| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Neuroendocrine Unit, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Dr. Anne Klibanski, Neuroendocrine Unit, Bulfinch 457B, Massachusetts General Hospital, 55 Fruit Street, Boston, Massachusetts 02114. E-mail: aklibanski{at}partners.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MEG3 was first identified as a human homolog to the mouse maternal imprinted gene Gtl2 (4, 5). The expression of Gtl2 was detected in the mouse embryo in a temporal and spatial manner from the one-cell stage onward, suggesting that Gtl2 is necessary for fetal and postnatal development (6). The Gtl2 gene gave rise to multiple cDNA isoforms, but no strong Kozak consensus sequence for translation initiation was detected in any of the ATG codons at the open reading frames; therefore, it has been suggested that this gene might function as an RNA (6). Recently, our laboratory found that MEG3 is strongly expressed in normal human tissues, particularly pituitary and brain; however, its mRNA is undetectable in pituitary tumors and human cancer cell lines (3). Importantly, MEG3a, one of the isoforms transcribed from the MEG3 gene, functions as a powerful cell growth suppressor. In colony formation assays, MEG3a caused a 70% decrease in cell colony numbers when transfected into human cancer cell lines, such as neuroglioma H4, breast adenocarcinoma MCF7, and cervical carcinoma HeLa. The MEG3 gene is located at chromosome 14q32.3, a locus that has been predicted to contain a tumor suppressor gene whose gain and loss have been associated with the pathogenesis and progression of several tumor types, such as meningiomas, nasopharyngeal carcinoma, colorectal carcinoma, and leukemia (7, 8, 9, 10). Together, these data are consistent with the hypothesis that MEG3 may play a role in pituitary tumorigenesis. In the present study, we investigated the mechanism associated with the loss of MEG3 expression in human pituitary tumors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Postmortem human normal pituitaries were obtained from the Harvard Brain Tissue Resource Center (Belmont, MA). Fourteen human clinically nonfunctioning tumors were obtained after transsphenoidal surgery with standard histological assessment demonstrating positive immunohistochemical staining for glycoprotein ß- and/or 
-subunits. This study was approved by the Massachusetts General Hospital subcommittee for human studies, and written consent was obtained from each patient who provided samples.
RNA extraction and RT-PCR
Total RNA from the tissues and cell lines was extracted using TRIzol reagent (Invitrogen Life Technologies, Inc., Carlsbad, CA) and was reverse transcribed using the RT System from Promega Corp. (Madison, WI) according to the manufacturer’s protocol. PCRs were performed under the following conditions: 94 C for 2 min, 94 C for 30 sec, 60 C for 30 sec, and 72 C for 1 min for 30 cycles, and 72 C for 10 min. All PCR primer sequences used in this study are shown in Table 1
. Cyclophilin A was used as an internal control (11).
|
Genomic DNA from normal human pituitaries and pituitary tumors without MEG3 expression was extracted using DNeasy Tissue Kit (QIAGEN, Valencia, CA). Genomic DNA from leukocytes was extracted using a Puregene DNA extraction kit (Gentra Systems, Minneapolis, MN). Genomic PCR was performed on all 10 exons of the MEG3 gene. The reactions were carried out using the following conditions: 94 C for 2 min, 94 C for 30 sec, 65 C [see Table 1
for annealing temperatures for PCR primers (Tms)] for 30 sec, and 72 C for 2 min for 30 cycles, and 72 C for 10 min.
