The imprinted gene, neuronatin (NNAT), is one of the most abundant transcripts in the pituitary and is thought to be involved in the development and maturation of this gland. In a recent whole-genome approach, exploiting a pituitary tumour cell line, we identified hypermethylation associated loss of NNAT. In this report, we determined the expression pattern of NNAT in individual cell types of the normal gland and within each of the different pituitary adenoma subtypes. In addition, we determined associations between expression and CpG island methylation and used colony forming efficiency assays (CFE) to gain further insight into the tumour-suppressor function of this gene. Immunohistochemical (IHC) co-localization studies of normal pituitaries showed that each of the hormone secreting cells (GH, PRL, ACTH, FSH and TSH) expressed NNAT. However, 33 out of 47 adenomas comprising, 11 somatotrophinomas, 10 prolactinomas, 12 corticotrophinomas and 14 non-functioning tumours, irrespective of subtype failed to express either NNAT transcript or protein as determined by quantitative real-time RT-PCR and IHC respectively. In normal pituitaries and adenomas that expressed NNAT the promoter-associated CpG island showed characteristics of an imprinted gene where ∼50% of molecules were densely methylated. However, in the majority of adenomas that showed loss or significantly reduced expression of NNAT, relative to normal pituitaries, the gene-associated CpG island showed significantly increased methylation. Induced expression of NNAT in transfected AtT-20 cells significantly reduced CFE. Collectively, these findings point to an important role for NNAT in the pituitary and perhaps tumour development in this gland.
Pituitary tumours are classified as either functional (hormone secreting) or non-functional and comprise ∼10% of diagnosed intracranial neoplasms (Asa & Ezzat 2002, 2005). Each of the differentiated cell types within the gland has the potential to develop a tumour and although the majority will remain benign, approximately one-third will show invasive and/or recurrent growth and a smaller proportion will become metastatic. Adenomas are frequently associated with significantly reduced life quality and irrespective of their hormone status morbidity can ensue through either compressive effects on brain structures and/or excessive hormone secretion (Ezzat et al. 2004).
Methylation-mediated or -associated gene silencing is a frequent finding in pituitary adenomas where most studies, with limited exceptions, have adopted candidate gene approaches of primary human adenomas (Zhang et al. 2003, Bahar et al. 2004, Zhao et al. 2005). Recently, our laboratory using a whole-genome approach found that the imprinted gene, neuronatin (Nnat), is expressed in normal mouse pituitary, but is not expressed in the mouse pituitary adenoma cell line At-T20. However, targeted, siRNA -mediated knockdown of DNA methyltransferase 1 (Dnmt1) in At-T20 cells induced re-expression in a time-dependent manner (Dudley et al. 2008). Re-expression was associated with partial demethylation of the Nnat genes CpG island, implying that silencing is mediated through DNA methylation (Dudley et al. 2008). Among several other genes identified in this murine model system, we also showed that NNAT was inappropriately silenced in association with CpG island methylation in a proportion of primary human pituitary tumours relative to normal human pituitary (Dudley et al. 2008).
In the pituitary, serial analysis of gene expression has identified NNAT as the fifth most abundant transcript after GH, pro-opiomelanocortin, PRL and an unknown transcript, and this and other studies also show that this transcript is not expressed at significant levels in most other adult tissues (Usui et al. 1997, Nishida et al. 2005, Suh et al. 2005). In the brain, NNAT is thought to participate in maintaining segment identity and the overall structure of the nervous system (Wijnholds et al. 1995). In addition, NNAT is also thought to be involved in both pituitary development and maturation (Aikawa et al. 2003). Taken together, these findings suggest an important role for NNAT in the adult and developing pituitary gland and that loss may have implications for pituitary tumorigenesis and perhaps that of other tumour types. Indeed, and in this context, decreased Nnat expression was identified in a screening study of rat cells in the progression from immortalization to transformation (Yoshioka et al. 2000). Although this study did not report the functional significance of Nnat loss in this transition, other studies of different tumour types have shown that NNAT may exhibit either a tumour suppressing or in some cases a tumour promoting role (Takeuchi et al. 1995, Asimakopoulos et al. 1996, Usui et al. 1997, Hu et al. 2002, Kuerbitz et al. 2002, Wood et al. 2002, Becker et al. 2005, Zhong et al. 2006).
Our initial findings of methylation associated silencing of NNAT in a small number of pituitary adenomas (Dudley et al. 2008) prompted us to address the unanswered questions of frequency of loss across different pituitary adenoma subtypes and expression patterns of NNAT, at the protein and transcript level, in individual cell types that comprise the normal human pituitary. These studies, in addition to allow us to determine the cell-type specificity of NNAT expression patterns in normal and tumour cells, also allowed us to determine associations between expression and CpG island methylation and finally, to gain further insight into the tumour suppressor function of this gene in a pituitary context.
