas analyzed demonstrated RASSF1A promoter hypermethylation ranging from 37.2–71.2%. The two normal parathyroid tissues analyzed were devoid of hypermethylation covering both the RASSF1A and APC promoters.

A standard curve was established for relative quantification using cDNA from tumour case CAD20, and the qRT-PCR was carried out in 25 µl reactions with 12.5 µl TaqMan Gene Expression Master Mix, 8.5 µl RNase free H2O, 1 µl TaqMan Gene Expression Assay and 3 µl cDNA. The thermocycling conditions were 50°C for 2 min, 95°C for 10 min followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. Two non-template controls were assessed as negative controls for each experiment, and all experiments were carried out in double. After quantification relative to the standard curve, all samples were normalized to their corresponding RPLP0 value and thereafter the normal parathyroid mean for each assay.

Statistical Analyses

Possible correlations between methylation density in the different promoters assayed as well as between individual promoters and clinical/genetic characteristics were evaluated using the Minitab 15 statistical software (Minitab Inc, PA, USA). Correlations were appraised by Pearson's correlation and differences between groups were verified using Fisher's exact test. P-values <0.05 were considered as statistically significant.

Supporting Information

Table S1.

doi:10.1371/journal.pone.0009472.s001

(0.03 MB XLS)

Acknowledgments

The authors are indebted to Dr. Klas Strååt for exceptional technical assistance, Dr. Robert Bränström for valuable contributions to the study design and to Ms. Elisabeth Ånfalk for excellent tissue handling.

Author Contributions

Conceived and designed the experiments: CJ NBK AV AH CL. Performed the experiments: CJ NBK FH. Analyzed the data: CJ NBK AV CL. Contributed reagents/materials/analysis tools: JN AH. Wrote the paper: CJ CL.