Microsatellite analysis of loss of heterozygosity (LOH)
Two microsatellite markers, D14S985 and WI-16835, located on 14q32, were used to analyze the potential allelic losses in pituitary tumors by real-time quantitative PCR. The sequence information for the two markers was obtained from the National Center for Biotechnology Information (GenBank accession no. Z52362 and G21236, respectively). D14S985 is mapped to the third intron of the MEG3 gene, and WI-16835 is downstream of MEG3. To optimize the real-time PCR conditions, serial dilutions were established for the DNA template from the normal pituitary. The DNA concentration was first determined by spectrophotometry, then checked by gel electrophoresis. Validation experiments were carried out for both D14S985 and WI-16835 primers with reference to the endogenous gene, somatostatin receptor type 2 (SSTR2). The optimal DNA concentration was determined when the efficiencies of the PCR for the microsatellite markers and SSTR2 were approximately equal, and PCRs gave the highest specificity and intensity. The real-time PCR was carried out in a total volume of 25 µl containing 25 or 12.5 ng genomic DNA, 12.5 µl SYBR Green Master Mix (Sigma-Aldrich Corp., St. Louis, MO), and 0.4 pmol forward and reverse primers. PCR was performed on the Smart Cycler II system (Cepheid, Sunnyvale, CA) by a hot start at 95 C for 2 min, followed by 40 cycles of a three-step PCR: 95 C for 5 sec, 66 C (see Table 1 for Tms) for 10 sec, and 72 C for 20 sec. The PCR products were quantified by measuring the fluorescence from the SYBR Green that was incorporated into the double-stranded DNA products at the end of each cycle. The threshold cycle number (Ct) at which fluorescence passed the threshold level was recorded. At the end of the reaction, the PCR products were analyzed on a 2% agarose gel to confirm the identities. The amounts of D14S985 and WI-16835 in each sample were normalized to SSTR2, and their relative amounts in tumors, compared with those in normal pituitary, were calculated by 2–
Ct, in which Ct = Ct (D14S985 or WI-16835) – Ct (SSTR2) and Ct = Ct (pituitary tumors) – Ct (normal pituitaries). The relative efficiency of microsatellite markers and SSTR2 was determined by evaluating Ct against serial dilutions of template. At least three repeats were performed for each real-time PCR.
Sodium bisulfite treatment and sequencing
One microgram of genomic DNA was treated with sodium bisulfite using the CpGenome DNA Modification Kit (Serologicals Corp., Norcross, GA) according to the manufacturer’s protocol. The 2.1-kb MEG3 promoter sequence upstream to the first exon was arbitrarily divided into four regions (see Fig. 4A). Hotstart PCRs were used under the following conditions: 95 C for 15 min, 94 C for 30 sec, 64.5 C (see Table 1 for Tms) for 30 sec, and 72 C for 2 min for 40 cycles, and 72 C for 10 min. PCR products were subcloned into a TOPO TA cloning vector (Invitrogen Life Technologies, Inc.), and five constructs representing each region from each sample were randomly selected for sequence analysis.
|
A synthetic 20-bp fragment containing a TATA box as a minimal promoter and a PmlI restriction site were cloned into a promoterless luciferase reporter vector, pGL3-Basic (Promega Corp.), to generate pGL3-TATA. Region 4 was cloned into pGL3-TATA to generate pGL3-TATA-R4 through the SacI and PmlI sites. Region 1, containing the native TATA and CCAAT boxes of the human MEG3 gene, was cloned into pGL3-Basic through the NheI and HindIII sites to generate pGL3-R1.
MCF7 and GH4 cells were seeded into 12-well tissue culture plates for transfection. For each well, 3 µl TransIT-LT1 transfection reagent (Mirus Corp., Madison, WI) was used together with 0.4 µg of reporter construct and 0.4 µg pCMV-ß-gal. After 16 h of transfection, cells were cultured in normal medium for another 24 h, and luciferase and ß-galactosidase assays were performed as previously described (12). The luciferase activity was normalized to the ß-galactosidase activity from the same transfection.