Materials and methods
Cell line and tissue samples
The mouse pituitary adenoma cell line AtT-20/D16v-F2 (AtT-20) was purchased from the European collection of cell cultures (Salisbury, UK). Semi-confluent (70–80%) cells were split 1:5 using 0.25% trypsin/0.53 mM EDTA (Sigma-Aldrich), into high glucose DMEM with glutamine and pyruvate (Biosera, Ringmer, UK), 10% foetal bovine serum (Biosera), 4 μg/ml gentamycin (Sigma-Aldrich) and 2 μg/ml ampicillin (Sigma-Aldrich). Monolayer cultures were incubated at 37 °C in a humidified atmosphere containing 5% CO2.
Human normal pituitaries and adenomas were obtained with informed consent and institutional ethical approval. Five normal pituitaries were obtained within 12 h of death, freeze fractured using the BioPulveriser 59013N (Biospec, Bartesville, OK, USA) to achieve a homogenous mixture of cells. The pulverized homogenate was stored at −80 °C.
The tumour cohort comprised a total of 47 adenomas. Subtypes were determined on the basis of clinical and biochemical findings before surgery, and on morphological and immunohistochemical (IHC) characterization of the tumour specimen. Out of the 47 adenomas included in the study, there were 11 somatotrophinomas, 10 prolactinomas, 12 corticotrophinomas and 14 non-functioning pituitary adenomas (NFPAs).
Nucleic acid extraction
Total RNA was extracted using the guanidinium thiocyanate–phenol–chloroform method as previously described (Chomczynski & Sacchi 2006). DNA was extracted from tissue using the Nucleon DNA extraction kit (Anachem, Bedfordshire, UK) according to the manufacturer's instructions. DNA extraction from whole blood was achieved using the Wizard Genomic DNA Purification Kit as recommended by the manufacturer (Promega). Nucleic acid purity and concentrations were assessed using a NanoDrop ND-1000 spectrophotometer (Labtech, Ringmer, East Sussex, UK).
cDNA synthesis and RT-PCR
M-MLV Reverse Transcriptase (Promega) was used to prepare cDNA template as previously described (Dudley et al. 2008). Quantitative real-time RT-PCR (RT-qPCR) using the relative standard curve method was performed on the ABI PRISM 7300 HT Sequence Detection System (Applied Biosystems, Cheshire, UK) using brilliant SYBR Green QPCR Master Mix solution (Stratagene, Amsterdam, The Netherlands). One microlitre cDNA was amplified under default conditions: 96 °C for 10 min followed by 40 amplification cycles of 95 °C for 30 s, 60 °C for 1 min and 72 °C for 1 min. Hypoxanthine-guanine phosphoribosyl transferase (HPRT) was amplified in parallel and used as an internal control to normalize the results and to allow for quantitative analysis of NNAT transcript levels. The primers used to amplify NNAT were 5′-CAACAGCGGACTCCGAGA-3′ and 5′-CATTCCAGGAACACCTGCA-3′. The primers used to amplify HPRT were 5′-CGGCTCCGTTATGGCG-3′ and 5′-GGTCATAACCTGGTCATCATCAC-3′. The calibrated RT-qPCR values for samples relative to the mean of five normal pituitaries were plotted as log to the base 2. An adenoma was deemed negative for NNAT expression in those cases where a threefold reduction in expression was observed.
To identify specific cell types within normal pituitary, sections were stained with antibodies directed against each of the major hormones secreted by this gland. Thus, as previously described (Simpson et al. 1999) with minor modifications, slides of formalin fixed, wax embedded normal human pituitary tissue were first re-hydrated and washed, serum blocked for 15 min at room temperature then incubated at room temperature for 30 min with antibodies directed against ACTH (mouse monoclonal, 1:2000, DAKO Cytomation, Ely, UK), GH (rabbit polyclonal, 1:6000, DAKO), FSH (mouse monoclonal, 1:5000, Serotec, Oxford, UK), PRL (mouse monoclonal, 1:2000, Serotec) or TSH (mouse monoclonal, 1:20 000, Serotec). Sections were then washed and incubated with biotinylated secondary antibody (Vector, Southgate, UK) for 30 min at room temperature. Antibodies were detected using the avidin–biotin complex (Vector) method using vector blue-stain as the chromogen.