References

  1. 1. Chandrasekharappa SC, Guru SC, Manickam P, Olufemi SE, Collins FS, et al. (1997) Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 276: 404–407.
  2. 2. Lemmens I, Merregaert J, Van de Ven WJ, Kas K, Zhang CX, et al. (1997) Construction of a 1.2-Mb sequence-ready contig of chromosome 11q13 encompassing the multiple endocrine neoplasia type 1 (MEN1) gene. The European Consortium on MEN1. Genomics 44: 94–100.
  3. 3. Carpten JD, Robbins CM, Villablanca A, Forsberg L, Presciuttini S, et al. (2002) HRPT2, encoding parafibromin, is mutated in hyperparathyroidism-jaw tumor syndrome. Nat Genet 32: 676–680.
  4. 4. Mosimann C, Hausmann G, Basler K (2006) Parafibromin/Hyrax activates Wnt/Wg target gene transcription by direct association with β-catenin/Armadillo. Cell 125: 327–341.
  5. 5. Rozenblatt-Rosen O, Hughes CM, Nannepaga SJ, Shanmugam KS, Copeland TD, et al. (2005) The parafibromin tumor suppressor protein is part of a human Paf1 complex. Mol Cell Biol 25: 612–620.
  6. 6. Yart A, Gstaiger M, Wirbelauer C, Pecnik M, Anastasiou D, et al. (2005) The HRPT2 tumor suppressor gene product parafibromin associates with human PAF1 and RNA polymerase II. Mol Cell Biol 25: 5052–5060.
  7. 7. Lin L, Zhang J-H, Panicker LM, Simonds WF (2008) The parafibromin tumor suppressor protein inhibits cell proliferation by repression of the c-myc proto-oncogene. Proc Natl Acad Sci U S A 105: 17420–17425.
  8. 8. Chen G, A J, Wang M, Farley S, Lee LY, et al. (2008) Menin promotes the Wnt signaling pathway in pancreatic endocrine cells. Mol Cancer Res 6: 1894–1907.
  9. 9. Mould AW, Duncan R, Serewko-Auret M, Loffler KA, Biondi C, et al. (2009) Global expression profiling of sex cord stromal tumors from Men1 heterozygous mice identifies altered TGF-beta signaling, decreased Gata6 and increased Csf1r expression. Int J Cancer 124: 1122–1132.
  10. 10. Groden J, Thliveris A, Samowitz W, Carlson M, Gelbert L, et al. (1991) Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66: 589–600.
  11. 11. Miyoshi Y, Ando H, Nagase H, Nishisho I, Horii A, et al. (1992) Germ-line mutations of the APC gene in 53 familial adenomatous polyposis patients. Proc Natl Acad Sci U S A 89: 4452–4456.
  12. 12. Zeki K, Spambalg D, Sharifi N, Gonsky R, Fagin JA (1994) Mutations of the adenomatous polyposis coli gene in sporadic thyroid neoplasms. J Clin Endocrinol Metab 79: 1317–1321.
  13. 13. Hayes MJ, Thomas D, Emmons A, Giordano TJ, Kleer CG (2008) Genetic changes of Wnt pathway genes are common events in metaplastic carcinomas of the breast. Clin Ca Res 14: 4038–4044.
  14. 14. Esteller M, Sparks A, Toyota M, Sanchez-Cespedes M, Capella G, et al. (2000) Analysis of adenomatous polyposis coli promoter hypermethylation in human cancer. Cancer Res 60: 4366–4371.
  15. 15. Virmani AK, Rathi A, Sathyanarayana UG, Padar A, Huang CX, et al. (2001) Aberrant methylation of the adenomatous polyposis coli (APC) gene promoter 1A in breast and lung carcinomas. Clin Ca Res 7: 1998–2004.
  16. 16. Juhlin CC, Haglund F, Villablanca A, Forsberg L, Sandelin K, et al. (2009) Loss of expression for the Wnt pathway components adenomatous polyposis coli (APC) and glycogen synthase kinase 3-β (GSK3-β) in parathyroid carcinomas. Int J Oncol 34: 481–492.
  17. 17. Hsi ED, Zukerberg LR, Yang WI, Arnold A (1996) Cyclin D1/PRAD1 expression in parathyroid adenomas: an immunohistochemical study. J Clin Endocrinol Metab 81: 1736–1739.
  18. 18. Tahara H, Smith AP, Gaz RD, Arnold A (1996) Loss of chromosome arm 9p DNA and analysis of the p16 and p15 cyclin-dependent kinase inhibitor genes in human parathyroid adenomas. J Clin Endocrinol Metab 81: 3663–3667.
  19. 19. Donninger H, Vos MD, Clark GJ (2007) The RASSF1A tumor suppressor. J Cell Sci 120: 3163–3172.
  20. 20. Geli J, Kiss N, Karimi M, Lee JJ, Bäckdahl M, et al. (2008) Global and regional CpG methylation in pheochromocytomas and abdominal paragangliomas: association to malignant behavior. Clin Ca Res 14: 2551–2559.
  21. 21. Lee JJ, Geli J, Larsson C, Wallin G, Karimi M, et al. (2008) Gene-specific promoter hypermethylation without global hypomethylation in follicular thyroid cancer. Int J Oncol 33: 861–869.
  22. 22. Shivakumar L, Minna J, Sakamaki T, Pestell R, White MA (2002) The RASSF1A tumor suppressor blocks cell cycle progression and inhibits cyclin D1 accumulation. Mol Cell Biol 22: 4309–4318.
  23. 23. Fackler MJ, McVeigh M, Evron E, Garrett E, Mehrotra J, et al. (2003) DNA methylation of RASSF1A, HIN-1, RAR-beta, Cyclin D2 and Twist in in situ and invasive lobular breast carcinoma. Int J Cancer 107: 970–975.
  24. 24. Jerónimo C, Henrique R, Hoque MO, Ribeiro FR, Oliveira J, et al. (2004) Quantitative RARbeta2 hypermethylation: a promising prostate cancer marker. Clin Ca Res 10: 4010–4014.