Treatment with 5-aza-2'-deoxycytidine
MCF7 and HeLa cells were seeded into 60 x 15-mm tissue culture dishes and cultured in DMEM containing 2 and 5 µM 5-aza-2'-deoxycytidine (Sigma-Aldrich Corp.) for 6 d. Cells cultured in the absence of 5-aza-2'-deoxycytidine were used as a control. The medium with different concentrations of 5-aza-2'-deoxycytidine was changed every day. RT-PCR was performed to detect MEG3 mRNA expression as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RT-PCR analysis was used to examine MEG3 expression in 14 clinically nonfunctioning tumors and three normal pituitary tissues. MEG3 was strongly expressed in all normal pituitaries examined (Fig. 1). In contrast, 13 of 14 tumors showed no MEG3 expression (Fig. 1). Cyclophilin A was used to verify the cDNA integrity in all samples examined (Fig. 1).
|
To investigate whether the absence of MEG3 mRNA expression in pituitary tumors is due to allelic deletion of the MEG3 gene, microsatellite analyses were performed on the normal pituitaries and 13 pituitary tumors without MEG3 expression, using two markers flanking a 16-kb region on chromosome 14q32.3 containing the MEG3 locus. SSTR2, a subtype of somatostatin receptor that was identified in all pituitary adenomas and normal pituitary tissue (13), was used as an endogenous reference. Validating experiments were performed to verify that the microsatellite markers and the SSTR2 gene have approximately the same efficiency of PCR amplification, i.e. similar change in Ct against serial dilutions of the template (data not shown). Real-time PCR results are given as the numbers of PCR cycles that passed a certain level of fluorescence (Ct). The normalized concentration of the target with respect to the endogenous reference is calculated by Ct = Ct (target) – Ct (reference), and the relative concentration for the microsatellite marker DNA in a sample compared with that in the normal pituitaries was calculated by 2–Ct, in which Ct = Ct (normal pituitary or tumor) – Ct (average normal pituitary). If both the normal pituitaries and the tumors have the same number of alleles, the relative concentration of the microsatellite markers by this method should be close to 1. If LOH exists in the tumors, there should be at least a 50% reduction in the relative concentration of the markers in the tumors. To mimic the loss of one allele, control real-time PCR experiments were performed using 25 ng and half (12.5 ng) the amount of the template from the normal pituitaries or tumors. A representative result is shown in Table 2. Although the relative concentrations obtained using 25 ng templates were close to 1, those obtained using 12.5 ng templates were well below 0.5. Figure 2A shows a representative real-time PCR amplification plot. When 25 ng genomic DNA from different samples were used as the template in real-time PCR, the differences among the respective Ct were close to 0.5 cycles (Fig. 2A). When comparing 12.5 and 25 ng template, the differences in Ct were close to three cycles (Fig. 2A), indicating a clear difference in genomic DNA presented in real-time PCR. These results indicate that this assay is sensitive enough to detect LOH. As shown in Table 3, when 25 ng genomic DNA was used as a template, the relative concentrations of both microsatellite markers in all tumors were close to 1. Figure 2B showed a representative agarose gel for marker WI-16835. Although there is a clear difference in the band intensity of the PCR products from those with different amounts of template (25 and 12.5 ng, respectively), no obvious band intensity changes were observed in the PCR products from the normal pituitary and pituitary tumors using 25 ng genomic DNA as template. These data clearly demonstrate that there is no allele loss at the MEG3 gene locus in human nonfunctioning pituitary tumors. In addition, we obtained matched blood samples and extracted DNA from the leukocytes of three patients. We performed similar analyses comparing the genomic DNA from tumor and leukocyte samples. As shown in Table 4, the relative concentrations of two microsatellite markers in tumors are comparable to those in the matched leukocytes from the same patient, confirming no LOH in the tumors. The PCR products were run on an agarose gel, and no changes in relative band intensity (microsatellite markers to SSTR2) were detected (Fig. 3).