To observe co-localization of NNAT with each pituitary hormone, the slides were then taken through an antigen retrieval step, serum blocked for 15 min at room temperature and then incubated with an antibody directed towards NNAT (rabbit polyclonal, 1:1000, Abcam, Cambridge, UK). NNAT antibody was detected using the EnVision DAB system (DAKO Cytomation) with Dako REAL 3,3-diaminobenzidine tetrahydrochloride (DAB; DAKO Cytomation) as the chromogen. Finally, the slides were haematoxylin counterstained with a 5 min immersion in 5% copper sulphate to intensify the brown DAB stain, then de-hydrated and mounted with Pertex substitute (BDH, Lutterworth, UK) and visualized using the Nikon Eclipse e600 camera. Specificity of the antibody to NNAT was assessed by incubation with 25 μg NNAT immunizing peptide (Abcam) prior to the staining procedure described above.
Co-localization of NNAT with each of the individual pituitary hormone antibodies used in this study was observed as a merging of the blue-(hormone) and the NNAT-(brown) specific stains to produce an intensity of colour that varied between dark blue and dark brown.
A total of 21 adenomas, that included specimens of each of the pituitary adenoma subtypes, were stained for NNAT immunoreactivity using the EnVision DAB system as described above.
DNA methylation analysis
Extracted DNA samples (3 μg) were mutagenized by treatment with sodium bisulphite solution, as previously described (Grunau et al. 2001). Following PCR amplification using primers specific for converted DNA, the methylation status of the CpG island associated with exon 1 of NNAT was assessed, by several different methodological approaches, including combined bisulphite restriction analysis (COBRA), single molecule sodium bisulphite sequencing or methylation sensitive pyrosequencing. Primers for COBRA and sodium bisulphite sequencing analysis were 5′-TATGGAAAGAGTAGATGGATTAT-3′ and 5′-AATTAAAAACCCTCTCCCACC-3′. The primary amplification was followed with a nested PCR with the upstream primer described above and the downstream, nested primer 5′-TCAAACTTACCTACAACAACA-3′. In this way, the derived amplicons encompassed a total of 25 CpG sites.
For COBRA, equal amounts of sodium bisulphite converted PCR product were incubated in the presence and absence of the restriction enzyme, BSTUI (New England Biolabs, Hertfordshire, UK) at 60 °C for 2 h as previously described (Simpson et al. 2004). Digested fragments were visualized by ethidium bromide staining of DNA electrophoresed through a 1.5% agarose gel. Post-digestion of the PCR product, and in the absence of methylation, a 218 bp amplicon was apparent. Methylation, at one or more of the four BSTUI (CGCG) sites within the amplicon yielded a profile of fragments, 75, 56, 39, 27 and 21 bp of which the larger fragments, 75 and 56 bp can be visualized on the gel.
Single molecule sodium bisulphite sequencing
JM109 cells were transformed with gel purified sodium bisulphite PCR products using the T-Easy vector system (Promega), according to the manufacturer's instructions. Blue white selection of individual clones was used prior to sequence analysis on the ABI Prism 3100 Genetic Analyzer (Applied Biosystems, Warrington, UK). In each case (normal pituitaries and adenomas), at least five and typically ten discrete clones were subjected to sequence analysis.
In order to determine the sensitivity of methylation sensitive pyrosequencing, the known standards of methylated DNA were prepared and analysed using this technique. NNAT gene-specific DNA was amplified from blood-derived DNA using primers 5′-GATCTCGGCAAACCCTCTTT-3′ and 5′-TTTCAACACAGTAGTGCCCC-3′ at an annealing temperature of 55 °C, resolved on a 1% agarose gel, excised and recovered using the Gene Clean II kit (Anachem) according to the manufacturers instructions and divided into two fractions. One fraction, comprising 100 μg PCR amplified DNA in a final volume of 100 μl was in vitro methylated with the enzyme SssI, 3 μl SssI methylase (NEB, Luton, UK) and 10 μl S-adenosyl methionine (NEB). The other fraction was sham in vitro methylated by omitting the SssI enzyme. The in vitro methylated and sham in vitro methylated NNAT DNA were mixed in ratios representing 0, 25, 50, 75 and 100% methylation and subjected to methylation sensitive pyrosequencing.
Methylation sensitive pyrosequencing
Prior to pyrosequencing, sodium bisulphite-modified DNA was amplified with primers, 5′-GATCTCGGCAAACCCTCTTT-3′ and 5′-TTTCAACACAGTAGTGCCCC-3′. Amplified DNA was subjected to pyrosequencing with primer 1, 5′-TTATTTTTTTTTATAAATAT-3′ and primer 2, 5′-GGTGGGTGGGTATTTAAGG-3′ that encompassed 10 CpG dinucleotides. Pyrosequencing was performed as previously described (Dudley et al. 2008).