  25. 25. Fischer JR, Ohnmacht U, Rieger N, Zemaitis M, Stoffregen C, et al. (2006) Promoter methylation of RASSF1A, RARbeta and DAPK predict poor prognosis of patients with malignant mesothelioma. Lung Cancer 54: 109–116.
  26. 26. He M, Vanaja DK, Karnes RJ, Young CY (2009) Epigenetic regulation of Myc on retinoic acid receptor beta and PDLIM4 in RWPE1 cells. Prostate 69: 1643–1650.
  27. 27. Björklund P, Akerström G, Westin G (2007) Accumulation of nonphosphorylated beta-catenin and c-myc in primary and uremic secondary hyperparathyroid tumors. J Clin Endocrinol Metab 92: 338–344.
  28. 28. Feinberg AP, Vogelstein B (1983) Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature 301: 89–92.
  29. 29. Ogino S, Nosho K, Kirkner GJ, Kawasaki T, Chan AT, et al. (2008) A cohort study of tumoral LINE-1 hypomethylation and prognosis in colon cancer. J Natl Cancer Inst 100: 1734–1738.
  30. 30. Miki Y, Nishisho I, Horii A, Miyoshi Y, Utsunomiya J, et al. (1992) Disruption of the APC gene by a retrotransposal insertion of L1 sequence in a colon cancer. Cancer Res 52: 643–645.
  31. 31. Yang AS, Estécio MR, Doshi K, Kondo Y, Tajara EH, et al. (2004) A simple method for estimating global DNA methylation using bisulfite PCR of repetitive DNA elements. Nucleic Acids Res 32: e38.
  32. 32. Belshaw NJ, Elliott GO, Foxall RJ, Dainty JR, Pal N, et al. (2008) Profiling CpG island field methylation in both morphologically normal and neoplastic human colonic mucosa. Br J Cancer 99: 136–142.
  33. 33. Segditsas S, Sieber OM, Rowan A, Setien F, Neale K, et al. (2008) Promoter hypermethylation leads to decreased APC mRNA expression in familial polyposis and sporadic colorectal tumours, but does not substitute for truncating mutations. Exp Mol Pathol 85: 201–206.
  34. 34. Shen WJ, Dai DQ, Teng Y, Liu HB (2008) Regulation of demethylation and re-expression of RASSF1A gene in gastric cancer cell lines by combined treatment of 5-Aza-CdR and NaB. World J Gastroenterol 14: 595–600.
  35. 35. Juhlin C, Höög A, Yakoleva T, Leibiger I, Leibiger B, et al. (2006) Loss of parafibromin expression in a subset of sporadic parathyroid adenomas. Endocr Relat Cancer 13: 509–523.
  36. 36. Van der Auwera I, Van Laere SJ, Van den Bosch SM, Van den Eynden GG, Trinh BX, et al. (2008) Aberrant methylation of the Adenomatous Polyposis Coli (APC) gene promoter is associated with the inflammatory breast cancer phenotype. Br J Cancer 99: 1735–1742.
  37. 37. Tominaga Y, Tsuzuki T, Uchida K, Haba T, Otsuka S, et al. (1999) Expression of PRAD1/cyclin D1, retinoblastoma gene products, and Ki67 in parathyroid hyperplasia caused by chronic renal failure versus primary adenoma. Kidney Int 55: 1375–1383.
  38. 38. Chen J, Röcken C, Lofton-Day C, Schulz HU, Müller O, et al. (2005) Molecular analysis of APC promoter methylation and protein expression in colorectal cancer metastasis. Carcinogenesis 26: 37–43.
  39. 39. Zysman M, Saka A, Millar A, Knight J, Chapman W, Bapat B (2002) Methylation of adenomatous polyposis coli in endometrial cancer occurs more frequently in tumors with microsatellite instability phenotype. Cancer Res 62: 3663–3666.
  40. 40. Römermann D, Hasemeier B, Metzig K, Göhring G, Schlegelberger B, et al. (2008) Global increase in DNA methylation in patients with myelodysplastic syndrome. Leukemia 22: 1954–1956.
  41. 41. DeLellis RA, Lloyd RV, Heitz PU, Eng C (2004) Tumours of endocrine organs. In: DeLellis RA, Lloyd RV, Heitz PU, Eng C, editors. World Health Organization classification of tumour pathology and genetics, Eds. Lyon: IARC Press. pp. 124–132.
  42. 42. Juhlin CC, Villablanca A, Sandelin K, Haglund F, Nordenström J, et al. (2007) Parafibromin immunoreactivity – its use as an additional diagnostic marker for parathyroid tumor classification. Endocr Relat Cancer 14: 501–512.
  43. 43. Farnebo F, Teh BT, Kytölä S, Svensson A, Phelan C, et al. (1998) Alterations of the MEN1 gene in sporadic parathyroid tumors. J Clin Endocrinol Metab 83: 2627–2630.
  44. 44. Farnebo F, Kytölä S, Teh BT, Dwight T, Wong FK, et al. (1999) Alternative genetic pathways in parathyroid tumorigenesis. J Clin Endocrinol Metab 84: 3775–3780.
  45. 45. Kiss NB, Geli J, Lundberg F, Avci C, Velazquez-Fernandez D, et al. (2008) Methylation of the p16INK4A promoter is associated with malignant behavior in abdominal extra-adrenal paragangliomas but not pheochromocytomas. Endocr Relat Cancer 15: 609–621.
  46. 46. Geli J, Kiss N, Lanner F, Foukakis T, Natalishvili N, et al. (2007) The Ras effectors NORE1A and RASSF1A are frequently inactivated in pheochromocytoma and abdominal paraganglioma. Endocr Relat Cancer 14: 125–134.