|
|
|
|
|
Hypermethylation of cytosine-guanine dinucleotides (CpGs) in the MEG3 promoter region in pituitary tumors
The promoter region of MEG3 is rich in CpG dinucleotides. To examine the methylation pattern, the 2.1-kb 5'-flanking region of the MEG3 gene containing a total of 112 CpG dinucleotides, was arbitrarily divided into four regions (Fig. 4A). We treated genomic DNA from normal pituitary and pituitary tumors with sodium bisulfite and examined the methylation pattern in this 2.1-kb promoter region. In our study, all cytosines other than those in CpGs were converted to thymine (data not shown), excluding the possibility of the presence of cytosines in CpGs due to incomplete treatment. Figure 4B showed a representative methylation pattern of the first 16 CpGs in region 4 of the genomic DNA from a normal pituitary and a tumor. In normal pituitary, the majority of CpGs are unmethylated, whereas in tumors, most of the CpGs are methylated. We examined the methylation status of each CpG in the 2.1-kb promoter region of MEG3 gene by this method, using genomic DNA from normal pituitaries and clinically nonfunctioning tumors lacking MEG3 expression. The methylation status of each CpG was quantified by the percentage of methylated CpGs among all PCR products analyzed. In region 1, 10 of 30 (33.3%) CpGs were hyper-methylated in tumors (n = 11) compared with normal pituitaries (n = 2; Fig. 4C). In region 4, nine of 33 CpGs (27.3%), which were clustered within the first 16 CpGs (56.3%), were hypermethylated in tumors (n = 13) compared with normal pituitaries (n = 3; Fig. 4C). In regions 2 and 3, only one of 49 CpGs were hypermethylated in tumors (data not shown). In conclusion, we observed more methylated CpGs in the genomic DNA from tumors than in those from normal pituitary, especially in regions 1 and 4 in the 5'-flanking region of MEG3 gene.
MEG3 expression in cancer cells treated with demethylation reagent
To explore the relationship between promoter methylation and MEG3 gene silencing, the methylation inhibitor, 5-aza-2'-deoxycytidine, was used to demethylate genomic DNA. MCF7 cells were used, because we have previously demonstrated that MCF7 cells do not express MEG3, and MEG3a inhibits cell growth in this well-validated cell line (3). As shown in Fig. 5, MEG3 mRNA was detected after treatment of MCF7 cells with two concentrations of 5-aza-2'-deoxycytidine for 6 d. As a control, no MEG3 expression was detected in the absence of 5-aza-2'-deoxycytidine treatment. Similar results were observed in HeLa cells (data not shown).
|
To test whether regions 1 and 4, identified by hypermethylation in tumors, are functionally important, promoter assays were performed on reporter plasmids containing these two regions in a rat pituitary-derived cell line, GH4, in addition to MCF7 cells. In transient transfection experiments, region 1 displayed promoter activities 111 and 347 times higher than those of the control plasmid pGL3-Basic lacking eukaryotic promoter and enhancer (Fig. 6A, n = 4; Fig. 6B, n = 5) in GH4 and MCF7 cells, respectively. Region 4 enhanced reporter expression from a minimal TATA promoter by 104- and 229-fold (Fig. 6A, n = 4; Fig. 6B, n = 5), respectively. These data clearly demonstrate that regions 1 and 4 contain functionally important sequences for gene expression.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Real-time PCR methods are increasingly widely used for identifying genomic copy number changes (18, 19, 20, 21). Using this method, we detected at least a 50% decrease in the relative concentration when half the amount of the template was used. It is not unusual to detect more than a 50% decrease using real-time PCR for detecting genomic microdeletions. Ponchel et al. (18) showed a ratio of 0.25 for detecting a known genomic deletion using SYBR-based real-time PCR. The ratio is likely to be associated with the kinetics of PCR with different primers and methods. Using this method, we did not find any LOH in the nonfunctioning tumors. Furthermore, no difference was observed in the relative concentrations of two microsatellite markers between genomic DNA from the tumors and matched leukocytes in three patients. These results are consistent with previous studies showing that genomic alterations occur much less often in nonfunctioning tumors than in hormone-secreting tumors (22) and that no LOH was detected in chromosome 14 in nonfunctioning tumors (23). We also examined the potential homozygous exon deletion and mutations of the MEG3 gene that may lead to silencing of MEG3 expression. No deletions or mutations were detected. Therefore, a genomic abnormality is unlikely to be responsible for the loss of MEG3 expression in clinically nonfunctioning tumors.