Loss of heterozygosity analysis
For loss of heterozygosity (LOH) studies, the microsatellite marker D20S478 was employed that maps proximal (1.08 Mb) to the NNAT gene. Matched blood and tumour DNA from eight samples was extracted as previously described and subjected to PCR. The PCR parameters followed the outline; 94 °C for 5 min, 40 cycles of 94 °C for 30 s, 55 °C for 1 min and 72 °C for 30 s and 72 °C for 5 min. PCRs were performed with 1 μl DNA and 20 pmol oligonucleotides for D20S478; 5′-CCAAAGACAAGTTTGTTTATCTG-3′ and 5′-TGTTATAGCAGCAATAGAAAACTAA-3′. Products were visualized by PAGE and silver staining as previously described (Simpson et al. 2000).
The complete coding sequence of NNAT (GeneBank ID NM181689) was obtained from the Mammalian Gene Collection (MGC: 3353944, Cambridge, UK) and cloned into a ‘TET on’ system using the tetracycline-inducible vector, pcDNA 4/TO (Invitrogen) to generate a NNAT expressing construct pcDNA 4/TO-NNAT (TO-NNAT). Empty vector controls (TO-MCS) did not contain the coding sequence for NNAT. Constructs were transfected into AtT-20 cells that had been established in our laboratory to express the Tet repressor construct, pcDNA 6/TR (Invitrogen) and designated, AtT-20TRC1 cells. For selection, the growth media of cells transfected with pcDNA 6/TR were supplemented with 6 μg/ml blasticidin, and for TO-NNAT and TO-MCS, 0.5 μg/ml bleocin. Transfectants were generated as previously described (Frost et al. 1999, Bahar et al. 2004). Stable-transfected AtT-20 cells expressing the Tet repressor (AtT-20-TRC1) and either the NNAT (TO-NNAT) construct or empty vector control (TO-MCS) was designated TR/TO-NNAT and TR/TO-MCS respectively.
Transfected cells were seeded at 2×104 cells/ml in DMEM supplemented with tetracycline reduced FCS in 60 mm Petri dishes with a gridded base (Corning Costar, Koolhovenlaan, The Netherlands) with or without the addition of 0.5 μg/ml tetracycline (Sigma) which was replaced at 2 day intervals. Cells were cultured at 37 °C until distinct colonies could be visualized following which the number of colonies were counted on a Nikon Microscope (×4 magnification) and photographed using a Leica confocal microscope (Milton Keynes, Buckinghamshire, UK). Colony forming efficiency (CFE) was calculated by counting the number of colonies over 100 squares on each plate.
The χ2 statistical test was used to determine the correlation between loss of NNAT expression as determined by RT-qPCR with IHC and also the association of loss of NNAT expression with promoter hypermethylation. Statistical significance was accepted at P<0.05.
IHC localization of NNAT in normal pituitary
Co-localization of NNAT, in post-mortem normal pituitaries, with individual pituitary hormones as determined by IHC was used to identify the different cell types that expressed NNAT. Within the anterior pituitary gland (AP), somatotrophs comprise ∼50% of the cells with the remaining cells, lactotrophs, gonadotrophs, corticotrophs and thyrotrophs comprising ∼10–20% each (Horvath & Kovacs 1988, Yeung et al. 2006). Irrespective of hormonal status of individual cells, ∼40% of cells within the AP-stained positive for NNAT protein (Fig. 1, panel A). The specificity of the NNAT antibody was confirmed in pre-adsorption studies outlined in Materials and Methods and is shown in Fig. 1, panel B.
Staining was apparent within clusters of cells and rarely scattered. Immunopositivity for NNAT was apparent within the cytoplasm and also on the outer cell membranes. Figure 1, panels C–L, shows normal pituitary stained for each of the pituitary hormones and the double staining of pituitary hormone and NNAT. These co-localization studies show that NNAT is expressed by each of the cell subtypes of the AP. NNAT was found to co-localize with ∼40% of somatotrophs, ∼30% of corticotrophs, ∼50% of thyrotrophs, ∼30% of lactotrophs and ∼20% of gonadotrophs.
NNAT expression status in pituitary adenomas
A total of 47 pituitary adenomas were analysed by real-time RT-qPCR for NNAT transcript expression. Overall, 33 out of 47 adenomas (70%) failed to express or showed barely detectable NNAT expression relative to the mean of expression in normal pituitaries. By subtype, 12 out of 14 non-functioning adenomas, 5 out of 11 somatotrophinomas, 6 out of 12 corticotrophinomas and 10 out of 10 prolactinomas showed loss or significantly reduced expression of NNAT (Fig. 2A).
In total, 21 adenomas comprising representative numbers of each pituitary adenoma subtype were assessed for the presence of NNAT by IHC (Fig. 2B). Of these, eight out of nine that expressed NNAT, as determined by RT-qPCR, stained positively for NNAT. Eleven out of twelve adenomas that showed loss or barely detectable expression of NNAT, as determined by RT-qPCR, did not express IHC detectable NNAT (Fig. 2B). There was a significant correlation between loss of NNAT expression, as determined by IHC and loss or significantly reduced expression as determined by RT-qPCR, P=0.002.
Methylation status of the NNAT genes CpG island in normal pituitaries and pituitary adenomas
The proximal promoter region of the NNAT gene was first identified using the CisRED database of highly conserved sequence elements in the proximal promoters of genes from multiple species (http://www.ensembl.org/Homo_sapiens/Gene/Regulation?g=ENSG00000053438) and we mapped a CpG island to this region with the CpG island Finder program (http://www.ebi.ac.uk). Sodium bisulphite sequencing of this CpG island showed, in normal pituitary, a methylation pattern where ∼50% of individual molecules were densely methylated and in the remaining molecules, this change was infrequent (Fig. 3A). Similar findings of ∼50% methylation were apparent using the related technique of COBRA analysis in normal pituitary (Fig. 3B).
In contrast to sodium bisulphite sequencing, COBRA analysis and pyrosequencing require less converted DNA. For adenomas where DNA quantity and quality were frequently limiting, these techniques allowed us to analyse a larger number for their methylation status. COBRA was performed on 14 adenomas, of which 11 did not express and 3 did express NNAT as determined by RT-qPCR. COBRA analysis showed that 10 out of 11 adenomas that did not express NNAT transcript were methylated and apparent as the disappearance of the unmethylated 218 bp amplicons (P=0.002). The three tumours that expressed NNAT transcript were not methylated as determined by this technique (Fig. 4A).
A limited number (four) of adenomas from the above COBRA analysis had sufficient DNA available to perform sodium bisulphite sequencing (Fig. 4B). In these cases, in four adenomas that did not express NNAT, a significant increase in methylation relative to that observed in normal pituitary was apparent and thus confirmed the COBRA data.
A total of 26 adenomas provided sufficient converted DNA for pyrosequencing. However, prior to this analysis, we designed experiments to confirm that this method was an accurate way of determining methylation profiles (Fig. 5A). Following this validation, 5 normal pituitaries, 4 adenomas that expressed NNAT and 22 adenomas that did not express NNAT as determined by RT-qPCR were analysed across 10 individual CpG dinucleotides. For seven out of nine specimens, that comprised normal pituitaries and NNAT expressing tumour, a methylation profile consistent with that of an imprinted gene was apparent (∼50% methylation; P=0.015 and Fig. 5B). However, 20 out of 22 tumours that fail to express NNAT showed a significant increase in methylation that showed some variation between 75 and 100% (P=0.007 and Fig. 5B).
Taken together, each of these techniques showed in a large number of adenomas that gene silencing was associated with CpG island methylation. In addition, where methylation was determined by more than one technique, the findings were concordant.
Loss of NNAT expression in pituitary tumours is not associated with LOH
The possibility that loss of NNAT expression was through the loss or deletion of the unmethylated allele and that the methylation patterns that we observed represented preferential retention of the imprinted, methylated, allele was investigated through LOH analysis in a limited number of specimens. Ten adenomas relative to matched blood DNA were screened using the polymorphic microsatellite marker, D20S478, closely linked to NNAT on chromosome 20q. Eight adenomas (three expressing and five non-expressing) were informative. Using previously published criteria (Miller & Yuan 1997), we found no evidence for loss (data not shown).
The robust expression of NNAT in normal human pituitary and the methylation associated silencing in the majority of human pituitary adenomas investigated, suggested an important function for the protein product of this gene in pituitary tumorigenesis. Therefore, we assessed the tumour suppressor activity of this gene/protein in At-T20 cells by CFE assays. These experiments were performed on TR/TO-NNAT cells relative to empty vector control TR/TO-MCS cells in the presence and absence of tetracycline. Tetracycline induced expression of NNAT in AtT-20 cells reduced CFE by ∼60% (Fig. 6). These findings reinforce those from our previous study, where IPTG (isopropyl β-d-1-thiogalactopyranoside) induced expression of NNAT reduced proliferation as determined by growth curve analysis (Dudley et al. 2008).