Sequence information showed that the MEG3 promoter region is rich in CpG dinucleotides that are generally underrepresented in the human genome (24). Our data demonstrate that the MEG3 promoter region in tumor cells contains significantly higher numbers of methylated CpGs than those in the normal pituitaries, especially in regions 1 and 4, located immediately in front of and approximately 1.6–2.1 kb upstream of the first exon, respectively. These two regions clearly demonstrated strong promoter activities in activation of reporter gene expression. The expression of MEG3 in human cancer cell lines induced by a demethylation reagent indicates that methylation is related to MEG3 gene transcription. Taken together, these data suggest that hypermethylation of regions 1 and 4 is an important mechanism associated with the loss of MEG3 expression in human pituitary tumors.
To study the imprinting controls, Wylie et al. (25) reported a hemimethylation pattern of the putative promoter region of the MEG3 gene in fetal tissues. However, according to our cDNA sequence (3), two of three regions in their study are located within the first intron of the MEG3 gene and are approximately 0.4 and 1.5 kb, respectively, downstream from the first exon. Murphy et al. (26) showed unmethylation of two regions of the MEG3 gene, about 150 and 200 bp in size and located within regions 1 and 4 described here, in patients with maternal uniparental disomy of chromosome 14 and hypermethylation in patients with paternal uniparental disomy. These data are consistent with the concept that MEG3 is a maternally imprinted gene. This allelic methylation difference may explain the hemimethylation pattern observed in region 4 of the genomic DNA from the normal pituitaries in our study. For the first time, we have thoroughly examined the methylation pattern of up to 2.1 kb in the MEG3 promoter region in human tumor cells. In this study we did not distinguish the maternal and paternal alleles of the MEG3 gene. Instead, we focused on investigating the difference in methylation patterns between normal pituitary and tumors, and we observed an increase in DNA methylation of the MEG3 regulatory region in tumors, such as regions 1 and 4.
Several putative transcription factor-binding sites were found in both region 1 and region 4, including those for cAMP response element-binding protein, activating protein-2, c-Ets-1, and Rel/nuclear factor-
B, which are known to be involved in cell cycle regulation, cell proliferation, and tumor invasion (27, 28, 29, 30). It will be important to investigate whether these transcription factors will bind to the corresponding cis-elements and whether DNA methylation blocks such binding, resulting in the loss of MEG3 expression. Also, a typical TATA and CCAAT box was found in region 1, suggesting that this is an authentic eukaryotic promoter.
Recently, methylation-associated silencing was found in several genes involved in pituitary tumorigenesis. For example, Bahar et al. (31) found that methylation of the promoter region is associated with loss of expression of the p53-induced growth inhibitor, GADD45
, in pituitary tumors and the mouse corticotrope cell line AtT20. Other genes include p16, Rb, and death-associated protein kinase genes (32). As mentioned by Bahar et al. (31), it has not been established whether hypermethylation of genes involved in pituitary tumorigenesis can cause pituitary tumor formation. However, the tumor type- and gene-dependent methylation patterns of these pituitary genes indicate that methylation may be a potentially important mechanism for pituitary tumor pathogenesis (31).
Hypermethylation of CpGs has been detected as an early event in cancer patients using DNA released from apoptotic cancer cells to sputum, blood, and urine samples with high sensitivity (24). Potentially, it will be important to study the methylation pattern of the genes involved in pituitary tumorigenesis to generate an accurate methylation profile that may allow early detection of pituitary tumors. These studies may also facilitate the development of therapeutic strategies targeting DNA methylation for pituitary tumors, especially for clinically nonfunctioning tumors that respond poorly to medical therapy and typically require neurosurgical intervention (33).