Despite data that show NNAT to be one of the most abundant transcripts present in the mature pituitary and its expression in the different cell types that characterize this gland had not thus far been described (Nishida et al. 2005). Employing IHC co-localization, we show that each of the discrete hormone secreting cell types within the anterior gland express NNAT. In contrast to these findings, expression of NNAT as determined by RT-qPCR showed that a significant proportion of primary tumours, irrespective of subtype, showed barely detectable expression of the NNAT transcript relative to normal pituitaries. In addition, IHC in situ detection of NNAT in representative examples of each adenoma subtype, confirmed these observations. In some cases, tumour and infiltrating or juxtaposed normal pituitary cells were apparent on single sections and in these, IHC frequently revealed the absence (tumour cells) and presence of NNAT protein (normal cells) respectively providing further confidence in our findings. The expression of NNAT in each of the differentiated cell types of the AP and loss in a substantial proportion of adenomas suggests an important role for this protein within this gland and most likely in tumour evolution.
Only a single previous study has reported expression of NNAT in pituitary adenomas and in this case, these investigators employed northern blotting (Usui et al. 1997). In contrast to our own findings, each of the major pituitary adenoma subtypes, with the exception of prolactinomas, but including, non-functional, ACTH and GH secreting adenomas expressed NNAT. The differences between these and our own observations are not entirely clear; however, this previous report only investigated two adenomas of each subtype, whereas our own study comprised a considerably larger number of each. In addition, a further explanation might be that northern blotting requires large amounts of RNA and the extraction of this nucleic acid may have included juxtaposed normal pituitary gland with tumour material.
Our initial identification and focus on NNAT was on the basis of its demethylation associated re-expression following siRNA mediated knockdown of the maintenance methylase enzyme DNMT1 in AtT-20 cells (Dudley et al. 2008). In addition, in the same study, we also showed in a small number of primary human pituitary adenomas, methylation-mediated silencing of NNAT as determined by COBRA analysis and pyrosequencing. However, insufficient adenomas were investigated to determine subtype-specific expression patterns. In this subsequent report, we have investigated the methylation status of this gene's CpG island in a larger cohort of adenomas that comprises each of the major subtypes. For many of the adenomas included in the investigation, the limited amount of tumour DNA available for study and or its integrity post-sodium bisulphite conversion required use of several different but related methodological approaches. Thus, for normal pituitaries, using either sodium bisulphite sequencing of individual molecules, COBRA analysis or pyrosequencing, we show a methylation pattern similar to that apparent for imprinted genes, which is differential methylation. In these cases, sodium bisulphite sequencing and pyrosequencing of NNAT showed that approximately half of the molecules examined were densely methylated and this change was infrequent in the remaining molecules. A similar pattern of methylation in normal pituitary and as determined by sodium bisulphite sequencing has recently described for the MEG3 gene which is also imprinted (Zhao et al. 2005). Using this and related techniques for tumours that failed to express NNAT as determined by RT-qPCR and IHC, a significant increase in methylation of this gene's CpG island was apparent, whereas tumours that expressed NNAT showed a methylation pattern similar to that observed in normal pituitaries.
The MEG3 gene represents the first imprinted gene identified as inappropriately methylated in pituitary tumours that showed loss or significantly reduced expression (Zhao et al. 2005). In addition, a previous study suggested through functional analysis, MEG3 to be a bona fide tumour suppressor (Zhang et al. 2003). Thus, NNAT represents a further example of an imprinted gene that is silenced in association with methylation of the normally expressed and unmethylated allele in pituitary tumours and which appears to act as a tumour suppressor gene. In contrast to MEG3, where methylation associated silencing appears to be confined, predominantly to NFPAs in the gonadotroph lineage (Zhao et al. 2005), methylation associated silencing of NNAT does not appear to show particular subtype specificity.
Previous reports, in this case of various primary leukaemias and their derived cell lines, show that they may harbour either, genetic or epigenetic aberrations that target the chromosome 20q12 region. This region, that also harbours the NNAT gene within a micro-imprinted domain, has been shown, in this tumour type, to incur either LOH or, more frequently, epigenetic (CpG island methylation) aberration (Kuerbitz et al. 2002 and references therein). Indeed, our own results of aberrant methylation, that presumably targets the expressed paternal allele of NNAT, could be explained by LOH in the 20q12 region that specifically targets this allele. Microsatellite analysis, employing a marker in close proximity to NNAT, albeit in a small number of informative tumours, relative to matched blood samples, did not show losses and thus suggest that genetic mechanisms are not responsible for loss of NNAT in these adenomas. Our search for informative single nucleotide polymorphisms in the NNAT gene was not successful (data not shown) and, therefore, precluded a more definitive exclusion of genetic aberrations.
Genome-wide identification of genes silenced through epigenetic mechanisms, and in particular through CpG island methylation, has been investigated in multiple different tumour types (Kuerbitz et al. 2002, Liang et al. 2002, Suzuki et al. 2002, Foltz et al. 2006, Ibanez de Caceres et al. 2006). These studies frequently employ pharmacological strategies to ‘unmask’ silent genes through inhibition of the DNA maintenance methylase enzyme, DNMT1. Our own genome-wide studies that led to preliminary identification of multiple genes including NNAT used a genetic (siRNA) ‘unmasking’ strategy that targeted the maintenance methylase, DNMT1. Interestingly, NNAT was also recently identified in a genome-wide pharmacological ‘unmasking’ approach in non-small cell lung carcinoma (NSCLC) cell lines following drug treatment with a DNA methylase inhibitor (Zhong et al. 2006). Similar to our own studies, these investigators also reported loss of NNAT in primary tumours although the methylation status of the NNAT genes CpG island in these tumours was not investigated. In addition, this study also showed that ectopic expression of NNAT in a NSCLC cell line inhibited growth in colony forming assays (Zhong et al. 2006). Our own studies, in this case of tetracycline induced expression of NNAT in AtT-20, also showed a significant reduction in CFE and reinforce our previous report that IPTG-induced NNAT expression inhibited cell proliferation as determined by growth curve analysis (Dudley et al. 2008).
The present study is the first report that describes the co-localization of NNAT to each of the major hormone secreting cell types of the anterior human pituitary gland. We also show that NNAT expression is decreased in a significant proportion of pituitary adenomas in association with promoter hypermethylation and that loss is irrespective of adenoma subtype. Although our functional analysis shows that re-expression in a pituitary cell line has a significant effect on the ability of these cells to form colonies, it is likely that this gene and its protein product will have multiple other functions within the pituitary.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
This study was supported by the Samantha Dickson Brain Tumour Trust and Brain Tumour UK.
The authors wish to thank Jim Richmond and Philip Whitby for their excellent technical assistance.
Aikawa S, Kato T, Elsaesser F & Kato Y 2003 Molecular cloning of porcine neuronatin and analysis of its expression during pituitary ontogeny. Experimental and Clinical Endocrinology and Diabetes 111 475–479.
Asimakopoulos FA, Gilbert JG, Aldred MA, Pearson TC & Green AR 1996 Interstitial deletion constitutes the major mechanism for loss of heterozygosity on chromosome 20q in polycythemia vera. Blood 88 2690–2698.
Bahar A, Simpson DJ, Cutty SJ, Bicknell JE, Hoban PR, Holley S, Mourtada-Maarabouni M, Williams GT, Clayton RN & Farrell WE 2004 Isolation and characterization of a novel pituitary tumor apoptosis gene. Molecular Endocrinology 18 1827–1839.
Becker M, Sommer A, Kratzschmar JR, Seidel H, Pohlenz HD & Fichtner I 2005 Distinct gene expression patterns in a tamoxifen-sensitive human mammary carcinoma xenograft and its tamoxifen-resistant subline MaCa 3366/TAM. Molecular Cancer Therapeutics 4 151–168.
Chomczynski P & Sacchi N 2006 The single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction: twenty-something years on. Nature Protocols 1 581–585.
Dudley KJ, Revill K, Whitby P, Clayton RN & Farrell WE 2008 Genome-wide analysis in a murine Dnmt1 knockdown model identifies epigenetically silenced genes in primary human pituitary tumors. Molecular Cancer Research 6 1567–1574.
Ezzat S, Asa SL, Couldwell WT, Barr CE, Dodge WE, Vance ML & McCutcheon IE 2004 The prevalence of pituitary adenomas: a systematic review. Cancer 101 613–619.
Foltz G, Ryu GY, Yoon JG, Nelson T, Fahey J, Frakes A, Lee H, Field L, Zander K & Sibenaller Z 2006 Genome-wide analysis of epigenetic silencing identifies BEX1 and BEX2 as candidate tumor suppressor genes in malignant glioma. Cancer Research 66 6665–6674.
Frost SJ, Simpson DJ, Clayton RN & Farrell WE 1999 Transfection of an inducible p16/CDKN2A construct mediates reversible growth inhibition and G1 arrest in the AtT20 pituitary tumor cell line. Molecular Endocrinology 13 1801–1810.
Grunau C, Clark SJ & Rosenthal A 2001 Bisulfite genomic sequencing: systematic investigation of critical experimental parameters. Nucleic Acids Research 29 E65–E75.
Hu Y, Ippolito JE, Garabedian EM, Humphrey PA & Gordon JI 2002 Molecular characterization of a metastatic neuroendocrine cell cancer arising in the prostates of transgenic mice. Journal of Biological Chemistry 277 44462–44474.
Ibanez de Caceres I, Dulaimi E, Hoffman AM, Al-Saleem T, Uzzo RG & Cairns P 2006 Identification of novel target genes by an epigenetic reactivation screen of renal cancer. Cancer Research 66 5021–5028.