In summary, we have for the first time comprehensively examined and compared the methylation pattern of the MEG3 promoter region in normal human pituitary tissues and clinically nonfunctioning pituitary tumors. We found hypermethylation in two functionally important regions of the MEG3 promoter in tumor cells. Furthermore, treatment with a methylation inhibitor restores MEG3 expression in human cancer cell lines. We conclude that epigenetic change is an important mechanism associated with MEG3 silencing in human pituitary tumors.
|
Footnotes |
|---|
First Published Online January 11, 2005
Abbreviations: CpG, Cytosine guanine dinucleotide; Ct, threshold cycle number; LOH, loss of heterozygosity; MEG3, maternal expressed gene 3; SSTR2, somatostatin receptor type 2; Tm, annealing temperature for PCR primers.
Received September 20, 2004.
Accepted January 4, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B is the answer: role of Rel/NF-B in the regulation of apoptosis. Oncogene 22:8961–8982[CrossRef][Medline]
, in human pituitary adenomas, is associated with CpG island methylation. Oncogene 23:936–944[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  X. Zhu, S. L. Asa, and S. Ezzat Fibroblast Growth Factor 2 and Estrogen Control the Balance of Histone 3 Modifications Targeting MAGE-A3 in Pituitary Neoplasia Clin. Cancer Res., April 1, 2008; 14(7): 1984 - 1996. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Zhou, Y. Zhong, Y. Wang, X. Zhang, D. L. Batista, R. Gejman, P. J. Ansell, J. Zhao, C. Weng, and A. Klibanski Activation of p53 by MEG3 Non-coding RNA J. Biol. Chem., August 24, 2007; 282(34): 24731 - 24742. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Barlier, J.-F. Vanbellinghen, A. F. Daly, M. Silvy, M.-L. Jaffrain-Rea, J. Trouillas, G. Tamagno, L. Cazabat, V. Bours, T. Brue, et al. Mutations in the Aryl Hydrocarbon Receptor Interacting Protein Gene Are Not Highly Prevalent among Subjects with Sporadic Pituitary Adenomas J. Clin. Endocrinol. Metab., May 1, 2007; 92(5): 1952 - 1955. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Zhu, K. Lee, S. L. Asa, and S. Ezzat Epigenetic Silencing through DNA and Histone Methylation of Fibroblast Growth Factor Receptor 2 in Neoplastic Pituitary Cells Am. J. Pathol., May 1, 2007; 170(5): 1618 - 1628. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Boikos and C. A. Stratakis Molecular genetics of the cAMP-dependent protein kinase pathway and of sporadic pituitary tumorigenesis Hum. Mol. Genet., April 15, 2007; 16(R1): R80 - R87. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Shorts-Cary, M. Xu, J. Ertel, B. K. Kleinschmidt-Demasters, K. Lillehei, I. Matsuoka, S. Nielsen-Preiss, and M. E. Wierman Bone Morphogenetic Protein and Retinoic Acid-Inducible Neural Specific Protein-3 Is Expressed in Gonadotrope Cell Pituitary Adenomas and Induces Proliferation, Migration, and Invasion Endocrinology, March 1, 2007; 148(3): 967 - 975. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W E Farrell Pituitary tumours: findings from whole genome analyses. Endocr. Relat. Cancer, September 1, 2006; 13(3): 707 - 716. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, unmethylated CpGs. C, The methylation status of each CpG in regions 1 and 4 is quantified by the percentage of methylated CpGs among all PCR products analyzed. For each CpG, a t test is used to analyze the statistical significance of the difference in the percentage of methylated CpGs between normal pituitaries and tumors. The positions of the first and last CpGs are indicated. Values are the mean ± SE. *, P < 0.05 between normal pituitaries and tumors. 
, Percentage of methylated CpGs from tumor samples; 
, percentage of methylated CpGs from normal pituitaries.