Kuerbitz SJ, Pahys J, Wilson A, Compitello N & Gray TA 2002 Hypermethylation of the imprinted NNAT locus occurs frequently in pediatric acute leukemia. Carcinogenesis 23 559–564.
Liang G, Gonzales FA, Jones PA, Orntoft TF & Thykjaer T 2002 Analysis of gene induction in human fibroblasts and bladder cancer cells exposed to the methylation inhibitor 5-aza-2′-deoxycytidine. Cancer Research 62 961–966.
Miller MJ & Yuan BZ 1997 Semiautomated resolution of overlapping stutter patterns in genomic microsatellite analysis. Analytical Biochemistry 251 50–56.
Nishida Y, Yoshioka M & St-Amand J 2005 The top 10 most abundant transcripts are sufficient to characterize the organs functional specificity: evidences from the cortex, hypothalamus and pituitary gland. Gene 344 133–141.
Simpson DJ, Bicknell JE, McNicol AM, Clayton RN & Farrell WE 1999 Hypermethylation of the p16/CDKN2A/MTSI gene and loss of protein expression is associated with nonfunctional pituitary adenomas but not somatotrophinomas. Genes, Chromosomes and Cancer 24 328–336.
Simpson DJ, Hibberts NA, McNicol AM, Clayton RN & Farrell WE 2000 Loss of pRb expression in pituitary adenomas is associated with methylation of the RB1 CpG island. Cancer Research 60 1211–1216.
Simpson DJ, McNicol AM, Murray DC, Bahar A, Turner HE, Wass JA, Esiri MM, Clayton RN & Farrell WE 2004 Molecular pathology shows p16 methylation in nonadenomatous pituitaries from patients with Cushing's disease. Clinical Cancer Research 10 1780–1788.
Suh YH, Kim WH, Moon C, Hong YH, Eun SY, Lim JH, Choi JS, Song J & Jung MH 2005 Ectopic expression of neuronatin potentiates adipogenesis through enhanced phosphorylation of cAMP-response element-binding protein in 3T3-L1 cells. Biochemical and Biophysical Research Communications 337 481–489.
Suzuki H, Gabrielson E, Chen W, Anbazhagan R, van Engeland M, Weijenberg MP, Herman JG & Baylin SB 2002 A genomic screen for genes upregulated by demethylation and histone deacetylase inhibition in human colorectal cancer. Nature Genetics 31 141–149.
Takeuchi S, Bartram CR, Wada M, Reiter A, Hatta Y, Seriu T, Lee E, Miller CW, Miyoshi I & Koeffler HP 1995 Allelotype analysis of childhood acute lymphoblastic leukemia. Cancer Research 55 5377–5382.
Usui H, Morii K, Tanaka R, Tamura T, Washiyama K, Ichikawa T & Kumanishi T 1997 cDNA cloning and mRNA expression analysis of the human neuronatin. High level expression in human pituitary gland and pituitary adenomas. Journal of Molecular Neuroscience 9 55–60.
Wijnholds J, Chowdhury K, Wehr R & Gruss P 1995 Segment-specific expression of the neuronatin gene during early hindbrain development. Developmental Biology 171 73–84.
Wood WM, Sarapura VD, Dowding JM, Woodmansee WW, Haakinson DJ, Gordon DF & Ridgway EC 2002 Early gene expression changes preceding thyroid hormone-induced involution of a thyrotrope tumor. Endocrinology 143 347–359.
Yoshioka N, Inoue H, Nakanishi K, Oka K, Yutsudo M, Yamashita A, Hakura A & Nojima H 2000 Isolation of transformation suppressor genes by cDNA subtraction: lumican suppresses transformation induced by v-src and v-K-ras. Journal of Virology 74 1008–1013.
Zhang X, Zhou Y, Mehta KR, Danila DC, Scolavino S, Johnson SR & Klibanski A 2003 A pituitary-derived MEG3 isoform functions as a growth suppressor in tumor cells. Journal of Clinical Endocrinology and Metabolism 88 5119–5126.
Zhao J, Dahle D, Zhou Y, Zhang X & Klibanski A 2005 Hypermethylation of the promoter region is associated with the loss of MEG3 gene expression in human pituitary tumors. Journal of Clinical Endocrinology and Metabolism 90 2179–2186.
Zhong S, Fields CR, Su N, Pan YX & Robertson KD 2006 Pharmacologic inhibition of epigenetic modifications, coupled with gene expression profiling, reveals novel targets of aberrant DNA methylation and histone deacetylation in lung cancer. Oncogene 19 2621–2634.